FISCAL GOOD PRACTICE NOTE 2 Assessing the Fiscal Cost of Subsidies and Fiscal Impact of Reform Sudarshan Gooptu  CONTENTS Acknowledgments iii About the Author iii Acronyms and Abbreviations iv 1. Motivation 1 2. Typical Questions to Address 2 3. Conceptual Framework 3 4. Why Do Fiscal Risks Matter in the Energy Sector? 14 5. How Do We Assess the Fiscal Impacts of the Fiscal Risks? 16 6. Fiscal Impacts and State-Owned Enterprises 20 7. Fiscal Risks from Energy Sector Public-Private Partnerships 22 8. Lessons: Fiscal Impact and Fiscal Risk Assessments 24 Endnotes 25 References 26 i GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM BOXES Box 1: Putting Egypt’s Energy Price Reforms in a Macroeconomic-Fiscal Context 9 Box 2 : Fiscal Space, Fiscal Sustainability, and Solvency Assessments Require Scenario Analyses and Stress Testing 11 Box 3: Toolkit for Fiscal Risk Assessment from Contingent Liabilities 23 TABLES Box Table 1.1: Energy Price Increases 10 Table 1: Hypothetical Example—Elements of a Fiscal Framework Model 13 Table 2: The Fiscal Risks Matrix on the Liabilities Side 18 Table 3: Fiscal Hedge Matrix: Assets and Contingent Financing 19 FIGURES Figure 1: Macroeconomic Linkages between Public Spending, Taxes, Financing, and Debt 7 Figure 2: Elements of a Fiscal Framework 7 Figure 3: Adding Subsidies to the Core Integrated Fiscal Framework 8 Box Figure 1.1: Budgeted Health, Education, Social Protection, and Energy Subsidy Spending 10 Figure 4: Realization of Contingent Liabilities under Uncertainty 16 Figure 5: Fiscal Risks from Explicit and Implicit Subsidies to Energy SOEs 21 Figure 6: Fiscal Risk Matrix: How SOEs Can Contribute to Fiscal Impacts in Future 21 ii ACKNOWLEDGMENTS ACKNOWLEDGMENTS This is the second in the series of 10 good practice notes under the Energy Sector Reform Assessment Framework (ESRAF), an initiative of the Energy Sector Management Assistance Program (ESMAP) of the World Bank. ESRAF proposes a guide to analyzing energy subsidies, the impacts of subsidies and their reforms, and the political context for reform in developing countries. This good practice note draws on analytical tools that have been developed and are being used in the recent operational work by country economists and fiscal experts in the Macroeconomics and Fiscal Management Global Practice (MFM) of the World Bank. Thanks go out to Chadi Bou Habib, Ibrahim Saeed Chowdhury, Simon Davies, Shireen Mahdi, and Wael Mansour for their valuable comments and inputs to earlier drafts of this note, and to Yameng Wang for diligent analytical support. This note also draws on the presentations that were prepared and delivered by Aart Kraay at the World Bank MFM Forum (2017), which skillfully summarized the literature on fiscal and debt sustainability and fiscal policy as it pertains to developing countries. All errors that remain are the sole responsibility of the author. ABOUT THE AUTHOR Sudarshan Gooptu is the global lead for Fiscal Policy in the Macroeconomics and Fiscal Management Global Practice of the Equitable Growth, Finance and Institutions Practice Group Vice Presidency at the World Bank. He has a Ph.D. in economics from the University of Illinois, Urbana-Champaign, Illinois (USA), and has been in the World Bank in various operational, research, and managerial roles since 1988. iii GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM ACRONYMS AND ABBREVIATIONS DSA debt sustainability analysis DSF Debt Sustainability Framework ESRAF Energy Subsidy Reform Assessment Framework ESMAP Energy Sector Management Assistance Program FY fiscal year GDP gross domestic product IMF International Monetary Fund INDC Intended Nationally Determined Contributions IPP independent power producer LIC low-income country LPG liquefied petroleum gas MAC market access countries MTBF medium-term budget framework MTFF medium-term fiscal framework P-FRAM PPP Fiscal Risks Assessment Model PPP public-private partnership REER real effective exchange rate RIR real interest rate SOE state-owned enterprise SPV special purpose vehicle iv 1. MOTIVATION 1 1. MOTIVATION The objective of this good practice note is to This note focuses on the costs of subsidies outline the ingredients of an assessment of to the government. One important form the fiscal impacts of energy subsidies in an of subsidies consists of direct budgetary economy from the aggregate fiscal perspective transfers from the government to either of the government. It demonstrates the consumers or producers, which are recorded interrelations between the fiscal balance, in the government public sector budget. For its financing, and impact on key debt and instance, with the justification of social benefits, fiscal sustainability indictors. As discussed governments often establish consumer in the Energy Sector Reform Assessment prices for energy that are below reference Framework (ESRAF) Good Practice Note prices (prices that would have prevailed in a on the definition of energy subsidies (Good competitive market, or the cost of efficient Practice Note 1), energy subsidies may be production if a competitive market does provided through various channels on the not exist), and then compensate the energy production and consumption sides, and suppliers for the difference (also referred may generate contingent liabilities—explicit to as the price gap) between the reference or implicit—for a government that must be prices and the government-controlled prices.1 monitored and managed as part of overall Another subsidy delivery mechanism is macroeconomic management. provision of the subsidy benefits directly to end-users, typically households. In addition, a ESRAF defines an energy subsidy as a government may provide subsidies in the form deliberate policy action by the government of tax exemptions to energy service providers, that specifically targets electricity, fuels, or tax credits for investment, or allowing the district heating and that reduces the net cost energy-related public utilities and national of energy purchased, reduces the cost of oil companies to run arrears on their debt energy production or delivery, increases the service and other payment obligations to the revenues retained by energy suppliers, or has government. any combination of these three effects. ESRAF also covers non-energy use of oil, gas, and coal, The global financial crisis of 2008–09 such as natural gas used as a feedstock for demonstrated how countries, whether net fertilizer manufacture and naphtha and liquefied energy exporters or importers, had to adopt petroleum gas (LPG) used as feedstocks in countercyclical fiscal policies that eroded petrochemicals. Subsidies are not always their fiscal buffers, which caused them to paid for by the government. Consumers may face situations of increased macroeconomic subsidize producers, producers may subsidize vulnerability and threats to their debt consumers, and financiers and other actors not sustainability. As fiscal deficits rose, policy linked to energy consumption or production, makers were facing calls to protect their including those outside the country, may be existing poverty reduction initiatives and covering the costs of subsidies. existing subsidy schemes. In some countries, there were political pressures to increase energy subsidies to protect households and 2 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM strategic investments or firms from these reforms have been explicitly integrated into mounting macroeconomic shocks. While the climate change policy packages outlined many governments have embarked on subsidy in their Intended Nationally Determined reforms, substantial subsidies still remain. Such Contributions (INDCs) in 13 countries under is the case in a significant part of the Middle the 2015 Paris Agreement.3 Assessing the fiscal East. Meanwhile, a key ingredient for arriving impacts of energy subsidy reforms (in fossil at a market-based solution to the climate fuels, including electricity and district heating change challenge is by removing subsidies relying on them) from the macroeconomic on fossil fuels and reflecting the economic, context is, therefore, an essential prerequisite social, and environmental costs of carbon- for beginning to implement fossil fuel subsidy intensive activities in their prices.2 Fossil fuel reforms. 2. TYPICAL QUESTIONS TO ADDRESS When assessing the fiscal impact of energy • Does the medium-term fiscal framework subsidy reforms from the macroeconomic (MTFF) of the government incorporate the perspective, policy makers typically grapple fiscal and budgetary implications of the with the following questions: government’s energy subsidies and their reforms? This includes assessing the fiscal • How important is the energy sector-related and debt trajectories during the phasing public spending in the macroeconomic out of energy subsidies, and the impact context of the country? This includes looking of energy price adjustments on the fiscal at not only direct subsidy programs that situation of the government during the one sees as a line item in a government’s transition to fully market-based energy budget, but also subsidies on the production pricing mechanisms. What is the fiscal side, and any sovereign guarantees and impact of alternative transition paths to concession arrangements accorded to the energy subsidy reforms, and what are their private sector. implications for debt dynamics? • How much revenue is the government losing • How much fiscal space does the government due to energy subsidies? Examples include have to finance the transition period in the fiscal concessions provided to energy energy subsidy reform program, and is there suppliers, lower dividends transferred by sufficient room for provision of adequate state-owned energy suppliers and lower mitigation measures while maintaining fiscal corporate income taxes paid by all energy sustainability? suppliers to compensate for the financial losses from consumer price subsidies that • Where there are consumer price subsidies, are not reimbursed by the government, and given volatile fuel prices, which energy underpricing of goods and services provided subsidy poses the largest risk to fiscal by the government to energy suppliers. sustainability in a country, fixed price, indexed price, or fixed nominal subsidy? 