71815 v1 BACKGROUND PAPER 15 (PHASE I) Financing Public Infrastructure in Sub-Saharan Africa: Patterns and Emerging Issues Cecilia Briceño-Garmendia, Karlis Smits, and Vivien Foster JUNE 2008 © 2009 The International Bank for Reconstruction and Development / The World Bank 1818 H Street, NW Washington, DC 20433 USA Telephone: 202-473-1000 Internet: www.worldbank.org E-mail: feedback@worldbank.org All rights reserved A publication of the World Bank. The World Bank 1818 H Sreet, NW Washington, DC 20433 USA The findings, interpretations, and conclusions expressed herein are those of the author(s) and do not necessarily reflect the views of the Executive Directors of the International Bank for Reconstruction and Development / The World Bank 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 publication is copyrighted. Copying and/or transmitting portions or all of this work without permission may be a violation of applicable law. The International Bank for Reconstruction and Development / The World Bank encourages dissemination of its work and will normally grant permission to reproduce portions of the work promptly. For permission to photocopy or reprint any part of this work, please send a request with complete information to the Copyright Clearance Center Inc., 222 Rosewood Drive, Danvers, MA 01923 USA; telephone: 978-750-8400; fax: 978-750-4470; Internet: www.copyright.com. All other queries on rights and licenses, including subsidiary rights, should be addressed to the Office of the Publisher, The World Bank, 1818 H Street, NW, Washington, DC 20433 USA; fax: 202-522-2422; e-mail: pubrights@worldbank.org. About AICD This study is a product of the Africa Infrastructure Country Diagnostic (AICD), a project designed to expand the world’s knowledge of physical infrastructure in Africa. AICD will provide a baseline against which future improvements in infrastructure services can be measured, making it possible to monitor the results achieved from donor support. It should also provide a better empirical foundation for prioritizing investments and designing policy reforms in Africa’s infrastructure sectors. AICD is based on an unprecedented effort to collect detailed economic and technical data on African infrastructure. The project has produced a series of reports (such as this one) on public expenditure, spending needs, and sector performance in each of the main infrastructure sectors—energy, information and communication technologies, irrigation, transport, and water and sanitation. Africa’s Infrastructure—A Time for Transformation, published by the World Bank in November 2009, synthesizes the most significant findings of those reports. AICD was commissioned by the Infrastructure Consortium for Africa after the 2005 G-8 summit at Gleneagles, which recognized the importance of scaling up donor finance for infrastructure in support of Africa’s development. The first phase of AICD focused on 24 countries that together account for 85 percent of the gross domestic product, population, and infrastructure aid flows of Sub- Saharan Africa. The countries are: Benin, Burkina Faso, Cape Verde, Cameroon, Chad, Côte d'Ivoire, the Democratic Republic of Congo, Ethiopia, Ghana, Kenya, Lesotho, Madagascar, Malawi, Mozambique, Namibia, Niger, Nigeria, Rwanda, Senegal, South Africa, Sudan, Tanzania, Uganda, and Zambia. Under a second phase of the project, coverage is expanding to include as many other African countries as possible. Consistent with the genesis of the project, the main focus is on the 48 countries south of the Sahara that face the most severe infrastructure challenges. Some components of the study also cover North African countries so as to provide a broader point of reference. Unless otherwise stated, therefore, the term “Africa� will be used throughout this report as a shorthand for “Sub-Saharan Africa.� The World Bank is implementing AICD with the guidance of a steering committee that represents the African Union, the New Partnership for Africa’s Development (NEPAD), Africa’s regional economic communities, the African Development Bank, the Development Bank of Southern Africa, and major infrastructure donors. Financing for AICD is provided by a multidonor trust fund to which the main contributors are the U.K.’s Department for International Development, the Public Private Infrastructure Advisory Facility, Agence Française de Développement, the European Commission, and Germany’s KfW Entwicklungsbank. The Sub-Saharan Africa Transport Policy Program and the Water and Sanitation Program provided technical support on data collection and analysis pertaining to their respective sectors. A group of distinguished peer reviewers from policy-making and academic circles in Africa and beyond reviewed all of the major outputs of the study to ensure the technical quality of the work. The data underlying AICD’s reports, as well as the reports themselves, are available to the public through an interactive Web site, www.infrastructureafrica.org, that allows users to download customized data reports and perform various simulations. Inquiries concerning the availability of data sets should be directed to the editors at the World Bank in Washington, DC. Contents Summary iv Public infrastructure spending: the headlines v The anatomy of public spending v General government expenditure vii Budget efficiency viii The hidden cost of utilities’ inefficiencies ix Emerging messages x 1 Motivation for this study 1 2 Description of tools: data and methodology, key definitions 5 General government expenditures sourced from budget documents 7 Expenditures of public, nonfinancial corporations, sourced from financial accounts 8 Introducing a country typology 10 3 Public infrastructure spending: headlines 11 Country efforts and purchasing possibilities 11 Anatomy of public infrastructure spending 13 4 General government expenditures 23 The macro outlook and its fiscal implications for infrastructure 23 Prioritizing within the infrastructure budget envelope 28 Budget efficiency: execution and maintenance 31 5 The hidden costs of utilities’ inefficiencies 39 6 Conclusions and policy implications 52 References 54 Appendixes 57 Appendix 1. Sector scope, functional classification 58 Appendix 2. Sources of data on infrastructure expenditures 59 Appendix 3. Country groups 59 Appendix 4. Primary fiscal balances 60 Appendix 5. Net change in central government budget: breakdown by source 61 Appendix 6. Net change in central government budget: breakdown by use 62 Appendix 7. Expenditure on main road network 63 Appendix 8. Variance around the trend line of road expenditure 64 Appendix 9. Contributions to QFCs (country aggregates)—water sector 65 Appendix 10. Contributions to QFCs (country aggregates)—power sector 66 Appendix 11. Water: efficiency and production indicators, 2006 67 Appendix 12. Electricity: efficiency and production indicators, 2006 69 Appendix 13. Annual maintenance and preservation expenditures, 2001–05 70 Appendix 14. Average annual maintenance expenditures on main road network, 2001–05 71 iii Summary T o be credible, any plan for scaling up infrastructure in Africa must rest on a thorough evaluation of how fiscal resources are allocated and financed. Because in every plausible scenario the public sector retains the lion’s share of infrastructure financing, with private participation remaining limited, a central purpose of such an evaluation is to identify where and how fiscal resources can be better used—if not increased—without jeopardizing macroeconomic and fiscal stability. The stakes are high, because the magnitude of Africa’s infrastructure needs carries a commensurate potential for misuse of scarce fiscal resources. We analyze recent public expenditure patterns to identify ways to make more fiscal resources available for infrastructure. We do this in three ways. First, we quantify the level and composition of public spending on infrastructure so as to match fiscal allocations to the particular characteristics of individual subsectors and to countries’ macroeconomic type (low-income fragile, low-income nonfragile, oil-exporting, and middle-income). Second, we evaluate public budgetary spending for infrastructure against macroeconomic conditions to get a sense of the scope for making additional fiscal resources available based on actual allocation decisions in recent years. And, third, we look for ways to make public spending for infrastructure more efficient, so as to better use existing resources. The Government Finance Statistics of the International Monetary Fund are neither comprehensive nor disaggregated enough to support an analysis of the fiscal costs of infrastructure for the period 2001–06. For that reason, our analysis is based on a new, standardized cross-country dataset of fiscal indicators for infrastructure that covers, but also extends beyond, spending from central government budgets. State- owned enterprises (SOEs) and extrabudgetary financing vehicles are also covered, as are private operators, as long as the assets they operate belong to the state or the operator continues to rely on public subsidies. Expenditure by subnational jurisdictions is only partially covered, however. Data are collected in such a way as to permit cross-classification by economic categories (including capital and current spending) as well as functional categories—information and communication technologies (ICT), power, roads, water, and sanitation. As far as possible, both budgeted and actual expenditures are recorded. Any exercise of this kind encounters data limitations. First, because it was not feasible to visit all subnational entities, some decentralized infrastructure expenditures probably have been underrepresented, with particular implications for the water sector. Second, it was not always possible to fully identify which items of the budget are financed by donors, and contributions by nongovernmental organizations (NGOs) to rural infrastructure projects are likely to have been missed completely. Third, it was not always possible to obtain full financial statements for all of the infrastructure special funds that we identified. Fourth, accurate recording of annual changes in fixed capital formation (capital expenditure) of SOEs remains a methodological challenge. Fifth, accurate measurement of existing public infrastructure stock will require further methodological development. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Public infrastructure spending: the headlines Most governments in Sub-Saharan Africa spend about 6–12 percent of their gross domestic product (GDP) each year on infrastructure, understood as comprising ICT, power, roads, water, and sanitation (figure A). Roughly half spend more than 8 percent of GDP, while only a quarter of countries spend less than 5 percent, the level commonly encountered among the countries of the Organisation for Economic Co-operation and Development. Cape Verde, Ethiopia, and Namibia spend well above 10 percent of their GDP on infrastructure. In the few middle-income countries of the region for which comparative information is available the level of public spending is known to be between 6 and 8 percent of GDP. Expressed as shares of GDP, Figure A. Fiscal flows devoted to infrastructure these fiscal efforts seem larger than when put in dollar terms. Most 20 600 countries of the region spend less 18 500 than $600 million a year on 16 Spending USD per capita infrastructure services—less than 14 Spending %GDP 400 $50 per person. Among landlocked 12 10 300 countries, whose infrastructure 8 needs tend to be particularly high, 200 6 the annual total is less than $30 per 4 100 capita. These annual expenditures 2 pale in comparison with the - - Namibia Zambia Malawi Senegal Benin Rwanda Mozambique Ghana Cameroon Kenya Chad Uganda Ethiopia amounts needed. An investment Tanzania South Africa Niger Nigeria Cape Verde Lesotho Cote d'Ivoire Madagascar budget of US$100 million purchases no more than about 100 MW of electricity generation, GDP Share (%) Spending per capita or 100,000 new household connections to water and sewerage, Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). or 300 kilometers of two-lane Note: Based on annual averages for the period 2001–05. paved road. The anatomy of public spending Most public spending on infrastructure in Sub-Saharan Africa passes through SOEs. SOEs have a particularly large role in the middle-income countries, where they account for over 70 percent of all public infrastructure spending. In Namibia, for example, 90 percent of expenditures on infrastructure are made by SOEs. In non-oil-exporting low-income countries, the share of expenditures realized by SOEs is close to 60 percent, or just below two-thirds of total infrastructure spending. The bulk of the fiscal resources that pass through SOEs go for current spending. Current spending includes spending on operations and maintenance, which is essential to harness the economic returns of capital. However, most of recorded current spending relates to so-called nonproductive expenses, namely wages and salaries. High levels of recurrent spending may indicate that operational inefficiencies are diverting resources away from investment. v FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Governments are the Figure B Public infrastructure spending by sector and institution most prominent financiers of infrastructure Investment investment in Sub-Saharan Africa. Except in the 3.0 SOEs 2.5 Gral Govern't middle-income countries, 2.0 governments are GDP Shares 1.5 responsible for between 1.0 80–90 percent of total 0.5 capital investment, - MIC MIC MIC MIC Oil Exporting Oil Exporting Oil Exporting Oil Exporting LIC-Fragile LIC-Fragile LIC-Fragile LIC-Fragile consistently allocating at LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-NoFragile least 80 percent of their infrastructure budgets to Water Power ICT Transport investment. In low-income countries that are aid- dependent or that export Current Spending SOEs oil, the prevalence of 3.0 Gral Govern't governments as investors 2.5 is driven by their role in channeling external funds 2.0 GDP Shares and/or natural resource 1.5 royalties. Most external development funds are 1.0 earmarked by donors for 0.5 investment. The dominant - role of the central Oil Exporting Oil Exporting Oil Exporting Oil Exporting LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-Fragile LIC-Fragile LIC-Fragile LIC-Fragile MIC MIC MIC MIC government as an investor is consistently found in most subsectors: Water Energy ICT Transport accounting for 80 percent of total public investment Source: AICD, Fiscal Baseline (2008). in transport and water supply, and about 40 percent in energy (figure B). The noticeable exceptions to this pattern are the ICT sector and, as noted, the middle-income countries. Even though capital budgets may fall far short of actual needs, on average, most countries are not able to spend more than one-third of the budgeted amounts. For a number of countries we were able to compare actual capital spending with the amounts originally budgeted. The budget execution ratios that emerged ranged from 28 percent (Benin) to 89 percent (Madagascar), with the average being 66 percent. This means that capital spending in the region might be 50 percent higher if only government agencies had the capability to spend all of the resources allocated to them. The problems behind the low execution rates include poor planning, deficiencies in project preparation, and delays in procurement. Budget execution ratios for current spending are, on average, a little higher. vi FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Transport and energy sectors together absorb the lion’s share of infrastructure spending—about 80 percent in low-income countries. The heavy spending on power is a response to the widely recognized power crisis on the continent. The efforts of the middle-income countries to support energy development contrast starkly in absolute spending terms with those of the poorer countries. Middle-income countries spend almost 5 times more on power than do aid-dependent low-income countries. Actual spending for water may be higher than shown here, because of difficulties in capturing spending data from municipal water utilities. Sectoral allocations differ markedly across different groups of countries. Aid-dependent countries tend to show relatively high levels of investment in roads and water, which together account for 80–95 percent of donors’ allocations to infrastructure in the region. Funds from donors make up about 50 percent of water spending and 25 percent of roads spending. By contrast, donors’ commitments to the energy sector have been low or inexistent in sharp opposition to the efforts of low-income countries that by themselves have been allocating close to 25 percent of their public infrastructure budgets to power to redress chronic underinvestment in that sector. General government expenditure For several years running, a favorable external environment (notably high commodity prices) and sustained domestic economic growth averaging at least 4.5 percent annually have expanded the resources available to the governments of Sub-Saharan Africa. The economies of oil-producing countries have grown at the fastest pace (up to 15 percent a year), for obvious reasons. Non-resource-intensive countries benefited from debt relief and successful policy reforms that offset the negative impacts of higher oil prices. Even heavily indebted poor countries (HIPC) grew at an annual average rate of 5.5 percent. Domestic revenues have been the largest source of additional funds for resource-intensive countries, whereas external grants played the most significant role for the poorest countries in the region. The favorable external environment helped many countries expand their budgets. In the period 2001– 05, Sub-Saharan governments’ budgets grew by almost 1.9 percent of GDP, with the regional average driven largely by increases in middle-income countries (table A). Not all countries benefited, however. Zambia’s budget contracted by more than 8 percent, while that of the Democratic Republic of Congo chalked up a 9 percent increase. The additional budgetary resources helped low-income aid-dependent countries to bolster capital investments, including infrastructure. As a share of GDP, capital investment increased in the low-income countries by more than 1 percent in 2002–05. About 40 percent of the additional resources were allocated to clearly favored infrastructure sectors. It is striking that the oil-exporters and middle-income countries decreased their investment despite having more fiscal resources available. The oil-exporting countries lowered their capital expenditures on average by 3.3 percent of GDP. In oil-exporting countries, the decrease in budgetary expenditure was largely absorbed by a significant reduction in infrastructure expenditures. To a large extent this reflects developments in Nigeria, where infrastructure expenditures decreased by 2.2 percentage points of GDP during the study period. The middle-income countries appear to have chosen to devote more resources to vii FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA maintenance. Most of their additional capital budget was allocated outside infrastructure, but not to health and education, as the table shows. Table A Net change in central government budgets by country group, financing source, and destination, 2001–06 % GDP Net central Financing sources Spending allocations government Of which Of which expenditure domestic Of which Of which health and Country group budget revenues donor grants infrastructure education Middle-income 4.08 3.40 (0.03) 0.02 0.13 Oil-exporting (3.73) 5.25 (0.07) (1.43) (0.34) Low-income, nonfragile 1.69 0.83 1.98 0.54 0.93 Low-income fragile 3.85 0.79 1.90 0831 0.43 Africa average 1.89 3.04 0.57 (0.14) 0.24 Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Averages weighted by national GDP. Totals may not add up. — = data not available. Budget efficiency Infrastructure stocks in many of the region’s countries are sorely in need of rehabilitation after years of poor maintenance. The percentage requiring rehabilitation ranges from 12 percent (Burkina Faso) to 48 percent (Democratic Republic of Congo)—the average for the survey group is 30 percent. Rehabilitation needs are significantly higher for rural infrastructure (35 percent) than for other types (25 percent), reflecting the difficulty of maintaining assets in isolated rural areas. Because rehabilitating assets is much more costly (in present-value terms) than maintaining them well, the magnitude of the rehabilitation backlogs indicates substantial inefficiency in lifecycle spending on infrastructure. Maintenance is the most challenging aspect of road spending. In environments characterized by weak fiscal management (nontransparent and politically dominated budget processes), assets often are neglected. Because maintenance yields little observable immediate benefit and is easily deferred, its budgetary allocations often are not protected by the executive or parliament. Furthermore, in Africa, donors have a dominant role in channeling funds to the sector. They earmark much of their funding, extended on concessional terms, for investment, which has the effect of making maintenance more costly than investment, because most maintenance funds must be raised domestically. Although the share of external financing that is allocated to road rehabilitation has increased in recent years, road spending in Sub-Saharan Africa is dominated by new construction, leaving maintenance a secondary priority. Roughly half of the countries in the sample have shortfalls of 40 percent or more in annual maintenance. Expenditure shortfalls are greater than 60 percent in Chad, Uganda, and Niger. Countries that have established well-functioning road funds tend to be more successful at maintaining their road networks and reducing the volatility of spending. viii FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA The hidden cost of utilities’ inefficiencies Reducing inefficiencies in infrastructure operations is perhaps the most practical and realistic way of making more resources available for infrastructure in the region. While most countries are devoting considerable effort to improving infrastructure, they are severely constrained in what they can spend. They have trouble raising domestic revenue and in reallocating revenue from other uses, which often requires structural reforms. By contrast, efficiency improvements can quickly enlarge governments’ availability of funds, allowing them to provide new services. Because spending on infrastructure consumes a significant share of GDP, even small efficiency gains can contribute large savings. For electricity, water supply, and, to some extent, telecommunications, we measure inefficiencies by quantifying their hidden costs. For the water and power sectors, hidden costs are estimated by using the end-product approach. The methodology identifies three relevant quasi-fiscal activities in utilities: underpricing (charging less than the economic cost of the good), undercollection (where bills are never sent or allowed to go unpaid), and excessive unaccounted losses (to leaks or theft, for example). Hidden costs are then estimated by comparing actual indicators Figure C Hidden costs for water and power utilities as share of GDP of a functioning SOE Water against ideal norms of cost- recovery, collection ratios, 4.5% and distribution losses. 