Report No. 71979-JO Hashemite Kingdom of Jordan Options for Immediate Fiscal Adjustment and Longer Term Consolidation November 2012 Poverty Reduction and Economic Management Department Middle East and North Africa Region Document of the World Bank Currency Equivalents (Exchange Rate Effective January 8, 2012) JD1 = US$1.412 Fiscal Year January – December Acronyms and Abbreviations Btu British thermal unit m3 Cubic meter cu ft Cubic feet MIC Middle-Income Country DoS Department of Statistics MoF Ministry of Finance DPL Development Policy Loan MoSD Ministry of Social Development DPR Development Policy Review MoWI Ministry of Water and Irrigation ERC Electricity Regulatory MT Metric ton Commission GoJ Government of Jordan MTEF Medium-Term Expenditure Framework GST General Sales Tax NAF National Aid Fund GWh Gigawatt hour NEPCO National Electric Power Company of Jordan HEIS Household Expenditure and O&M Operations and Maintenance Income Survey of 2008 HFO Heavy fuel oil o/w Of Which HH Household(s) PEP Public Expenditure Perspectives IBT Inclining Block Tariff PIMI Public Investment Management Efficiency Index IMF International Monetary Fund NREL National Renewable Energy Laboratory IPP Independent Power Producer SBA Stand-by Arrangement (IMF) JD Jordanian Dinar SCC Social Security Corporation JPRC Jordan Petroleum Refinery SOE State-Owned Enterprise Company JVA Jordan Valley Authority toe Ton of Oil Equivalent kg Kilogram TOR Terms of Reference kWh Kilowatt hour UMIC Upper-Middle-Income Country l Liter WAJ Water Authority of Jordan LPG Liquefied Petroleum Gas Vice President: Inger Andersen Country Director: Ferid Belhaj Sector Director: Manuela V. Ferro Sector Manager: Bernard G. Funck Task Team Leader: Chadi Bou Habib HASHEMITE KINGDOM OF JORDAN Options for Immediate Fiscal Adjustment and Longer Term Consolidation Table of Contents Acknowledgments........................................................................................................................................iv 1. INTRODUCTION ................................................................................................................................ 1 2. Reversing the Decline in Domestic Revenues .................................................................................. 6 2.1. Understanding Jordan’s Decline in Revenues to GDP from 2007 to 2011........................................ 6 2.2. Improving the Revenue System ....................................................................................................... 10 2.3 Options to Increase Revenues........................................................................................................... 12 3. EFFICIENCY GAINS AND SAVINGS BY LARGE CATEGORIES OF SPENDING ................................ 14 3.1. Wage Bill ......................................................................................................................................... 15 3.2. Capital Expenditure ......................................................................................................................... 20 3.3. Social Assistance/Social Spending .................................................................................................. 23 3.4. Public Pensions ................................................................................................................................ 25 4. FISCAL IMPLICATIONS OF THE ENERGY AND POWER SECTORS ................................................ 28 4.1. Types of Energy Subsidies............................................................................................................... 28 4.2. Recent Developments ...................................................................................................................... 29 4.3. Short-Term Options and Recommendations to Address Energy Subsidies in the Electricity Sector ................................................................................................................................................................ 35 4.4. Short-Term Options and Recommendations to Address Energy Subsidies in the Petroleum Sector ................................................................................................................................................................ 39 4.5. Long-Term Options to Address Energy Subsidies and Diversify Sources of Energy ..................... 42 5. DISTRIBUTIONAL IMPACT OF ELECTRICITY TARIFF AND CONSUMER SUBSIDY REFORMS ..... 44 5.1. Scenarios for Cost Recovery from Residential Users ................................................................. 44 5.1.1 Residential Electricity Usage Patterns under Existing Tariffs .................................................. 45 5.1.2 Residential Electricity Tariffs - Policy Reforms for Cost Recovery ........................................... 47 5.2. Food Subsidies ................................................................................................................................. 51 6. REDUCING THE DEFICITS OF OWN BUDGET AGENCIES: THE CASE OF THE WATER SECTOR 54 6.1. Characteristics of the Water Sector in Jordan.................................................................................. 54 6.2. Sources of Waste and Deficits in the Water Sector ......................................................................... 55 6.3. Options for Reform and Potential Savings ...................................................................................... 57 i 7. CONCLUSION .................................................................................................................................. 59 Appendices.................................................................................................................................................. 66 Appendix 1. Revenues, Expenditures, and Deficits in Jordan, 2011–13 (JD mil)) ............................. 66 Appendix 2. Flowchart for Implications of Exogenous Shocks in Jordan........................................... 67 Appendix 3. Tariffs Structure in Jordan .............................................................................................. 68 Tables Table 1.1 Summary of Additional Measures and Their Impacts on the Budget, 2012 and 2013 ............... (% of GDP) ...................................................................................................................................... 5 Table 2.1 Annual Cost of 2008 Tax Exemptions and Reduction Package................................................ 9 Table 2.2 Jordan’s Income Tax Rates before and after 2010 Reform..................................................... 10 Table 2.3 Savings from Planned Revenue Measures (percent of GDP) ................................................ 11 Table 3.1 Possible Savings from Freeze Measures, 2012 and 2013 (% GDP ........................................ 20 Table 4.1 Structure and Flows of Subsidies in Jordanian Energy Sector, 2007–12................................ 29 Table 4.2 Change in Petroleum Subsidies, 2007–12............................................................................... 29 Table 4.3 Jordan’s Consumption of Petroleum Products, 2008–11 (MT) ............................................... 32 Table 4.4 Estimation of Losses and Cost Recovery by Large Categories of Consumers for 2012......... 38 Table 4.5 Scenarios for LPG Subsidies Reform (JD mil) ....................................................................... 42 Table 5.1 Tariff Brackets and Household Consumption of Electricity................................................... 46 Table 5.2 Per Capita Expenditures by Decile ......................................................................................... 46 Table 5.3 Increases in Total Revenues (with zero elasticity).................................................................. 48 Table 5.4 Impact on Revenues of Increases with an Elasticity of -0.3 ................................................... 49 Table 5.5 Increases in Revenues from Reformed Tariff Structure.......................................................... 50 Table 5.6 Simulations of Cost Recovery from Subsidy Cuts.................................................................. 53 Table 7.1 Framework for the Selection of Policy Actions...................................................................... 61 Table 7.2 Matrix of policy Objectives and Actions ................................................................................ 63 Boxes Box 4.1 PPP Structure in Jordan Electricity Sector.............................................................................. 33 Box 7.1 Six proposed Criteria for the Selection of Short-Term Policy Measures................................ 59 Figures Figure 2.1 Domestic Revenues Trend, 2000–11 (%) ............................................................................. 7 Figure 2.2 Trends in Jordan’s Tax Revenues: Share of GDP, 2007–11 (%).......................................... 9 Figure 3.1 Budget Spending in Jordan as Share of GDP, 2000–11 (%)............................................... 14 Figure 3.2 Real Wage Bill and Historical Trends, 2000–13 (year-to-year change, %) ....................... 16 Figure 3.3 Civil Servants’ Wage Bill Composition, 2010–12 (%) ....................................................... 17 Figure 3.4 Compensation of Employees (% of total expenditures) ...................................................... 18 Figure 3.5 Overall PIMI: Decomposition by Subindex........................................................................ 21 Figure 3.6 Functional Classification of Capital Expenditure, 2008–12 (%) ........................................ 22 ii Figure 3.7 Social Assistance Spending by Major Programs, 2010–13 (US$ mil) ................................ 24 Figure 3.8 Public Pension Expenditures 2010 (% of GDP) ................................................................. 26 Figure 3.9 Pensions and Compensations, 2008–12 (JD mil and % of GDP) ....................................... 27 Figure 4.1 Structure of Jordan’s Power Generation Fuels (000s toe)................................................... 30 Figure 4.2 Average Price of Crude Oil and Natural Gas, 2005–11 (US$/mil BTU)............................. 