3. CONCEPTUAL FRAMEWORK 3 • Are fiscal impacts from uncertainty • What are the fiscal costs of eliminating fossil (especially from exchange rate, commodity fuel subsidies and supporting renewable price, and interest rate fluctuations) and energy and adoption of green technologies fiscal risks from the government’s contingent in a fiscally sustainable manner? These liabilities being incorporated in fiscal reforms will accompany a country’s climate sustainability when making informed fiscal change mitigation efforts as indicated in policy decisions for the implementation of their respective voluntary INDCs that they the energy subsidy reforms in the country? pledged under the 2015 Paris Agreement. 3. CONCEPTUAL FRAMEWORK Removing energy subsidies in a country, if effects on the poor). These reforms will also appropriately implemented, will typically take affect future fiscal deficits and associated a few years to undertake. Where subsidies public sector financing requirements which, take the form of consumer support borne by in turn, will affect a country’s gross debt and energy suppliers, government revenue from the affordability of the government’s overall the suppliers may increase after the subsidy public spending needs in the medium term. reforms. If the subsidies include consumer price subsidies, prices of goods and services The existing methodologies for assessing that use energy to produce and deliver it fiscal sustainability can be summarized in usually rise during the transition period and three main approaches: accounting, analytical, thereafter. This includes prices of goods and and empirical. services in the consumption basket of the The accounting approach uses identities poor, such as food and public passenger involving government revenues, expenditures, transportation services. Depending on which deficits, public debt levels, economic growth form of energy is being subsidized, poor and rates, and interest rates. Then it projects debt vulnerable groups might not be receiving and debt ratios going forward, and identifies subsidies directly from the government the forces driving the debt dynamics (debt before the energy subsidy reforms—the poor, dynamics decomposition). The assessment is especially in low- and lower-middle-income based on the level and growth of projected countries, do not own motorized vehicles or debt. Intuitively, debt sustainability implies backup power generators and therefore do that the accumulation of public debt is not not purchase gasoline or diesel, and many are excessive, and this translates into saying that also not connected to grid electricity—but the public debt is not growing “too fast” or after the energy subsidy reforms will need to that the level of public debt is not “too high.” be provided with appropriately targeted social For each scenario that we want to explore, protection to compensate them for indirect we can compute the corresponding debt effects of higher energy prices (see Note 3 for projection and debt dynamics decomposition. quantification of the impact of these indirect 4 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM The analytical approach goes further and the solvency condition. For the solvency establishes a formal condition of solvency, conditions mentioned before, the standard for example: indicators would be, respectively, the following: • The achievement of a stable ratio of debt • The debt-stabilizing primary balance. to gross domestic product (GDP). • The permanent adjustment in the structural • The intertemporal budget constraint. primary balance projected over a finite time horizon. • The achievement of an acceptable target for the debt-to-GDP ratio (such as the The debt-stabilizing primary balance is widely Maastricht 60% value). used in practice,for example, the indicator S2 used by the European Commission. This This has the advantage that the solvency indicator measures the permanent adjustment condition gives a well-defined meaning to the in the structural primary balance projected vague expressions “too fast” and “too high.” over an infinite time horizon—that is, very far Furthermore, the approach intends to answer into the future—that would be needed for the such policy questions as What should the debt-to-GDP ratio to satisfy the intertemporal government do to restore sustainability? By budget constraint). comparing the projected debt in the baseline scenario with the arbitrarily chosen debt path Similarly, the indicator S1 also used by that satisfies the solvency condition, one the European Commission measures the can derive the fiscal adjustment needed to permanent adjustment in the structural primary fill the gap between the two debt paths. If balance projected over a finite time horizon such a fiscal adjustment policy were to be that would be needed for the debt-to-GDP implemented, one would be able to restore ratio to reach a 60% value at the end of that sustainability—that is, the new public debt horizon. One can compute these indicators projection (incorporating the change in not only for the baseline scenario, but also fiscal policies into the baseline) would meet for the alternative ones. The fundamental building block of fiscal sustainability analysis is the public sector or government budget constraint (identity): Change in debt = interest payments – primary balance – seigniorage + GDP growth effect + inflation effect on domestic debt + foreign exchange revaluation effect The standard debt accumulation equation as share of GDP Yt, with lowercase letters denoting shares of gross domestic product (GDP), that is, xt ≡ Xt /Yt, is as follows: Dt Yt − 1 Dt − 1 Bt (1 + it ) Dt − 1 Bt = (1 + it ) − = − Yt Yt Yt − 1 Yt (1 + p t )(1 + gt ) Yt − 1 Yt (1 + rt ) dt = dt − 1 − bt (1 + gt ) 3. CONCEPTUAL FRAMEWORK 5 Y /P Where 1 + gt = t t is (gross) real GDP would imply an increasing probability of Yt − 1 /Pt − 1 growth rate. the government falling into a debt distress situation (for example, arrears, serious • D t i s domestic currency–denominated problems to rollover maturing debt, and nominal debt; defaults). Estimations are usually based on • it is (gross) nominal interest rate; the experiences of national governments undergoing debt distress with their specific • πt is the domestic inflation rate; circumstances related to macroeconomic developments, indebtedness, fiscal policies, • Bt is primary balance (that is, excluding and institutional and political conditions. debt service); and • Bt–it Dt–1 is overall balance (that is, including This approach is very useful to monitor ex ante debt service). the risk of debt distress going forward, given the current circumstances and developments. This equation is the basis for analyzing debt The World Bank-IMF DSF, for instance, relies dynamics4 on the empirical approach to determine thresholds for external debt (controlling for If the drivers of future debt dynamics (real indicators of institutional quality) and to interest rate rt, real growth rate gt, and primary classify countries according to their level of balance as share of GDP bt) are uncertain, there risk of debt distress. In this context, several is the need for stochastic fiscal sustainability studies report public debt thresholds around analysis rather than deterministic debt 80–90% of GDP, which would draw the line sustainability analysis (DSA). If the drivers are between the growth-enhancing and growth- endogenous, there is the need for empirical hampering effects of public debt. evidence on interactions between drivers of debt dynamics (that is, fiscal policy and Reinhart and Rogoff (2010) analyze 44 growth), as well as theory-based simulations of industrial and developing economies over alternative paths. Fiscal sustainability is more two centuries and find that the GDP growth than just whether present value condition rate for countries whose debt exceeds 90% of holds. The Debt Sustainability Assessment GDP is lower than that for low-debt countries. Framework (DSF) of the World Bank and the Kumar and Woo (2010) focus on 38 industrial International Monetary Fund (IMF) for low- and emerging economies in 1970–2007, and income countries (LICs) and DSA for market- also conclude that the debt threshold is around access countries (MAC-DSA) can be used for 90% of GDP. Caner, Grennes, and Köhler-Geib this purpose. These have been designed on (2010) find a lower threshold at 77% of GDP the basis of rich empirical evidence on debt based on the period 1980–2008. indicators and debt sustainability. The World Bank and the IMF are currently revising the Samples and statistical techniques differ LIC-DSF jointly, while the MAC-DSF will be across studies, implying robustness of the reviewed later in 2018.. conclusion that public debt thresholds are fairly high as a share of GDP. Studies also The empirical approach uses empirical report little correlation, if any, between growth