4.0% 3.5% For telecommunications 3.0% 2.5% utilities, we quantify the 2.0% hidden cost of labor 1.5% 1.0% redundancies by comparing 0.5% partial labor-productivity 0.0% Namibia Benin Mozambique Tanzania Senegal Burkina Faso Zambia Malawi South Africa Uganda Kenya Ghana Cote d'Ivoire Nigeria Sudan Ethiopia Rwanda Niger DRC Cape Verde Lesotho ratios of existing telecom Madagascar incumbents against world- class fixed-line providers in OECD countries. Mispricing Unaccounted Losses Collection Inefficiencies Quasi-fiscal activities in Power Africa represent average annual hidden costs of the 4.5% following (minimum) 4.0% 3.5% magnitudes: 0.5 percent of 3.0% GDP in the water sector 2.5% (figure C), 0.8 percent in 2.0% 1.5% the power sector, and 0.3 in 1.0% the telecom sector. The 0.5% 0.0% smaller economic size of Kenya Nigeria Benin Mozambique Tanzania Senegal Burkina Faso Zambia Malawi South Africa Madagascar Chad Uganda Ghana Ethiopia Rwanda Niger DRC Congo Cameroon Cape Verde Lesotho water utilities, together with skewed coverage in the Under-Pricing Unaccounted Losses Collection Inefficiencies Source: Authors’ own calculations ix using data from the AICD Database. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA sample because of decentralization and fragmentation, partially explains their lower hidden costs. Underpricing is the main source of hidden costs in both power and water utilities. Not only is underpricing inefficient, but the associated capital subsidies are hugely inequitable because access to these services is skewed toward the better off, with substantial shares of the poor remaining unconnected to the electrical grid and water supply network. In middle-income countries, unaccounted losses stand out as the greatest source of inefficiency for power utilities, particularly maintenance-deprived distribution networks. Aid-dependent countries show slightly higher levels of hidden costs relative to their peers, largely because of mispricing, and, in the water sector, poor collection practices. In the telecom sector, countries that have maintained state ownership of telecommunications incumbents, thereby deterring competition, not only are forgoing future tax revenues from expanded business activity but also are creating an additional burden of hidden costs from inefficiency (usually a bloated workforce). Such costs can exceed 0.3 percent of GDP. Emerging messages The countries of the region are devoting substantial shares of their GDP to infrastructure (6–12 percent when all sources are taken into account), but that does not amount to much in absolute terms, because the economies in question are small. On average, low-income countries are spending less than $50 per capita per year, with public investment being only a fraction of this. There is a marked division of labor between SOEs and central governments. While SOEs account for the bulk of infrastructure spending in most countries, they undertake very little capital spending. Most public investments for infrastructure continue to be made through central government budgets, with the resulting assets often transferred to SOEs for subsequent operation and maintenance. Despite a favorable budget environment, only aid-dependent countries seem to be allocating additional resources to infrastructure. The combination of a commodity boom and widespread debt relief has created substantial buoyancy in government budgets. In the case of aid-dependent countries, about 30 percent of the additional funds have been allocated to infrastructure. However, in middle-income countries almost none of the additional resources gleaned from the recent good years have gone for infrastructure. In oil-exporting countries infrastructure investment has actually fallen even as resource revenues have surged. Regardless of how windfall revenues are spent, governments in the region could substantially enlarge their fiscal space by redressing inefficiencies in infrastructure psending. Three major sources of inefficiency have been identified here: inattention to maintenance, failures to spend budgeted funds, and hidden costs. There is substantial direct and indirect evidence of undermaintenance, which leads to higher costs over the infrastructure lifecycle. On average, almost a third of the infrastructure assets of the countries of the region are in need of rehabilitation. With the present value of rehabilitating infrastructure exceeding the cost of preventive maintenance, it is easy to see that, over time, countries are spending more than they need to spend to preserve a fixed amount of infrastructure stock. x FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Second, very low ratios of execution of capital budgets point the way to an easy and budget-neutral increase in public investment—if only execution ratios can be raised. Addressing the causes of low budget execution deserves very serious attention, as solving the problem could increase public investment by 50 percent without any increase in budgeted resources. Moreover, until such deficiencies are addressed it will remain difficult to achieve higher levels of investment, even if more external resources are injected. Third, the hidden costs of utilities absorb some 1.8 percent of GDP, indicating a major potential dividend in return for the right set of actions. Underpricing is by far the largest contributor to hidden costs in power and water utilities, although, as noted, unzealous bill collection and distribution losses are also important. xi 1 Motivation for this study The coordinated efforts of African countries and the international community toward achieving the Millennium Development Goals (MDGs) has drawn attention to an enormous funding challenge. The cost of achieving the MDGs is estimated at 13 percent of average GDP in Sub-Saharan Africa (Sachs and others, 2004). Looking beyond the MDGs, recent research1 estimates Sub-Saharan Africa’s aggregate infrastructure needs—both new investment and operations and maintenance (O&M)—at $75 billion a year for 2006–15, or 11.7 percent of average GDP (table 1.1). Estimates for power alone are about $43 billion a year, or 7 percent of GDP, half for the investments needed to overcome chronic shortages and to propel trade in power. For low-income countries, infrastructure needs quickly add up to 20 percent of GDP (in fragile states, this reaches an impossible 70 percent). For middle-income countries, investment needs are two-thirds of O&M, while most low-income countries have investment needs 70–80 percent higher than O&M costs. Table 1.1 Sub-Saharan Africa infrastructure needs 2006–15, by sector Water supply and sanitation Energy ICT Transport Total US$ billion a year GDP share (%) Shares Middle income 17.92 6.62 4.89 80.93 0.95 13.23 100.00 Oil exporting 18.73 8.97 16.84 41.97 3.14 38.05 100.00 LIC-nonfragile 24.15 21.40 16.87 48.42 3.54 31.17 100.00 LIC-fragile 16.38 42.92 10.96 56.99 2.34 29.71 100.00 Africa 74.90 11.69 13.39 56.90 2.57 27.14 100.00 Source: Africa Infrastructure Country Diagnostic, 2008 Note: Averages weighted by country GDPs. Totals may not add up because of rounding. The private sector’s historically limited contribution to infrastructure provision and financing underscores the importance of the public sector in meeting infrastructure needs, at least for the foreseeable future. Despite efforts and good intentions, until recently the scale of private finance was not as great as anticipated in the 1990s and early 2000s (up to 2006). Nor has it extended beyond the more lucrative areas of infrastructure, such as telecommunications, power generation, railways, and ports—or the larger and wealthier economies, such as South Africa, Nigeria, and Kenya. Since the late 1990s, a number of Sub-Saharan African countries have raised private finance for traditionally state-funded infrastructure. But this has amounted to roughly 0.8 percent of GDP per year—minuscule when compared with the approximately 4 percent needed to fill the “infrastructure financing gap,� or, in other words, the funds still needed after factoring in known financing sources (cost recovery, public expenditure, official development assistance, and financing from the private sector and countries not in the Organisation for 1 Infrastructure needs estimates per sector within the Africa Infrastructure Country Diagnostic (2008). FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Economic Co-operation and Development, OECD). Private sector contributions are also hindered by stagnation in domestic capital accumulation, traceable to aggregate savings rates that are many times lower than in other developing regions. Average saving rates from national income accounts are 11 percent in Sub-Saharan Africa, compared with 20 percent in Latin America and the Caribbean, 18 percent in South Asia, 19 percent in the Middle East and North Africa, and 34 percent in East Asia and the Pacific. The characteristics that make infrastructure industries prone to government intervention—in the form of public ownership, regulation, or both—are also widely accepted. From the supply viewpoint, infrastructure service provision is characterized by large, fixed investments (usually site- and industry- specific) and sharply increasing returns to scale. From the demand viewpoint, infrastructure services range from those with the characteristics of private goods, such as telecommunications (for which prices could be set efficiently by markets), to those that are closer to public goods, such as rural roads. Shifting fiscal resources to and from infrastructure is not free of controversy. Easterly and Servén (2003) point out the growing evidence that, in developing countries, fiscal stabilization has been achieved by compressing productive public investment—notably in infrastructure—thus sapping the potential for long-run economic growth. The authors propose a shift away from the short-term preoccupation with fiscal deficits, toward a longer-term focus on fiscal solvency.2 But the International Monetary Fund (IMF, 2004) questions the assumption that public infrastructure spending in developing countries is necessarily growth-enhancing, and suggests that efforts to make more resources available for infrastructure should focus on reallocating spending within the current fiscal envelope. Recent empirical work finds that the growth-enhancing effects of public investment in infrastructure are not always greater than those of public investment in health and education (Estache and Muñoz, 2007). Such controversy is attributable to the very nature of infrastructure, whose characteristics complicate its treatment in public finances. • Infrastructure investments are large, lumpy, and infrequent; they often take more than one budget cycle to complete. They are therefore difficult to accommodate within a single budgetary cycle and much better suited to a medium-term expenditure framework. In addition, budget allocations for multiyear investment projects may not be sustained over time, thus delaying implementation and reducing projects’ eventual rate of return. Even worse, interruptions in funding may leave a country with a graveyard of incomplete public works that never materialize into a productive asset. • Infrastructure assets require sustained preventive maintenance. Failure to maintain such assets eventually leads to deterioration and the need for major rehabilitation, which costs considerably more in present-value terms than does preventative maintenance. Nevertheless, deterioration is a gradual process, and maintenance has low visibility, creating a permanent temptation to defer such spending to accommodate more politically rewarding expenditures. • Some efficiency gains from long-lived assets materialize only with time, as higher marginal returns begin to appear. This applies to assets constructed according to 20–30 year demand projections there 2 At around the same time, Blanchard and Giavazzi (2003) advanced a similar argument with respect to the European Union’s Stability and Growth Pact, which limited budget deficits to 3 percent of GDP and formed the macro-economic underpinning of European Monetary Union. 2 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA delivered low returns early in their life cycle. Other efficiency gains depend on the successful implementation of policy decisions aimed at a more efficient execution of multiyear projects (to ensure that costs are kept to a minimum and that countries begin to reap their benefits as soon as possible), and at making sure that periodic maintenance of relatively lightly used assets (such as new roads) is not postponed. The debate over the availability and allocation of fiscal resources has particular relevance for Sub- Saharan Africa, which is faced with burgeoning infrastructure needs, limited private participation, and meager public budgets. It is in this context that fiscal space becomes relevant. The IMF has defined “fiscal space� as the budgetary room that allows a government to provide resources for a desired purpose without imperiling its financial position.3 The concept focuses on the linkages between fiscal policy trade- offs and the availability of resources in the medium-term to ensure fiscal sustainability. There are several ways to make more fiscal resources available for development needs (including infrastructure), utilizing both domestic and external sources. The Development Committee of the World Bank Group (World Bank, 2006) has proposed a framework that defines four options for central governments: (i) raise additional tax revenues, (ii) increase public sector borrowing, (iii) obtain more international aid, and (iv) improve the efficiency of current expenditures. Making fiscal resources available has an important intertemporal component, since effective use of resources today leads to increased productivity, thus generating more resources to fund tomorrow’s policy choices. Countries that make more fiscal resources available by cutting development expenditures may undermine long-term growth, thereby restricting their fiscal space in the future. Aid-dependent countries face the challenge of using donor financing to remove bottlenecks to growth and, in so doing, to expand their future fiscal space. This paper analyzes recent spending patterns to identify ways to increase the availability of fiscal resources. We do this in three steps. First, we quantify the level and composition of public infrastructure expenditures. By examining how infrastructure spending is allocated across subsectors, institutions, and expense categories, we characterize past patterns and levels of fiscal allocations in terms of subsector specificities and country types (such as resource-rich, aid-dependent). This helps identify forward-looking options for fiscal expansion and other financing alternatives Second, we assess public spending on infrastructure against the background of overall fiscal resource availability. The initial fiscal conditions for infrastructure spending are framed in a macroeconomic context so as to give the reader a realistic sense of the scope for creating additional fiscal space by increasing budgetary allocation. Fiscal resources are broken down into three observable aspects: revenue effort, access to aid, and access to financial markets. Third, we look for ways to increase the efficiency of public infrastructure spending. Changing the composition and increasing the efficiency of public spending for infrastructure are two of the most practical ways of increasing governments’ resource envelope. This study provides a cross-country comparison of the efficiency of expenditures in the water, electricity, and communications sectors. In the 3 For details, see Heller and others (2006). 3 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA water and electricity sectors, potential efficiency gains are quantified by measuring the hidden costs of state-owned service providers. To achieve our purpose, we assembled a detailed database of public spending flows, the absence of which had prevented systematic evaluation of the quality and impact of public spending in infrastructure. 4 2 Description of tools: data and methodology, key definitions The Government Finance Statistics (GFS) compiled by the International Monetary Fund (IMF) do not provide a detailed or comprehensive picture of infrastructure spending. At present, the GFS constitutes the main source of cross-country data on public finance. But the information on infrastructure presents a number of problems, particularly for Africa. First, the GFS focuses on tracking general government expenditure, whereas a large share of infrastructure spending passes through nonfinancial public corporations (parastatals). Second, even within the category of general government spending, the GFS is limited in practice to central government spending, with little reporting of subnational and special funds—two other important channels of infrastructure spending.4 Lastly, the GFS does not break down infrastructure spending by subsector or expense category. In response to these limitations, we built a new database of standardized cross-country data that seeks to give a detailed yet comprehensive picture of infrastructure spending.5 Our analysis is based on a systematic, cross-country study of public spending on infrastructure both within and beyond the bounds of central government budgets. The data-collection process is based on a standardized methodology developed and explained in detail by Briceño-Garmendia (2007). To ensure the cross-country comparability of the data, detailed templates guided the data-collection process in the field. The methodology is designed to be comprehensive in the sense of covering all relevant budgetary and nonbudgetary areas of infrastructure spending. The collection of data on fiscal spending was grounded in an overview of the institutional framework for delivering infrastructure services in each of the countries while aiming at identifying all of the channels through which public expenditure on infrastructure flows. The work began with a detailed review of the central government budget. Thereafter, financial statements were collected from all the parastatals and special funds that had been identified in the institutional review. In countries where infrastructure service providers are highly decentralized (as in the case of municipal water utilities), it proved possible to collect financial statements only from the three largest service providers. Privatized infrastructure service providers were included if a majority of their shares remained government owned, or if they continued to depend on the state for capital or operating subsidies. Thus, telecommunications incumbents are typically included, whereas mobile operators are not. In some countries, local governments have begun to play an increasing role in infrastructure service provision. It was not possible to collect comprehensive expenditure data at the local government level. However, in some cases the central government produces consolidated local government accounts. Where these do not exist, an alternative source of information is the fiscal transfers from central to local governments, which are reported in the budget and on which local governments rely heavily, given 4 Based on IMF (2001), the public sector can be roughly divided into general government and public corporations. General government comprises central, state, and local governments. Public corporations can be grouped, according to the nature of their activities, into financial corporations (engaged in providing financial services for the market) and nonfinancial corporations (engaged in producing goods and nonfinancial services). 5 Soon available online at www.AfricaInfrastructure.org. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA limited alternative sources of revenue. In some cases, transfers are earmarked for infrastructure-related spending, whereas in others the share allocated to infrastructure could only be estimated.6 Data were collected in such a way as to permit both classification and cross-classification by economic and functional categories. That is, a matrix was established so that spending on each functional category could be decomposed according to the economic nature of the expense, and vice versa. Functional classification followed as closely as possible the 4-digit category or class level of the functional classification (COFOG) proposed in the IMF’s Government Financial Statistics Manual 2001 (GFSM, 2001),7 which allowed us to identify all the major infrastructure subsectors.8 The economic classification of expenses also followed the IMF framework, permitting us to distinguish to some extent between current expenditures, capital expenditures, and various subcategories thereof.9 The institutional scope of the study builds on the definition of the public sector spelled out in the system of national accounts described by the United Nations (1993), that is, all units of general government and all public corporations. The system of national accounts defines public corporations as those in which the public authorities are considered the owners by virtue of owning all, or a majority of, shares, equity, or other form of capital (table 2.1).10 Table 2.1 Institutional distribution of gross national expenditure General government Financial Nonfinancial Nonprofit institutions Household sector corporations corporations serving household sector A. Public expenditures B. Expenditures of D. Expenditures of F. Expenditures of G. Household current of which public public financial public nonfinancial nonprofit institutions and capital expenditures on corporations corporations serving households expenditures infrastructure of which expenditures sector on infrastructure C. Expenditures of E. Expenditures of private financial private nonfinancial corporations corporations Source: United Nations, 1993. Note: Shaded area reflects expenditures of public sector. 6 For a quick summary of sources of data and information, see table 2.1. 7 Definitions and explanations of the infrastructure cost elements figuring in the database can be found in Briceño- Garmendia (2007). 8 The main categories covered in the study are electricity (0435), road transport (0451), water transport (0452), railway transport (0453), air transport (0454), pipeline and other transport (0455), communication (0460), waste water management (0520), and water supply (0630). Irrigation spending is estimated as a share of agriculture (0421). 