31 Figure 4.3 Trends in Petroleum and Electricity Consumption in Jordan, 2006 Base Year .................. 33 Figure 4.4 Electricity Consumption by Users, 2008–11 (GWh)........................................................... 33 Figure 4.5 NEPCO’s Average Selling vs. Buying Tariffs.................................................................... 36 Figure 4.6 NEPCO’s Average Buying vs. End-Consumer Tariff, 2007–12 (JD/ kWh sold) ............... 37 Figure 4.7 Retroactive Fuel Surcharge from 2010 and 2011 for Electricity Generation, 2012–18...... 40 Figure 4.8 Crude Oil Movements, 2010 to Present (US$/barrel) ......................................................... 41 Figure 5.1 Household Expenditures on Subsidized Food Items and Household Receipts of Subsidies as a Share (%) of Their Budgets (by per Capita Expenditure Decile) ........................................................ 52 Figure 6.1 Water Subsidies and Transfers to Households (%)............................................................. 56 iii Acknowledgments This report was prepared by a team composed of Chadi Bou Habib (MNSED, TTL), Rome Chavapricha (MNSEG), Caroline Van Den Berg (MNSWA), Ndiame Diop, Eric Le Borgne, Wael Mansour, and Umar Serajuddin (MNSED). The team benefited from the technical guidance of Paolo Verme (MNSED) and from the helpful comments of Herwig Immervoll, (ECSH3), Jennifer Keller (PRMED), Naoko Kojo (PRMED), and Mario Mansour (IMF). The team also benefited from the helpful collaboration of Kristina Kostial, Dmitriy Rozhkov, Yasser Abdih, Andrea Gamba, Hui Jin, and Nkunde Mwase––all of the International Monetary Fund. The team acknowledges the collaboration and contributions from the Ministry of Finance in Jordan. We would like to thank in particular H.E. Secretary General Dr. Omar al Zoubi for his interest and precious and extensive comments and insights. The team also is grateful to the staff of the Ministry of Finance, who provided valuable data and insights. The contributions of the various persons met in Amman during the preparation of this study also are acknowledged. Finally, the team would like to extend their thanks to Ferid Belhaj (Country Director), Hedi Larbi (Country Director) and Bernard Funck (Sector Manager) for their strategic guidance and support throughout the preparation of this report. Alicia Hetzner who edited the report, and Sariette Jippe and Muna Abeid Salim provided excellent administrative support. iv 1. Introduction 1. Market-oriented reforms and an exceptionally favorable external environment in 2000-2008 propelled Jordan‘s economic growth to 6.7 percent over the last decade. This performance was better than MENA‘s average, which stood at 4.5 percent in that period. Jordan has consistently outperformed MENA in terms of GDP growth since the late 1990s in spite of high oil prices, which increased growth in the resource-rich countries. The only exception was during the recent global financial crisis in 2009-2010 when growth fell more sharply in Jordan, stopping the accelerated growth spells observed since 2004. This sharp decline in growth since 2009 is a sign of Jordan‘s weak growth sustainability. During this period, Jordan‘s fiscal situation worsened due to both structural and cyclical factors and fiscal policy proved unable to protect the country from large shocks. 2. Fiscal policy has been largely pro-cyclical, a pattern that has made it a major source of macroeconomic instability. For instance, while GDP growth averaged 8.1 percent in 2004-2008, the primary fiscal deficit excluding grants stood at 6.6 percent of GDP and the overall deficit excluding grants averaged 9.3 percent. Following commendable efforts to reduce public debt in 2000-2008, the latter is on the rise again (71 percent of GDP in 2011 in gross terms, against 67 percent in 2010). Because Jordan‘s capital account is open and the country is exposed to shocks, prudent macroeconomic management is crucial to reduce volatility. 3. The recent deterioration in Jordan’s public finances requires immediate fiscal adjustments to avoid a financial crisis. In 2011 Jordan’s budget deficit reached 12.7 percent of GDP, and the overall public sector deficit reached 18.6 percent of GDP. The latter included the National Electricity Production Company’s (NEPCO’s) deficit of 4.9 percent of GDP, and other autonomous agencies’ deficits totaling 1.0 percent of GDP. NEPCO’s deficit is related to the collapse in supply of Egyptian gas and the ensuing rise in the costs of electricity generation. The end-user electricity tariffs did not fully reflect the cost of back-up fuels. Hence, in 2011 NEPCO’s selling tariff to consumers was less than half the actual average buying tariff NEPCO paid to generators. Projections for 2012 set the budget deficit at 14.1 percent of GDP, and the overall public sector deficit at 21.1 percent of GDP. 4. To some extent, the current fiscal predicament is the consequence of Jordan’s policy response to the exogenous price shocks and economic swings of the last decade. The country first responded to the food and fuel price shocks of 2003–07 by increases in subsidies, made possible due to substantial revenue increases in 2004–07 fueled by strong economic growth. In 2008, these subsidies largely were eliminated. They were replaced, however, by salary increases for public and private sector employees and military personnel, both active and retired, and by a number of tax exemptions on consumption goods. 5. Looking at Expenditures, these are mostly non-discretionary, leaving limited room for adjustment aside from subsidies and capital expenditures. Looking at the economic classification of expenditures, we find that Military Expenditures, Compensation of Employees, Pensions, and Debt Service absorbed altogether 61 percent of total spending on average in 2009-2011, against 58 percent in 2006-2008. Military Expenditures, the most dynamic element, rose from 20 percent of total spending in 1 2006 to 26 percent in 2011. Debt service and pensions for accumulated rights are of a contractual nature and cannot be changed easily. 1 Similarly, changing salaries and compensations is a difficult process to implement due to the political cost it entails, and cannot be enforced beyond some levels and thresholds because of the painful welfare loss it generates. Capital spending and subsidies, which are categories of spending that can reversed in more ease, amounted together to 25 percent of total spending in 2009-2011 against 28 percent in 2006-2008. The share of capital spending declined to 16 percent of the total in 2011 from 21 percent in 2006 while the subsidy share rose. 6. The main source of fiscal trouble came from the revenue side. Indeed, while expenditures to GDP declined by 4.2 percentage points between 2007 and 2011, domestic revenues declined by 9.4 percentage points. As a fiscal stimulus in response to the global crisis, Jordan resorted in 2009 to large- scale tax exemptions and cuts and boosted capital spending. The revenue gap arising from the sales tax and customs exemptions alone is estimated at JD500 million. In addition, tax arrears have grown considerably, reaching JD1.7 billion in 2011. Forgone revenues due to the generous and complex investment incentive system were estimated at JD910 million in 2008-09. Finally, the tax reform of 2009- 10 resulted in additional income tax reductions. Although there might be redundancies between these arrears, revenue gaps, and forgone revenues, they altogether amount to JD3.1 billion, or 15 percent of GDP. 7. The Government has already begun to react by cutting both budgetary and overall public sector deficits. Measures adopted for 2012 are expected to cut these deficits by 3.3 percentage points to 10.7 percent of GDP and 17.7 percent of GDP, respectively, or JD2.4 billion and JD3.9 billion. These include revenue measures of 0.4 percent of GDP (increase in residency and money transfers fees; increase in taxes on cars; introduction of sales taxes on mobile phones and air conditioners; removal of tax exemption on agricultural inputs, increase in taxes on air tickets; tobacco and alcoholic products) and spending measures of 2.9 percent of GDP (e.g., cuts on capital spending; increase in fuel prices; and cuts on military and other current expenditures). 8. More is needed if the country is to avoid the unsustainable debt dynamics of the early 1990s when the debt-to-GDP ratio reached 220 percent, forcing Jordan to undertake painful adjustment. Jordan did not recover its 1988 real GDP per capita until 2006. 2 Even if all the measures committed by the 1F government were successfully implemented, the primary fiscal deficit of the central government would remain at a high 8.2 percent of GDP in 2012. This is significantly higher than the 3 percent of GDP in primary deficit that is needed to stabilize Jordan’s debt dynamics (Appendix 1). Adding NEPCO’s and other autonomous agencies’ deficits, a primary surplus of 4 percent of GDP at end-2012 would be needed, despite the recent increase in electricity tariffs. 9. Chapter 2 examines the revenue side. Much of the current fiscal stress is due to the decline in revenue mobilization in Jordan. Options to reverse the revenue trend with immediate actions on 1 Reform can be more easily implemented for future pensions with a modification of parameters defining the rights of future retirees. Hence, pensions can be more easily used for longer term consolidation rather than element of an immediate fiscal adjustment package. 2 World Bank, “Hashemite Kingdom of Jordan Development Policy Review: Improving Institutions, Fiscal Policies and Structural Reforms for Greater Growth Resilience and Sustained Job Creation.� Report No 70706-JO, Washington, DC, June 2012. 2 exemptions and tax arrears are the most equitable measures to enact. The chapter shows that, in addition to the 0.4 percent savings from governmental measures, space exists to collect 2.2 percent of GDP in 2012 on an annual basis and 2.1 percent in 2013, by stopping the accumulation of arrears, recovering old arrears, and removing all the exemptions granted in 2008 for general sales tax (GST) and customs. 