evidence to estimate critical threshold values and debt for low-debt countries, thus implying for the debt burden indicators that, if crossed, that, provided the thresholds are not breached 6 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM and adequate institutions are in place, debt produce and deliver and are consumed by the financing can safely be used to spur investment poor and vulnerable groups (or any others the and growth. government wishes to compensate as a result of the energy subsidy removal in the interim). Any energy subsidy program that was initially Using an integrated fiscal framework will help designed and implemented with a view toward to quantify the fiscal resource envelope and stabilizing energy prices for consumers and plan the resource allocation among competing producers will depend on the government’s spending programs. fiscal capacity to absorb fluctuations in energy prices at each stage in the supply chain down The foundation of the MTFF is the intertemporal to consumption. It needs to be considered in budget constraint of the government. The the context of budget allocation tradeoffs, MTFF is a basis for discussing economic tax policies, and public sector borrowing outlook and analyzing fiscal policy options. requirements. It helps make decisions on tax reform, revenue-sharing agreements, expenditure An integrated MTFF is needed to assess the planning and rationalization, fiscal rules, revenue and expenditure impact of energy borrowing framework, debt contracting, subsidy removal over time. This is a unified and restructuring. The MTFF goes beyond analytical framework that brings together published fiscal accounts concerning horizon the most salient outcomes of the institutions and scope, and deals with aggregate items and polices governing budget, financing, and generates multiyear projections of key intergovernmental fiscal system, and asset- debt, macroeconomic, and fiscal indicators to liability management in a country. It provides assess the implications of the economic cycle, a basis for assessing the performance of medium-term planning, structural changes, the economy and the public finances at an price and debt dynamics. This includes aggregate level. As a result, it integrates the examining disaggregated budget items, such fiscal situation and prospects of different as energy subsidies, contingent liabilities, as regions or tiers of government within the well as hidden liabilities (undertaken using off- country that will be influenced by the fiscal budget vehicles and quasi-fiscal operations effects of subsidy removal. This impact will of subnational governments to rescue local be positive—in terms of the fiscal resources SOEs), potential sources of debt, and spending saved from the energy subsidies previously obligations and their aggregate medium-term being provided to firms and households that fiscal impact. At the subnational level, an MTFF are removed as well as additional government involves a set of projections consistent with the revenue from energy suppliers previously specificities of the subnational economy and suffering financial losses created by consumer public finance institutions. Figures 1–3 show price subsidies not reimbursed by the the interactions among economic and fiscal government—and negative—in terms of the variables and the data needed to estimate increase in fiscal resources needed to support fiscal impacts under the baseline outlook additional social protection programs that may and alternative scenarios. In doing so, the need to be put in place to compensate for the MTFF can help determine the total amount increase in costs and retail prices of essential of resources available and allocation of these products and services that use energy to resources across spending agencies. 3. CONCEPTUAL FRAMEWORK 7 FIGURE 1: Macroeconomic Linkages between Public Spending, Taxes, Financing, and Debt Nominal and real GDP, activity indicators, prices of goods Macro Economic activity and assets. Tax bases. Personal income tax, corporate income tax, property tax, Revenues taxes and charges (excl. property tax), general subvention, subsidies and funds for current purposes, other current Budget (current & capital) revenues. Sales of assets, subsidies and funds for current purposes, other property revenues. Operating & Capital Budget Expenditures Salaries, guarantees and sureties, interests, other current expenditures (excl. salaries, guarantees and sureties, and (current & capital) interest). Property expenditure. Primary balance, borrowings, use of financial assets (e.g., use of budget surpluses from previous years use of unal- Sources of funds located funds, other use of assets and other revenues not related to debt). Debt service (principal amortization and interest from exist- Uses of funds Financing ing and new debts), other financing needs. Cash Flows a. Debt (% of revenue) Indicators b. Debt service plus guarantees and sureties (% of revenue). c. Current revenues plus property revenue (Fiscal Responsibility (excl. subsidies/funds) minus current expenditure (% of Law) revenue). d. 3-year average of indicator. Existing debts Debt stock and debt service (principal amortization and interest). Forecasts debt-by-debt. (incurred in the past) Debt Existing and New Debts New debts Borrowing, debt stock and debt service (principal (to be incurred in the amortization and interest). Forecasts debt-by-debt. projection horizon) FIGURE 2: Elements of a Fiscal Framework Core Model Uncertainties Portfolio Analysis Model (ALM) GDP External debt (or assets) Natural Instrument 1 hazard Instrument 2 shock ... Revenue Expenditure Domestic debt (or assets) Interest payments or investment incomes Instrument 1 Macro- Instrument 2 Fiscal balance economic ... shock Financing Derivatives, contingent credit lines, other products Interest rates & FX Debt (or asset) rates shock accumulation Ex: Sovereign Wealth Fund Note: FX = foreign exchange; ALM = asset liability management. 8 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM The MTFF has the ability to generate fiscal (domestic and external). To this end, the MTFF forecasts, typically multiyear projections (say helps assess the stability (smoothness) of fiscal 3, 5, or 10 years), to inform medium-term outcomes and the underlying fiscal position planning. It helps assess the impact of economic of government over time. It also informs the fluctuations, price dynamics, structural changes government about the sustainability of fiscal in the economy and institutions, contingent and financing policies, given the financial liabilities, potential sources of spending viability of projected expenditure programs obligations in the future, and investment and and/or debt accumulation, and examines operational and maintenance costs, and whether the reform trajectory is robust to highlights interactions among economic and shocks. fiscal variables. One needs to undertake scenario analyses and stress tests. This starts In countries with material dependence on with a view by the analyst of the baseline oil revenue for the government’s budget, for macroeconomic outlook and alternative instance, if there is a risk that international oil scenarios that account for alternative prices will remain below the highs recorded macroeconomic shocks and policy options the in 2014 and earlier, the application of a government may be considering to accompany medium-term fiscal framework (MTFF) would the energy sector reforms. This includes policy be especially timely in order to assess the decisions on tax reform, revenue-sharing need to raise non-oil revenues, to reduce agreements, expenditure planning and and gradually eliminate subsidies, adjust rationalization, fiscal rules, framework for asset- other public expenditures, and in general to liability management, and its borrowing policies flag when the spending on key expenditures FIGURE 3: Adding Subsidies to the Core Integrated Fiscal Framework GDP Global commodity price domestic economic price Demand Revenue elasticity wrt income Subsidy option Expenditure Core model Retail price Fiscal balance Demand elasticity wrt price Interest payments Financing Energy consumption Debt accumulation Fiscal cost of subsidy External debt Domestic debt Note: wrt = with respect to 3. CONCEPTUAL FRAMEWORK 9 is slipping. An MTFF helps to ensure fiscal especially affected by price volatility, as well discipline by making more apparent the as the effects of exchange rate fluctuations impact of current policies on the government on the size of energy subsidies. This resulted balance in the coming years. Likewise, the in increasingly unsustainable subsidy budgets existence of an MTFF may facilitate monitoring and the realization among policy makers that by providing benchmarks, against which phasing out subsidies would increase fiscal budgetary developments can be assessed space. Since 2014, Indonesia successfully over time and there is alignment with the reduced fossil fuel subsidies from 3.1% of GDP country’s development strategy and policy in 2014 to just 1% in 2016. This sharp drop priorities. Overall, a well-designed MTFF reflects the falling oil prices since 2014 and the should reflect the impact of past budgetary government’s policy aim of removing gasoline commitments, as well as the future cost of subsidies and limiting diesel subsidies, and new policy measures. The strengthening of replacing the kerosene price subsidy with the the MTFFs can efficiently complement the LPG subsidy by gradually taking subsidized introduction of other institutional reforms, kerosene