9 Current expenditures are broken down into compensation of employees, use of goods and services, consumption of fixed capital, interest, subsidies, grants and transfers, social benefits, and other current expenditure. Capital expenditures are broken down into buildings, structures, machinery, and equipment; other fixed assets; other capital expenditure: and transfers of capital expenditures to lower levels of government. 10 See United Nations (1993) for details. The system of national accounts also states that enterprises in which the government holds less than half of the shares may still be classified as public corporations if the government controls the business by influencing all principal aspects of management. 6 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA General government expenditures sourced from budget documents As far as possible, both budget estimates and actual expenditures were recorded for the period 2001– 05. There are usually three stages in the public expenditure process. First, resources are budgeted for particular purposes. Second, funds are released from the Ministry of Finance (MoF) to the responsible institutions. Finally, resources are spent by the recipient institutions. Expenditure patterns can differ substantially across these three different stages. While actual spending is the variable of greatest interest, and the main focus of the results presented in this report, it is also important to understand how infrastructure spending is affected by the budget execution process. To this end, both budgeted and actual expenditures are recorded wherever possible. Attempts were also made to capture release figures—but with very limited success. For countries that have not yet fully implemented GFSM 2001 standards, it was necessary to perform a line-by-line recoding of the budget in consultation with the corresponding line ministries. In several countries, national budget expenditures are grouped according to programs for each line ministry (such as in Zambia). In these cases, expenditures had to be disaggregated according to infrastructure subsectors. The advantage of using GFSM 2001 functional codes is that they reflect the purpose of the expenditure in a consistent pattern that is not distorted, for example, by changes in institutional responsibilities over time. Functional areas can be spread across many institutions and are subject to reallocation from one institution to another, contingent on changes over time in institutional responsibilities and institutional frameworks. It is difficult to ensure that expenses are analyzed and classified according to their economic use. On the one hand, there is no clear-cut separation between what should be considered a capital expense, a rehabilitation expense, or even a maintenance expense. So, capital and current expenditures can be mistakenly accounted under inappropriate budget categories by the lack of clear definition of the expenditure’s nature. There may also be deliberate misclassification done in order to increase the chances of budgetary approval.11 As much as possible, the economic classifications used in each country were remapped in accord with the GFSM 2001 framework to develop a common understanding across countries. Efforts were made to avoid double-counting of transfers across levels of government. There is a danger that transfers from the central government to parastatals, special funds, and subnational governments may be reported twice: once as a central government transfer and once as an expenditure by the recipient institution. Great care was taken to match up these line items across institutions and ensure that they were counted only once—as an expenditure made by the recipient institution. Similar care was taken to eliminate double-counting subsidies that originated as central government transfers to state- owned enterprises (SOEs). 11 Unfortunately, the GFMS 2001 leaves many economic categorizations open to interpretation, particularly with regard to the uses of the expenses that are most relevant to infrastructure services (for example, rehabilitation, operations, and maintenance). In this first report, the analysis will be limited to establishing the broad distinctions between current and capital expenses. 7 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Expenditures of public, nonfinancial corporations, sourced from financial accounts Creating standardized expenditure data for public nonfinancial corporations requires using principles established in the system of national accounts. Data on the expenditures of public nonfinancial corporations are very heterogeneous, so this component of the data set was essentially built from scratch. The challenge is to measure the sources and uses of parastatals’ income and changes in fixed assets (focusing on current and cumulative accounts) in a way that permits both classification and cross- classification by economic and functional category. According to the system of national accounts, corporations’ transactions can be grouped into current accounts (production, distribution of income, use of income), accumulation accounts (changes in assets and liabilities, changes in net worth), and balance sheets (stocks of assets and liabilities, and net worth). Recurrent expenditure flows can be determined by a relatively straightforward classification of entries in the annual financial statement accounts. Standardizing capital expenditures is not a straightforward exercise. According to the system of national accounts, investments can be classified as changes in financial or nonfinancial assets. In this report, however, we focus exclusively on changes in nonfinancial assets. In particular, we seek to measure tangible nonfinancial assets, excluding inventories, consumption of own capital, and production of other intermediary goods (table 2.2). Notwithstanding these efforts, it is important to recognize the inevitable data limitations in any exercise of this kind. These limitations should be borne in mind when interpreting the results of the analysis. First, it is likely that decentralized infrastructure expenditures may be improperly covered, with particular implications for the water sector. The study does not adequately capture spending by small, decentralized service providers (for example, municipal water utilities), or spending by local governments that is not funded through fiscal transfers. This Table 2.2 Classification of assets in the system of national concern does not affect countries with highly accounts centralized administrations or national water Gross fixed capital formation utilities (such as Benin, Cape Verde, Madagascar, Consumption of fixed capital Rwanda, and Uganda). However, in some Changes in inventories countries, local utilities and subnational Nonfinancial assets Acquisitions less disposals of governments have begun to play an increased role valuables in infrastructure services. Acquisitions less disposals of nonproduced, nonfinancial assets Second, it proved difficult to fully identify Financial assets externally financed public infrastructure Source: United Nations (1993). expenditures, since not all donor aid is reported in Note: Shaded area includes a subset of assets that will be used to national budget accounts. Aid disbursements are measure capital expenditure flows. most often channeled through the central government (via government agencies or line ministries) but in several cases might bypass government agencies, leaving no accounting record for the central government. An example of a direct expenditure is funds given directly to nongovernmental organizations to support small-scale infrastructure projects in rural communities. Official government budget accounts would accurately measure only funds that are channeled through the central government, leaving a 8 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA significant proportion of other external funds uncounted. For example, in aid-dependent Rwanda only 50 percent of donor aid is reflected in the national budget. Data on the execution of the budget—money actually spent out of or over budget estimates—are not available in all countries.12 In some countries, we had problems accounting for the execution of foreign- financed expenditures.13 Third, it was not always possible to obtain full financial statements for infrastructure special funds (for example, rural electrification funds or privatization funds, which have often been used for public investment). We found that the level of transparency and accountability in the spending of infrastructure special funds was highly variable across countries and sectors. In a significant number of cases, it was not possible to obtain financial statements for these funds. However, it should be noted that this problem generally did not apply to road funds. Only in Côte d’Ivoire was it impossible to obtain a financial statement on the road fund. Fourth, capturing annual changes in fixed capital formation (capital expenditure) of SOEs using financial statements represents a methodological challenge. For one thing, financial statements often do not distinguish clearly between expenditure on financial and nonfinancial assets, making it very hard to measure actual capital expenditures. In some cases, financial statements lacked information on levels or changes in spending on assets. Review of investment plans might provide a better understanding of capital expenditures, but that would involve considerably more effort, and it would be very difficult to make a standardized comparison across countries, since these investment plans do not always accurately measure actual expenditure levels. Secondly, while SOEs are “off budget,� important transactions between public budgets and these entities nevertheless occur with regularity, particularly with regard to on-lending, capitalization, debt swaps, transfers of already built assets, and various kinds of operating subsidies. Examples exist of public subsidies made to utilities to finance capital spending but recorded as current spending, as well as examples of the opposite: capital infusions that are used to pay wages or other current costs. Fifth, there are serious constraints in supplementing expenditure flows with statistics on asset stocks. For the purpose of this study, expenditure flows may be sufficient to give us a basic understanding of short-term fiscal sustainability and efficiency issues. However, for a more thorough analysis of infrastructure expenditures and growth, or to control for initial conditions in efficiency analyses, it is of the utmost importance to be able to account for public infrastructure stocks. Accurate measurement of existing public infrastructure stocks remains a very challenging task and a natural extension of the exercise presented here. 12 Proxies for actuals are used in Zambia (estimates), Côte d’Ivoire (releases), Senegal (releases), Mozambique (estimates), and Nigeria (estimates). 13 Proxies for externally financed actual expenditures are used in Ghana (estimates), Lesotho (estimates), Tanzania (estimates), and Benin (releases). 9 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Introducing a country typology This report organizes the Sub-Saharan country sample into four nonoverlapping groups, closely following IMF (2007): oil-exporting, low-income nonfragile, low-income fragile, and (non-oil-exporting) middle-income countries (table 2.3).14 Table 2.3 Typology of countries Alternative classification for low-income Oil-exporting Low-income Low-income Middle- income Low-income Low-income countries nonfragile fragile countries coastal landlocked Angola Benin Burundi Botswana Benin Burkina Faso, Cameroon Burkina Faso Central Africa Rep. Cape Verde Comoros Burundi Chad Ethiopia Ghana Comoros Lesotho Côte d’Ivoire Central Africa Rep. Congo Kenya Congo, Dem. Rep. Mauritius Eritrea Congo, Dem. Rep. of Equatorial Guinea Madagascar of Namibia, Gambia, The Ethiopia Gabon Malawi Côte d’Ivoire Seychelles Ghana Malawi Nigeria Mali Eritrea South Africa Guinea Mali Sudan* Mozambique Gambia,The Swaziland Guinea-Bissau Niger Niger Guinea Kenya Rwanda Rwanda Guinea-Bissau Liberia Uganda Senegal Liberia Tanzania Madagascar Zambia Uganda Sao Tome and Mozambique Zimbabwe Zambia Principe Sao Tome and Sierra Leone Principe Togo Senegal Zimbabwe Sierra Leone Tanzania Togo Source: Classification as proposed in IMF (2007), except for Sudan, which was added for completeness. We also explored an alternative classification as proposed by Collier and O’Connell (2006), who argue that countries’ economic behaviors are strongly affected by their endowment of natural resources and geography. Endowment of natural resources can so define a country ‘s potential that we classify oil- exporting countries as such even if they would otherwise qualify for another group according to their income or geography. Geography is reflected in the landlocked/coastal split. The landlocked condition increases the demand for—and importance of—transport services,15 as well as their costs. Higher transport costs have a usually unaccounted negative effect on private stock creation, growth outcomes, and returns of capital.16 14 See annex 3 for definitions. 15 See Ndulu (2004). 16 See Agénor and Moreno (2006), Limão and Venables (2001), and Collier and Gunning (1999). 10 3 Public infrastructure spending: headlines Country efforts and purchasing possibilities The vast majority of African countries have annual public expenditures on infrastructure17 in the order of 6–12 percent of GDP. Roughly half of the countries spend the equivalent of more than 8 percent of their GDP, while in three out of four African countries infrastructure public spending accounts for more than 5 percent of GDP.18 Only a quarter of countries spend less than 5 percent of their GDP on infrastructure (figure 3.1). By comparison, in most OECD countries, annual public expenditure on infrastructure is less than 5 percent of GDP. The level is 6.4 percent in Chile, 5.7 percent in Turkey, and 7.6 percent in Indonesia. Large fiscal efforts, as expressed in terms of share of GDP, are incommensurate with the meager amount of infrastructure services they can provide. In dollar terms, the majority of countries spend less than $600 million on infrastructure services—less than $50 per person per year—or less than $10 per person per year in investment. Annual expenditures pale when compared to the amounts needed. An investment budget of $100 million could purchase around 100 MW of electricity generation, or 100,000 new household connections to water and sewerage, or 300 kilometers of a two-lane paved road. Figure 3.1 Public spending on infrastructure across Africa (frequency for sample) 60 80 80 70 70 50 60 60 40 50 50 30 40 40 30 30 20 20 20 10 10 10 0 0 0 below 5 % 5-8% over 8% below 600 m$ 600-2000 m$ over 2000 m$ below 50 $ 50-400 $ over 400 $ As a % of GDP In m$ per year In $ per capita Source: Africa Infrastructure Country Diagnostic, Fiscal Database (2008). Based on data presented in annex 1, table 1 (separately bound). African governments differ greatly in their ability to afford the sums they spend for infrastructure (figure 3.2). The countries of Sub-Saharan Africa are faced with a dilemma. The tax base in most countries is too narrow to finance the most obvious infrastructure needs and too narrow to support subsidized access for poor consumers (Estache, 2007). That base cannot be increased, however, without economic growth, which in turn depends on the elimination of bottlenecks caused by inadequate infrastructure. 17 Consistently with the outlined methodology, infrastructure is understood as these 4 sectors: water and sanitation, power, transport (including roads) and ICT. 18 Excluding expenditures by local governments. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 3.2 Fiscal flows devoted to infrastructure 20 600 18 500 16 Total Fiscal Spending Spending USD per capita 14 Spending Spending %GDP 400 12 GDP share per capita 10 300 ( percent) ($) 8 MIC 8.86 550.51 6 200 Oil-exporting 5.44 76.15 4 100 LIC–nonfragile 8.74 34.95 2 LIC–fragile 3.44- 46.56 - - LIC–coastal 8.06 46.92 Namibia Zambia Malawi Senegal Benin Rwanda Mozambique Ghana Cameroon Kenya Chad Uganda Ethiopia Tanzania South Africa Niger Nigeria Cape Verde Lesotho Cote d'Ivoire Madagascar LIC–landlocked 8.39 27.36 Totals 24–country sample 7.74 231.74 Totals, excluding South Africa 6.74 39.70 GDP Share (%) Spending per capita Totals excluding South Africa and Nigeria 7.93 43.11 Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, tables 1 and 2 (separately bound). In low-income countries, fiscal efforts ranging from 4 percent of GDP in Rwanda and Côte d’Ivoire to more than 10 percent of GDP in Ethiopia translate into annual spending of less than $45 per capita. These averages mask enormous differences between countries and country groups. The lowest level of per capita spending on infrastructure is found in Rwanda (about $8 per capita per year). The lowest per capita expenditures on infrastructure are found in landlocked countries, which spend, on average, $28. Low-income countries, of course, have narrow tax bases and are generally unable to attract significant private investment to crowd out the lack of public resources. In middle-income countries (including South Africa), public infrastructure spending is about ten times higher in per capita terms than in low-income countries. Middle-income countries spend roughly $550 per year per person, while for low-income countries (including both oil-exporting and nonfragile states) this is roughly $45 per year. The disparities are more shocking at the country level. Rwanda’s per capita spending is 3 and 7 percent of total spending in Namibia and Cape Verde, respectively, while Niger and Malawi only managed to spend 2.5 percent of what South Africa does.19 A country’s income level is highly correlated with public infrastructure investment (figure 3.3). The emerging results for Africa simply confirm assumptions. As a ratio, public infrastructure investment is negatively associated with GDP per capita, while per capita infrastructure spending is positively related 19 Differences in per capita spending levels shrink when expenditures are adjusted for differences in purchasing power parity. (Poor countries tend to have higher purchasing power.) Expenditure levels measured in purchasing power parity (PPP) dollars per capita are more or less the same across different country groupings: resource- intensive countries, non-resource-intensive countries, and aid-dependent countries. 12 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA with GDP per capita. These relations have been documented in the literature and essentially underscore the endogenous nature of investment and wealth. Figure 3.3 Infrastructure investment and country income 60.0 Namibia 6.0 Annual Public Infrastructure Investment 50.0 5.0 Ethiopia Zambia Uganda Mozambique Senegal Annual Public Infrastructure Investment 40.0 4.0 Chad --GDP Shares-- Niger Senegal Nigeria -- US$ per capita -- 30.0 3.0 Benin Madagascar Ghana Zambia Malawi Namibia Rwanda Kenya South Africa Nigeria 2.0 20.0 Chad Cameroon Cameroon Benin Tanzania Uganda Mozambique Cote d'Ivoire Kenya 1.0 10.0 Ghana Cote d'Ivoire Niger Ethiopia Tanzania Rwanda Malawi Lesotho Lesotho - - - 1,000 2,000 3,000 4,000 5,000 6,000 - 1,000 2,000 3,000 4,000 5,000 6,000 GDP per Capita ($) GDP per Capita ($) Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Low-income countries, so to speak, are doubly penalized in their possibilities for buying infrastructure services. On the one hand, the association between infrastructure spending and national income does not imply causality, but it does confirm that poor countries are severely limited by their own purchasing power. On the other hand, low-income countries are more ill-equipped than middle-income countries in existent, functioning infrastructure assets. This suggests that even if low-income countries counted with the quantity and quality of infrastructure that middle-income countries have, they would be at a disadvantage when trying to move forward. But they are not even there. That is why the fiscal efforts of individual countries will not go far unless other players such as traditional donors, non-OECD financiers (such as India, China and Gulf States) and the private sector act coordinately. Anatomy of public infrastructure spending Institutional background On aggregate, state-owned enterprises (SOEs) channel the largest share of public infrastructure expenditures in Sub-Saharan Africa. SOEs have a particularly large role in the middle-income countries, where they account for over 70 percent of all public spending on infrastructure. In Namibia, which has the highest absolute and relative expenditure level, 90 percent of expenditures on infrastructure are made by SOEs. In non-oil-exporting, low-income countries, the share of expenditures realized by SOEs is close to 60 percent or just below two-thirds of total infrastructure spending (figure 3.4). Large differences between expenditure levels are driven, to a large extent, by differences in the spending levels of SOEs. This is a very important finding from the fiscal viewpoint for their implications for institutional development, fiscal balances, and infrastructure planning. First, SOEs’ main stakeholder and lender of last resort is the central government. Thus, while SOEs are expected to partially or totally 13 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA recover costs, they don’t, and their net operating positions affect the country’s fiscal balances when accumulated debts trigger bail-outs, capitalizations, or simply debt-swaps between the government and the SOEs. Secondly, SOEs are expected to function on a quasi-commercial basis, playing a key role in matching infrastructure supply with country needs as dictated by the nature of business, carrying out most of the investment, particularly if they handle the highest shares of public infrastructure spending and because they are charging user fees that should cover the enterprise’s capital and operations and maintenance (O&M) needs. Figure 3.4 Public infrastructure spending by institution 20 18 Nonfinancial Public Institutions On-Budget Composition of fiscal spending 16 Nonfinancial 14 public Shares on total ( percent) On-budget institutions GDP Shares (%) 12 MIC 19.3 80.7 10 Oil exporting 52.5 47.5 8 LIC–nonfragile 36.9 63.1 6 LIC–fragile 25.4 74.6 4 LIC–coastal 32.6 67.4 2 LIC–landlocked 47.2 52.8 - Totals 24–country sample 29.2 70.8 Niger Chad Malawi Uganda Nigeria Totals excluding South Africa 45.1 52.7 Zambia Benin Ghana Rwanda Lesotho Kenya Tanzania Ethiopia Senegal Namibia Cote d'Ivoire Cameroon Madagascar Cape Verde South Africa Mozambique Totals excluding South Africa and Nigeria 35.8 61.1 Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 2. However, the bulk of SOE resources goes to current spending; between 70–85 percent of total current in public infrastructure goes through SOEs. Current spending includes O&M spending. Though the latter is essential to harness the economic returns of capital, it is often recorded as a nonproductive expense because it does not directly translate into asset creation. In fact, the crudest proxy equates public investment with productive expenditure even though, admittedly, not all investment is productive and not all current expenditure is waste. Overall, however, very high levels of recurrent spending may indicate that operational inefficiencies are diverting resources away from investment. Oil-exporting countries are an exception for institutional spending patterns as the central government accounts for the most significant portion of infrastructure expenditures. Boosted by record-high oil revenues, central governments’ expenditures are covering 70 percent of total public spending. This figure is driven by developments in Chad and Nigeria, where central governments account for close to two- thirds of all infrastructure expenditures. The relative high importance (though not dominant role) of central governments’ expenditures in low-income countries—accounting for about 40 percent of public infrastructure spending—partly derives from their in channeling external development funds. In aid- 14 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA dependent countries, the share of externally funded expenditures absorbs more than half of all central government spending on infrastructure. Governments are the Figure 3.5 Public infrastructure spending, functional and economic breakdown most prominent financiers of infrastructure investment Investment SOEs in Sub-Saharan Africa. 