10. Chapter 3 presents options available on the spending side, including a more proactive freeze on wages, allowances, and pensions in addition to additional prioritization of capital spending. Indeed, wages, and especially allowances, have increased rapidly in 2011. For the purpose of immediate fiscal adjustment, the government may need to examine options to freeze the wage bill; bring many allowances to their 2010 nominal levels and maintain some others at their 2011 levels; and remove bonuses. A less desirable option could be to delay or freeze pensions. Also, some room is left for additional prioritization of capital spending. In sum, annual savings from these measures would reach 1.2 percent of GDP in 2012 and 2.09 percent of GDP in 2013. 11. Chapter 4 considers measures to stop the losses in the power sectors and to recover their past accumulated losses. If the fuel subsidy for power generation were eliminated, this measure would generate annual savings of 1.0 percent of GDP in 2012 (assuming an average oil price of US$101.8 per barrel).3 In parallel, the Government would have to pay NEPCO an additional 2.7 percent of GDP in 2012 and 2.6 percent of GDP in 2013. These payments would cover (a) cost recovery for the public sectors’ power consumption, that is, 17.6 percent of overall electricity consumption; (b) upfront compensation for the share of public sector consumption in NEPCO’s losses in previous years; and (c) a subsidy after tariff reform to cover the residual deficit of NEPCO related to preferential tariffs for households (HH) in the lower consumption brackets. 12. Chapter 5 outlines several scenarios of power tariff reforms with a focus on households. Chapter 5 also shows that 75 percent of food subsidies could be saved, and the remaining 25 percent targeted more efficiently to the poor. These savings would amount to an annual 0.6 percent of GDP and 0.7 percent of GDP in 2012 and 2013, respectively. 13. Chapter 6 examines possible scenarios to reform the water sector and water tariffs. Annual savings from these measures could reach 0.2 percent of GDP in 2012 and 0.18 percent in 2013. 14. The objective of the above measures is both to curb public sector borrowing requirement and to bring down the primary fiscal deficits to a debt-stabilizing level. Table 1.1 summarizes the additional measures that can be initiated in the short term and their potential impacts. For 2012, Table 1.1 shows the results both annually and for the last quarter of the year. Thus, even if the measures were implemented only in the last quarter of the year, they could still save up to JD300 million in overall deficits (1.4 percent of GDP). With additional measures of the type described below, the debt stabilizing primary deficit could 3 The assumptions on oil price are taken from “Jordan – Request for a Stand-By Arrangement�, International Monetary Fund, July 2012, available on the website of the Ministry of Finance http://www.mof.gov.jo/admin/Upload/Optional/Bundle%20Version%20Request%20for%20a%20Stand- By%20Arrangement%20-%201%20-%20DMSDR1S.PDF. 3 be reached in 2013.4 Foreign grants may help too but should be treated as windfall that, in case it occurs, would provide the Government with additional fiscal space without changing the course of reform. 15. Fiscal adjustment does not need to be regressive, to the contrary. The tax burden has been declining since 2007 due to a mix of loose tax administration, generous exemptions, and policy reform choices. The well-established and well-connected private sector is most likely to be the part of the society that benefits from this declining revenue pressure. Expenditures remain high partly due to the system of expensive fuel, food, and power subsidies, all largely regressive. Another reason behind high expenditures is the constant demand for public employment and public funds to compensate for the lack of private sector job opportunities. In the meantime, higher deficits translate directly into high demand for domestic financing by the public sector. In this constrained financial environment, those who are the most likely to be crowded out from access to credit are young entrepreneurs and potential newcomers. Hence, by loosening fiscal pressure and granting generous tax exemptions to well-established private sector and foreign investors, Jordan’s current system ultimately is crowding out another part of the domestic private sector. In this context, fiscal reform could potentially help private sector growth while putting into contribution those segments of the society that largely benefitted from the declining revenue pressure. 4 Table 1.1 does not take into account the removal of fuel subsidies and the introduction of the cash compensation scheme decided on November 13, 2012. Indeed, preliminary estimations show that this major move would generate additional savings of about 0.12 percent on the overall public sector deficit for the last month and a half of 2012, and of about 1 percent of GDP for the year 2013, which would be a substantial contribution to overall fiscal adjustment. 4 Table 1.1 Summary of Additional Measures and Their Impacts on the Budget, 2012 and 2013 (% of GDP) 2012 2012 5 2013 % of GDP Annual Last Annual basis quarter basis Additional freeze on hiring, salaries, and allowances 0.78 0.20 1.20 Additional savings from pensions 0.62 0.15 0.80 Additional savings from capital spending 0.00 0.00 0.27 Cutting food subsidies by 75% 0.62 0.16 0.70 Adjusting water tariffs 0.20 0.05 0.18 Removal of fuel subsidy to power sector 1.00 0.25 0.00 Total additional savings from spending 3.22 0.80 3.16 Additional revenues due to zero accumulation of arrears in 0.90 0.90 0.83 corresponding year Additional revenues from recovery of stock of arrears over 8 years 0.96 0.96 0.89 Removal of 2008 tax exemptions 0.37 0.09 0.34 Total additional revenues 2.23 1.95 2.06 Total additional fiscal savings 5.44 2.75 5.22 Budget balance before additional measures -10.72 -10.72 -9.34 Budget balance after additional measures -5.28 -7.97 -4.11 Subsidizing the residual of NEPCO after tariff reform -0.51 -0.13 -0.48 Government to pay power bill at cost recovery -1.25 -0.31 -1.22 Paying Government’s share of NEPCO's accumulated losses -0.95 -0.95 0.00 Budget balance after additional measures and settlements with -7.99 -9.36 -5.81 NEPCO Debt service 2.57 2.57 3.05 Primary budget balance after additional measures and -5.42 -6.79 -2.76 settlement with NEPCO Debt stabilizing primary balance -3.00 -3.00 -2.78 Gap between debt stabilizing primary balance and new primary -2.41 -3.78 0.02 balance JD mil Budget balance without additional measures -2,384 -2,384 -2,241 Additional fiscal savings 1,210 612 1,254 Payments to NEPCO -603 -309 -407 Remaining financing gap -1,777 -2,081 -1,395 Projected grants 934 934 921 Financing from domestic or foreign markets -843 -1,147 -474 Source: Ministry of Finance (MoF); IMF-SBA documents; World Bank staff scenarios and calculations. 5 Assuming that the GoJ starts implementing fiscal adjustment measures before the end of the last quarter of 2012. 5 2. Reversing the Decline in Domestic Revenues 16. The steady decline of the share of Jordan’s domestic revenues to GDP has contributed strongly to the country’s financial strains. Since 2007, Jordan’s domestic revenues have shrank by 9.4 percentage points of GDP. Given the build-up of fiscal deficits, it is necessary to reform the revenue system both to meet the immediate objectives of fiscal adjustment and to restore longer term fiscal sustainability. 17. Immediate results can be obtained by reducing certain tax expenditures and sales tax exemptions and by selectively increasing taxes and excises on specific commodities. According to the Government’s estimates, such measures can bring up to 0.4 percent of GDP in additional annual revenues for 2012 and 0.6 percent for 2013. Additional revenue could be generated. Indeed, Jordan must deal with a stock of tax arrears that by end-2011 was estimated at JD1.7 billion, or 8.3 percent of GDP. In addition, the revenue gap from customs and sales tax is estimated at JD500 million (2.4 percent of 2011 GDP), including the exemption packages granted in 2008 and 2009. Immediate fiscal adjustment in Jordan can be reinforced by JD495 million (2.2 percent of GDP) on an annual basis in 2012 if two strong, swift measures are taken. The first is to stop the accumulation of tax arrears and to gradually recover the outstanding stock. The second measure is to totally remove the 2008 package of tax exemptions. Many of these measures still can be implemented in the last quarter of 2012 and still generate JD433 million or1.9 percent of GDP in 2012. 18. In the medium term, Jordan should look into the cost of the investment promotion packages, estimated at JD910 million in 2008 (4.4 percent of 2011 GDP), and also revisit income tax rates and exemption thresholds. Finally, initiatives with longer term impacts on the tax base and on the sufficiency and equity of the tax system could be considered immediately. Such initiatives include introducing capital gains taxes on the gains realized in the real estate and capital markets, increasing the existing property tax, or increasing the tax on interest income. 2.1. Understanding Jordan’s Decline in Revenues to GDP from 2007 to 2011 19. Jordan’s domestic revenues have declined steadily as share of GDP (Figure 2.1). During 2009– 11, domestic revenues declined from an average of 29.2 percent of GDP in 2006–08 to an average of 22.6 percent. The minimum was 20.5 percent in 2011; and the peak, 29.9 percent in 2007. The strongest decline occurred in the taxes related to the real estate sector (Real estate tax and registration fees). These taxes declined from 2.4 percent in 2006–08 to 1.2 percent of GDP in 2009–11. Over the same periods, the share of sales tax declined from 11.5 percent of GDP to an average of 10.1 percent of GDP, and the share of income tax declined from 3.9 percent from 3.7 percent of GDP. The downward trend in domestic revenues shows that, in the past few years, Jordan's policy choices wasted considerable public sector resources. These policies also enabled loosening tax pressures to the benefit of well-established groups, individuals, and sectors. Finally, these policies increased the pressure to finance deficits, which had dried out precious financial resources for other groups of potential entrepreneurs. 