out of the market. Mexico’s gasoline such as the introduction of an expenditure and diesel subsidy reforms were outlined in rule or top-down budgeting. Fiscal policy the 2014 Hydrocarbon Law and are currently decisions, such as the fuel taxation regime underway with its announced gradual price that accompanies the energy price adjustment adjustment mechanism (IEA 2016). mechanism (as in Mexico’s energy subsidy reforms in 2017), can affect the pace and The Arab Republic of Egypt is another country magnitude of any decline in subsidy costs or where the government in 2014 committed to increase in government revenue as part of an an ambitious plan to achieve large reductions energy subsidy reform program in a country. in energy subsidies. Following four annual electricity price reforms and three fuel price Energy subsidy reforms in Indonesia and increases, energy subsidies fell steeply from Mexico were triggered by a combination of 6.5% of GDP in fiscal year (FY) 2014 (July to declining production of fossil fuels, rising June) to 3% in FY2016 (see box 1), which in demand, exchange rate devaluation, and turn generated important savings that helped large fluctuations in fuel prices. Being major shift the government budget toward social oil and gas producers, both countries were sectors. BOX 1: PUTTING EGYPT’S ENERGY PRICE REFORMS IN A MACROECONOMIC- FISCAL CONTEXT The government in 2014 committed to an ambitious plan to eliminate energy subsidies, which needed a mid-course correction because of changes in the macroeconomic framework. The goal was to progressively drive energy subsidies down to a target of 0.5% of GDP by FY2019, leaving only limited support for LPG and electricity to benefit low-income consumers. Between July 2014 and July 2017, the government implemented four annual electricity price reforms and three fuel price increases (box table 1.1). As a result, energy subsidies fell steeply from 6.5% of GDP in FY2014 to 3% in FY2016. Subsidies were set to fall further to 2.5% of GDP in FY2017, but the substantial depreciation of the Egyptian pound and box continues next page 10 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM resulting increases in the cost of energy BOX TABLE 1.1: Energy Price Increases production meant that subsidies instead rose toward 3.9% of GDP in FY2017. Year-on-year changes Cost Cumulative Given macroeconomic challenges, the (percent) recovery changes (%) by cabinet has approved the deferral of 2014 2015 2016 2017 (%) FY2017/18 the cost recovery target for electricity until FY2022. Energy subsidies are now Diesel 64 0 31 55 232 64 projected to decline to 3.2% of GDP in Gasoline 80 78 0 47 55 306 73 FY2018 and 1.4% in FY2019, continuing Gasoline 92 41 0 35 43 170 67 downward thereafter, thanks to the planned tariff adjustment. Also, in the Gasoline 95 7 0 0 6 14 89 absence of any energy price reforms LPG 0 0 88 100 275 35 since 2014, it is conservatively estimated Automotive that subsidies would have been higher 144 0 45 25 344 83 natural gas by EGP 256 billion in FY2018, raising the energy subsidy bill to 8.9% of GDP. Electricity 31 19 33 40 190 79 Electricity tariffs in FY2018 are already higher than had originally been targeted BOX FIGURE 1.1: Budgeted Health, Education, Social for FY2019, according to original plans Protection, and Energy Subsidy Spending of FY2014, demonstrating government 400 commitment to energy subsidy reform. Social protection Increase The commitment to containing energy 350 in energy Education + health subsidy is also shown in the shift of In billions Egyptian pound subsidy 300 the government’s approach to energy without 250 256 2014–17 price setting. Starting in 2016, instead of 191 price announcing a trajectory for electricity 200 reform~ prices, the government committed to a 150 subsidy target in its MTFF agreed with 46 52 Actual the IMF, and will adjust prices annually 100 140 139 133 energy to meet this target. Going forward, the 50 (20) 98 80 (9) (8) subsidy (13) (10) petroleum sector has committed to - applying automatic fuel price indexation FY14 FY15 FY16 FY17 FY18 budget* budget* to reduce the impact of external factors on the subsidy target. ~World Bank estimate. *MoF budget of energy subsidy.  Consistent energy price adjustments Note: Numbers in parentheses are in billion U.S. dollars. have yielded important savings that helped shift the government budget toward social sectors. Prior to reform, energy subsidies exceeded the budget for education, health, and infrastructure combined, and were almost three times the budget for government investment. Price reforms have generated savings that contributed to reducing fiscal deficit while allowing for additional government expenditure on education, health, and social protection. Government spending on health and education outstripped spending on energy subsidies for the first time in FY2015 (see box figure 1.1) and has continued to do so. At the same time, spending on social protection is budgeted to almost double in FY2018, as an explicit measure to mitigate the negative social impact of depreciation and energy price increases. 3. CONCEPTUAL FRAMEWORK 11 Financial risks for government and its state- whether public resources available to finance owned enterprises (SOEs) (such as power infrastructure and social spending and promote utilities and national oil companies) from economic development are affordable over fluctuations in exchange rates and interest rates the medium term at an aggregate level and (domestic or external) will pose vulnerabilities helps the authorities monitor compliance and may require budget support. High levels of with fiscal rules and spending targets. DSA SOE debt pose risks to fiscal sustainability and assesses the financial viability of debt-financed stability. One can also examine if the expected budget deficits. It monitors compliance with current expenditures by the government public debt rules and targets for debt and (given by existing policies and trends) are borrowing (see box 2). adequate to fund the needed services in health and social assistance. Otherwise an even Assessing sustainability, in simple terms, larger budgetary savings must be achieved in requires one to form a view about how ensuing years. The implicit social contract— outstanding stocks of liabilities are likely reflecting political concerns about “fair share of to evolve over time. These, in turn, depend sacrifices”—requires increasing tax collection on macroeconomic and financial market and reducing selected expenditures (such as developments that are, by their very nature, energy subsidies going to the rich), and its uncertain and variable (for example, the costs fiscal tradeoffs can be well illustrated using of rolling over debt). It also depends on taking such a country-specific MTFF. a country-specific view about how much fiscal adjustment is politically and socially feasible Fiscal sustainability of energy subsidies and in the near term (that is, how much primary links to debt sustainability and medium-term surplus can be realistically observed in any budget-fiscal framework. The MTFF assesses given year). BOX 2: FISCAL SPACE, FISCAL SUSTAINABILITY, AND SOLVENCY ASSESSMENTS REQUIRE SCENARIO ANALYSES AND STRESS TESTING Fiscal Space World Bank Global Economic Prospects (2015) states that fiscal space “is the availability of budgetary resources for a specific purpose . . . without jeopardizing the sustainability of the government’s financial position or the sustainability of the economy.” The IMF Staff Paper “Assessing Fiscal Space” (2016) states that fiscal space “in general refers to room for undertaking discretionary fiscal policy relative to existing plans without undermining fiscal sustainability.” Assessment of fiscal sustainability requires not just knowing whether fundamental solvency condition holds. Knowledge of likely future paths of debt in baseline and alternative scenarios is also needed, as well as stories around the likely macroeconomic consequences of those paths. Solvency requires the present value of real future primary balances to be equal to the current level of real debt. Examples of typical scenario analyses and stress tests in an MTFF are provided below. box continues next page 12 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM Baseline Scenario • Outlook deemed more likely to happen. • Current legislation versus current policies (no policy change). Alternative Scenarios and Fiscal Risk • Slowdowns in economy-wide growth, income, and employment lead to revenue decline and spending increase (automatic stabilizers). • Fluctuations in commodity prices imply risks for revenue base. • Slow tariff adjustments in SOEs imply revenue decline and potential pressures for higher budget subsidies. • Rising prices (such as oil and commodities) may increase budget-funded subsidies and operating costs. • Wages and pension liabilities (age-related spending) may rise. • Natural disasters pose reconstruction and recovery costs (earthquake, droughts, floods, climate change). • Increases in cost of borrowing lead to higher debt servicing costs. • Incorrect modeling, optimism in forecast of revenue, and lower anticipated spending items in an MTFF may imply that the exogenous parameters and projections might not be accurate (model uncertainty). Debt sustainability also depends on the surrounding fuel prices and GDP growth. behavior of the balance sheets and revenue- The stochastic DSA incorporates past expenditure balances of several different parts co-movements in key macroeconomic of the economy—the government, the banking variables