3.0 Gral Govern't Excluding middle-income 2.5 countries, governments 2.0 carry out between 80 and 90 GDP Shares percent of total capital 1.5 investment, consistently 1.0 allocating 80 percent of 0.5 their infrastructure budgets to investment. This pattern - LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-Fragile LIC-Fragile LIC-Fragile LIC-Fragile Oil Exporting Oil Exporting Oil Exporting Oil Exporting MIC MIC MIC MIC can be driven in low- income and oil-exporting countries by the Water Power ICT Transport government’s role in channeling external development funds and Current Spending SOEs commodities’ royalties, 3.0 Gral Govern't respectively. The majority 2.5 of external development funds are earmarked for 2.0 GDP Shares investment. The central 1.5 government is a dominant 1.0 investor in most infrastructure subsectors: 80 0.5 percent of total public - investment in transport and Oil Exporting Oil Exporting Oil Exporting Oil Exporting LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-NoFragile LIC-Fragile LIC-Fragile LIC-Fragile LIC-Fragile MIC MIC MIC MIC water supply, and about 40 percent in energy despite the recent retrenchment of Water Energy ICT Transport public sector donors (figure 3.5). A noticeable exception Source: AICD, Fiscal Baseline (2008). to this pattern is information and communication technology (ICT) and, as alluded, middle-income countries. Half of the countries invest between 3 and 4 percent of GDP; Mozambique, Uganda, Ethiopia, and Cape Verde invest over 4 percent of their GDP. These numbers are spearheaded by investments in transport and to some extent energy, certainly led by the central government itself. Cape Verde is the most salient case of high investment effort essentially doubling regional averages, with reliance on SOE investment, including the notorious publicly owned TACV, the Cape Verdian airline. 15 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 3.6 Public infrastructure capital investment by sector and institution 7 6 WSS Power ICT Transport 5 GDP shares (%) 4 3 2 1 - Lesotho Kenya Cote d'Ivoire Mozambique Tanzania Cameroon Senegal Cape Verde South Africa Ghana Chad Uganda Nigeria Benin Ethiopia Rwanda Niger Malawi Zambia Namibia Madagascar 7 General Government SOEs 6 5 GDP shares (%) 4 3 2 1 - Benin Ghana Rwanda Niger Lesotho Kenya Chad Ethiopia Senegal Namibia Malawi Uganda Cameroon Nigeria Zambia Madagascar Cape Verde South Africa Mozambique Tanzania Cote d'Ivoire Source: AICD, Fiscal Baseline (2008) Note: No data were available for SOE investment in Côte d’Ivoire, Madagascar, Chad, Senegal, and Mozambique. Sectoral background The transport and energy sectors account for the lion’s share of infrastructure spending. For low- income countries these two sectors combined absorb about 80 percent of their infrastructure expenditures 16 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA (figure 3.7). The heavy spending on power is a response to the widely recognized power crisis on the continent. The efforts of middle-income countries to support energy development contrast starkly, in fiscal terms, with those of the poorer countries. Middle-income countries spend almost 5 times more than what aid-dependent and heavily indebted poor countries (HIPC) spend on power. Low shares for water can be partly attributed to difficulties in capturing expenditure data from municipal water utilities. Figure 3.7 Public infrastructure spending by sector 20 WSS 18 Composition of fiscal spending ICT 16 Shares on total Power 14 ( percent) WSS Power ICT Transport Transport MIC 12.9 21.3 24.0 41.9 GDP Shares (%) 12 10 Oil-exporting 14.1 44.6 7.6 33.6 8 LIC–nonfragile 11.5 41.1 14.9 32.4 6 LIC–fragile 12.6 61.8 – 25.6 4 LIC–coastal 8.4 39.8 21.3 30.5 2 LIC–landlocked 18.3 45.8 6.1 29.8 - Average, Africa 12.5 30.7 19.1 37.7 Benin Niger Ghana Namibia Nigeria Tanzania Senegal Ethiopia Mozambique Lesotho Cote d'Ivoire Average, excluding South Africa 12.6 43.1 12.0 32.4 Average, excluding South Africa and Nigeria 12.3 38.5 17.3 31.9 Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 2. Most infrastructure expenditures on energy are made by SOEs and used for current expenditure. Only in countries that have large shares of foreign-donor-financed expenditures on electricity (Senegal and Tanzania) and in resource-intensive countries (Chad and Nigeria) do direct expenditures by central governments remain large (figure 3.7). The minor role of SOEs in capital investment may be partially explained by ambiguities in asset ownership. But inefficient pricing policies are probably the most important cause of the shortage of investment funds. 17 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 3.8 Aggregate expenditures on energy: sources and uses 7 Budget Nonfinancial Enterprises 6 5 GDP Shares (%) 4 3 2 1 - Namibia Benin Malawi Zambia Chad Niger Ghana Kenya Nigeria Tanzania Senegal Ethiopia Rwanda Mozambique Uganda Cameroon Madagascar Cape Verde South Africa Lesotho Cote d'Ivoire 7 Investment Current 6 5 GDP Shares (%) 4 3 2 1 - Lesotho Chad Uganda Kenya Nigeria Benin Rwanda Niger Cameroon Senegal Malawi Zambia Cape Verde Namibia Ghana Cote d'Ivoire Ethiopia Mozambique Tanzania South Africa Madagascar Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 4. Data limitations prevent us from fully capturing aggregate expenditures on water and sanitation, but a rough estimate is slightly more than 1 percent of GDP. Subnational entities account for a large share of overall spending on water and sanitation, which complicates cross-country comparisons of spending levels. The data provided in figure 3.9 represent the lower bound of spending on water supply and sanitation services. The largest expenditures for water supply and sanitation are recorded in Niger, Namibia, and Ghana. 18 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 3.9 Aggregate expenditures on water and sanitation: sources and uses 3.5 Budget Nonfinancial Enterprises 3.0 2.5 GDP Shares (%) 2.0 1.5 1.0 0.5 - Kenya Niger Chad Nigeria Uganda Lesotho Cote d'Ivoire Benin Cameroon Rwanda Malawi Senegal Zambia Ethiopia Ghana South Africa Tanzania Namibia Cape Verde Madagascar Mozambique 3.5 Investment Current 3.0 2.5 GDP Shares (%) 2.0 1.5 1.0 0.5 - South Africa Lesotho Kenya Cote d'Ivoire Benin Ethiopia Mozambique Niger Tanzania Cameroon Senegal Zambia Cape Verde Namibia Madagascar Chad Uganda Ghana Nigeria Rwanda Malawi Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 3. While the public sector retains an active role in providing communication services in several cases, that role has been taken over by the private sector in most countries. In recent decades, many countries have privatized publicly owned telecommunications, but communication services are still provided by SOEs in Benin, Cameroon, Chad, Ethiopia, Ghana, Kenya, Mozambique, Namibia, and Zambia. For the African countries with public incumbents, cost-recovery and successful productive spending mostly rely on SOE governance and operational efficiency. Over 90 percent of public expenditure is channeled through SOEs, and from that amount only 25 percent goes to capital. Ethiopia and Namibia are the only 19 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA countries where central government expenditures on communication amount to a significant share of GDP (0.33 percent and 0.47 percent, respectively) (figure 3.10). Figure 3.10 Aggregate expenditures on communications: sources and uses 3.5 Budget Nonfinancial Enterprises 3.0 2.5 GDP Shares (%) 2.0 1.5 1.0 0.5 - South Africa Lesotho Kenya Nigeria Ethiopia Benin Niger Mozambique Senegal Tanzania Cameroon Zambia Malawi Cape Verde Namibia Madagascar Uganda Chad Ghana Sudan Rwanda 3.5 Investment Current 3.0 2.5 GDP Shares (%) 2.0 1.5 1.0 0.5 - Cameroon Cape Verde South Africa Lesotho Kenya Ghana Sudan Nigeria Ethiopia Benin Niger Rwanda Mozambique Senegal Tanzania Zambia Malawi Namibia Madagascar Uganda Chad Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 5. Countries with publicly-owned telecom utilities are Benin, Cameroon, Chad, Ethiopia, Ghana, Kenya, Mozambique, Namibia, South Africa, and Tanzania. 20 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 3.11 Aggregate expenditures on transport: sources and uses 12 Budget Nonfinancial Enterprises 10 8 GDP Shares (%) 6 4 2 - Niger Malawi Senegal Benin Ethiopia Nigeria Ghana Kenya Uganda Chad Lesotho Cote d'Ivoire Namibia Zambia Cameroon Tanzania Rwanda South Africa Cape Verde Mozambique Madagascar 12 Investment Current 10 8 GDP Shares (%) 6 4 2 - Niger Malawi Senegal Benin Ethiopia Nigeria Ghana Kenya Uganda Chad Lesotho Cote d'Ivoire Namibia Zambia Cameroon Tanzania Rwanda South Africa Cape Verde Mozambique Madagascar Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, table 6. Central government spending on the transport sector differs from other sectors. Funds are primarily focused on investment. The exception to this pattern is Cape Verde, where there is considerable spending on the air transport SOE. For the rest of the countries, the profile of transport spending is dominated by roads. Overall, roads absorb between 60-80 percent of total transport budgets. That might be behind the higher ratios of current-to-investment spending found in middle-income countries such as Tanzania and Kenya, all with better records in road maintenance. Also, the very high importance of roads within the transport sector underscores the impact of donors’ agendas. The limited evidence available indicates the heavy dependence of roads investment on foreign funding, which ranges from just over 50 percent in a country such as Senegal to almost 90 percent in a country such as Rwanda. The volatility of official 21 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA development assistance (ODA) flows contributes to the volatility of public investment in the sector. Thus, the very high ratio of road investment to GDP in Chad in 2003 –05, in Tanzania in 2000, and in Madagascar in 2004–05, are all associated with short-lived surges in ODA. 22 4 General government expenditures The macro outlook and its fiscal implications for infrastructure African countries have seen their economies grow at a solid 4 percent annual average in recent years, with half of the countries in the sample achieving growth rates in excess of 4.5 percent for the period 2001–05. The highest growth rates are observed in oil-exporting countries, which have benefited from rising oil prices. The completion of the oil pipeline in Chad pushed the average annual growth rate in that country over 15 percent. Côte d'Ivoire, in a postconflict situation, is the only country that has experienced a growth slowdown. Low-income countries that benefited from debt relief and successful policy reforms managed to offset the negative effects of higher oil prices; heavily indebted poor countries (HIPC) saw annual average growth rates in excess of 5.5 percent. Particular robust growth rates were observed in Tanzania and Mozambique. The middle-income countries grew more slowly than the other groups, at an annual average growth rate below 4.3 percent. But growth rates in Sub-Saharan Africa trail those in other developing regions with higher income levels. Over the last five years, economic growth in Africa has lagged considerably behind growth in other developing regions, notably East Asia, Europe, and Central Asia. Average annual growth rates in per capita GDP were only 2.5 percent in Sub-Saharan Africa, compared with 7.3 percent in East Asia and 5.24 percent in Europe and Central Asia. Furthermore, with the notable exception of South Africa, average per capita income remains below $600, which is about seven times lower than in Latin America and the Caribbean and four times lower than in Europe and Central Asia. These disparities underscore the size of the development challenge facing Sub-Saharan African countries. Furthermore, most of these countries are likely to fall short of the Millennium Development Goals (MDGs). The World Bank’s Global Monitoring Report for 2005 notes that in order to meet key MDGs by 2015, the region needs average annual growth rates of more than 7 percent over the next decade. But beyond the MDGs, the region faces important hurdles to achieve sustained growth. Inadequate public infrastructure (notably energy and transport) has been identified as a critical bottleneck to sustained growth in Sub-Saharan countries. Empirical work on African countries that explicitly models infrastructure as a growth variable finds that infrastructure strongly supports economic growth.20 Studies that address the issue of reverse causation (higher output creates higher demand for infrastructure) are scarce, but a few examine African countries. Using a sample of 52 countries and addressing the issue of reverse causation, Canning and Bennathan (2000) find rates of return to investment in electrical generating capacity of around 40 percent. For most countries, this rate was roughly the same as the return on noninfrastructure capital. But for the 11 African countries in the sample, the average rate of return to generating capacity was 53 percent—1.3 times higher than the African rate of return to noninfrastructure capital. Even higher rates of return were found for investment in paved roads, the African average being 69 percent, about 1.7 times higher than returns from noninfrastructure capital. 20 For details see Estache and Wodon (2006) and references therein. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Furthermore, Sachs (2004) has identified high transport costs as one of the five structural reasons for persistent poverty traps in Sub-Saharan Africa. Efforts to provide basic infrastructure and meet development needs are frustrated by weak public finances. Compared with other developing regions, public financing capabilities in Sub-Saharan Africa are characterized by weak domestic revenue collection, limited borrowing capacity, and a high reliance on external grants. Countries are prone to negative fiscal balances, especially in low-income countries. Domestic revenue generation remains weak in comparison to other developing countries; nonetheless, domestic revenues constitute the most important source of fiscal resources, particularly for oil-exporting and middle-income countries (table 4.1). Table 4.1 Primary fiscal balance (share of GDP) Revenues Budgetary expenditures, excl. Primary balance, Domestic revenues Grants debt service Primary balance excl. grants MIC 23.62 0.06 21.21 2.46 2.41 Oil–exporting 35.38 0.19 30.55 5.02 4.83 LIC–nonfragile 15.22 4.84 21.27 (1.21) (6.04) LIC–fragile 12.22 1.01 12.26 0.98 (0.03) LIC–coastal 15.50 3.25 19.23 (0.48) (3.74) LIC–landlocked 13.24 5.79 20.45 (1.41) (7.20) Africa average 23.29 1.33 22.57 2.06 0.72 Excl. South Africa 23.34 2.55 24.31 1.59 (0.97) Excl. South Africa and Nigeria 15.72 3.62 19.75 (0.41) (4.03) Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Net change over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. Yet, external grants play a significant role as a budgetary financing source in the poorest countries of Sub-Saharan Africa. External grants account for about one-third of the fiscal resources available in low- income countries. The share of foreign grants in the overall fiscal resource envelope ranges from more than 40 percent in Rwanda to less than 5 percent in the middle-income countries of the sample. Foreign grants exceed 30 percent of fiscal resources in eight countries of the sample. Excluding South Africa and Nigeria, foreign grants account, on average, for more than 3.5 percent of GDP in the study countries. The highest levels of foreign grants are found in low-income countries recovering from conflict and in landlocked countries. The favorable external environment has helped countries expand their available budgets. In the period 2001–05, central government budgets grew by almost 1.9 percent of GDP, driven largely by increases in middle-income countries (table 4.2). Paired with continued progress in economic reforms in some countries (such as Ghana and Tanzania), the positive international environment strengthened domestic investment and increased capital inflows, easing pressure on fiscal policy, fostering economic growth, and increasing fiscal space in most countries of the region. Changes in available budgets differ significantly 24 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA across nations, ranging from a 9 percent increase in the Democratic Republic of Congo to a contraction of more than 8 percent in Zambia.21 Table 4.2 Net change in central government budget: breakdown by financing sources (share of GDP) Financing sources Net expenditure budget Domestic revenues Oil revenues Grants Net borrowing MIC 4.08 3.40 – (0.03) 0.71 Oil–exporting (3.73) 5.25 7.74 (0.07) (8.92) LIC–nonfragile 1.69 0.83 – 1.98 (1.12) LIC–fragile 3.85 0.79 0.61 1.90 1.16 LIC–coastal 2.81 0.87 0.17 1.06 0.88 LIC-landlocked 0.75 0.73 – 3.59 (3.56) Africa average 1.89 3.04 1.67 0.57 (1.71) Excl. South Africa (0.33) 2.66 3.20 1.09 (4.07) Excl. South Africa and Nigeria 1.49 0.93 0.19 1.54 (0.99) Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Net change over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. In oil-exporting countries, additional resources from record-high oil prices supported debt-reduction efforts. A significant portion of oil-price-triggered increases in domestic revenues has been used to pay back debt, narrowing in tandem the net budget available for nondebt expenditures. This arrangement clearly makes remaining debt more sustainable in the long run, as reflected in the decrease in net borrowing, perhaps at the risk of postponing productive investment in infrastructure. In non-oil-exportion, low-income countries, domestic revenues have increased only marginally, but the slow pace has been compensated by increases in foreign grant inflows. For the other low-income countries, and to some extent middle-income countries, further increases in domestic revenue increases are contingent to institutional reforms to make revenue collection more effective and widen the tax base. Sub-Saharan Africa can only somewhat rely on favorable external conditions, as those seen in recent years. Marginal propensity to invest in infrastructure In Sub-Saharan Africa, fiscal adjustments might be achieved though capital investments, and to some extent, cuts in infrastructure investments. While more research is needed to prove this as a pattern, the infrastructure investment driving fiscal balance behaviors are not a new phenomenon (figure 4.1). The macroeconomic literature has long acknowledged that fiscal adjustment episodes tend to include disproportionate public investment cuts.22 For example, Calderón and Servén (2004), based on averages 21 See annex 11 for details. 22 Serven (2006) and Hicks (1991) summarizes the facts on Latin American and other developing countries. For industrialized counties, see also Roubini and Sachs (1989); De Haan, Sturn, and Sikken (1996) document the experience of industrialized countries. 25 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA for eight Latin American countries, show that cuts in infrastructure investment amounted to about 40 percent of the observed fiscal adjustment between the early 1980s and late 1990s. This was remarkable because public infrastructure investment represented less than 25 percent of overall public expenditure in Latin American countries. The other side of this adjustment pattern is the volatility in annual capital expenditures. Figure 4.1 Changes in infrastructure investment, primary surplus, and government budgets 2.5 2.5 Ethiopia Ethiopia Madagascar 2.0 2.0 Changes in Infrastructure Investment (GDP Shares) Changes in Infrastructure Investment (GDP Shares) 1.5 1.5 Tanzania Tanzania 1.0 1.0 Uganda Ghana Cameroon Uganda Rwanda Cameroon Rwanda Ghana 0.5 Mozambique 0.5 Cote d'Ivoire Mozambique Malawi Cote d'Ivoire Malawi Niger Benin Chad Chad Benin Niger - South Africa Namibia Lesotho Lesotho - Namibia South Africa (10) (5) Senegal - 5 10 15 Senegal Kenya (8) (6) (4) (2) - 2 Kenya 4 6 8 10 (0.5) Cape Verde (0.5) Cape Verde (1.0) (1.0) (1.5) (1.5) (2.0) Zambia Zambia (2.0) Nigeria Nigeria (2.5) (2.5) Changes in Primary Surplus (GDP Shares) Changes in Government Budget (GDP Shares) Source: AICD, Fiscal Baseline (2008), processed. A vast majority of African countries show strong volatility of infrastructure investment despite the multiyear nature of infrastructure projects and the potential effect of time lags. Changes in the overall budget envelope are correlated with changes in infrastructure investment, consistent with the well- documented procyclical behavior of infrastructure investment (figure 4.2). In this context, half of the countries in the sample decreased their capital expenditures,23 and with the exception of Chad, Namibia, and Uganda, cuts in capital spending were accompanied by cuts in capital spending on infrastructure.24 Thus, the additional budgetary resources helped low-income countries to bolster capital investments. Capital investment as a share of GDP increased, in the low-income countries in 2002–05, by more than 1 percent of GDP, about 40 percent of which went toward addressing infrastructure bottlenecks (table 4.4). It is striking that the oil-exporting countries and middle-income countries decreased investment despite having more fiscal resources available. Middle-income countries seem to have chosen to strengthen maintenance. Capital expenditures in oil-exporting countries—led by Nigeria—decreased on 23 For details see annex 11. 24 In Chad, a decrease in budgetary resources and a decrease in capital expenditures as shares of GDP were accompanied by an increase in nominal GDP owing to the significant increase in oil revenues after completion of the oil pipeline. When measured as a share of non-oil GDP, Chad’s fiscal resources and capital expenditures increased. 