6 Figure 2.1 Domestic Revenues Trend, 2000–11 (%) 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0 -5.0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Share of GDP (%) Growth (%) Source: MoF; World Bank staff calculations 20. The decline in revenues to GDP is due to a combination of cyclical and structural factors. The origins of this decline can be traced back to five major developments: (i) slowdown in economic activity since the end of 2008, (ii) tax cuts initiated in 2008 and expanded in 2009 as part of an overall fiscal stimulus package, (iii) backload of tax exemptions for promoting investment, (iv) 2009–10 tax reform that reduced tax rates and increased exemption thresholds, and (v) backload of tax arrears related to flaws in tax administration and tax compliance. While (i) can be related to the cyclical decline in economic activity due to the global economic downturn, (ii), (iii), (iv), and (v) reflect structural policy choices that have major economic and fiscal implications. While separating the roles of different factors in reducing the share of revenues to GDP is not an easy task, overall revenue trends provide interesting hints. 21. The numerous and diverse tax cuts and exemptions in Jordan have been costly because they were in place for a considerable time. The most popular and recurrent exemptions can be grouped in three main categories. x The first category exempts general sales tax (GST) or customs on consumption goods and services, commodities, and intermediate goods for industries and manufacturing. These exemptions take the form of a reduction of the tax rate or the application of a zero rate. x The second category focuses on corporate tax and customs exemptions for private firms and typically benefits foreign direct investment (FDI), especially for real estate and infrastructure projects. Such exemptions typically are granted by special legislation or through the investment promotion law. x The special economic zones (SEZs) and the legislation supporting them offer tax exemptions and reduce other fees, especially those related to labor. 7 x Reforms in the latter two categories are structural and can be considered medium-term objectives, especially since they require legislative changes. Short-term actions are possible only on exemptions of goods and services. 22. The 2008 tax cuts have reinforced the impact of the economic slowdown on the buoyancy of revenues. These tax cuts (first category described in paragraph 21 above) were intended to stimulate demand and reduce domestic prices. The 2008 tax cuts have reduced revenues by an estimated JD103 million on an annual basis, or 0.7 percent of 2008 GDP (Table 2.1). These cuts included the removal of customs duties and GST for 128 items. Among these items were (a) GST exemptions on 13 basic food items, (b) zero rate for 15 food items, (c) removal of GST for non-tourist restaurants, and (d) temporary exemption from GST for retailers with turnover of less than JD100,000 per year. To attenuate the transmission of the rise in oil prices to cement prices, a special tax exemption was granted for cement. Other exemptions were granted for energy-saving equipment and for agriculture and livestock inputs. The GST on internet services and steel was decreased. World Bank staff estimated that, in a 3–4 year horizon, these exemptions would generate additional financing needs that would raise the debt to GDP ratio by an additional 2 percentage points.6 Furthermore, these tax subsidies were regressive because non-poor households received approximately 2.5 times the amount of subsidies received by poor households. 7 23. Following the global crisis of 2008, tax exemptions were used as a tool of fiscal stimulus, and the list of exemptions was expanded. With the outbreak of the summer 2008 crisis, the Government decided to exempt additional goods. The use of tax exemptions as a key element of the fiscal stimulus package meant that at least 200 items were added to the list of exempted goods, contributing to the slowdown in GST revenues. The revenue gap from GST and customs is estimated at JD500 million (2.4 percent of GDP). Over the same period, customs represented 9.6 percent of tax revenues, but due to Jordan’s gradual implementation of the WTO agreements, they were declining steadily. Overall, tax proceeds dropped from 20.4 percent of GDP in 2007 to 14.9 percent of GDP in 2010. Figure 2.2 illustrates this drop and accentuates the contribution of GST and customs to this decline. The real estate sector later benefited from tax exemptions in 2010 that contributed to reducing revenues from the sector by 5.5 percent in 2011. 6 World Bank, “Impact of Oil and Food Shocks on Fiscal Budget� (Washington, DC, 2009, unpublished). 7 World Bank, Jordan Subsidies Note, 2011 includes a comprehensive list of 2008 exemptions. 8 Table 2.1 Annual Cost of 2008 Tax Exemptions and Reduction Package Tax Total Amount revenue revenue GDP Tax exemptions and reductions (JD mil) (%) (%) (%) GST and customs exemption on 13 basic items and zero rate of 15 items 17.3 0.60 0.40 0.11 GST exemption for non-tourist restaurants 5.1 0.18 0.12 0.03 Special tax exemption on cement 5.1 0.18 0.12 0.03 Temporary tax exemption on retailers with turnover < JD100,000 2.0 0.07 0.05 0.01 Customs tax exemption for energy saving equipment –13 items 6.3 0.23 0.14 0.04 GST exemption for agriculture and livestock inputs = 35 items. 20.3 0.74 0.46 0.13 o/w previously taxed at 4% = 30 items 17.6 0.64 0.40 0.11 o/w previously taxed at 16% = 5 items 2.6 0.09 0.06 0.02 GST zero rate for energy-saving equipment = 48 items 8.0 0.29 0.18 0.05 GST decrease for internet services = 1 item (16% to 8%) 2.5 0.09 0.06 0.02 GST decrease for steel products =1 item (16% to 8%) and exemption for 16.0 0.58 0.37 0.10 other steel products = 2 items Total 102.6 3.72 2.34 0.66 Source: MoF; World Bank staff calculations. Figure 2.2 Trends in Jordan’s Tax Revenues: Share of GDP, 2007–11 (%) Source: MoF; World Bank staff calculations. 24. Jordan is also forgoing large amounts of revenues due to both investment promotion initiatives and inadequate tax administration and enforcement. Jordan has implemented a complex system of tax exemptions of products and sectors related to investment promotion. This system has significantly reduced the visibility, efficiency, and effectiveness of the tax system. Total forgone revenues related to investment promotion packages or tailored legislation tripled between 2005 and 2008, the year in which 9 they are believed to have reached their maximum. According to MoF estimates,8 in 2008 these exemptions and reductions reached over JD910 million. This amount was equivalent to 21 percent of total domestic revenues collected that year and to 4.4 percent of the GDP of 2011. Moreover, inadequate tax administration and loose enforcement of collection resulted in substantial accumulation of arrears, estimated to have reached JD1.7 billion (8.3 percent of GDP) at end-2011. 9 25. The 2010 tax reform further weakened revenue buoyancy. The tax reform simplified the tax code and removed many specific rates and treatments to groups, sectors, or companies that had been burdening tax administration. However, in practice, the high exemption thresholds on individual incomes have exempted over 85 percent of Jordanian households from taxation. Moreover, the highest bracket of individual income is now taxed at 14 percent, which is low even by Middle East and North Africa’s standards. As a consequence, income tax collected from individuals decreased by 15 percent in 2010, by 3 percent in 2011, and by 12 percent in the first 4 months of 2012 compared to the same period in 2011. Table 2.2 compares the structure of tax rates prior to the 2010 reform to the new structure that resulted from the 2010 reform. Table 2.2 Jordan’s Income Tax Rates before and after 2010 Reform Taxes/Exemptions Before 2010 reform After 2010 reform Individual Income Tax Rate (progressive, upper 30% 14% boundary) Exemption Threshold for Individuals (JD, annual) 12,000 12,000 Exemption Threshold for Household (JD, annual) 12,000 24,000 Income Tax for Banks and Financial Institutions 35% 30% Income Tax on Exchange, Brokerage, Insurance, and 35% 24% Leasing Companies Income Tax on Telecommunication Companies 25% 24% Income Tax for Retail and Services Firms 25% 14% Income Tax Rate for Other Businesses 15% 14% Source: World Bank staff, based on Jordanian tax legislation. Note: Prior to reform, an additional 3% in earmarked taxes was added to all corporate and business rates. 2.2. Improving the Revenue System 26. In 2010 the Government of Jordan began to take measures to improve the revenue system. It became increasingly obvious to the authorities that relying on foreign grants to finance expenditures without addressing the serious flaws in the revenue mobilization system is not sustainable. It also became clear that the preferential tax policies distorted relative prices, harmed equity, and had limited effectiveness to offset weak demand (tax cuts, in general, have a significantly lower fiscal multiplier than expenditure measures). The 2010 tax reform began to address some distortions related to the generous exemptions system. Based on the new law, in 2010 the cabinet issued a decision to eliminate exemptions that had no specific duration or amount and had been issued by previous cabinets. This decision does not 8 Communication of His Excellency Dr. Bassem El Salem, Minister of Finance, with the Prime Minister on July 13, 2009, No. 12/1/27/16062. 9 Findings from July 2012 IMF mission. Figure is available in IMF’s Memorandum of Economic and Financial Policies. 10 apply to exemptions established by legislation or international agreements. All exemptions decided by the cabinet and benefiting a specific agent or group of agents were eliminated, generating JD50 million in additional revenues in 2010. 