in a country-specific manner. These system, and the corporate and household variables will affect the amount of energy sectors—which are linked with one another by subsidies that a government may have to actual and contingent liabilities. Incorporating provide in any given year due to fluctuations these factors, though theoretically desirable, in oil and commodity prices, real effective may be practically difficult, given the availability exchange rates (REERs), real interest rates, or of consistent and reliable information. Hence, a output growth in the economy. This analytical significant investment must be made in putting approach helps to set out a probabilistic in place monitoring and reporting systems that path of government debt and allows for compile these data on a regular basis in order comparison of the impacts of realization of for it to be used for assessing sustainability. various contingent liabilities versus reform Table 1 shows a hypothetical example of the scenarios in a world with uncertainty.5 There kind of country-specific information that is is an important qualification, however. This needed in this regard. approach is helpful where the government is not materially dependent on oil and FISCAL SUSTAINABILITY UNDER gas revenue. Otherwise, the complexity of UNCERTAINTY calculating fiscal revenue from oil and gas may make it impractical to capture stochastic Fiscal sustainability under uncertainty elements on the revenue side. addresses volatility and uncertainty 3. CONCEPTUAL FRAMEWORK 13 TABLE 1: Hypothetical Example—Elements of a Fiscal Framework Model Year t t+1 … T Economic activity GDP (nominal) million LCU GDP (nominal) growth rate % Real GDP growth rate % GDP deflator growth rate % CPI Inflation % LCU per Exchange rates GBP unit of BAM LCU per EUR unit of EUR LCU per USD unit of USD Revenue Revenue million LCU Taxes million LCU Social security contributions million LCU Grants million LCU Other revenue million LCU Primary Primary (non-interest) expenditure million LCU expenditure Wages and salaries million LCU Social contributions million LCU Subsidies million LCU to public corporations million LCU to private enterprises million LCU Grants million LCU Social benefits million LCU Other expense million LCU Net acquisition of nonfinancial assets million LCU Interest Interest payments million LCU expenditure Total expenditure Total expenditure million LCU Primary fiscal Primary fiscal balance million LCU balance Overall fiscal Overall fiscal balance million LCU balance Cyclically adjusted Cyclically adjusted primary balance million LCU Cyclically adjusted overall balance million LCU 14 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM p Looking at the same conceptual framework, X t = b + ∑ fj X t − j + ut j=1 recall the basic debt accumulation equation (Kraay 2017), where b and fj are more parameters to be (1 + rt) estimated. dt = dt–1 – bt (1 + gt) An example of this approach is a stochastic We then define vector of uncertain future fiscal sustainability model based on a drivers of debt Zt = (rt , gt , bt)' spreadsheet developed by the World Bank that can generate scenarios and stress tests under For a given initial debt ratio d0, the uncertain uncertainty for a wide range of circumstances, future path of dt depends on the uncertain including to assess the fiscal impact of energy future path of Zt: subsidy reforms, as well as fiscal impact of { Z t }t =T t=1 → { d t }t =T t=1 public-private partnerships (PPPs), sovereign guarantees on energy projects, concession We then use econometric model of dynamics arrangements entered into by energy utility of drivers of debt in Zt to simulate {Zt}t with private partners, and natural disasters t=1 , =T for example, panel vector autoregression in a country. It addresses volatility and estimates in Hevia (2012) or undertaking Monte uncertainty surrounding oil prices and GDP Carlo Simulations on the correlated variables. growth. This analytical tool incorporates past co-movements in oil and commodity prices, p q Z t = m + ∑ B j Z t − j + ∑ Γj Xt–j + et the REER, real interest rate (RIR), and output j=1 j=1 growth. This approach helps to set out a probabilistic path of government debt and We also estimate the parameters m , Bj, Γj. allows for comparison of realizations of various Because the exercise is purely predictive, contingent liabilities versus reform scenarios identification of causal effects is not required. in a world with uncertainty. “Fan charts” and cumulative probability distributions allow Xt is vector of exogenous global variables (for the analyst to simulate the impacts of these example, world growth, terms of trade, and shocks with key fiscal and debt sustainability- oil prices) that affect debt drivers in Zt related variables. 4. WHY DO FISCAL RISKS MATTER IN THE ENERGY SECTOR? Contingent liabilities pose fiscal risks that need energy sector poses in a country. If state- to be identified and managed as well. There owned energy firms are making large losses are potential future subsidies, beyond those and considered too big or too important to currently being financed by the government, fail, if operational inefficiencies in the sector that need to be identified and quantified in lead to mounting debts by all energy firms order to get a fuller picture of the fiscal costs and threaten the delivery of essential energy and macroeconomic vulnerabilities that the services, or if currency depreciation makes 4. WHY DO FISCAL RISKS MATTER IN THE ENERGY SECTOR? 15 it impossible for energy firms to pay back Going beyond estimating aggregate fiscal risks foreign debts when their revenues are in local to pointing out the specific sources of such risks currency, the government may eventually have and what can be done about them is important. to step in and rescue the firms. This requires deep sector or utility knowledge and collaborative efforts among specialists with Similarly, unpredictable weather conditions different areas of expertise. For example, several in countries that depend on hydropower recent country-specific studies have aimed to or natural disasters that impact energy link fiscal risks, investment in energy, consumer availability and prices will pose fiscal risks for a tariffs, and economic performance (Mansour government. In countries with a large share of and others 2016). Sectoral models in the hydropower in the power mix, such as in east Comoros, Kosovo, and Lebanon allowed policy Africa, droughts pose a considerable financial makers to understand better the implications challenge: hydropower is often the least-cost of policy decisions on these key variables. In source of electricity, whereas emergency diesel all cases, the models developed responded to generation to make up for lost hydropower is country-specific context and policy questions. the most expensive, and yet raising electricity In the Comoros case, the questions were the tariffs to capture the sudden increase in the following: cost of power generation is seldom, if ever, politically feasible. In the absence of very high • What are the subsidy needs, and will ongoing power tariff increases, the price gap grows, as reforms narrow the utility’s financing gap, do contingent liabilities for the government. thereby reducing fiscal risk? This makes it imperative to look for ways to In the Kosovo case, the questions were the improve the government’s fiscal position over following: the medium term, while at the same time bearing some of the fiscal costs of energy • How do investment and subsidy choices subsidy reforms in the transition period to impact required power imports, exports, fully market-determined pricing mechanisms and consumer tariffs, and what are the likely for energy. New infrastructure investments fiscal impacts of these choices? using innovative PPPs may also create claims on future public resources, calling for a careful In the Lebanon case, the question was as assessment of associated fiscal risks and for follows: contingency planning by the fiscal authorities • What is the expected economic loss in terms in a government. of growth resulting from the deficiency of the electricity sector? 16 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM 5. HOW DO WE ASSESS THE FISCAL IMPACTS OF THE FISCAL RISKS? Fiscal risks need to be assessed as part of a In practice, to understand how important the multi-dimensional effort and will require a look contingent liabilities that can arise from energy at both sides of a government’s balance sheet. subsidies (as opposed the subsidies currently A convenient analytical approach that can be financed by the government) may be in the used to identify and catalogue the various overall fiscal situation of the government, sources of fiscal risks for the government the first step is to look at both sides of the from energy sector subsidies is using the government balance sheet. This fiscal analysis government’s balance sheet. needs to cover the entire portfolio of direct (defined below) and contingent liabilities, as Theoretical foundations can explain the well as assets, and the revenue base. This, in realization of contingent liabilities by modifying turn, requires that the government identify, the domestic debt accumulation equation to classify, and assess its fiscal risks so that it can reflect risk that government may assume regularly generate reliable estimates of future additional liabilities lt = Lt /Yt with probability payments that may result from its past and pt, such that pending liabilities. Informational, political, and ⎧(1 + rt) institutional challenges need to be overcome ⎪(1 dt–1 – bt , with probability 1 – pt ⎪ + gt) in this regard, for which an active role of senior dt = ⎨ policy makers in the government is critical. ⎪(1 + rt) dt–1 – bt + lt , with probability pt ⎪ ⎩(1 + gt) On the liabilities side, this implies estimating Figure 4 shows that either the contingent the fiscal risks matrix (see Table 2). This fiscal risk materializes (right panel) or it does delineates the liabilities of the government not (left panel), and the debt accumulation in terms of a two-by-two matrix. Liabilities equation holds in both states of the world. can be categorized as direct or contingent, FIGURE 4: Realization of Contingent Liabilities under Uncertainty Contingent Liability Realization Scenario Baseline Scenario (Guarantees and PPP Project X financial liability) 80 Debt (% of GDP) 120 Explicit liabilities (% of GDP) 110 70 100 60 90 80 50 70 40 60 50 30 40 20 30 20 10 10 4 5 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 6 7 8 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 0 0 0 0 0 0 0 0 0 0 0 0 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 1 5 10 25 50 75 90 95 99 Baseline scenerio 5. HOW DO WE ASSESS THE FISCAL IMPACTS OF THE FISCAL RISKS? 