26 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA average by 3.3 percent as a share of GDP. In other words, the favorable external environment did not cause sharp or even proportional increases in expenditures in oil-exporting countries. Table 4.3 Net change in central government budgets: breakdown by economic use (share of GDP) Economic uses Net expenditure Other o/w Capital o/w budget Wages current Infrastructure expenditure Infrastructure MIC 4.08 (0.11) 4.95 (0.02) (0.75) 0.04 Oil–exporting (3.73) (1.82) (0.23) 0.03 (1.69) (1.46) LIC–nonfragile 1.69 0.05 0.49 (0.00) 1.15 0.54 LIC–fragile 3.85 0.79 2.49 0.09 0.58 0.22 LIC–coastal 2.81 0.16 1.00 0.04 1.65 0.45 LIC–landlocked 0.75 0.21 0.58 (0.03) (0.04) 0.55 Africa average 1.89 (0.39) 2.68 0.00 (0.39) (0.14) Excl. South Africa (0.33) (0.63) 0.41 0.02 (0.10) (0.31) Excl. South Africa and Nigeria 1.49 0.09 0.69 0.01 0.71 0.47 Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Net change over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. Infrastructure allocation across social sectors Only low-income countries—particularly aid-dependent countries—have allocated additional fiscal resources to clearly favored infrastructure sectors. Roughly one-third of low-income additional resources were allocated to infrastructure. The education and health sectors were benefited (table 4.5). In middle- income countries, most of the additional budget was allocated to sectors other than infrastructure and social. In oil-exporting countries, the decrease in budgetary expenditure was largely absorbed by a significant reduction in infrastructure expenditures. To a large extent this reflects developments in Nigeria, where infrastructure expenditures decreased by 2.2 percent of GDP during the study period. As previously mentioned, this might be largely explained by debt-reduction efforts. 27 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Table 4.4 Net change in central government budget: breakdown by functional category (share of GDP) Functional categories Net expenditure budget Infrastructure Education Health Other sectors MIC 4.08 0.02 0.12 0.01 3.93 Oil–exporting (3.73) (1.43) (0.17) (0.17) (1.96) LIC–nonfragile 1.52 0.54 0.69 0.24 0.07 LIC–fragile 3.85 0.32 – 0.43 3.11 LIC–coastal 2.60 0.48 0.57 (0.03) 1.57 LIC–landlocked 0.75 0.52 0.56 0.80 (1.13) Africa average 1.85 (0.14) 0.19 0.05 1.75 Excl. South Africa (0.41) (0.29) 0.28 0.08 (0.48) Excl. South Africa and Nigeria 1.37 0.49 0.40 0.23 0.26 Source: AICD, Fiscal Database, 2008. Note: Net change over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. Data on education are not available for Burkina Faso, Democratic Republic of Congo, Nigeria, or Tanzania. In conclusion, for most countries in Sub-Saharan Africa, the infrastructure sectors have benefited from additional allocations following the expansion of countries’ fiscal budgets. This might be the result of explicit policy making. It remains to be seen whether these gains in resources for infrastructure are adequate to remove the bottlenecks to growth—and whether, ultimately, they are affordable. It may well be that the tax base of most African countries is simply too narrow to permit them to cure their infrastructure problems by investing incrementally greater shares of resources from their modest fiscal space. Long-term reliance on a favorable external environment is probably not an option. Prioritizing within the infrastructure budget envelope About half of the infrastructure budget goes to the transport sector, particularly roads, which account for the second-highest share of aggregate infrastructure spending (including that provided by the nonfinancial sector) and the single largest item of central government spending. In low-income countries, power captures close to one-fourth of the infrastructure budget. This is somewhat higher than the 20 percent allocated to water and sanitation. In middle-income countries, water is undoubtedly the second- most-important focus of government; most, if not all, spending on energy is let out of the budget to nonfinancial public enterprises. 28 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.2 Central government budget by sector and use Budget Infrastructure Spending ICT 3.5 Power 3.0 WSS 2.5 Transport GDP Shares 2.0 1.5 1.0 0.5 - Oil Exporting LIC-Fragile Oil Exporting LIC-Fragile Oil Exporting LIC-Fragile MIC MIC MIC LIC-NoFragile LIC-NoFragile LIC-NoFragile Total Government Budget Investment Current Spending Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Detailed data presented in annex 1, tables 3–6. Budget prioritization is strongly influenced by donors’ agendas, particularly for roads and water. Donors tend to have a bias toward investment over maintenance and significantly favor spending on water and roads, particularly for rural areas. On average, roads and water together make up between 80 and 95 percent of donors’ allocation to the region (table 4.5). In most non-oil exporting, low-income countries, donors provide 50–60 percent of what is spent on water and sanitation. In the case of roads, donor funding accounts, on average, for one-fourth of total spending (table 4.6). In terms of shares of GDP, road spending is highest in Ethiopia, Lesotho, Mozambique, and Chad, all countries with significant donor inflows for road rehabilitation. Donors have not focused on the energy sector, while governments are allocating increasing amounts of domestically raised resources to the sector. The divergence between donors’ agendas and countries’ priorities is manifested in the spending profiles of the energy sector. Donors have kept a very low profile in the sector in an environment of low cost-recovery, low affordability, and very high risks (table 4.5). On the contrary, the governments of low-income countries have allocated close to 25 percent of their budgets to tackle the ongoing crisis of chronic underinvestment. 29 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Table 4.5 Donors’ intervention in Africa: budget relevance and composition Share of gov’t spending financed by donors Composition of donor allocations Overall Ratios in percent Budget Energy WSS Roads Total Energy Water Roads MIC 2.0 0.2 2.3 4.4 100 5 21 74 Oil–exporting 5.9 1.1 2.2 11.0 100 7 7 87 LIC–nonfragile 35.9 38.1 50.6 31.5 100 18 24 57 LIC–fragile – – – – – – – – LIC–coastal 39.6 39.4 62.2 33.4 100 22 26 52 LIC–landlocked 26.2 32.1 29.6 24.9 100 9 20 71 Africa average 18.8 10.6 24.0 21.8 100 17 23 61 Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Sub-Saharan countries might be paying a high price for relative overspending on investment in roads and water to the detriment of other priority sectors and overall maintenance. While budget allocation across sectors is as much an economic decision as a political one, undoubtedly the cost of funds is one of the most important considerations for governments when allocating funds, and donor money is, by design, concessional. The rational decision is to take advantage of cheap money. Furthermore, Sub-Saharan countries are paying a higher price for funds to the energy sector— admittedly in crisis—due to the retrenchment of donors. Each dollar raised and spent by a Sub-Saharan government has a social value premium of almost 20 percent (Warlters and Auriol, 2005). This captures the incidence of that tax on the society’s welfare (due to changes in consumption patterns) and the administrative costs of raising the money, among other things. 25 In comparison, official development assistance (ODA) and funding from the World Bank’s International Development Association (IDA) are subsidized—with respect to domestically raised funds—26 at about 70 and 55 percent, respectively.27 Funds from countries outside the Organisation for Economic Co-operation and Development (OECD), such as India and China, are approaching some African countries with quasi-concessional lending. The subsidy ratio for India and China funds is estimated at around 25 percent, while for Arab funding it is 50 percent.28 25 The marginal cost of public funds measures the “change in welfare associated with raising an additional unit of tax revenue� (Warlters and Auriol, 2005). 26 Note that, except for grants, ODA and IDA loans have to be repaid by governments using domestically raised resource. That established the basic price tag for money. 27 This is based in the standard concessional terms For instance, IDA loans charge zero interest (though 0.75 percent service charge) with 10 years of grace (Foster and others, 2008) 28 India, China and the Gulf States on average charge 4 percent, 3.6 percent, and 1.5 percent of interest and grant a 4–year grace period (Foster and others, 2008). 30 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.3 Costs of capital by funding source Cost of Capital relative to the Marginal Cost of Public Funds Public 1.2 Private 1.1 India 0.9 China 0.9 Arabs 0.6 ODA 0.5 IDA 0.3 Grants 0.0 0.0 0.2 0.4 0.6 0.8 1.0 1.2 Source: Average marginal cost of public funds as estimated by Warlters and Auriol (2005). Cost of Equity for Private Sector as in Sirtaine and other (2005) and Estache and Pinglo (2004). For the rest, author’s calculations. Budget efficiency: execution and maintenance Budget execution Infrastructure services provided directly through government budgets (or special extrabudgetary funds) face institutional challenges that can often prevent spending money even if available. Administrative and procedural difficulties can often prevent budget allocations from being ultimately released and realized. The ratios of released to budgeted expenditure and of realized to budgeted expenditure are usually know as budget-variation or budget-execution ratios. Central governments’ budget-variation ratios help measure the efficiency of public administration. Execution ratios of less than 100 percent indicate that not all of the budgeted funds were spent, either because of inefficiencies in implementation or competing pressures for the original allocation that arose during the fiscal year. Execution ratios of more than 100 percent indicate that the amount spent was greater than that budgeted, which most likely indicates a cost overrun or policy realignment. In the aggregate, African countries are not able to spend one-third of their capital budgets and one- fourth of their recurrent budgets (figure 4.4). Poor timing of project appraisals and late releases of budgeted funds because of procurement problems often prevent resources from being used in the budget year. Delays in in-year fund releases are also associated with poor project preparation, leading to changes in the original terms agreed on with contractors (such as changes in deadlines, technical specifications, budget, costs, and so on). In other cases capital budgets are reallocated to current expenditures because of political or social pressures. 31 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.4 Budget-variation ratios for capital and recurrent spending Budget Execution Ratios for Capital Budget Execution Ratios for Current Spending Madagascar 89.0 Cameroon 137.62 Ethiopia 82.7 Madagascar 115.92 Cameroon 73.8 Chad 98.17 Uganda 73.3 Ethiopia 82.47 Malawi 69.4 Ghana 82.34 Average 65.5 Kenya 65.5 Kenya 63.8 Average 72.32 Niger 60.8 Niger 69.79 Chad 60.7 Malawi 67.16 Ghana 53.3 Uganda 64.74 Benin 27.8 Benin 33.96 - 20.0 40.0 60.0 80.0 100.0 - 20.00 40.00 60.00 80.00 100.0 120.0 140.0 160.0 0 0 0 0 Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008). Note: Based on annual averages for the period 2001–05. Maintenance: the case of roads The roads sector is the most salient example of a service provided by the government. Roads are typically provided without user fees, since tolling is only economic above a relatively high minimum traffic threshold, making pricing issues less relevant. Whether due to political appeal or donor influence, there is now a generalized sense that investment in the road network in Africa might be overextended. The issue is that high levels of investment can be a problem if they displace essential maintenance of existing assets, since it is much more costly over time to rehabilitate or replace poorly maintained assets than to keep up a sound preventive maintenance regime. In Sub-Saharan Africa, the share of assets in need of rehabilitation is a striking 30 percent overall, reaching more than 40 percent for rural roads. Uganda, Nigeria, Rwanda, and the Democratic Republic of Congo have the highest shares of rehabilitation backlogs (figure 4.5). Close to half of the assets of railways and rural roads are in urgent need of rehabilitation. But roads deserve special attention. From the financial viewpoint, roads are primarily a responsibility of governments and in most cases road budgets are the most prominent single item in national budgets in Sub-Saharan Africa. Transportation is a proven key ingredient of sustainable development, and roads, for the foreseeable future, are the main transport mode in the region. 32 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.5 Infrastructure assets in need of rehabilitation (percent of total) 60% 50% Average rehabilitation index 50% 40% Average rehabilitation index 40% 30% 20% 30% 10% 20% 0% 10% Railways Rural Roads Average Rural average Main Roads Rural Water Urban Water Generation Non-rural average Irrigation 0% Cote d'Ivoire Burkina Faso Cape Verde Madagascar Tanzania Cameroon Mozambique Kenya Lesotho South Africa Ghana Chad Namibia Ethiopia Niger Benin Malawi Zambia Sudan Senegal Uganda Nigeria Rwanda DRC Source: AICD. Rehabilitation ratios only underscore the need for paying increased attention to road maintenance deficiencies, and the most challenging aspect of investing in roads is providing for their maintenance. In environments characterized by weak fiscal management practices (nontransparent and politically dominated budgeting processes), assets are rarely maintained. This is particularly true when, as in the case of long-life road networks, maintenance has little observable benefits in the short term, and therefore its budgetary allocations are not naturally protected by the executive level and the parliament.29 Furthermore, in the context of Africa, donors have a dominant role in channeling funds to the sector. Most of this funding, earmarked to investment, is concessional by nature, making the cost of additional investments cheaper than the cost of maintenance, usually raised domestically. Not surprisingly, there is a bias toward capital spending at the expense of maintenance. Except in the middle-income countries, more than two-thirds of spending on roads goes to capital expenditures; maintenance remains a secondary spending priority.30 Oil-exporting countries, especially Nigeria and Chad, spend around 80 percent of all related expenditures on capital (table 4.6). In other low-income countries, such as Ethiopia, Senegal, and Zambia, the spending on capital is 12, 24, and 36 thousand dollars per kilometer respectively—well above the continent’s average, 10 thousand dollars per kilometer. These amounts are also more than 15 times higher than the average capital spending in middle-income countries. These startling differences between middle-income countries and low-income countries can be attributed to differences in initial conditions—middle-income countries already have an established road network. However, from the perspective of fiscal policy, how each country allocates resources toward maintenance and capital expenditures has important intertemporal implications. Countries with a high 29 Gwilliam and Shalizi (1996) suggest another method might be to reduce the expected rate of return on projects resulting in fewer projects crossing the hurdle rate, since inadequate maintenance leads to lower levels of return. 30 All expenditures from road funds and current expenditures by central and local governments are assumed to be for maintenance (current expenditure), and all foreign-financed expenditures and capital expenditures by central and local governments are assumed to be for capital rehabilitation. Furthermore, we assume that all foreign-funded and central government capital expenditures are used on the primary road network. Central governments’ current expenditures (adjusted for overhead) are assumed to be transfers to local governments for maintenance of the secondary network (local roads). The distribution of road fund spending between primary and secondary networks is based on an RMF matrix. 33 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA share of capital spending on roads will have to allocate greater resources for maintenance in the future, thus straining public finances. An increase in capital expenditures undoubtedly puts additional fiscal pressure to provide larger resources on road maintenance to preserve rehabilitated road networks. Table 4.6 Economic classification of expenditures on road transport Capital expenditure Current spending Shares on total ( %) Overhead (incl. rehabiliation) (incl. maintenance) MIC 0.54 4.96 93.95 Oil–exporting 5.77 81.03 13.20 LIC–nonfragile 2.95 56.64 35.72 LIC–fragile 2.94 65.95 0.37 LIC–coastal 4.03 59.90 36.07 LIC–landlocked 1.01 55.39 17.58 Africa average 2.35 37.37 57.17 Excl. South Africa 4.12 67.33 22.70 Excl. South Africa and Nigeria 2.77 56.87 32.06 Source: AICD, Fiscal Database, 2008. Note: Annual averages over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. High capital expenditure on roads may be justified by large rehabilitation backlogs. Using the RONET model, it is possible to produce detailed estimates of the rehabilitation requirements for each country’s road network, taking into account the current distribution of network condition, and working toward a target of clearing the current rehabilitation backlog within a five-year period. On that basis, the rehabilitation requirements can be compared with the current levels of capital expenditure to determine whether these are high enough to eliminate the rehabilitation backlog within a reasonable period of time. It is important to note that this calculation is only illustrative, and is based on the assumption that the entire capital budget is devoted to network rehabilitation. While rehabilitation does tend to dominate capital spending, other types of investment do take place, including upgrading road categories or adding new roads; available data do not make it possible to know the exact split. However, the calculation is helpful in illustrating whether current levels of capital expenditure would be high enough to address the rehabilitation problem if they were fully allocated to rehabilitation works. 34 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.6 Capital expenditure as a percentage of rehabilitation requirements Percentage deviation from requirement 400% 350% 300% 250% 200% 150% 100% 50% 0% Cameroon Madagascar Mozambique Cote d'Ivoire Lesotho Tanzania Niger Nigeria Senegal Malawi Uganda Zambia Rwanda Ghana Chad Benin Kenya Ethiopia -50% -100% Rehabilitation Rehabilitation adjusted for capital execution Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008), AICD RONET Analysis, 2008. Note: Based on annual averages for the period 2001–05. When compared to norms, roughly half of the countries have shortfalls of 40 percent or more in annual maintenance (figure 4.7). It is relevant to compare current maintenance expenditures with the norm for Africa. Using the RONET model, it is possible to produce detailed estimates of the routine and periodic maintenance requirements for each country’s road network, taking into account the current distribution of network condition. On that basis, the maintenance requirements can be compared with the current levels of maintenance expenditure to determine whether these are high enough to preserve the network in good condition. It is important to note that this calculation is only illustrative, and is based on the assumption that the entire maintenance budget is spent on maintenance works at efficient unit costs. However, the calculation is helpful in illustrating whether current levels of maintenance would be high enough to preserve the network if they were efficiently spent. 35 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 4.7 Maintenance expenditure as a percentage of requirements 600% Maintenance expenditure as % 500% 400% requirements 300% 200% 100% 0% Madagascar Senegal Uganda Niger Malawi Rwanda Mozambique Chad Lesotho Namibia Nigeria Ghana Tanzania Ethiopia Kenya Benin Cameroon Zambia South Africa -100% Routine plus periodic maintenance Routine maintenance Source: Africa Infrastructure Country Diagnostic, Fiscal Baseline (2008), AICD RONET Analysis, 2008. Note: Based on annual averages for the period 2001–05. Maintenance shortfalls are more pronounced in countries without road funds and fuel levies. In cases such as Chad, Nigeria, Uganda, Niger, and Senegal, maintenance spending is less than half the normative requirements. Underspending on maintenance is evident in low-income countries (particularly the resource-rich) and in countries with difficult geographical environments and terrain. Middle-income countries tend to spend substantially above the maintenance norm. However, the problem of underspending on maintenance is by far the most pronounced in countries that lack a road fund and hence a fuel levy. Among countries with fuel levies, those with high levies do substantially better than those with low ones. If the definition of maintenance is narrowed to consider only the routine aspects of maintenance, the costs are reduced by about one half. Yet around a quarter of the countries in the sample are not devoting adequate resources to cover even routine maintenance activity. The most important vehicle for domestic funds are the so-called road funds, which are designed to sequester funds for use on maintenance. Domestic financing for roads is channeled mainly through a mix of central government transfers to local governments and spending from road funds. Road fund expenditures vary significantly in proportion and absolute amount from country to country, but, on average, less than one-fifth of total spending for roads is channeled through road funds. Most funding passes through central governments’ main budgets, either as transfers to regional governments or as capital expenditure. 