27. The deterioration of the fiscal situation in 2011 pushed the Government to remove more tax exemptions. The Government decided to end the tax cut on real estate transactions starting in January 2012. As a consequence, revenues from the registration of real estate transactions rose by 29 percent in the first 4 months of 2012 compared to the same period of 2011. On an annual basis, this rate corresponds to an additional JD70 million (0.3 percent of GDP) collected from the real estate sector. 28. The Government is pursuing a set of additional measures expected to generate 0.4 percent of GDP in 2012 and 0.6 percent in 2013 (Table 2.3). A MoF technical committee has been investigating the list of remaining exempted goods and services. The committee’s main task is to examine the social and economic justifications for the exemptions. The long list includes the 128 items totally or partially exempted or zero-rated in 2008, plus the 200 items exempted in 2009, plus over 110 items that had benefited from special treatment prior to 2008, for a total of 438. Among these, the removal of the 2008 exemption on some agriculture inputs is a direct consequence of the investigation conducted by the technical committee. The investigation found that mining companies that use the same inputs benefit greatly from this exemption. The increased excise on alcohol and tobacco and raising the tax on cars from 80 percent to 90 percent fulfill the objectives of sufficiency and, to some extent, of equity. 29. However, not all of these measures meet the standards of sufficiency, equity and neutrality. For instance, the increases in residency fees for non-Jordanians and on money transfers may be inherently inequitable since they de facto increase the burden on low-income, unskilled foreign labor. The increase of sales tax on mobile phones and air conditioners could be distortive for these two promising sectors in Jordan. The Government rather might consider removing exemptions, or increasing rates, on luxury goods. These options are being considered but have been implemented only partially. One last measure that has been debated but does not seem to have materialized is the increase in mining exploration fees by US$2 per ton, which would bring JD60 million of additional revenues (0.27 percent of GDP) in 2012 (MoF estimate). Table 2.3 Savings from Planned Revenue Measures (percent of GDP) Measures 2012 2013 Increase Resident Fees for Non-Jordanians 0.03 0.04 Increase in Money Transfers Fees 0.05 0.08 Increase Tax on Cars 0.09 0.13 Introduce Sales on Mobile Phones and Air-Conditions 0.04 0.06 Remove Exemptions on Agricultural Usage Items 0.07 0.11 Increase Tax on Air Tickets 0.09 0.13 Increase Tax on Alcoholic and Tobacco Products 0.05 0.07 Total 0.42 0.62 Source: IMF, Stand-by Arrangement Package, based on information provided by GoJ, July 2012 11 30. The design of some of the recent measures could also be improved. Some measures listed in Table 2.3 could have been bolder, especially the increase of taxes on cars, on alcohol and tobacco, and, to a lesser extent, on air tickets. In fact, the social or developmental impacts, and even the economic relevance, of many exemptions are questionable. In any case, indirect tax incentives have proved very prone to abuse. 10 In many cases, targeted transfers and subsidies can be more efficient in fulfilling the economic or developmental objective set for tax exemptions. Moreover, the regressive nature of tax exemptions has been proved even for the 13 basic food items, and World Bank staff found that less than 10 percent of the exemption amount actually benefits the poor.11 31. Furthermore new exemptions are still being granted, despite the mounting fiscal stress. In the summer of 2012, tax exemptions were extended to certain categories of cheese and steel. Also, the 2-year tax credit on GST still persists. As of July 2012, sixty-six companies benefited from this generous credit with zero nominal interest rate and negative real interest rate. Despite the low tax burden on tobacco, high consumption of cigarettes in Jordan, and high levels of fiscal deficit, further increase of excise on tobacco and cigarettes are not currently being considered. 32. For the longer term, the Government is planning to address the large tax arrears and the forgone revenues from investment promotion, and to reform the income tax law. From 2008 to 2012, tax arrears have more than doubled: going from JD800 million to JD1.7 billion. The Government of Jordan is willing to engage in a comprehensive tax administration reform to provide the resources and to implement the processes to recover these large arrears. 33. For the longer term, and regarding forgone revenues related to investment promotion, any modification of an exemption granted under a specific law would require legal amendment. Nevertheless, the Government has taken several steps to bring these type of exemptions under control: (a) all exemptions granted under Cabinet decisions are being revised, (b) the authorities are making sure that beneficiaries whose exemptions have expired are complying with the tax law; (c) tax exemptions no longer can be granted to investors by specific laws or decisions; and (d) exemptions can be granted only when based on the texts of the income tax law and on tax laws in general. 2.3 Options to Increase Revenues 34. In the short term the Government of Jordan may consider the following actions: a. Stop immediately the accumulation of tax arrears, which seems to have occurred at an average annual rate of JD200 million since 2008. GoJ may mobilize all human and institutional resources to enforce this collection at any time of the year. Such a move would reduce the deficit by an additional 0.9 percent of GDP in 2012, b. Recover the accumulated stock of arrears. One possible scenario would be to recover these arrears over a period of eight years, similar to the one used in the lengthiest scenario for 10 H.H. Zee, J.G. Stotsky, and E. Let, “Tax Incentives for Business Investments: A Primer for Policy Makers in Developing Countries,� World Development 30 (9): 1497–516, September 2002. 11 World Bank, Jordan Subsidies Note, 2011. 12 recovering NEPCO’s accumulated losses. A swift move on recovering accumulated arrears would bring JD212 million per year, or an estimated 0.96 percent of GDP in 2012, c. Stop all 2008 exemptions, which would generate an additional saving of JD82 million or 0.4 percent of GDP on annual basis in 2012 (JD20 million and 0.1 percent of GDP for the last quarter of the year), d. Stop granting any new tax exemptions of any nature in the future, e. Suspend all types of tax credits. 35. In the longer term, and in addition to measures to restore the sufficiency and efficiency of the existing system, Jordan has space both to include new categories of revenues and taxpayers in the tax; and to increase the taxation of other categories of revenue. In fact, it worth noting that: x In a middle-income economy such as Jordan, capital gains, property, and interest incomes are relatively highly concentrated. x In Jordan, corporate income, wages, and consumption are the main sources of tax revenue. Thus, targeting capital gains, interest income, and real estate wealth would enhance the feeling of equity, especially since these categories are practically exempt from taxation. x Economically speaking, capital gains taxation and interest income taxation target income categories so would have the same impact as any income taxation. Beyond equity considerations, taxing property is a way for GoJ both to capture part of the appreciation in property value that results from its policies and to have a policy tool in a market in which supply is largely inelastic. More generally, Jordan attracts large amounts of foreign wealth to its capital market, real estate, and banks deposits. For this reason, taxing capital gains, properties, and interest rates enables policymakers to redirect part of this wealth to benefit the resident population, especially the wage earners and small business owners. It is these groups who are hit by the inflationary impact of these foreign inflows in good years and suffer from the overall impact of a downturn in bad years. 13 3. Efficiency Gains and Savings by Large Categories of Spending 36. Following a steady decline since the mid-2000s, public spending bounced back strongly in 2011. The primary source of Jordan’s current fiscal predicament lies in the 9.5-percentage-point decline in domestic revenues between 2007 and 2011. A spending surge that year exacerbated the situation and expenditures increased from 30.4 percent of GDP in 2010 to 33.2 percent in 2011. With that, the budget deficit reached an unprecedented 12.7 percent of GDP. If no measures are taken, Jordan’s 2012 budget deficit figures are expected to remain in the same order of magnitude. 37. Current expenditure has accounted for most of the recent surge in spending. After being contained for two years, it jumped in 2011 by 2.8 percentage points of GDP. Within current spending, the substantial increases were in wages, salaries, pensions, compensations, and the military. After hitting 8.5 percent in 2009, capital spending declined to 5.1 percent of GDP in 2010 and remained contained in 2011. Figure 3.1 Budget Spending in Jordan as Share of GDP, 2000–11 (%) 45% 40% 35% 30% 25% 20% 15% 10% 5% 0% 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 Current Spending Capital Spending Source: Based on MoF monthly reports for the corresponding years, www.mof.gov.jo; World Bank staff calculations. 38. The government has begun to address the situation. The measures already decided and being implemented by the Government are expected to generate savings of 1.4 percent of GDP and 1.3 percent of GDP in 2012 and 2013, respectively. If fully implemented, the additional options presented in this chapter could generate additional savings of up to 1.6 percent of GDP in 2012 on a yearly basis (0.4 percent for the last quarter) and 2.5 percent in 2013. 