17 and simultaneously, whether they are explicit most prevalent cases of subnational energy or implicit. subsidies are typically for district heating. Direct liabilities of a government are the Contingent liabilities are also a convenient legal and contractual obligations that the political tool, since there is no immediate government will have to honor in any event. effect on the government’s budget and they Typical direct liabilities of a government leave the eventualities of higher public debt include government debt and expenditures and fiscal costs for the future. They are also as stated in the budget act each year, but used as a form of providing “hidden” state they also include non-discretionary long-term assistance in a country (for example, to a failing expenditure obligations of the government, and inefficient state-owned energy firm). such as the civil service wage bill, pension, and future recurrent cost of public investment Implicit liabilities of the government are projects and fully or partially rescued commitments that are typically based on parastatals. If there are any long-term purchase political announcements, public expectations, agreements based on law and contracts, they and possible interest group pressures in are also part of government direct liabilities. countries. They include primarily losses, Such long-term obligations often involve non-guaranteed obligations, arrears, or the private sector. Typical examples of such deferred maintenance of autonomous or obligations are power purchase agreements semi-autonomous SOEs in the energy sector. with independent power producers (IPPs) and Since some of these entities are deemed special purpose vehicles (SPVs) set up for PPP critical for delivery of public services (such arrangements, under which the government as affordable energy supply), a government is the main purchaser of services for its own may be expected by the public for moral or use or as an input to provide another service, political reasons to eventually pay for these or on behalf of final consumers. These explicit non-guaranteed debts, arrears, and deferred contingent liabilities of a government are maintenance of some SOEs. For the same commitments that are based on law and reason, losses, non-guaranteed debt, and contracts, which primarily include explicit arrears of off-budget liabilities and local government guarantees and financial risk governments are implicit contingent liabilities associated with direct government liabilities of the state government. These contingent and explicit guarantees. The most obvious liabilities can become the direct liabilities of the contingent liability of this kind is the explicit central government if some uncertain event is loan guarantees issued to state-owned energy triggered where the original entity or debtor firms. Country experience suggests that these is unable to meet that payment obligation. explicit contingent liabilities can quickly and On the assets side, the different sources significantly raise the government’s debt-to- of potential revenues that can cover the GDP or deficit-to-GDP ratios, when unforeseen government obligations are catalogued in the events occur. Another common source of fiscal hedge matrix (see table 3; Polackova- contingent liability is on account of PPPs and Brixi and Mody 2002). These sources of off-budget entities. For instance, in Nigeria, financial safety are also either explicit or the Lagos State Government has in the past implicit, direct or contingent. Direct sources issued a guarantee to debt services payable include sources that the government can by the state utility concessions company. The 18 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM employ by its legal power to raise income products, for example, through corporate from its existing or tangible assets. The direct income tax reductions to the producers and implicit sources represent those sources distributors of the petroleum products. While that the government can draw on from its contingent implicit sources of financial security existing assets, but are not in their direct are typically not available to the government, control at the time and will require a special unless some uncertain event occurs, even circumstance to access. This latter source then these sources would require a special would then offset the governments’ fiscal justification by the government to use them. risks to a limited extent only.6 Contingent explicit sources include measures that the Once these two matrixes are filled with government can legally use to raise revenues country-specific items, the government can from sources other than its own assets. These identify the exact scope of the fiscal analysis can be reduced by subsidizing petroleum and management of fiscal risks, including those emanating from the energy sector. TABLE 2: The Fiscal Risks Matrix on the Liabilities Side Direct Liabilities Contingent Liabilities Explicit liabilities • Foreign and domestic sovereign • Guarantees for borrowing and obligations (Legal obligation debt of sub-national governments and SOEs. no choice) • Budget expenditures—both in • Guarantees for trade and exchange rate the current fiscal year and those risks. legally binding over the long • Guarantees for private investments term (civil servant salaries and (PPPs). pensions) • State insurance schemes (deposit insurance, private pension funds, crop insurance, flood insurance, war-risk insurance). • Unexpected compensation in legal cases related to disparate claims. Implicit liabilties • Future public pensions if not • Defaults of sub-national governments and (Expectations— required by law SOEs on nonguaranteed debt and other political decision) obligations. • Social security schemes if not required by law • Liability clean-up in entities being privatized. • Future health care financing if not required by law • Bank failures (support beyond state insurance). • Future recurrent cost of public investments • Failures of nonguaranteed pension funds, or other social security funds. • Environmental recovery, natural disaster relief. Note: These liabilities refer to fiscal authorities, not the central bank. Source: Polackova-Brixi (1998). 5. HOW DO WE ASSESS THE FISCAL IMPACTS OF THE FISCAL RISKS? 19 TABLE 3: Fiscal Hedge Matrix: Assets and Contingent Financing DIRECT CONTINGENT SOURCES (based on the stock (dependent on future events, such as value of Financial Safety of existing assets) generated in the future) Explicit • Asset recovery (such • Government revenues from natural resource as workouts, sales of extraction and sales Based on nonperforming loans, state government legal • Government customs revenues equity sales) powers, such as ownership, the right • Tax revenues less • Proceeds from privatization to raise taxes, and of SOEs and other public }} New tax expenditures to be introduced in other revenues resources the future • Recovery of government }} Revenues already earned from forward loan assets (for example, sales (such as commodity forward sales) resulting from earlier direct government lending }} Costs of hedging instruments and re-insurance purchased by government to protect tax revenue Implicit • Stabilization and • Profits of SOEs contingency funds* Based on • Contingent credit lines and financing government indirect • Positive net worth of central commitments from international financial control bank institutions. • Current account surpluses across currencies. * Can be designed as general or specific-purpose funds under direct or indirect control of government. Source: Polackova-Brixi and Schick (1998), p. 26. 