36 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Table 4.7 Average expenditures for roads of which of which road GDP share ( %) Total spending external fund MIC 0.31 0.01 0.06 Oil–exporting 1.31 0.14 0.21 LIC–nonfragile 2.36 0.74 0.51 LIC–fragile 0.66 – – LIC–coastal 1.81 0.61 0.53 LIC–landlocked 2.51 0.62 0.22 Africa average 1.03 0.22 0.17 Excl. South Africa 1.76 0.43 0.33 Excl. South Africa and Nigeria 1.97 0.61 0.47 Source: AICD, Fiscal Database, 2008. Note: Annual averages over the period 2001–06. Averages weighted by countries’ GDP. Totals may not add up. Implementation of second-generation road funds seems to help protect resources for road maintenance. Although it is possible to compute an optimal maintenance requirement (various maintenance norms for any type of road), implementing a mechanism that ensures the availability of sufficient funding to meet those requirements over the life cycle of the road network is a major challenge, not only in developing countries but in many developed countries as well. Finding the right balance between new investment and maintenance has been also the focus of extensive debate.31 In fact, during the last decade, the majority of the countries of Sub-Saharan Africa made efforts to establish so-called second-generation road funds, 32 based on relaying charges from users of roads33 (rather than earmarking a proportion of general tax revenues) while incepting a separate administrative agency built upon principles of commercializing road maintenance by collecting user charges. Countries such as Ghana, Namibia, and Zambia that have made genuine efforts toward implementing second-generation road funds spend on average well above $2,000 per kilometer per year on maintenance of major networks.34 Road maintenance in these countries does not differ significantly from estimated maintenance expenditures in South Africa. Expenditures per kilometer of primary network are below $500 in Niger and Rwanda. With the notable exception of Malawi, maintenance expenditures via road funds have increased in the majority of countries, but at a lower pace than overall road expenditure. These increases on maintenance expenditures are significantly smaller than increases in capital expenditures financed by external donors 31 Gwilliam and Shalizi (1996), Benmaamar (2006), and references cited therein. 32 The predecessor to this mechanism were “first generation road funds,� which were characterized by the earmarking of certain road-related taxes and charges to a special off-budget account. Technically, however, these funds are nothing more than a separate budget line item with no separate oversight. 33 In most cases, these are indirect charges for access (various vehicle registration fees) and use (fuel levies), but there are a few examples of direct charges (e.g., toll roads). 34 Because of data limitations, we were able to clearly identify maintenance expenditures only for the main network. Estimating expenditure on maintenance on the secondary network is difficult, because it is not possible to obtain a precise estimate of the maintenance expenditures of local governments. Different degrees of data accuracy prevent reliable cross-country comparisons of maintenance expenditures on secondary roads. 37 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA or central government. The share of external financing, most of which is allocated to road rehabilitation, has increased in every country that receives external financing. Central government allocations to roads have increased in the majority of countries, but have decreased in Malawi, Madagascar, and Nigeria. Another important achievement of road funds has been to avoid the volatility of maintenance expenditures. In addition to supporting higher levels of road maintenance expenditure, there is some evidence that road funds have also helped to improve the predictability of road maintenance expenditure. Expenditure data for 2001–05 show that volatility of road fund expenditures (measured by calculating the standard deviation around the trend line) was only half that of expenditures arising from external funding and one-third that of central government allocations.35 Moreover, the volatility of road fund expenditures appears to be lower in countries that have made efforts to ensure the independence of their road funds and increased the proportion of revenues channeled directly into the road funds. In conclusion, countries in Sub-Saharan Africa are allocating a significant share of GDP to road sector expenditures, which have only increased in most countries. This resource allocation to roads is driven in part by large donor funding inflows in aid-dependent countries and robust domestic revenues in oil-exporting countries. The vast majority of this expenditure is allocated to capital expenditures, leaving maintenance a secondary priority. Countries that have made efforts at establishing well-functioning road funds tend to be more successful at ensuring optimal provision of maintenance expenditures and reducing the volatility of expenditure flows. 35 For details see annex 14. 38 5 The hidden costs of utilities’ inefficiencies There are myriad approaches to the measurement of efficiency. The most widely accepted are based on the idea that rational economic agents will pursue maximum output from a given set of inputs.36 In his seminal work, Farrell (1957) introduced the concept of production frontiers—or a possible set of optimal outputs—that provide a basis for measuring efficiency.37 His approach to efficiency measurement relies on the estimation of a production frontier for which inputs and outputs can be correctly measured. Technical efficiency refers to the inputs needed to produce a given level of output using a specific technology. Allocative efficiency measures the extent to which factors of production are used in optimal proportion, given their prices or economic returns. Accurate measurement of efficiency is extremely data-intensive and requires detailed disaggregation and precise measurement of both inputs and associated outputs. The efficiency of public spending for infrastructure services involves the simultaneous production of multiple outputs. For instance, electricity generation will have no benefit if access to the network is not created in parallel. Similarly, delivering an improved water supply implies a multioutput production function. On the input side, expenditures on infrastructure services are natural proxies for the inputs of the production functions. In the case of services provided by public enterprises, however, money flows do not capture the costs of inefficient administration or poor policy decisions, which are frequently hidden from central government accounting. These so-called hidden costs have fiscal implications and reveal the implicit use of resources across sectors and, to some extent, across economic agents. Thus, public expenditures are not always what they seem from the input or cost perspective and should be reassessed for efficiency. Hidden costs are estimated by monetizing the costs of four quasi-fiscal activities (QFAs): mispricing, unaccounted-for losses, undercollection of invoiced amounts, and labor redundancies. Hidden costs are important to gauge. Not only do they give a sense of the scope, scale, and opportunity cost of inefficient operations, but they also help to pinpoint the sources of inefficiency, which may be policy or operational in nature. From a macro perspective, estimating hidden costs is essential for any accurate assessment of a country’s budget. The majority of utilities’ hidden costs are ultimately financed by subsidies, direct or indirect. In efficiency analysis, adding hidden costs to the level of public spending provides a more realistic proxy of public resource utilization for infrastructure provision, both within and across countries. QFAs disguise the size and scope of the public sector in general and have sizeable macroeconomic effects beyond the cash-flow shocks they produce in the public sector. A permanent source of concern in public finance management, QFAs can: (i) increase financial instability, (ii) be a highly distorting factor in resource allocation, and (iii) potentially create substantial contingent liabilities (Mackenzie and Stella, 1996). They also create distorted incentives in the economy, leading to overconsumption and waste (Petri and Taube, 2003). From a macro perspective, QFAs affect the overall public sector balance without explicitly affecting the budget. More often than not, QFAs escape parliamentary scrutiny, distort the size 36 Or the converse, to produce a specific level of output at the least cost. 37 For a survey of the literature on production frontiers, see Forsund and others (1980) and references cited therein. FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA of the public sector, and derail economic policy making. To allocate scarce resources properly, it is critical to know how they are actually allocated, and this is masked by QFAs. When utilities engage in QFAs, conventional measures of borrowing requirements based on financial activities are misleading. In terms of fiscal discipline, it is critical to have an accurate estimate of the magnitude of the budget and its financing. In the presence of QFAs, the overall budget deficit (or budgetary spending in general) is understated. Cumulative debts and triggered contingent liabilities lead to ad hoc periodic capitalization and debt swaps that usually bypass parliamentary resource-allocation approvals. This implies that, in a very inefficient and nontransparent manner, fiscal resources are allocated without reference to development plans and without any economic or strategic rationale. These lumpy and ad hoc fiscal interventions divert funds from priority projects, funneling scarce resources into uses that bring zero or negative return. The hidden costs of QFAs partially quantify the often unintended and usually nontransparent transfers from the producer to the consumer, while also yielding a sense of the opportunity cost of inefficient operations at the sector and enterprise level. Hidden costs include all utility operations that could in principle be duplicated by specific budgetary measures in the form of an explicit tax, subsidy, or other direct expenditure (Mackenzie and Stella, 1996). Such costs are self-perpetuating: the absence of financial sustainability and inappropriate expansion of maintenance, rehabilitation, and service lead to further waste and create more incentives for utilities to hide their poor financial situation and seek more bailouts. Identifying and quantifying the sources of hidden costs in utilities is prerequisite to getting them on the policy agenda and eventually eliminating them. Special attention has been paid to this issue in the energy sector and in transition economies (Petri and Taube, 2003; Saavalainen and ten Berge, 2006; and Chivakul and York, 2006), motivated by the heavy burden imposed on these economies by policies that pursued universal access to electric power regardless of cost.38 For infrastructure sectors the most comprehensive attempt to estimate hidden costs outside the energy sector has been made by Ebinger (2006), who estimates the hidden costs for power, natural gas, water supply, and, to some extent, railways. She finds that for a sample of around 15 transition countries in Europe and Central Asia, QFCs averaged 4.4 percent of GDP for power, 1 percent GDP for gas, and 1.2 percent for water. Identifying and reducing inefficiencies in infrastructure operations is perhaps the most practical and realistic way to free up resources for infrastructure in Africa. Most countries are making huge efforts to improve their infrastructure. However, they are severely constrained by their ability to afford the quantities of infrastructure needed to stimulate economic growth. This first attempt to estimate hidden costs in Africa’s utilities uses, for the water and power sectors, the most common approach for quantifying hidden costs—the so-called end-product approach. Described in detailed by Ebinger (2006),39 the methodology identifies three relevant QFAs in utilities: mispricing, undercollection, and excessive unaccounted-for losses. It estimates hidden costs by comparing actual indicators of a functioning state- owned enterprise (SOE) against ideal norms of cost-recovery, collection ratios, and distribution losses. For the telecom utilities we quantify the hidden cost of labor redundancies by comparing partial labor 38 In transition economies, QFCs in the power and gas sector combined have been estimated, as a share of GDO, at between 10 percent (Kyrgyz Republic, Turkmenistan) and 20 percent (Tajikistan, Uzbekistan, Azerbajan). For details, see Petri and Taube, 2003; and Saavalainen and ten Berge, 2006. 39 See also, Petri, Gunter, and Tsyninki, 2002. 40 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA productivity ratios of existing African telecommunications incumbents against world-class fixed-line providers in the member countries of the Organisation for Economic Co-operation and Development. Hidden costs in water and power utilities Quasi-fiscal Figure 5.1 Hidden costs of water and power utilities (share of GDP) activities in Africa entail Water average annual hidden costs of 0.6 percent of 4.5% GDP in the water sector, 4.0% and 1.9 percent in the 3.5% 3.0% power sector. These 2.5% overall aggregates mask 2.0% differences across 1.5% 1.0% sectors and among 0.5% countries (figure 5.1). 0.0% Namibia Benin Mozambique Tanzania Senegal Burkina Faso Zambia Malawi South Africa Uganda Kenya Ghana Cote d'Ivoire Nigeria Sudan Ethiopia Rwanda Niger DRC Cape Verde Lesotho Madagascar Relative to GDP, hidden costs for power utilities are more than double Mispricing Unaccounted Losses Collection Inefficiencies those for water utilities. The smaller economic Power size of water utilities, together with their 4.5% miscoverage in the 4.0% 3.5% sample (due to 3.0% decentralization and 2.5% 2.0% fragmentation), partially 1.5% explains the lower 1.0% apparent losses. In the 0.5% 0.0% water sector, hidden Kenya Nigeria Benin Mozambique Tanzania Senegal Burkina Faso Zambia Malawi South Africa Madagascar Chad Uganda Ghana Ethiopia Rwanda Niger DRC Congo Cameroon Cape Verde Lesotho costs amount to no more than 1.5 percent of GDP except in the Under-Pricing Unaccounted Losses Collection Inefficiencies Democratic Republic of Source: Authors’ own calculations using data from the AICD Database. Congo (2.6 percent GDP), while in the power sector hidden costs are close to zero in South Africa, about 0.2 percent of GDP in Benin, and more than 4 percent of the GDP in Malawi. Mispricing is the main source of hidden costs in both power and water supply utilities (table 5.1). In other words, the main source of hidden costs for African utilities is linked to policy decisions. The implication is not that subsidies for the poor should be eliminated across the board, but rather that subsidies and fiscal transfers must be made explicit and accounted for in budgetary appropriations. The second main source of hidden costs is collection inefficiencies—or failure to collect bills—a cost that is 41 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA inextricably tied up with related to managerial and administrative practices. For that reason, political will and capacity building must play a key role in any approach to the problem. Table 5.1 Typology of quasi-fiscal activities and hidden costs Water sector Power sector Collection Unaccounted Collection Unaccounted Mispricing inefficiencies for losses Mispricing inefficiencies for losses Lesotho Uganda Mozambique Kenya Sudan Mozambique Senegal Mozambique Uganda Mozambique Zambia Ghana Niger Nigeria Nigeria Rwanda Tanzania Cape Verde Benin Malawi Malawi More than 33% of hidden costs from Nigeria Tanzania Nigeria More than 66% of hidden costs Rwanda Zambia Burkina Faso South Africa Rwanda DRC Benin Senegal Cape Verde More than 66 % of hidden costs from More than 33% of hidden costs from Malawi Ghana Kenya Côte d'Ivoire Burkina Faso Tanzania Niger Uganda Cameroon Lesotho Chad Madagascar Madagascar DRC from Kenya South Africa Namibia Benin Ethiopia Source: AICD Database. Aid-dependent countries show slightly higher levels of hidden costs relative to their African peers. The chief causes are mispricing and, in the water sector, collection inefficiencies. Both sources of inefficiency are distortionary and nontransparent mechanisms of transferring resources to users—and are very costly from the fiscal viewpoint. In resource-rich countries, mispricing is also the predominant source of hidden costs, whereas in middle-income countries unaccounted-for losses are the salient factor for power utilities. The latter are quite worrisome because they are associated with maintenance-deprived distribution networks. What these numbers suggest is that, consciously or unconsciously, African governments are using public utilities for quasi-fiscal purposes. The composition of hidden costs implicitly reveals rational policy decisions that are somehow related to the overall characteristics of the economy. Tariff regimes that admittedly do not allow for cost recovery (and that thus provide an implicit subsidy to consumers) have existed alongside official tolerance of arrearages in customer payments or low collection rates (an implicit tax on utilities), particularly in oil-exporting countries. Furthermore, tolerance of pilferage (an implicit subsidy of customers) adds to deferred maintenance and underinvestment (implicit borrowing from future taxpayers) as the main causes of excessive losses in distribution utilities. 42 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Table 5.2 Average quasi-fiscal hidden costs and their drivers according to different typologies Total hidden Unaccounted Collection cost losses Underpricing inefficiencies GDP share (Share of total hidden costs) WATER MIC 0.47 12.26 76.65 11.09 Oil–exporting 0.13 14.27 37.72 26.70 LIC–nonfragile 0.50 25.92 51.26 22.83 LIC–fragile 0.95 14.14 76.01 9.85 LIC–coastal 0.56 23.46 64.82 11.72 LIC–landlocked 0.60 24.49 39.23 36.28 Africa average 0.43 16.37 62.04 17.07 Excl. South Africa 0.39 19.86 48.34 23.15 Excl. South Africa and Nigeria 0.49 19.94 49.42 18.40 POWER MIC 0.02 96.42 0.67 0.24 Oil–exporting 1.46 24.94 39.48 29.35 LIC–nonfragile 1.78 34.77 43.54 21.69 LIC–fragile 1.43 8.45 - 22.29 LIC–coastal 1.26 32.13 33.01 15.70 LIC–landlocked 2.52 26.44 40.84 32.72 Africa Average Ctry 0.84 62.33 19.62 12.99 Excl. South Africa 1.60 27.93 37.53 24.84 Excl. South Africa and Nigeria 1.68 28.69 39.26 18.32 Source: Authors’ calculations using data from the AICD, Fiscal Database, 2008. Note: Annual averages over the period 2001–06. Averages weighted by countries’ GDPs. Totals may not add up. Collection rates: tolerance of end-users’ arrears? Utilities seem to be tolerant of nonpayment even though they cannot fully recover those losses via cross-subsidies. Utilities in Africa do not “officially� report significant inefficiencies due to collection problems. About 60 percent of African power and water utilities report collection ratios higher than 90 percent of total bills, with only 10 percent of the utilities reporting collection rates below 60 percent. These official figures are seemingly contradicted by household data, based on which up to 40 percent (power) and 60 percent (water) of households do not pay their utility bills (figure 5.2). Regardless, whether the latter figures compound nonpayment by illegal connections and therefore overstate the problem, the two sides of the story are difficult to reconcile. On the one hand, the data underscore failures to enforce payments. On the other hand, collection ratios inconsistent with nonpayments might just reflect cross-subsidization. Thus, cross-subsidies and relaxed collection practices seem to be joined (“If large users pay their bills and small users leave unpaid bills, what is the problem?�). By design or by default, neither collection index is a good proxy for miscollection or aggregate transfers from the producer to the consumer. 43 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 5.2 Collection and payment rates Water Power 70.0 60.0 60.0 50.0 50.0 40.0 40.0 30.0 30.0 20.0 20.0 10.0 10.0 0.0 0.0 <60 60<>90 <90 <60 60<>90 <90 Ranges (%) Ranges (%) Reported by Utility Implicit Reported by HHs Reported by Utility Implicit Reported by HHs Source: Authors’ own calculation using data from the AICD Database. A more useful assessment of collection inefficiencies compares collected bills against tariff regimes in place. A quick and dirty implicit collection measure is calculated as the ratio between collected revenues and effective tariffs. The implicit collection brings a more solid ground to the story of hidden cost by miscollection when allowing benchmarking of this type of inefficiencies against the aim of policy makers (decision variable tariff regime and level). Estimates show that about 35 percent (water) and 45 percent (power) of African utilities have implicit collection ratios higher than 90 percent, while 75 percent (water) and 90 percent(power) have implicit collection rates lower 60 percent of their potential bills. Frequency charts of implicit ratios show a less extreme distribution of payers and no payers. Hidden costs associated with collection inefficiencies are relatively small, averaging 0.1 percent of GDP for water utilities and 0.4 percent GDP for power. In water, only Zambia and Malawi have hidden collection costs of around 0.4 percent of GDP, perhaps the highest in region. In power, Burkina Faso, Uganda, Ghana, and Niger have hidden collection costs of more than 1 percent of GDP. 44 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Box 5.1 Subsidies for large electricity users: the case of Zambia The average effective power tariff in Zambia, at 3 cents per kWh, is one of the lowest in Africa. This current level does not even allow for recovering operating costs, needless to say total costs. Zambia has one of the lowest average costs of the region (due to the combination of hydropower technologies with excess generation capacity). This situation of inefficient pricing policies gets compounded by exceptionally favorable prices that the power utility ZESCO gives to mining companies, particularly Copperbelt Energy Corporation (CEC). A signed long-term agreement set mining tariffs at 2 cents per kWh, not only below cost-recovery but even one-third lower than the effective tariff for an average residential customer (100 kWh per month). As the mining sector is the recipient of 50 percent of total ZESCO sales, this translates into a conservative estimate of $30 million in annual subsidies, with a projected cumulative deficit of $926 million over the next 10 years. The Zambia case is not unique to the region. Until the year 2003, Ghana’s power distribution company VRA, was engaged in a long-term agreement with Volta Aluminum Company. VALCO was for a while VRA’s most important customer, consuming one-third of its power generation and benefiting from an electricity preferential price estimated to be half of the cost-recovery level. Source: (i) Zambia Electricity Regulator Board, 2008 Press Statement on the ERB Decision on ZESCO Application to Revise Electricity Tariffs, other Charges, Fees and Penalties (ii) World Bank (2008), Zambia Growth Infrastructure and Investments: A Role for Public Private Partnership (iii) Chivakul, M. and York, R,M 2006 Implications of Quasi-Fiscal Activities in Ghana, International Monetary Fund Working Paper, Washington D.C. Unaccounted losses: paying today for yesterday’s neglect? Two-thirds of African power utilities have transmission and distribution losses more than twice higher than the acceptable norm for the sector.40 The water sector has a U-shape frequency chart. About one quarter of the utilities achieves losses within the international norm in clear contrast to a 60 percent of the sector suffering of losses more than twice higher than the acceptable norm of the sector (figure 5.3). Excessive power transmission and distribution losses and/or excessive nonrevenue water levels are a predictable consequence of poor maintenance and rehabilitation as much as inadequate billing and metering practices. They are also a very crude evidence of the consequence of poor project appraisal (preference of investment projects over maintenance). Deferring spending on maintenance or/and misinvesting is perhaps the most perverse and difficult to track financing mechanism of utilities’ cash deficits. Mismaintenance and investment can disguise today’s poor financial positions but not the permanent decay of the service—in quantity and quality—for future customers. In fact, provisioning for subsequent maintenance and replacement investment is the most efficient way of fully grabbing the economic returns of existing assets. 