14 39. This chapter examines a menu of short-term additional expenditure savings options. Section 3.1 proposes options for containing the wage bill as it has increased beyond inflation, productivity, and GDP growth in recent years. Resulting savings in the last quarter of 2012 could equal up to 0.2 percent of GDP. Section 3.2 examines savings options from existing capital spending. This section shows that there is little room to generate beyond the 0.7 percent of GDP savings expected by the Government for 2012. This option proposes to reallocate these savings toward public investment that would have high impact on future employment and economic activity growth. Section 3.3 highlights the small social spending in the Kingdom and the need to enhance these programs and improve the targeting mechanisms. Section 3.4 looks at the inflated public pension and points to potential savings from implementing the suggested reforms in the temporary social security law. Although pension freeze is legally difficult to implement, but such a painful measure can be part of last resorts move in case fiscal situation worsens further. Resulting savings over the coming two years could equal up to 0.62 percent of GDP in 2012 (0.15 percent for the last quarter). Spending on energy, electricity, and consumption subsidies are left for other chapters in the report. 3.1. Wage Bill 40. Despite times of low growth, the civil and military wage bill has increased rapidly in the last years. Expressed in constant terms, the overall wage bill 12 has increased at a higher rate than the real GDP growth of the Kingdom, especially in the recent years (see Figure 3.2 for historical trends). Figures from the Ministry of Finance (MoF) indicate that remunerations for civil servants and military personnel have grown at an average real rate of 7.1 percent since 2007, compared to 5.1 percent real growth in economic activity. The 2012 budget envisages that the wage bill would increase by 1.9 percent. Since 2009, the Government has adopted a formal policy of freezing public sector hiring and filling only vacant positions. However, the next paragraph shows that exemptions to the policy have reduced its effectiveness and encouraged the expansion of the wage bill. 41. The education and health sectors and the security apparatus have not been subjected to any hiring freeze. Vacancies are impeding good health service delivery due to the lack of specialists in public hospitals and clinics. However, such is not the case for the education sector, in which the exemption is used as a loophole. After being recruited to teach in the public schools, new teachers often shift to other posts in the civil service. Exemptions also are extended to the defense sector. Defense employees constitute 7.0 percent of the total labor force. Hence, Jordan stands as an outlier since this number is much higher than the 3.4 percent average for all MENA countries and several folds above the 0.7 percent average for global upper middle income countries (UMIC). 13 12 In this report, “wage bill� includes the wages, salaries, and allowances for civil servants in addition to military spending that show under current expenditures in the “Government’s Financial Bulletin,� http://www.mof.gov.jo/ar/Pages.php?page=subOpPage&id=83. It is very likely that much of this current military spending in fact is spending on wages and salaries. 13 World Development Indicators, 2012. Figures are for 2010. 15 Figure 3.2 Real Wage Bill and Historical Trends, 2000–13 (year-to-year change, %) 35 30 25 20 15 10 5 0 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 -5 Real GDP Inflation Total Wage Bill Civil Servants Wage Bill Source: Government Financial Bulletin, http://www.mof.gov.jo/ar/Pages.php?page=subOpPage&id=83; General Budget Directorate, www.gbd.gov.jo; Department of Statistics, www.dos.gov.jo; IMF-SBA; projections for 2012 and 2013; World Bank staff calculations. Note: 1. Projected figure for 2013 is based on the indicative numbers of the official Medium-Term Expenditure Framework (MTEF) highlighted in the 2012 Budget. 2. Budget revisions were announced by the Minister of Finance through several public statements in May 2012. However, they do not include the estimated wage bill. 42. In part as a result, military spending has now reached as much as 9.6 percent of GDP, a level unparalleled particularly among countries that are at peace with all their neighbors. In comparison upper middle income countries spend an average 2.0 percent of GDP on defense and security. Jordan even stands out among the high spending MENA countries where the average is 6 percent of GDP.14 Not extending the freeze to the defense sector thus will limit significantly the impact of the hiring freeze policy. The political sensitivity of the hiring freeze in the defense sector must be weighed against the high risks entailed by continuing high fiscal deficits year after year. This political sensitivity may not be linked to the defense function per se. Public sector employment, especially in the military, has been one of the important determinants of household welfare growth.15 In practice, governments have used public sector employment, especially in the military and security forces, as a tool for income redistribution. Under the current fiscal constraints, such a tool has reached its limits. 14 Figure for Jordan is for 2011 and is based on the “Government’s Financial Bulletin,� http://www.mof.gov.jo/ar/Pages.php?page=subOpPage&id=83. Figures for MENA and MIC averages are for 2010 and are taken from the World Development Indicators 2012. 15 W. Mansour, “The Patterns and Determinants of Household Welfare Growth in Jordan 2002-2010�, World Bank, Policy research Working Paper No 6249, Washington, DC, October 2012. 16 43. Numerous allowances are provided as part of the wages of civilian personnel in the Jordanian public sector. These allowances add up to more than 2.5 times the salaries bill. Allowances constitute a large share of Jordan’s civil servants’ wage bill 16 (Figure 3.3). These allowances are scattered among 10 categories, the largest being the cost of living expenses, additional allowances, basic allowances, and bonuses. On average, in 2010–11, these represented respectively 37 percent, 16 percent, 9 percent, and 3 percent of the total civil servants wage bill. In 2011 these allowances were increased by as much as 17.4 percent, whereas base salaries increased by only 7.3 percent. Figure 3.3 Civil Servants’ Wage Bill Composition, 2010–12 (%) 100% 90% 80% 70% 72% 74% 74% 60% 50% 40% 30% 20% 28% 26% 26% 10% 0% 2010 2011 Budget-2012 Basic Salaries Allowances Source: The 2012 Budget Law, www.gbd.gov.jo 44. Efforts have been made to curb the growth of allowances in the 2012 budget. If this target is kept, it is estimated that allowances17 will grow by 2.9 percent, compared to 3.6 percent for the civil servants’ base salaries bill. Despite this slowdown, additional measures are needed. Consolidating these allowances and reducing, or at least freezing them could generate needed savings. Freezing allowances at the nominal level of 2011 could generate up to JD21 million (approximately 2 percent of 2012 budgeted expenditures) on a yearly basis (JD5.3 million for the last quarter of the year). Additionally, freezing allowances at the 2010 levels and eliminating bonuses would generate up to JD156 million, equivalent to 0.7 percent of GDP on a yearly basis (0.18 percent for the last quarter of the year). 16 No detailed breakdown of the military wage bill is available. 17 According to the budget classification, 10 categories of allowances are identified. These are: personal living expenses, family living expenses, basic allowance, overtime, transport allowance, transport fees, fieldwork, bonuses, additional allowances, and other allowances. Unfortunately, no further breakdown of each category is available to the authors. 17 45. Looking at similar comparator countries, Jordan’s wage bill remains large (Figure 3.4). Furthermore, on average, wages are higher in the public sector than in the private sector.18 This is particularly the case at the lower end of the range. This imbalance reduces the incentive for joining the private sector, raises reservation wages, and puts additional pressure on public employment and, consequently, the wage bill. In addition, public sector employees receive benefits, such as health insurance and pension, which are much more generous than in the private sector. Figure 3.4 Compensation of Employees (% of total expenditures) 35.0 33.0 30.0 25.3 25.0 23.5 22.4 20.0 15.0 10.0 5.0 0.0 UMIC World Jordan MENA Source: World Development Indicators, 2012. Note: For Jordan: MoF Public Bulletin. All figures are for 2009. 46. Moreover, the increased public wage bill is not necessarily linked to increased productivity in the public sector. Using employment and value-added data from the Department of Statistics (DoS) and comparing the increase in public sector employment productivity to the growth in the public sector wage bills, results show a visible mismatch. While public sector productivity declined by an average of 0.6 percent during 2000–09, the real wage bill grew by an annual average of 5.2 percent during the same period. This data reinforces the argument to contain a wage bill that continuously adds to the fiscal burden yet is not justified by any productivity growth. 47. Recent reforms in public sector wages have been costly. In 2011 Government of Jordan (GoJ) launched a comprehensive reform program to tackle civil service employment. 19 This reform has not yet generated the required savings. As a result of this salary and pension alignment between civil servants and 18 According to the National Employment Strategy, conducted jointly by GoJ and the World Bank (2012), in 2009 the average monthly wage in the public sector was JD412, compared to JD338 in the private sector. 19 The reform objectives are four-fold (a) to consolidate the compensation schemes of independent institutions and defining unified governance and wage policy principles across these institutions; (b) to align civil service and own- budget agencies’ salaries to increase equity; (c) to increase productivity, re-examining terms of reference for all functions of civil service employment; and (d) to identify vacant positions that need to be filled through new recruitment or reallocation of staff, and vacant positions that need to be eliminated. 