20 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM 6. FISCAL IMPACTS AND STATE-OWNED ENTERPRISES Energy subsidy reforms are often carried out growth prospects. This has implications for the as part of broader energy sector reforms. design, speed, and success in implementation Energy SOEs are often among the largest of a country’s efforts to manage the SOEs in many countries, and the SOE fiscal macroeconomic vulnerabilities that it may risk analysis should apply to the energy firms be exposed to because of commodity price as well. To this end, one needs to look at SOE fluctuations and their effects on fiscal revenue, fiscal risks emanating from the energy firms. In and its impact on inclusive growth over the addition, SOE restructuring will impose fiscal medium to long term. pressures—especially for loss-making SOEs and those with already large outstanding In a decentralized economy, the provincial liabilities. and municipal governments may have to spend more on special retraining programs for Sovereign guaranteed debt of state-owned displaced SOE workers and provincial social energy entities and related SOEs is not all safety net commitments. The province’s ability that one needs to monitor and manage from to generate plausible subnational revenues the fiscal risks perspective. One also needs to and expenditure projects are important for take into account other sources of fiscal risks assessing and maintaining fiscal sustainability from SOEs’ operations, such as the operating of these added fiscal costs. Investments in losses expected, reported liabilities, deferred urban infrastructure and quality of life, reform maintenance, and payment arrears by SOEs of government-owned energy firms, and the (for example, fuel cost owed by a state-owned associated social protection mechanisms will utility to a state oil company). These total require the mobilization of substantial financial obligations need to be covered on top of any resources, for instance, by ending subsidies future losses. In addition, SOE liabilities may to loss-making locally administered public translate into explicit and implicit fiscal costs utilities and divesting SOE shares and assets for the national or provincial governments, or managing them more efficiently. the latter particularly so in the case of district heating. In natural-resource-rich developing Wider and faster transformation of SOEs countries, SOEs often dominate the mining can place greater demands on social safety or hydrocarbon sector, and anything they do nets. Additional reforms to improve the related to revenue and spending will most fiscal sustainability and efficiency of key likely have macroeconomic implications. Oil programs, including unemployment insurance production and the concentrated nature as or minimum subsistence payments in urban well as the size of the government revenue areas and pensions, need to be undertaken it generates create demand for low-priced in countries. petroleum products from citizens in the form of Figure 5 illustrates the ways in which the consumer price subsidies. One needs to follow state may be called upon to provide fiscal the transmission mechanisms of these sectors support to state-owned energy suppliers and as they affect the macroeconomic situation other associated SOEs in a country. Figure of the country and its medium- to long-term 6. FISCAL IMPACTS AND STATE-OWNED ENTERPRISES 21 6 highlights the importance of examining where there has been lack of progress on cost balance sheets of the individual SOEs in pass-through mechanisms. For hydrocarbon order to design an appropriate fiscal risk exporters, a significant part of energy subsidies management strategy. Just looking at the may be carried on the balance sheets of operating statements and cash flows will upstream hydrocarbon producers, cross- present only a partial picture. subsidizing below-market provision of fuels downstream. SOE liabilities may translate into Energy SOEs are exposed to both exchange explicit and implicit fiscal costs for a national rate and fuel price fluctuations, which make or subnational government. cost recovery fragile, especially in countries FIGURE 5: Fiscal Risks from Explicit and Implicit Subsidies to Energy SOEs SOE • Carries out noncommercial • Capital injections • High-debt levels objectives • Bail outs Can result in • Arrears (vis-à-vis tax May require • Incurs losses (technical) authorities, suppliers, • Cleanups of SOE • Mispricing inter-SOE) balance sheets Exacerbated by State • Allows arrears to • Moral hazard Mandates SOE to carry accumulate • Reduced fiscal space out NCOs (sometimes • Subsidizes lending to SOE • High borrowing costs unremunerated by State) • Guarantees SOE debts • Fiscal vulnerability FIGURE 6: Fiscal Risk Matrix: How SOEs Can Contribute to Fiscal Impacts in Future Operating Balance sheet statement Revenues Direct liabilities Contingent liabilities Direct Tax payments Dividend flows Explicit • Direct subsidies and • Sovereign guarantees obligations transfers, including to for SOE debt cover noncommercial • Contracts (including obligations of SOEs PPPs) with guarantee • On-lent loans to SOEs clauses (e.g., securing loan repayments) Implicit • Arrears owed to the • Default of unguaranteed obligations State SOE debt (bail outs, capital injections) • Inter-enterprise arrears • Cleanup of SOE liabilities, arrears 22 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM 7. FISCAL RISKS FROM ENERGY SECTOR PUBLIC- PRIVATE PARTNERSHIPS Energy SOEs may enter into PPPs in their 1 | Who initiates the project? The impact of investment and service delivery decisions (for main fiscal indicators (that is, deficit and example, through build-operate-transfer and/or debt) varies depending on the public entity concession agreements with private partners ultimately responsible for the project (such over a multiyear period), which have associated as central, local governments, and SOEs). direct and contingent fiscal costs. Assessment 2 | Who controls the asset? Simple, of fiscal impacts from such arrangements standardized questions assist the user in requires combining sector-specific and making an informed decision about the economy-wide analyses of fiscal risks. government’s ability to control the asset. The PPP-Fiscal Risks Assessment Model The funding structure of the project is (P-FRAM), developed by the IMF and the what determines its implication on main World Bank, is an analytical tool that can fiscal aggregates. P-FRAM allows for three be used to assess the potential fiscal costs funding alternatives: (a) the government and risks arising from such PPP projects. In pays for the asset using public funds; (b) many countries, investment projects have the government allows the private sector been procured as PPPs not for efficiency to collect fees directly from users of the reasons, but to circumvent budget constraints asset (such as tolls); or (c) a combination and postpone recording the fiscal costs of of the two. providing infrastructure services. Some 3 | Does the government provide additional governments have ended up procuring s u p p o rt to th e p ri vate p ar t ne r? projects that either could not be funded within Governments can not only fund PPP their budgetary envelope, or that exposed projects directly, but they can also provide public finances to excessive fiscal risks. To a variety of support to the private partner, address these concerns, P-FRAM has been including guarantees, equity injections, developed as an analytical tool to quantify and tax amnesties. the fiscal implications of PPP projects. It is designed to be used mostly by PPP units in 4 | PPP fiscal risk matrix. Directed by a ministries of finance. sequence of questions regarding project characteristics, P-FRAM identifies a set In practice, assessing a PPP project involves of possible explicit or implicit contingent both gathering specific project information and liabilities, inviting the user to present making judgments about the government’s information on the likelihood and impact role at critical stages of the project cycle. of each risk, and on the mitigation measure P-FRAM provides a structured process for in place. Once project-specific and gathering information for a PPP project in macroeconomic data are entered, P-FRAM a simple, user-friendly, spreadsheet-based automatically generates standardized platform, following a four-step decision tree, outcomes, which include (a) project cash as follows: 7. FISCAL RISKS FROM ENERGY SECTOR PUBLIC-PRIVATE PARTNERSHIPS 23 flows; (b) fiscal tables and charts, both in that country at the time or could be on a cash and accrual basis; (c) debt addressed in the medium term. sustainability analyses with and without the PPP project; and (d) sensitivity analysis Having catalogued sources of fiscal risks and of main fiscal aggregates to changes in looked at fiscal impacts of energy subsidies, macroeconomic and project-specific SOE, and PPP fiscal risks using these analytical parameters. The P-FRAM “heat map” is tools, the results can then be consolidated to generated by the model, which provides see their overall impacts on fiscal sustainability a visual depiction of the fiscal risks from under uncertainty. Box 3 provides a summary the selected PPP projects, its likely fiscal of this toolkit for fiscal risk assessments from impact, whether there is a contingency contingent liabilities. This analysis can be plan to address or mitigate it, and if it conducted at national and subnational levels requires priority action by the authorities of government. BOX 3: TOOLKIT FOR FISCAL RISK ASSESSMENT FROM CONTINGENT LIABILITIES Fiscal Risks and Fiscal Hedge Matrix helps identify which fiscal risks are large, how their size can be affected so as to begin to design fiscal risk mitigation strategies and reform measures to minimize these risks. Integrated Fiscal Framework, the subject of this ESRAF good practice note, helps quantify the fiscal resource envelope and plan the resource allocation among competing spending programs as and after energy subsidy reforms are implemented. P-FRAM Model is applied to specific cases to identify cash flow and actuarial effects, as well as a “heat map” of various fiscal risks the project faces on the basis of these PPP contracts (at national and subnational levels). Stochastic Fiscal Sustainability Assessment incorporates uncertainty into the standard DSA. It also looks at aggregation of fiscal risks in a probabilistic and endogenous analytical framework, and generates “fan charts” and cumulative probability distributions of key variables (such as debt-to-GDP ratio). 