40 Norms of good performing networks: 10 percent transmission and distribution for power, 20 percent of nonrevenue water for water. 45 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 5.3 Excessive distribution losses 70% 60% 50% 40% 30% 20% 10% 0% r r r r 0% he he he he ig ig ig ig -1 H H H H +/ % 0% 0% + 00 0% -3 -6 -1 10 % % % 10 30 60 Deviation from the Norm in Percent Water Power On average, hidden costs associated with unaccounted losses amount are a low 0.1 percent of GDP in the water sector, contributing by less than 20 percent to total cost inefficiencies. Only the Democratic Republic of Congo sticks out with 0.5 percent of GDP in hidden costs. For Mozambique unaccounted losses are the main source of inefficiencies, accounting for close to 50 percent of its hidden costs. In power, however, hidden costs average 0.5 percent of GDP, reaching their highest level in Niger, Ghana, and Uganda. Tariff regimes: aiming at cost-recovery? With the notable exception of Cape Verde and Uganda, tariff regimes in the water sector do not allow for cost recovery. Clearly, tariffs can be—and actually are—used as a conduit for subsidies by policy makers. From a fiscal perspective, the budget deficit (or spending) is underestimated. From a sector perspective, about 80 percent of AFrican countries (19 out of a 24 country sample) have in place cost- recovery policies to recoup overall operations and maintenance (O&M) and some investment. Nonetheless, for the large majority of water utilities, tariffs did not meet various investment-cost threshold levels. At the highest block prices, where cost-recovery is more feasible, only one-third of water utilities meet capital costs41 while if focusing in average (across user groups) effective tariffs only four water utilities (out of 40 in the sample) meet simultaneously operation, maintenance and capital costs 41 Seven utilities in South Africa, and Namibia, together with utilities in Cape Verde, Benin , Niger, Rwanda, Senegal, Burkina, Lesotho and Côte d’Ivoire. 46 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA (figure 5.4). This suggests that investment costs are either completely assumed or subsidized by the governments (central and/or local), communities, and/or NGOs.42 Figure 5.4 Average effective tariffs against historical average costs Water sector Power sector 4.50 0.35 4.00 0.30 Average Effective Tariff ($/m3) Average Effective Tariff ($/kWh) 3.50 0.25 3.00 2.50 0.20 2.00 0.15 1.50 0.10 1.00 0.05 0.50 0.00 0.00 0.00 1.00 2.00 3.00 4.00 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 Historical Average Costs ($/m3) Historical Average Costs ($/kWh) Source: Banerjee and others, AICD 2008. Source: Briceño-Garmendia and Shkratan, AICD 2008. Artificially low tariffs not only create additional economic costs but also induce over-consumption and waste of already insufficient water supply resources. Altogether, in the water sector, hidden costs due to underpricing (or mispricing) amount to an average of 0.4 percent for Africa in GDP terms and about 55 percent of total hidden costs in the sector. The hidden costs due to mispricing top 2 percent of GDP in the Democratic Republic of Congo but also reach over 0.5 percent of GDP for Malawi, Madagascar, Senegal, and Côte d’Ivoire.43 Admitting the existence of affordability issues that will continue if not addressed,44 the emerging policy issue is that these hidden costs accumulate from the fiscal viewpoint, eventually inducing instability and ad-hoc spending. The African power sector is characterized by higher prices and even higher costs. In the power sector, 6 out of 20 countries in the sample have effective tariffs aligned with average historic costs (figure 5.5). Power mispricing entails hidden costs equivalent to about 1 percent of GDP or 60 percent of total hidden costs in the sector. This is more than twice the cumulative mispricing hidden costs from the water sector (0.4 percent of GDP). Malawi (3.3 percent of GDP), Zambia (2.3 percent), Niger (1.7 percent) and Cameroon (1.6 percent) stick out as the countries creating the highest hidden costs to their economies due to mispricing.45 Nonetheless, the main concern in power is that historic costs are extremely high by international standards, being mostly driven by extremely operational costs (Foster and others, 2008). Given that investment levels carried out by utilities are known to be low, cost-recover tariffs for 42 For details Banerjee, S., Foster V, at al., 2008. 43 For detailed calculations see annex 15. 44 Thoroughly covered in AICD sister piece by Banerjee, S, Wodon, Q., at al., 2008 45 For detailed calculations see annex 16. 47 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA admittedly high operational inefficiencies implicitly assume governments would/should finance capital requirements (further tariff increases might just essentially be unaffordable or infeasible). Box 5.2 Distributional incidence of capital subsidies to power and water Widespread underpricing of power and water by utilities in Sub-Saharan Africa prevents the utilities from recovering their capital costs, or even their full operating costs, from their operating revenues. To make up the deficits, governments routinely provide substantial capital subsidies to the utilities. Subsidies for basic services such as water and electricity would be socially benign if access to those services were more evenly distributed. In fact, however, access is low and sharply skewed to wealthier households. This means that the capital subsidies are highly regressive—they benefit the rich disproportionately. Access to power and water provided by utilities is almost nonexistent across the bottom three-fifths of the household income distribution in Sub-Saharan Africa—with the result that about 80 percent of the utilities’ customers are drawn from the wealthiest 40 percent of households. Although the precise rate of subsidy distribution does not depend on access alone but also on levels of consumption and tariff structures, the highly inequitable distribution of subsidy suggested by the patterns of access is not substantially modified by tariff structures, often because the subsidized tariff bands are so wide that many well-off households fall within them. 100% 80% 60% 40% 20% 0% Q1 Q2 Q3 Q4 Q5 Piped Water Electricity An AICD team studied subsidy incidence for power utilities in 18 Sub-Saharan countries and for water utilities in 15 countries. They devised a parameter () to designate the share of the utility subsidy captured by the poor relative to their share in the population. Thus, if the poor represent half of the population and capture half of the subsidy, the parameter has a value of one. Values lower than one indicate a regressive distribution of the subsidy with the poor capturing less than their share of the population, and vice versa. The findings are sobering. In all cases studied the value of  lies well below one; ranging from as low as 0.01 to a high of 0.78. In fact, the average value of  across all countries and for both services turns out to be only 0.33 for both power and water. This means that the poor capture a share of the subsidy that is only a third of their share in the population. Sources: Banerjee and others, 2008; Wodon and others, 2008. Investment plans based on more efficient existing operational technologies might lead to lower costs, and lower cost-recovery power tariffs. A recent costing exercise of investment requirements for the East Africa and the South Africa power pools (EAPP and SAPP) provides a reference for what incremental costs could be were investment plans carried out under reasonable (benchmark) unit cost assumptions (Econ Analysis, 2008). One would expect incremental costs under investment expansions to be higher than historical costs because capital premiums are essentially nonexistent for historical costs. However, with few exceptions, in both trade and nontrade scenarios, incremental costs are significantly lower than 48 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA historical costs. In the trade scenario all countries but Madagascar and Ethiopia would face lower average costs, reduced by about 40 percent on average (figure 5.5). Salient beneficiaries of trade would be Zambia, Uganda, and Malawi, which while facing increased average costs if carrying power investment in isolation, would reduce about 25 percent of their current average costs if expanding with trade. Figure 5.5 Percent improvement of “reasonable� cost investment plans over historic average costs 100 80 60 40 20 0 -20 -40 -60 -80 -100 Rwanda Zambia Uganda Tanzania Malawi Lesotho Kenya Ethiopia Mozambique Madagascar South Africa No-trade Trade Source: Authors’ own calculation using data from the AICD database. Hidden costs can be reduced by one-fourth just by improving power technologies and therefore reducing production costs. These results suggest that current mispricing hidden costs are not a disguise for postponed investment plans but rather a cover up for very high operational inefficiencies. In the hopeful scenario of (good) scaling-up the sector, reduced “mispricing hidden costs� can potentially be a significant source of fiscal space or resources in the power sector. In summary, mispricing hidden costs in power (0.8 percent of GDP) are almost double the 0.5 percent GDP hidden costs in water. While these aggregate levels might not seen outrageously high from a macro perspective, they certainly are a substantial source of fiscal space. From a sectoral viewpoint, they hint at very important issues that claim immediate attention. For instance, long-term sector sustainability would like require revising tariff schedules and reducing cost inefficiencies by improving the quality of investments and technologies. In conclusion, tapping inefficiencies created by quasi-fiscal activities in power and water utilities could increase the value for money and availability of fiscal resources by about 1.5 percent of GDP in an average country. This translates into an estimate of over $9.5 billion across Sub-Saharan Africa. The feasibility of raising tariffs toward cost-recovery levels is contingent on political and social considerations. However, prices should start reflecting costs in order not to induce waste and overconsumption, and if subsidies are needed they should be explicit, either in transfers to the utility or, even better, subsidies targeted to those who actually need them. In the case of collection inefficiencies, policy and administrative measures can bring short-term results and immediate cash-in benefits. Obviously, this requires political commitment, first, in netting out payment arrears between government agencies and utilities, and then creating without demagogy the culture of paying for services. Reducing 49 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA distribution inefficiencies passes through a period where more spending is inevitable before more resources are made available. Distribution losses are a consequence of neglect because of poor maintenance, lack of rehabilitation, and/or poor metering and billing, all of which require timely and targeted spending to overcome. Telecommunications utilities as social buffers It is not uncommon that public utilities are used as social buffers redistributing wealth via excessive employment. This is an evident sign of labor indiscipline. The dollar value of labor redundancies—or hidden costs of excessive employment—is estimated to amount to up to 0.4 percent of GDP (Benin, Mozambique and South Africa) or cost in excess that can reach $400 per subscriber (Chad) (table 5.4). These estimates take as a norm the number of subscribers per employees seen in OECD fixed-line operators. Table 5.4 Monetary value of labor redundancies (annual average, 2001–06) OECD benchmark 700 main lines per employee Main lines per employee % of GDP $ per subscriber Benin 20 0.40 91.0 Cameroon 15 0.15 47.9 Chad 63 0.14 488.9 Ethiopia 90 0.07 4.7 Kenya 19 0.00 0.4 Mozambique 138 0.41 69.6 Namibia 22 0.38 10.5 South Africa 17 0.43 0.3 Tanzania 30 0.31 123.2 Average 46 0.26 92.9 Source: Authors’ own calculation using data from the AICD Database. Estimates only include countries with state-owned telecom utilities. These striking results for labor inefficiencies underscore the importance of strengthening external governance mechanisms that can impose discipline on SOEs behavior. The relation between increased governance and lower labor costs attributed to inefficiencies is direct (figure 5.6). 50 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Figure 5.6 Link between governance and cost of labor redundancy 600 Cost of Labor Redundancy 500 (US$/subscriber) 400 300 200 100 y = -5.513x + 248.26 0 0 20 40 60 80 100 Governance Score Source: AICD. In conclusion, telecommunications labor inefficiencies are almost 0.3 percent of GDP on average, equivalent to $2.1 billion. While public incumbents do not have to play an important role in providing telephony services, the hidden costs of labor redundancies would be borne by governments if and when the explosion of small private cellular providers continues and prices continue their downward trend. If the regulatory framework is such to deter competitors and induce discriminatory practices, the hidden labor costs would very likely then be transferred to users via higher tariffs. 51 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA 6 Conclusions and policy implications Sub-Saharan African countries are devoting substantial shares of GDP to infrastructure, but that does not amount to much in absolute terms. Countries typically devote 6-12 percent of their GDP to infrastructure, when all sources are taken into account. While these shares look relatively large compared to middle- and high-income countries, due to very low levels of GDP, they amount to very little in absolute terms. On average, low-income countries are spending less than $50 per capita per year, with public investment being only a fraction of this. There is a marked division of labor between state-owned enterprises (SOEs) and central government. While SOEs account for the bulk of infrastructure spending in most countries, they undertake very little capital spending. Most public investment on infrastructure continues to be undertaken through the central government budget, with assets often transferred to SOEs for subsequent operation and maintenance. Despite a favorable budget environment, only aid-dependent countries seem to allocate additional resources to infrastructure. The combination of a commodity boom and widespread debt relief has created substantial buoyancy in government budgets. In the case of aid-dependent countries, about 30 percent of the additional envelope has been allocated to infrastructure. However, in middle-income countries almost none of the additional budget has gone to infrastructure. Moreover, in oil-dependent countries, infrastructure investment has actually fallen substantially even as resource revenues have surged. Budget envelopes could be substantially increased by addressing substantial inefficiencies. Three major sources of inefficiency have been identified in infrastructure spending: undermaintenance, budget- execution failures, and hidden costs. There is both direct and indirect evidence of undermaintenance leading to higher lifecycle costs of infrastructure. On average, around 30 percent of the region’s infrastructure assets are in need of rehabilitation. Given that the present value of rehabilitating infrastructure is substantially higher than that associated with a sound preventive maintenance regime, this finding already indicates that, over time, countries are spending more than they should to preserve a given infrastructure stock. In the case of the road sector, it is possible to isolate maintenance expenditure with greater precision than elsewhere and to compare it to rigorously determined country-specific benchmarks. Here it becomes apparent that roughly half of the countries show shortfalls of 40 percent or more in maintenance spending. Capital budget-execution ratios of only 66 percent on average indicate potential for a budget-neutral 50 percent increase in public investment. Deficiencies in planning, project preparation, and public procurement conspire to create delays that prevent countries from spending more than two-thirds of the public investment allocated to infrastructure in the budget. Addressing the causes of low budget execution deserves very serious attention, since this alone could increase public investment by 50 percent without any increase in budgeted resources. Moreover, until such deficiencies are addressed it will remain difficult to achieve higher levels of investment even if more external resources are injected. The hidden costs of power, water, and telecommunications utilities absorb around 1.8 percent of GDP, presenting a major incentive to address inefficiency. A combination of underpricing, undercollection of revenues, and very high distribution losses relative to technical norms lead to high 52 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA hidden costs of power and water. Underpricing is by far the largest contributor in both sectors, although inefficiencies are also important. The dividend attached to solving these problems is very high—1.5 percent of GDP, and even more in some cases. Increasing efficiency would free up substantial resources for public investment without increasing budget allocations. 53 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA References Agénor, Pierre-R., and B. Moreno-Dodson. 2006. “Public Infrastructure and Growth: New Channels and Policy Implications.� Mimeo, World Bank, Washington, DC. Banerjee, S., V. Foster, Y. Ying, H. Skilling, and Q. Wodon. 2008. Achieving Cost Recovery, Equity and Efficiency in Water Tariffs: Evidence From African Utilities. Africa Infrastructure Country Diagnostic (AICD). Banerjee, S., Q. Wodon, A. Diallo, T. Pushak, H. Uddin, C. Tsimpo, and V. Foster. 2008. Access, Affordability and Alternatives: Modern Infrastructure Services in Sub-Saharan Africa. AICD. Benmaamar, M. 2006. Financing of Road Maintenance in Sub-Saharan Africa. Washington, DC: World Bank. Blanchard and Giavazzi. 2003. “Improving the Stability and Growth Pact Through a Proper Accounting of Public Investment.� Mimeo. Briceño-Garmendia, C. 2007. Fiscal Cost of Infrastructure Provision: A Practitioner’s Guide. AICD. Briceño-Garmendia, C., and M. Shkaratan. 2008. Cost Recovery and Efficiency in Electricity Tariffs: Evidence from African Utilities. AICD. Calderón, C., and L. Servén. 2004. Trends in Infrastructure in Latin America. Policy Research Working Paper 3400, World Bank, Washington, DC. Canning, D., and E. Bennathan. 2000. The Social Rate of Return on Infrastructure Investments. Policy Research Working Paper 2390, World Bank, Washington, DC. Chivakul, M., and R. York. 2006. Implications of Quasi-Fiscal Activities in Ghana. International Monetary Fund (IMF) Working Paper, Washington, DC. Collier, P., and S. O’Connell. 2006. Opportunities and Choices in Explaining African Economic Growth. Oxford: Center for Study of African Economies. Collier and Gunning. 1999. “Why has Africa Grown Slowly?� Journal of Economic Perspectives 13(3). De Haan, J., J. Sturm, and B. Sikken. 1996. “Government Capital Formation: Explaining the Decline.� Weltwirtschaftliches Archiv 132 (1): 55–74. Easterly, W., and L. Servén. 2003. The Limits of Stabilization: Infrastructure, Public Deficits and Growth in Latin America. World Bank, Stanford University, Washington, DC. Ebinger, J. 2006. Measuring Financial Performance in Infrastructure. Working Paper Series, World Bank, Washington, DC. Econ Analysis. 2008. Costing Power Infrastructure Investment Needs in Sub-Saharan Africa. AICD. Estache, A. 2007. Managing the Interaction Between Infrastructure and Fiscal Policy for Faster Growth in Africa. Background Note, AICD. 54 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Estache, A., and M. E. Pinglo. August 2004. “Are Returns To Private Infrastructure In Developing Countries Consistent With Risks Since the Asian Crisis?� Policy Research Paper 3373, World Bank, Washington, DC. Estache, A., and R. Muñoz. 2007. IMF Financial Programming versus the Growth Enhancing Power of Education, Health and Infrastructure Investment: Lesson from Senegal and Uganda. AICD. Estache, A., and Q. Wodon. 2006. Infrastructure and Poverty in Sub-Saharan Africa. Washington, DC: World Bank. Farrell, M. J. 1957. “The Measurement of Productive Efficiency.� Journal of the Royal Statistical Society, Series A (General) 120(3). Forsund, F., C. A. K Lovell, and P. Schmidt. 1980. “A Survey of Frontier Production Functions and of Their Relationship to Efficiency Measurement.� Journal of Econometrics 13: 5–25, North- Holland Publishing Company. Foster, V., et al. 2008a. “Africa’s Power Supply Crisis: Unraveling the Paradoxes.� Chapter Contribution to IMF Africa Regional Economic Outlook, Draft, Spring 2008. Foster, V., W. Butterfield, and C. Chen. 2008b. Building Bridges: China’s Growing Role as Infrastructure Financier for Africa. Public-Private Infrastructure Advisory Facility (PPIAF). Gupta, S., K. Honjo, and M. Verhoeven. 1997. “The Efficiency of Government Expenditure: Experiences from Africa.� IMF Working Paper, Washington, DC. Gwilliam, K., and Z. Shalizi. 1996. Road Funds, User Charges and Taxes. Washington, DC: World Bank. Heller, P. S., M. Katz, et al. 2006. “Making Fiscal Space Happen: Managing Fiscal Policy in a World of Scaled-Up Aid.� IMF Working Paper 270, Washington, DC. Herrera, S., and G. Pang. 2005. “Efficiency of Public Spending in Developing Countries: An Efficiency Frontier Approach.� Policy Research Working Paper 3645, World Bank, Washington, DC. Hicks, N. 1991. “Expenditure Reductions in Developing Countries Revisited.� Journal of International Development 3(1): 29–37. IMF. October 2007. Regional Economic Outlook: Sub-Saharan Africa. Washington, DC: IMF. ———. 2004. “Public Investment and Fiscal Policy.� Board Paper Prepared by Fiscal Affairs Department, IMF, Washington, DC. ———. 