18 independent institutions, an additional JD83 million have been added to the wage allocations of the 2012 budget. An additional JD85 million and JD90 million are expected in 2013 and 2014, respectively.20 Savings are envisaged to begin in 2015, when public sector employment will be streamlined. However, there is a risk for these programs to add to the budget deficit in the short to medium term without being able to generate the consolidations and savings expected in the medium to long term. 48. Potential savings. In the short run, policy measures could focus on containing the wage bill within an overall effort of fiscal adjustment. Immediate short-term measures include: x Tighten the implementation of the hiring freeze for all public sector jobs, including education, health, and, more importantly, army and security forces x Cancel redundant vacant positions x Freeze the wage bills for 2012 and 2013 taking the nominal values of 2011 as benchmark. Table 3.1 shows the potential savings generated had the freeze accounted for the 2011 number of employees and the average wage of the employee. x Eliminate “Bonuses� and “Other allowances� (refer to footnote 10), budgeted in 2012 at JD54.9 million (0.25 percent of GDP). x Maintain the allowances labeled “Overtime� and “Fieldwork Allowances� at their 2011 nominal levels (JD16.9 million) since this is lower than 2010 (JD20.3 million). x Bring the remaining six allowances categories down to their 2010 nominal levels (528 million or 2.8 percent of GDP). These are budgeted in 2012 for JD643 million (2.9 percent of GDP). Savings generated for the above measures are summarized in Table 3.1. The Government already has committed to curb military recurrent spending, most likely a military wage bill, by 0.68 of GDP and 0.64 percent of GDP, respectively, in 2012 and 2013. The figures in Table 3.1 are inclusive of these measures. Moreover, figures for 2012 are on an annual basis. Except for the savings on the military wage bill, the additional measures proposed here are unlikely to be implemented before the last quarter of 2012. Therefore, to get a more accurate estimate of the outcome of these additional measures in 2012, the corresponding figures in Table 3.1 should be divided by 4. In the longer term, GoJ could focus on increasing public sector labor productivity and improving service delivery. Accomplishing these will require continuing the civil servant reforms including defining TORs and improving the competitive hiring process with the objective of attracting highly qualified and market-competitive staff. Needless to say that, total freeze on wages and hiring and reduction in allowances cannot be maintained beyond the objective of immediate fiscal adjustment over one to two years. In some sectors, a prolonged freeze would lead to a leakage or a shortage in highly qualified staff that may affect the quality of service delivery to the economy and population. Hence, salary freezes and cuts are among the measure that the Government may be compelled to implement if the fiscal indicators continue on their downward trend. 20 General Budget Department, Budget Law for 2012, GoJ, Amman, www.gbd.gov.jo 19 Table 3.1 Possible Savings from Freeze Measures, 2012 and 2013 (% GDP All public sector - incl. defense 2012 2013 Salary freeze 0.44% 0.56% Hiring freeze 0.32% 0.60% Combined effect 0.76% 1.16% Cuts in 2011 civil servant allowances 0.70% 0.65% Total effect 1.46% 1.81% Civil servants - basic salary item Salary freeze 0.01% 0.01% Hiring freeze 0.03% 0.05% Combined effect 0.04% 0.06% Civil servants - allowances item Allowances freeze 0.02% 0.10% Hiring freeze 0.07% 0.13% Combined effect 0.09% 0.23% Source: 2012 Budget Law (www.gbd.gov.jo); World Bank staff calculations. 3.2. Capital Expenditure 49. The current period of fiscal stress compels the Government to generate savings from all levels of public spending, including from capital expenditures. However, putting an excessive share of the adjustment on capital spending might have negative consequences on long-term growth. Indeed, capital spending has proven to be volatile in Jordan and very much dependent on external grants. The share of capital expenditure to GDP has fluctuated from an average high of 9.2 percent during 1990–95 to a low of 6.7 percent in during 2005–11. Investment spending typically is pro-cyclical; in other words, important cuts occur during periods of low economic growth. Rather than reduce recurrent spending and tackle structural rigidities, governments have resorted to decreasing capital spending as the primary policy tool for implementing fiscal consolidation programs. This again has been the case since 2009. By so doing, Jordan is (a) giving away an important policy instrument that could be used to counterbalance the slowdown in economic activity driven by any decline in external or domestic demand, and (b) jeopardizing long-term growth prospects by giving up on growth-enhancing investments while maintaining high levels of recurrent spending. 20 Figure 3.5 Overall PIMI: Decomposition by Subindex Source: Dabla-Norris and others (2011) (see footnote 11 for full reference). Note: PIMI = Public Investment Management Efficiency Index. 50. Inefficiencies in public investment in Jordan are relatively mild. Constructing a Public Investment Management Efficiency Index (PIMI) for 71 countries in the World,21 Dabla-Norris and others (2011) ranked Jordan among the relatively good performers, especially among middle-income countries (Figure 3.5). 22 However, additional improvement can be introduced to enable the Kingdom to reach the efficiency of economies such as Tunisia, South Africa, or Brazil. The World Bank DPR (2012) found that the positive impact of capital spending in Jordan seems to manifest itself with 4-quarter lags and reaches its peak only after 3 years. Such a lag in the output response to fiscal policy can be reduced with the implementation of reforms linked to project selection and budgeting (including quality of spending) and project evaluation. For Jordan, therefore, the question is not linked to the efficiency of public spending but is a question of quantity and quality. In the first instance, allocations of public investment have not been enough to sustain the high growth objectives of the Kingdom. On the contrary, in the past three years, allocations have decreased. In a second instance, the sectors that have benefitted from public investment lately are not necessarily sectors with high economic returns. These outcomes reveal the need to reprioritize the allocations 51. The quality of capital expenditure is not conducive to growth and employment. The distribution of capital expenditure by functional classification reveals concerns over the quality of such spending and its ability to support growth and employment creation, especially under the allocations set for the 2012 21 PIMI is composed of 4 components related to the project cycle: Appraisal, Selection, Implementation, and Evaluation. Jordan’s weakest components are evaluation and appraisal. 22 E. Dabla-Norris, J. Brumby, A. Kyobe, Z. Mills, and C. Papageorgiou, “Investing in Public Investment: An Index of Public Investment Efficiency.� IMF Working Paper, (International Monetary Fund, Washington, DC, 2011). 21 budget (Figure 3.6). The figures indicate a continued decline of spending on “Economic Affairs� from 2.8 percent of GDP in 2009 to 1.5 percent in the 2012 budget. “Economic Affairs� encompasses spending on the infrastructure of most backbone services and sectors that contribute significantly to GDP growth and employment in the Kingdom, 23 and therefore should not be reduced further. Conversely, more allocations have been directed toward security and social sectors that are not necessarily productive from an economic returns/growth perspective. In particular, Figure 3.6 highlights an increase in capital spending on defense, public order and safety, and housing. For fiscal consolidation purposes, cuts can be made in these items and the savings generated reallocated toward boosting spending on the “economic affairs� item. This report commends GoJ’s decision to reduce land acquisitions in the 2012 budget and to review the price mechanisms for land valuations. Figure 3.6 Functional Classification of Capital Expenditure, 2008–12 (%) Distribution Share of GDP 100% 9.0 8.5 7.4 6.6 10.2 8.4 90% 8.0 0.6 80% 14.5 16.0 15.3 12.6 15.7 7.0 1.4 70% 15.8 17.5 19.6 6.0 60% 19.0 13.4 0.5 5.0 1.6 50% 0.9 0.3 0.5 0.8 0.4 40% 4.0 1.0 0.8 38.2 36.4 33.4 0.6 32.8 35.4 0.9 0.7 30% 3.0 2.8 0.9 20% 2.0 2.3 6.7 9.6 1.9 1.9 1.5 7.0 9.7 8.1 10% 1.0 0.6 9.2 9.6 5.8 5.9 5.9 0.4 0.5 0.4 0.4 0% 0.6 0.8 0.3 0.3 0.3 2008 2009 2010 2011 Budget-2012 0.0 2008 2009 2010 2011 Budget-2012 General Public Service Defense General Public Service Defense Public Order and Safety Economic Affairs Public Order and Safety Economic Affairs Environmental Protection Housing and Community Amenities Environmental Protection Housing and Community Amenities Health Recreation, Culture and Religion Health Recreation, Culture and Religion Education Social Protection Education Social Protection Source: MoF and World Bank staff calculations. Note: General Public Services = Spending on research, foreign affairs, and other nonclassified items. 52. Potential savings. In the short run and under fiscal austerity conditions, Jordan has limited options to expand public investment. Therefore, the country is forced to improve the quality of capital spending. The Government of Jordan has made two decisions to achieve this objective: x Enforce the Government’s 2012 freeze on land acquisition and extend it to 2013. According to the Government, stopping acquisitions could generate savings of up to JD51 million (0. 23 percent of GDP) in 2012,24 and JD50 million (0.21 percent of GDP) in 2013. The fiscal stress in Jordan is believed to require a more radical freeze of land acquisition, which would bring an additional JD53 million of savings (or 0.22 percent of GDP) in 2013. 25 Savings would be 23 According to the budget classification in Jordan, the “Economic Affairs� header includes spending on Agriculture and Fishery, Mining, Manufacturing, Construction, Energy and Fuel, Telecommunications, Transport, Trade, and General Labor. 24 In May 2012, GoJ decided to stop acquisitions made by Ministry of Finance and Ministry of Transportation. Per MoF, these measures could be maintained in 2013. 