24 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM 8. LESSONS: FISCAL IMPACT AND FISCAL RISK ASSESSMENTS To appropriately assess the fiscal impact • Develop reform scenarios based on of energy subsidy reforms in a country, the relevant policy questions, potential sector following steps should be considered: reforms, and any potential investment plans (including through PPPs), and incorporate • Examine income and expense statements them into an integrated fiscal framework. of energy SOEs and identify budgetary flows between government (national or • Examine implications of energy subsidy subnational) and the SOEs. removal scenarios on medium-term debt and fiscal sustainability indicators (deterministic • Identify all other sources of energy subsidies. DSA) and compare with an appropriate • Catalogue fiscal risks stemming from energy country debt sustainability thresholds as sector by constructing relevant “fiscal risks appropriate. matrix” for government (to include sources, • For middle-income countries and countries such as PPPs, sovereign guarantees, and off- with market access (and better data), budget or quasi-fiscal spending in sector). undertake fiscal sustainability analysis under uncertainty in addition (that is, a stochastic DSA) as appropriate. ENDNOTES ENDNOTES 1 See Good Practice Note 1 by Masami Kojima. It should be noted that the price-gap approach does not capture the full fiscal cost of energy subsidies. The price-gap approach may capture price support to consumers or producers, but fails to capture many other forms of subsidies that do not have a clearly identifiable effect on the prices being examined, such as underpricing of access to government-owned land, shifting of risk burdens from producers to consumers or the government, and below-market provision of loans. 2 Rentschler and Bazilian (2016) note that between 1980 and 2010, 36% of global carbon emissions were driven by fossil fuel subsidies. They cite another study by Schwanitz and others (2015), which found that reforming fossil fuel subsidies by 2020 could reduce global carbon emissions by 6.4% in 2050, and if 30% the savings were reinvested in renewables and energy efficiency, this would further take this reduction in carbon emissions to 18%. 3 These are Burkina Faso, China, the Arab Republic of Egypt, Ethiopia, Ghana, India, Morocco, New Zealand, Senegal, Sierra Leone, Singapore, United Arab Emirates, and Vietnam.. 4 Seigniorage embodies the impact of monetary policy on debt build-up in a country. It is defined as the change in money supply as a share of GDP. (It can be estimated by the ratio of change in reserve money to gross domestic product). 5 This typically can supplement other fiscal risk assessment tools in the World Bank Group Toolkit, such as “The Fiscal Risks Matrix,” the World Bank-IMF PFRAM, and PROST model for pensions, among others, to get a more holistic picture of the fiscal impacts of energy subsidies and their reform path. 6 Another contingent financing source for energy SOEs is the withholding of dividends by them to make up for losses, including subsidies. 25 GOOD PRACTICE NOTE 2: ASSESSING THE FISCAL COST OF SUBSIDIES AND FISCAL IMPACT OF REFORM REFERENCES Aguinaga, Paulina, Charles Ncho-Oguie, and Jean-Pascal Nganou. 2015. Review of Oil-Price Subsidies: Lessons from Uganda. In background paper for UGANDA Country Economic Memorandum, “Economic Diversification and Growth in the Era of Oil and Volatility.” Chapter 2, World Bank Report No. 97146-UG, Washington, D.C., June. Burnside, Craig. 2005. “Assessing New Approaches to Fiscal Sustainability Analysis.” In Fiscal Sustainability in Theory and Practice: A Handbook. Washington, D.C.: World Bank. Cabrera, Maynor Vinicio, Jaime Aristy-Escuder, Blanca Moreno-Dodson, and Miguel Eduardo Sánchez Martín. 2016. Fiscal Policy and Redistribution in the Dominican Republic: An Analysis Based on the Commitment to Equity Methodology, for 2013. Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/827851478242582427/ Fiscal-policy-and-redistribution-in-the-Dominican-Republic-an-analysis-based-on-the- commitment-to-equity-methodology-for-2013 Caner, Mehmet, and Thomas J. Grennes, and Friederike (Fritzi) N. Köhler-Geib. 2010. Finding the Tipping Point—When Sovereign Debt Turns Bad. Rochester, NY: SSRN. https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1612407 Coady, David, Ian W. H. Parry, Louis Sears, Baoping Shang. 2016. How Large Are Global Energy Subsidies? http://www.imf.org/en/publications/wp/issues/2016/12/31/ how-large-are-global-energy-subsidies-42940 Hernandez Ore, Marco Antonio, Luis Alvaro Sanchez, Liliana D. Sousa, and Leopoldo Tornarolli (eds.). 2017. Fiscal and Welfare Impacts of Electricity Subsidies in Central America. Washington, D.C.: World Bank. IEA (International Energy Agency). 2016. Fossil Fuel Subsidy Reform in Mexico and Indonesia. Paris: OECD/IEA. https://www.iea.org/publications/freepublications/ publication/PartnerCountrySeriesFossil_Fuel_Subsidy_Reform_Mexico_ Indonesia_2016_WEB.pdf ———. 2016. World Energy Outlook 2016. Paris: OECD/IEA. https://www.iea.org/ publications/freepublications/publication/world-energy-outlook-2016---executive- summary---english-version.html ———. 2017. Tracking Fossil Fuel Subsidies in APEC Economies. Paris: OECD/IEA. https:// www.iea.org/publications/insights/insightpublications/tracking-fossil-fuel-subsidies-in- apec-economies.html IMF (International Monetary Fund). 2016. Assessing Fiscal Space: An Initial Consistent Set of Considerations. IMF: Washington, D.C. http:// www.imf.org/en/publications/policy-papers/issues/2017/01/13/ pp5080-assessing-fiscal-space-an-initial-consistent-set-of-considerations Kraay, Aart. 2017. “Presentation in MFM’s Fiscal Sustainability and Fiscal Risks Analyses Course.” Macroeconomics and Fiscal Management Global Practices, World Bank, Washington D.C. May. 26 REFERENCES Kumar, Manmohan S., and Jaejoon Woo. 2010. Public Debt and Growth. IMF Working Paper No. WP/10/174. Fiscal Affairs Department. Washington, D.C.: IMF. https://www.imf.org/ external/pubs/ft/wp/2010/wp10174.pdf Mansour, Wael, Shireen Mahdi, Simon Davies, and Chadi Bou Habib. 2016. Assessing Policy Trade-Offs with Easy-to-Use Models in PERs—Examples from the Energy Sector. Macroeconomics and Fiscal Management Practice Notes No. 14. Washington, D.C.: World Bank Group. http://documents.worldbank.org/curated/en/748971467995618392/ pdf/106478-BRI-PUBLIC.pdf Polackova-Brixi, Hana. 1998. Contingent Liabilities—A Threat to Fiscal Stability. PREM Notes, no. 9, November. https://www.internationalbudget.org/wp-content/uploads/Contingent- Liabilities-A-Threat-to-Fiscal-Stability.pdf Polackova-Brixi, Hana, and Sudarshan Gooptu. 2002. “Dealing with Contingent Liabilities in Indonesia and Thailand.” In Hana Polackova Brixi and Allen Schick (eds.), Government at Risk: Contingent Liabilities and Fiscal Risk, Chapter 11. Washington, D.C., and New York: World Bank and Oxford University Press. Polackova-Brixi, Hana, and Ashoka Mody. 2002. “Dealing with Government Fiscal Risk: An Overview.” In Hana Polackova Brixi and Allen Schick (eds.), Government at Risk: Contingent Liabilities and Fiscal Risk, Chapter 1. Washington, D.C., and New York: World Bank and Oxford University Press. Polackova-Brixi, Hana, and Allen Schick (eds.). 2002. Government at Risk: Contingent Liabilities and Fiscal Risk. Washington, D.C., and New York: World Bank and Oxford University Press. ISBN 0-8213-4835-3. Reinhart, Carmen M., and Kenneth S. Rogoff. 2010. Financial Crash to Debt Crisis. NBER Working Paper No. 15795; JEL No. F3,H6,N10. Cambridge, Mass.: NBER. Rentschler, Jun, and Morgan Bazilian. 2016. “Reforming Fossil Fuel Subsidies: Drivers, Barriers and the State of Progress.” Climate Policy (June 24). http://dx.doi.org/10.1080/1 4693062.2016.1169393. Schwanitz, V. J., F. Piontek, C. Bertram, and G. Luderer. 2014. “Long-Term Climate Policy Implications of Phasing Out Fossil Fuel Subsidies.” Energy Policy 67:882–894. Sovacool, Benjamin K. 2017. “Reviewing, Reforming, and Rethinking Global Energy Subsidies: Towards a Political Economy Research Agenda.” Ecological Economics Journal 135 (May):150–63. Published by Elsevier. http://dx.doi.org/10.1016/j. ecolecon.2016.12.009 World Bank. 2015. World Bank Global Economic Prospects: A Fragile Recovery. http://www. worldbank.org/en/publication/global-economic-prospects 27 Energy Subsidy Reform Assessment Framework LIST OF GOOD PRACTICE NOTES NOTE 1 Identifying and Quantifying Energy Subsidies NOTE 2 Assessing the Fiscal Cost of Subsidies and Fiscal Impact of Reform NOTE 3 Analyzing the Incidence of Consumer Price Subsidies and the Impact of Reform on Households — Quantitative Analysis NOTE 4 Incidence of Price Subsidies on Households, and Distributional Impact of Reform — Qualitative Methods NOTE 5 Assessing the readiness of Social Safety Nets to Mitigate the Impact of Reform NOTE 6 Identifying the Impacts of Higher Energy Prices on Firms and Industrial Competitiveness NOTE 7 Modeling Macroeconomic Impacts and Global externalities NOTE 8 Local Environmental Externalities due to Energy Price Subsidies: A Focus on Air Pollution and Health NOTE 9 Assessing the Political Economy of Energy Subsidies to Support Policy Reform Operations NOTE 10 Designing Communications Campaigns for Energy Subsidy Reform