2001. Government Finance Statistics Manual. Washington, DC: IMF. (www.imf.org/external/pubs/ft/gfs/manual/). Limão, N., and A. Venables. 2001. “Infrastructure, Geographical Disadvantage, Transport Costs, and Trade.� The World Bank Economic Review 15(3): 451–479. Makenzie, G. A., and P. Stella. 1996. Quasi-Fiscal Operations of Public Financial Institutions. Washington, DC: IMF. 55 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Ndulu B. J. 2004. “Infrastructure, Regional integration and Growth in Sub-Saharan Africa: Dealing with the Disadvantage of Geography and Sovereign Fragmentation.� Background Paper Prepared for the Commission for Africa, unpublished. Park, B., L. Simar, and C. Weiner. 2000. “The FDH Estimator for Productivity Efficiency Scores: Asymptotic Properties.� Economic Theory 16: 855–877. Petri, M., T. Gunter, and A. Tsyninki. 2002. “Energy Sector Quasi-Fiscal Activities in Countries of the Former Soviet Union.� IMF Working Paper, Washington, DC. Petri, M., and G. Taube. 2003. “Fiscal Policy Beyond the Budget.� Emerging Markets Finance and Trade 39(1). Roubini, N., and J. Sachs. 1989. “Government Spending and Budget Deficits in the Industrial Countries.� Economic Policy 8: 99–132. Saavalainen, T., and J. ten Berge. 2006. “Quasi-Fiscal Deficits and Energy Conditionality in Selected CIS Countries.� IMF Working Paper, Washington, DC. Sachs, J., J. McArthur, G. Schmidt-Traub, M. Kruk, C. Bahadur, M. Faye, and G. McCord. 2004. “Ending Africa’s Poverty Trap.� Brookings Papers on Economic Activity. Servén, L. 2005. “Fiscal Discipline, Public Investment and Growth.� Mimeo, World Bank. Sirtaine, S., M. E. Pinglo, J. L. Guasch, and V. Foster. 2005. How Profitable Are Infrastructure Concessions In Latin America? Empirical Evidence and Regulatory Implication. PPIAF. United Nations. 2001. 1993 System of National Accounts. New York: United Nations. (http://unstats.un.org/unsd/sna1993/). Warlters, M., and E. Auriol. August 2005. “The Marginal Costs of Public Funds in Africa.� Policy Research Paper 3679, World Bank, Washington, DC. World Bank. 2005. Global Monitoring Report. Oxford University Press. World Bank. 2006. Fiscal Policy for Growth and Development: An Interim Report. Development Committee, Washington, DC: World Bank. 56 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendixes Appendix 1. Sector scope, functional classification Appendix 2. Sources of data for data for expenditures of infrastructure Appendix 3. Country groups Appendix 41. Primary fiscal balances Appendix 5. Net change in central government budget: breakdown by source Appendix 6. Net change in central government budget: breakdown by use Appendix 7. Expenditure on main road network Appendix 8. Variance around the trend line of road expenditure Appendix 9. Contributions to QFCs (country aggregates)—water sector Appendix 10. Contributions to QFCs (country aggregates)—power sector Appendix 11. Water: efficiency and production indicators, 2006 Appendix 12. Electricity: efficiency and production indicators, 2006 Appendix 13. Annual maintenance and preservation expenditures, 2001–05 Appendix 14. Average annual maintenance expenditures on main road network, 2001–05 57 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 1. Sector scope, functional classification Included in cross-country Code Category comparison 704 Economic affairs 7042 Agriculture, Forestry, Fishing and Hunting Irrigation Systems (a share of 70421 Agriculture) no 7047 Other Industries 70474 Multipurpose development projects no 7043 Fuel and energy 70431 Coal and other solid mineral fuels yes 70432 Petroleum and natural gas yes 70433 Nuclear fuel yes 70434 Other fuels yes 70435 Electricity yes 70436 Nonelectric energy yes 7045 Transport 70451 Road transport yes 70452 Water transport no 70453 Railway transport no 70454 Air transport no 70455 Pipeline and other transport no 7046 Communication 70460 Communication yes 705 Environmental protection 7052 Waste water management yes 70520 Waste water management yes 706 Housing and community amenities 7063 Water supply 70630 Water supply yes 58 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 2. Sources of data on infrastructure expenditures General government sector Nonfinancial corporations sector Source of documents Consolidated budget of general Financial statements of public nonfinancial government (or separate budget books of corporations: each government unit): Balance sheet Recurrent budget Income statement Development budget Cash flow statement of which external development budget Special funds Stages Estimates, releases, actuals Actuals Institutional categories Central government, local government Economic classification GFMS2001 Source Ministry of Finance SOE Sources of expenditures Externally funded (donor) expenditures Not available (e.g., what share of are identified separately expenditures by borrowing) Appendix 3. Country groups46 Oil-exporting countries Countries where net oil exports make up 30 percent or more of total exports. Oil exporters are classified as such even if they would otherwise qualify for another group Middle-income countries Countries that are not oil exporters and have per capita income higher than $905, according to 2006 gross national income per capita as calculated by the World Bank. Non-fragile low-income countries Countries that are not oil exporters and have per capita income equal or lower than $905 and a score higher than 3.2 on the Country Policy and Institutional Performance Assessment of the World Bank, following the classification in the 2007 Global Monitoring Report. Non-fragile low-income countries The remaining countries of the sample. 46 Adapted from IMF (2007) p.4. 59 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 4. Primary fiscal balances Budgetary expenditures Primary excl. debt balance excl. GDP share Domestic revenues Grants service Primary balance grants Benin 16.26 2.05 19.09 (0.79) (2.84) Burkina Faso 11.86 5.28 20.50 (3.37) (8.65) Cameroon 16.63 0.41 13.40 3.64 3.24 Cape Verde 22.36 7.48 31.42 (1.57) (9.06) Chad 8.94 3.18 16.88 (4.77) (7.94) Congo, Dem. Rep. 8.64 1.92 10.46 0.10 (1.82) Côte d'Ivoire 13.81 0.61 13.05 1.37 0.76 Ethiopia 16.95 4.99 25.27 (3.32) (8.32) Ghana 20.90 5.27 24.69 1.48 (3.79) Kenya 18.48 1.42 18.82 1.08 (0.34) Lesotho 45.59 2.59 43.75 4.43 1.85 Madagascar 10.29 5.03 25.10 (9.79) (14.81) Malawi 18.92 8.66 28.08 (0.50) (9.16) Mozambique 12.79 9.49 24.46 (2.18) (11.67) Namibia 31.28 0.12 32.76 (1.36) (1.48) Niger 10.79 5.76 18.39 (1.84) (7.60) Nigeria 41.63 - 35.24 6.39 6.39 Rwanda 12.77 9.48 22.31 (0.05) (9.54) Senegal 18.09 1.82 20.88 (0.97) (2.79) South Africa 23.24 - 20.67 2.57 2.57 Tanzania 11.42 5.07 17.43 (0.94) (6.01) Uganda 9.33 5.97 16.85 (1.55) (7.52) Zambia 17.93 9.65 25.62 1.96 (7.70) Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP Note: Annual averages over the period 2001–06 Averages weighted by countries’ GDP. Totals may not add up. 60 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 5. Net change in central government budget: breakdown by source CG net expenditure of which oil GDP share budget Domestic revenues revenues Grants Net borrowing Benin (1.07) 0.42 - 1.06 (2.54) Burkina Faso (0.10) 0.63 - (0.89) 0.16 Cameroon 0.05 1.11 0.15 0.22 (1.28) Cape Verde 0.73 (0.08) - (1.91) 2.72 Chad (6.59) 1.27 4.22 (2.75) (5.11) Congo, Dem. Rep. 9.06 3.63 - 4.84 0.59 Côte d'Ivoire 1.54 (0.47) 0.88 0.59 1.42 Ethiopia (0.23) (0.21) - 0.93 (0.95) Ghana 7.14 5.80 - 2.11 (0.77) Kenya 1.12 (1.00) - 1.21 0.90 Lesotho (3.26) 6.57 - (3.06) (6.77) Madagascar 2.87 0.77 - 1.85 0.25 Malawi 7.53 7.71 - 5.28 (5.46) Mozambique (3.87) 1.15 - (4.01) (1.01) Namibia (1.89) 1.27 - (0.10) (3.06) Niger 1.83 (0.57) - 2.38 0.03 Nigeria (4.68) 6.80 10.44 - (11.48) Rwanda 3.37 0.59 - 5.15 (2.38) Senegal 4.73 1.42 - (0.00) 3.32 South Africa 4.32 3.45 - - 0.87 Tanzania 6.64 1.43 - 3.10 2.11 Uganda (2.79) (0.66) - (0.32) (1.82) Zambia (6.38) (1.05) - 17.43 (22.76) Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Net change, 2001–06. Averages weighted by countries’ GDP. Totals may not add up. 61 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 6. Net change in central government budget: breakdown by use CG net expenditure of which Capital of which GDP share budget Wages Other current Infrastructure expenditure infrastructure Benin (1.07) 1.06 (2.24) 0.04 0.11 0.20 Burkina Faso (0.10) 0.13 (0.74) - 0.51 - Cameroon 0.05 (0.58) 0.63 (0.01) 0.01 0.71 Cape Verde 0.73 1.89 (0.87) (0.04) (0.29) (0.50) Chad (6.59) (1.21) (2.31) 0.21 (3.08) 0.18 Congo, Dem. Rep. 9.06 2.32 3.90 - 2.84 - Côte d'Ivoire 1.54 0.11 1.86 0.13 (0.43) 0.32 Ethiopia (0.23) 0.22 (2.25) (0.04) 1.80 2.16 Ghana 7.14 (0.00) 1.26 (0.09) 5.88 0.81 Kenya 1.12 (0.03) (0.17) (0.00) 1.31 (0.30) Lesotho (3.26) (1.17) 0.98 0.10 (3.08) 0.01 Madagascar 2.87 0.04 (0.14) (0.02) 2.98 2.04 Malawi 7.53 0.74 3.25 (2.11) 3.54 0.30 Mozambique (3.87) 0.38 0.06 0.01 (4.31) 0.44 Namibia (1.89) 0.60 (0.98) (0.04) (1.51) (0.01) Niger 1.83 (0.31) (0.66) 0.26 2.80 0.18 Nigeria (4.68) (2.36) (0.27) 0.04 (2.05) (2.20) Rwanda 3.37 (1.35) 0.81 0.09 3.91 0.73 Senegal 4.73 0.23 1.95 (1.61) 2.55 (0.20) South Africa 4.32 (0.13) 5.17 (0.02) (0.71) 0.04 Tanzania 6.64 0.25 2.80 1.00 3.59 1.13 Uganda (2.79) (0.30) 0.98 (0.08) (3.47) 0.87 Zambia (6.38) (0.79) 2.33 0.73 (7.91) (1.97) Source: AICD, Fiscal Database, 2008; IMF Statistical Appendixes, WB DDP. Note: Net change, 2001–06. Averages weighted by countries’ GDP. Totals may not add up. 62 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 7. Expenditure on main road network Length of main network Maintenance expenditures Rehabilitation/capital expenditures (***) (**) (km) Annual Averages in $ per km Benin 4,735 $ 3,015.81 $ 4,306.67 Burkina Faso 10,231 n.a. n.a. Cameroon 11,008 $ 2,608.66 $ 5,823.50 Chad n.a. n.a. Côte d'Ivoire 13,291 n.a. $ 9,016.45 Ethiopia 9,721 $ 2,599.32 $ 19,524.64 Ghana 11,139 $ 2,339.67 $ 11,557.48 Kenya 22,341 $ 2,872.55 $ 1,176.95 Lesotho 3,005 $ 1,255.09 $ 8,353.15 Madagascar 9,599 $ 1,450.28 $ 11,851.26 Malawi 3,444 $ 1,960.15 $ 6,952.26 Mozambique 9,808 $ 2,371.65 $ 12,209.19 Namibia 16,362 $ 3,462.30 $ 166.70 Niger 6,055 $ 493.59 $ 4,137.90 Nigeria $ 16,963.74 Rwanda 2,836 $ 491.75 $ 6,466.14 Senegal 4,780 n/a $ 24,937.63 South Africa (*) 76,288 $ 2,074.64 $ 708.23 Tanzania 28,730 $ 1,433.37 $ 4,162.02 Uganda 9,171 n/a $ 14,367.39 Zambia 3,290 $ 3,029.45 $ 36,348.70 Averages (weighted according to the size of network) Resource-intensive countries $ 2,852.39 $ 13,243.52 Non-resource-intensive (excluding South Africa) $ 2,001.27 $ 9,973.38 Aid dependent $ 2,020.88 $ 11,342.76 HIPC countries (end of 2005) $ 1,985.69 $ 11,470.62 Middle-income countries $ 2,134.58 $ 684.84 Average $ 2,076.51 $ 8,384.69 Average excluding South Africa $ 2,077.23 $ 10,769.11 Source: AICD Fiscal Baselines, 2007. Primary network length data from RONET. South Africa, expenditures of SANRAL. Note: Based on annual averages for the period 2001–05 n.a. - time series were not long enough to calculate the standard errors. (*) Excludes local governments ) (**) Includes expenditures of roads fund plus current expenditures of central government (current expenditures excluding transfers to local municipalities) (***) Capital expenditures of central government and externally funded expenditures 63 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 8. Variance around the trend line of road expenditure Current expenditures (*) Capital expenditures (**) Road funds Central government External Central government Benin 16.50 44.01 36.51 62.39 Cameroon 23.91 9.87 23.17 19.17 Chad 2.06 21.84 33.42 79.40 Côte D'Ivoire n.a. 57.44 n.a. 12.86 Ethiopia 3.59 n.a. n.a. 13.48 Ghana 10.89 36.77 19.51 20.08 Kenya 4.14 3.61 n.a. 12.85 Madagascar 27.12 43.12 24.65 16.92 Malawi 10.02 78.18 n.a. 45.17 Namibia 4.00 n.a. n.a. n.a. Niger 19.59 n.a. n.a. n.a. Rwanda 21.79 16.16 38.18 50.64 Senegal n.a. 102.67 18.74 57.93 Tanzania n.a. 4.54 28.83 8.09 Uganda n.a. n.a. n.a. n.a. Mozambique 7.92 13.02 32.32 23.25 Nigeria n/a 47.36 n.a. 15.10 Lesotho 11.18 9.42 9.25 35.66 Zambia n.a. 12.05 25.65 Weighted averages (according to size of network) Resource-intensive countries 13.55 38.29 27.18 23.05 Non-resource-intensive 10.22 22.02 29.59 27.59 (excluding South Africa) Aid dependent 11.13 22.07 31.80 28.06 Weighted average (size of 10.59 26.61 29.30 26.36 network) Simple average 12.52 33.34 26.46 31.16 Source: AICD Fiscal Baselines, 2007 Note: Based on annual averages for the period 2001–05. n.a. - time series were not long enough to calculate the standard errors. (*) Includes expenditures of roads fund plus current expenditures of central government (current expenditures excluding transfers to local municipalities) (**) Capital expenditures of central government and externally funded expenditures 64 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 9. Contributions to QFCs (country aggregates)—water sector Collection Total hidden Unaccounted inefficiencies Unaccounted Collection cost losses 2/ Underpricing 3/ 4/ losses Underpricing inefficiencies Share of Country GDP (%) Share of total QFC (%) Share of GDP (%) Tanzania 0.1 33.1 35.8 31.1 0.0 0.0 0.0 Nigeria 0.2 19.7 45.7 34.6 0.0 0.1 0.1 Benin 0.2 15.1 73.0 11.9 0.0 0.1 0.0 Cape Verde 0.3 15.3 0.0 84.7 0.0 0.0 0.2 Ethiopia 0.3 26.8 70.8 2.4 0.1 0.2 0.0 Uganda 0.3 27.7 0.0 72.3 0.1 0.0 0.2 Namibia 0.3 0.1 73.5 26.4 0.0 0.2 0.1 Kenya 0.3 28.2 60.0 11.8 0.1 0.2 0.0 Burkina Faso 0.3 0.0 22.6 77.4 0.0 0.1 0.2 Rwanda 0.3 29.7 34.4 35.9 0.1 0.1 0.1 Sudan 0.4 26.1 27.3 46.5 0.1 0.1 0.2 Niger 0.4 0.0 95.2 4.8 0.0 0.4 0.0 South Africa 0.5 12.5 77.0 10.4 0.1 0.4 0.0 Lesotho 0.5 17.0 83.0 0.0 0.1 0.4 0.0 Mozambique 0.5 49.2 50.4 0.4 0.3 0.3 0.0 Côte d'Ivoire 0.8 2.3 97.7 0.0 0.0 0.8 0.0 Senegal 0.9 0.2 99.8 0.0 0.0 0.9 0.0 Madagascar 0.9 20.3 79.7 0.0 0.2 0.7 0.0 Zambia 1.1 28.1 32.7 39.1 0.3 0.3 0.4 Malawi 1.2 23.6 43.0 33.4 0.3 0.5 0.4 Ghana 1.2 42.7 30.6 26.7 0.5 0.4 0.3 Congo, D.R. 1.3 40.8 27.2 32.0 0.5 0.4 0.4 Source: Authors’ own calculations using data from the AICD Database. 2/ Unaccounted Losess = (End-User Consumption) * (Average Cost Recovery Price)* (Total Loss Rate - Normative Loss Rate)/(1-Normative Loss Rate) 3/ Underpricing = End-User Consumption * (Average Cost Recovery Price - Average Actual Tariff). 4/ Collection Inefficiencies = End-user Comsumption* Average Actual Tariff* (1 - Collection rate) 65 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 10. Contributions to QFCs (country aggregates)—power sector Collection Unaccounted Underpricing Inefficiencies Unaccounted Collection Total Hidden Cost Losses 2/ 3/ 4/ Losses Underpricing Inefficiencies Country Share of GDP (%) Share of total QFC (%) Share of GDP (%) South Africa 0.00 100.0 0.0 0.00 0.00 0.00 0.0 Benin 0.24 61.7 0.0 38.26 0.15 0.00 0.1 Kenya 0.47 60.4 39.6 0.00 0.28 0.19 0.0 Mozambique 0.67 45.1 14.0 40.91 0.30 0.09 0.3 Chad 1.00 28.5 71.5 0.00 0.29 0.72 0.0 Cape Verde 1.18 26.1 45.3 28.56 0.31 0.53 0.3 Madagascar 1.31 28.9 71.1 0.00 0.38 0.93 0.0 Lesotho 1.37 22.7 61.9 15.48 0.31 0.85 0.2 Nigeria 1.42 26.1 33.4 40.51 0.37 0.47 0.6 Burkina Faso 1.47 24.6 0.0 75.38 0.36 0.00 1.1 Rwanda 1.74 23.2 76.8 0.00 0.40 1.33 0.0 Ethiopia 1.78 30.0 70.0 0.00 0.54 1.25 0.0 Uganda 1.85 40.8 0.0 59.17 0.75 0.00 1.1 Cameroon 2.22 28.2 71.8 0.00 0.62 1.59 0.0 Zambia 2.33 3.5 96.5 0.00 0.08 2.25 0.0 Tanzania 2.41 24.7 75.3 0.00 0.59 1.82 0.0 Senegal 2.47 24.9 50.4 24.73 0.62 1.24 0.6 Ghana 3.14 25.7 0.0 74.32 0.81 0.00 2.3 Niger 3.51 22.8 49.3 27.96 0.80 1.73 1.0 Malawi 4.36 17.4 76.6 6.0 0.8 3.3 0.3 DRC 4.66 27.5 0.0 72.51 1.28 0.0 3.4 Source: Authors’ own calculations using data from the AICD Database. 2/ Unaccounted Losses = (End-User Consumption) * (Average Cost Recovery Price) * (Total Loss Rate - Normative Loss Rate) / (1-Normative Loss Rate) 3/ Underpricing = End-User Consumption * (Average Cost Recovery Price - Average Actual Tariff) 4/ Collection Inefficiencies = End-user Comsumption* Average Actual Tariff* (1 - Collection rate) 66 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 11. Water: efficiency and production indicators, 2006 Average cost End user recovery Average consumption tariff effective tariff NRW Implicit collection Country Utility (M3/year) (USD/m3) (USD/m3) 5/ (%) ratio (%) 6/ Benin SONEB 22,977,000 1.10 0.85 24 95 Burkina Faso ONEA 33,817,240 1.15 1.03 18 60 Cameroon SNEC — 1.06 0.67 35 — Cape Verde ELECTRA 2,931,781 3.24 4.20 31 81 Chad STEE — 1.06 0.29 35 33 Côte d'Ivoire SODECI 128,853,000 1.03 0.09 22 100 DRC REGIDESO 137,052,336 1.10 0.91 41 75 Ethiopia ADAMA 2,660,048 0.72 0.39 43 89 Ethiopia AWSA 52,395,813 0.72 0.33 37 96 Ethiopia DIRE DAWA 2,565,138 0.58 0.24 22 96 Ghana GWC 130,463,393 1.06 0.71 53 63 Kenya KIWASCO 1,800,000 1.53 0.81 71 79 Kenya MWSC 12,516,164 0.89 0.67 38 21 Kenya NWASCO 97,869,588 0.56 0.24 38 100 Lesotho WASA 11,213,684 1.10 0.58 28 100 Madagascar JIRAMA 63,756,027 0.87 0.29 34 100 Malawi BWB 14,185,900 0.97 0.40 51 100 Malawi CRWB 4,722,000 0.81 0.32 17 57 Malawi LWB 23,600,000 0.63 0.74 22 57 Mozambique AdeM Beira 3,771,240 0.91 0.52 60 100 Mozambique AdeM Maputo 25,432,162 1.13 0.52 62 100 Mozambique AdeM Nampula 2,447,001 0.75 0.52 44 98 Mozambique AdeM Pemba 1,214,471 0.93 0.52 45 82 Mozambique AdeM Quilimane 978,095 0.82 0.52 35 100 Namibia Oshakati Municipality 1,031,884 1.84 1.98 21 81 Walvis Bay Namibia Municipality 3,626,811 1.06 0.97 16 100 Windhoek Namibia Municipality 17,500,000 2.48 1.73 14 85 Niger SPEN 33,791,483 0.86 0.49 19 96 Nigeria Borno — 1.06 0.67 35 31 Nigeria FCT 17,300,000 1.06 0.53 80 31 Nigeria Kaduna 63,690,000 1.06 0.22 21 31 Nigeria Katsina 31,000,000 0.46 0.26 14 24 Nigeria Lagos 50,405,903 0.66 0.67 57 31 Nigeria Plateau 14,660,000 1.06 0.67 23 31 Rwanda ELECTROGAZ 9,780,000 0.91 0.67 38 63 Senegal ONAS nap 1.06 0.67 35 87 67 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Senegal SDE 99,640,165 1.25 0.50 20 100 South Africa Cape town metro 215,660,000 1.61 0.47 18 100 Drakenstein South Africa Municipality 13,464,003 1.10 0.35 12 100 eThekwini Metro South Africa (Durban) 200,000,000 1.96 1.42 32 79 South Africa Joburg 328,246,350 1.90 0.32 31 45 Khartoum Water Sudan Corporation 150,000,000 0.68 0.50 40 38 South Darfur Water Sudan Corporation 3,700,000 0.89 0.86 49 59 Upper Nile Water Sudan Corporation 1,788,500 1.13 0.80 29 48 Tanzania DAWASCO — 1.06 0.61 35 30 Tanzania DUWS 5,110,000 0.82 0.54 30 63 Tanzania MWSA 7,432,028 0.59 0.47 49 73 Uganda NWSC 38,000,000 1.00 1.01 34 53 Zambia LWSC 35,346,257 0.67 0.53 54 43 Zambia NWSC 73,799,000 0.60 0.35 37 33 Zambia SWSC 8,136,000 0.70 0.40 55 33 — = data not available. 68 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 12. Electricity: efficiency and production indicators, 2006 Historical normative cost System Volume recovery tariff Average losses Implicit electricity billed ($) – average effective tariff (% electricity collection ratio (GWh/year) total cost (USD/kwH) 1/ produced) (%) 2/ Benin SBEE 506.6 0.14 0.15 18.25 95 Burkina Faso SONABEL 537.5 0.24 0.31 24.68 63 Cameroon AES SONEL 3263.7 0.14 0.06 30.79 100 Cape Verde Electra 161.7 0.23 0.20 17.32 90 Chad STEE 78.0 0.84 0.30 33.20 100 Côte d'Ivoire CIE 5741.9 0.34 — — 77 DRC Congo SNEL 4091.8 0.07 0.10 40.00 42 Ethiopia EEPCO 2069.2 0.09 0.02 40.00 100 Ghana VRA 3762.0 0.13 0.15 25.44 55 Kenya KPLC 4379.0 0.14 0.13 18.10 53 Lesotho LEC 353.6 0.11 0.08 20.00 89 Madagascar JIRAMA 753.7 0.17 0.10 23.74 100 Malawi Escom 1055.0 0.10 0.04 23.00 86 Mozambique EDM 1306.7 0.09 0.09 25.00 84 Namibia NamPower 3363.0 0.00 — 12.00 — Namibia NORED 131.4 — — 18.00 — Niger NIGELEC 333.1 0.43 0.26 27.00 61 Nigeria PHNC 21401.9 0.08 0.05 30.00 52 Rwanda ELECTROGAZ 157.4 0.38 0.20 23.00 100 Senegal SENELEC 1710.0 0.24 0.18 21.19 84 South Africa ESKOM 256959.0 0.06 0.08 10.03 100 Sudan NEC — — — — — Tanzania TANESCO 2628.0 0.16 0.07 26.00 100 Uganda UEDCL 1806.4 0.12 0.14 36.00 63 Zambia ZESCO 3516.0 0.08 0.03 12.00 100 Source: AICD Database. — = data not available. 1/ Average Effective Tariff = (Effective Residential Tariff at 100 kwh)*{Residential Share on Consumption / Residential Share on Revenue} 2/ Implicit Collection Ratio = (Average Revenue/Average Effective Tariff) or (Reported Non Payment) 69 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 13. Annual maintenance and preservation expenditures, 2001–05 Km of primary in fair and good Maintenance expenditure (**) Preservation norms (***) condition (US$ millions) (US$ millions) Benin 911.32 $ 16.43 $ 27.20 Cameroon 2,372.09 $ 43.05 $ 57.82 Chad 862.42 $ 3.88 $ 23.49 Côte D'Ivoire 4,545.94 $ 0.68 $ 61.28 Ethiopia 4,122.02 $ 43.82 $ 57.11 Ghana 6,284.60 $ 71.73 $ 103.03 Kenya 6,657.12 $ 110.36 $ 122.96 Lesotho 676.72 $ 9.61 $ 22.62 Madagascar 3,804.92 $ 19.15 $ 47.53 Malawi 2,567.75 $ 15.67 $ 35.42 Mozambique 4,819.70 $ 67.73 $ 75.94 Namibia 5,589.24 $ 101.81 $ 85.45 Niger 3,250.09 $ 8.24 $ 29.86 Rwanda 945.43 $ 5.79 $ 8.57 Senegal 2,141.99 $ 20.00 $ 47.53 South Africa 59,665.02 $ 390.54 $ 537.79 Tanzania 4,100.02 $ 128.48 $ 123.11 Uganda 2,119.06 $ 24.87 $ 91.10 Zambia 3,989.00 $ 86.06 $ 65.31 Notes: Primary network length and road quality data from RONET. South Africa, expenditures of SANRAL Source: AICD Fiscal Baselines, 2007 (**) Includes expenditures of roads fund plus current expenditures of central government (current expenditures excluding transfers to local municipalities) (***) RONET data 70 FINANCING PUBLIC INFRASTRUCTURE IN SUB-SAHARAN AFRICA Appendix 14. Average annual maintenance expenditures on main road network, 2001–05 Length of main Road fund expenditure Maintenance expenditures Rehabilitation network (km) on main roads (*) ($) on main roads (**) ($ per km per year) expenditures (***) ($) Benin 4,735 $ 14,279,046 $ 3,015.81 $ 20,390,875 Burkina Faso 10,231 n.a. n.a. n.a. Cameroon 11,008 $ 28,714,978 $ 2,608.66 $ 64,102,491 Chad n.a. n.a. n.a. $ 100,431,165 Côte d'Ivoire 13,291 n.a. n.a. $ 119,835,431 Ethiopia 9,721 $ 25,267,135 $ 2,599.32 $ 189,792,767 Ghana 11,139 $ 26,061,377 $ 2,339.67 $ 128,737,666 Kenya 22,341 $ 64,176,053 $ 2,872.55 $ 26,294,362 Lesotho 3,005 $ 3,770,923 $ 1,255.09 $ 25,097,111 Madagascar 9,599 $ 13,921,492 $ 1,450.28 $ 113,762,596 Malawi 3,444 $ 6,750,000 $ 1,960.15 $ 23,940,957 Mozambique 9,808 $ 23,262,139 $ 2,371.65 $ 119,752,630 Namibia 16,362 $ 56,650,000 $ 3,462.30 $ 2,727,596 Niger 6,055 $ 2,988,755 $ 493.59 $ 25,055,751 Rwanda 2,836 $ 1,394,509 $ 491.75 $ 18,336,585 Senegal 4,780 n.a. n/a $ 119,199,012 Tanzania 28,730 $ 41,180,461 $ 1,433.37 $ 119,573,801 Uganda 9,171 n/a n/a $ 131,767,055 Zambia 3,290 $ 9,966,882. $ 3,029.45 $ 119,587,211 Arithmetic average $ 2,097.22 $81,576,948 South Africa (*) 76,288 $ 158,270,950.42 $ 2,074.64 $ 119,199,012.08 Notes: Primary network length data from RONET. South Africa, expenditures of SANRAL Source: AICD Fiscal Baselines, 2007 Notes: Primary network length data from RONET. South Africa, expenditures of SANRAL (*) Excludes local governments) (**) Includes expenditures of roads fund plus current expenditures of central government (current expenditures excluding transfers to local municipalities) (***) Capital expenditures of central government and externally funded expenditures 71