25 Savings for 2012 and 2013 are calculated based on the 2012 Budget Law estimates for total land acquisitions. 22 less if projects under the “Economic Affairs� category in the functional classification of Public Expenditure are exempted, especially projects related to backbone infrastructure. These last have important impacts on economic growth. x Enforce Government’s freeze on projects under the Prime Minister and on Other Capital Expenditures. According to Government estimates, this freeze would generate JD104 million (0.47 percent of GDP). For 2013, one option would be to freeze spending on items not classified as “Economic Affairs� at the 2011 nominal level.26 This measure would generate savings of JD107 million (0.5 percent of GDP) in 2013. 27 Savings could be directed primarily toward reducing the overall fiscal deficit but also could be reallocated to projects under “Economic Affairs�––bringing the spending on the latter to 2.2 percent of GDP. 3.3. Social Assistance/Social Spending 53. Jordan has a comprehensive social assistance system that provides cash assistance to vulnerable groups in the Kingdom. 28 Jordan offers several social assistance programs whose aims are to alleviate poverty and help the most vulnerable groups in the society, especially women, children, the disabled, and poor families. The majority of these programs are administered by the Ministry of Social Development (MoSD), National Aid Fund (NAF), and Zakat Fund. Other smaller programs are run by the Royal Court and the Prime Minister’s Office. 54. Although these social programs remain small, they offer many services that reach across the whole Kingdom. Indeed, social assistance reached JD115 million dollars in 2011, 29 equivalent to 1.7 percent of total budget expenditure or 0.6 percent of GDP. A breakdown of these programs and their magnitude is depicted in Figure 3.7. The NAF program is the largest and represents over 76 percent of total social assistance spending. These programs offer mostly cash assistance as well as additional services such as healthcare, education, and rehabilitation support and in-kind transfers. They target specific groups that GoJ considers poor or vulnerable. 26 According to the current allocations under the 2012 budget, if such measure is adopted, the bulk of the cuts (JD51.4 million) would have to come from housing and community amenities projects. 27 It is assumed that the overall envelope for capital spending for 2013 increases by 12.8 percent, as reported in the indicative figures of the Medium-Term Expenditure Framework (MTEF) of the 2012 budget. 28 This section does not tackle large social spending items such as subsidies for food, fuel, or electricity. These will be addressed in different segments of the report. The primary interest here is the smaller social assistance programs devised to alleviate poverty and administered by specific units in the Government. 29 Budget Law for 2012 (www.gbd.gov.jo). This number does not include the Zakat Fund programs. The Zakat Fund is financed through the annual contribution (Zakat) to the poor imposed by Islamic law and tradition. It is managed by the Ministry of Religious Affairs. 23 Figure 3.7 Social Assistance Spending by Major Programs, 2010–13 (US$ mil) 180 160 8% 7% 9% 9% 140 18% 15% 18% 14% 120 100 80 60 77% 76% 76% 74% 40 20 0 2010 2011 Budget-2012 Proj-2013 NAF MoSD Others Source: Budget Law for 2012 (www.gbd.gov.jo); World Bank staff calculations. Note: MoSD spending excludes the NAF and wages and administrative costs. “Others� = the social programs in the Royal Court and in the Prime Minister’s Office. 55. Despite being small, the social assistance programs suffer from waste and inefficiencies, especially the NAF cash transfers. Despite reforms to increase the efficiency of the NAF, significant leakages persist, with only 15 percent of beneficiaries classified as poor according to the national poverty line. 30 The main driver of such waste is linked to the subjectivity of the eligibility testing for award of the assistance. This testing does not allow for impartial comparison of household welfare across differing geographical locations. Additionally, there is a lack of synergy with the DoS criteria for household expenditures that determine poverty level in the Kingdom. In effect, an assessment of families’ welfare may vary among field offices, or even among NAF officers handling applications. The lack of a live database of applicants and beneficiaries accentuates the problem. A World Bank assessment of the NAF system found that the 2005 NAF budget was more than double what was needed to eliminate chronic poverty but failed to do so due to the inefficient targeting mechanism adopted. 31 56. Options to be considered. The social assistance programs are very small to be squeezed further under an immediate fiscal adjustment plan and therefore any further reduction can jeopardize the services offered to these vulnerable groups of the society. 32 However, there are significant efficiency gains to be made through improving the targeting mechanisms as to reduce waste where benefits go to those that do not deserve it. Additionally, GoJ may consider using some of the generated savings from the wage bill, water, pensions and electricity to expand these social programs so that it enables their services to reach out to the largest pool of the poor and vulnerable. 30 World Bank, “Jordan Poverty Update,� Washington, DC, November 2009. 31 World Bank, “Jordan Poverty Update,� 2009. 32 The amounts allocated for these programs in the 2012 budget already are roughly similar in nominal terms to the 2011 allocations (Figure 3.8). 24 Short-term reform measures to improve efficiency can include: x Using better targeting mechanisms, especially for cash transfers. The proxy means-testing mechanism supported by the World Bank could be implemented in the short run, especially since it has been under preparation for several years and all of its elements have been completed (formula, questionnaire, and some piloting). x Eliminating cross-transfers among social programs and from social programs to other institutions. This ban will improve the transparency of the accounts and avoid the misuse of funds. 3.4. Public Pensions 57. Jordan’s social security system provides old age, disability, and survivors’ pensions that cover 36 percent of the labor force at a cost of 5.7 percent of GDP. 33 Figure 3.8 highlights some international comparisons. The conclusions of analyses conducted by the Jordanian Social Security Corporation (SSC) 34 and the World Bank 35 reaffirmed the urgent need to introduce reforms to make the pension system sustainable in the long term. In addition, there is a need to substantially increase the scope of the social security programs in the areas of maternity benefits, unemployment, and health insurance. 58. The SSC has prepared an ambitious reform proposal that addresses the most critical issues. These reforms are embedded in a new temporary Social Security law, approved by GoJ in March 2010. Major features include to (a) adjust pension benefit formulas and eligibility conditions to ensure financial sustainability and equity; mainly increasing the early retirement age from 45 to 55, stopping the early retirement option for new labor market entrants, and indexing pensions to inflation; (b) expand coverage by extending the mandate to contribute to firms of fewer than 5 employees; (c) introduce savings mechanisms to protect workers against income loss due to unemployment; (d) eliminate a significant disincentive to hire women by introducing maternity benefits that finance maternity leave, which currently are paid by employers; and (e) introduce social health insurance mechanisms. 33 World Bank, Recovery under Global Uncertainty Development Policy Loan, Program Document, Report No. 49971-JO (Washington, DC, September 2009). This DPL includes all types of pensions. The public pension was estimated at 4.1 percent of GDP in 2010. 34 O. Razzaz, “Social Protection: Lessons from Jordan� (SSC, GoJ, Amman 2010). 35 World Bank, “Country Partnership Strategy for the Hashemite Kingdom of Jordan for the Period FY12–FY15,� Report No. 58114-JO (Washington, DC, February 2012). 25 Figure 3.8 Public Pension Expenditures 2010 (% of GDP) 8 7.2 7 6.3 6.1 6 5.5 5 4.0 4.1 4 3.4 3 3 2.2 1.7 1.9 2 1.3 1.5 1.0 1 0.6 0.7 0 Source: IMF, “The Challenge of Pension Reform in Advanced and Emerging Economies� (Fiscal Affairs Department, Washington, DC, December 2011). 59. Amidst the comprehensive reform of Social Security, public sector pensions have been growing steadily over the past years (Figure 3.9). According to MoF’s recent figures, pensions and compensations increased by 15.8 percent in 2011 and reached JD862 million (equivalent to 12.7 percent of budgetary expenditures and 4.2 percent of GDP). Another increase of 15.9 percent is envisioned in the 2012 budget, raising the cost of the pensions and compensations to GDP by 0.3 percentage point by the end of the year (to 4.5 percent). These large increases are a direct result of (i) the subsequent salary increases, most of which occurred starting in 2008; (ii) the mergers of beneficiaries from the old pension system with the new one; and (iii) the latest public sector salary restructuring, which started in 2010. On a positive note, prior to the new legislation, pensions used to increase ad hoc with no clear guidance or parameters on the determinants of the increase, its timing, and its frequency. This lack of guidelines created high uncertainty for both pensioners and system administrators. This anxiety has lessened with the adoption of the new law in which the pension will be raised annually according to an inflation measure based on a price index of elderly consumers. To avoid perverse incentives, indexation will begin at the normal age of retirement. However, the current drift in public spending might require Jordan to postpone any measure that would additionally burden its already strained public finances. Hence, in case the fiscal situation continues to deteriorate, Jordan may need to give-up on indexation and on any annual increase in individual pension disbursements to beneficiaries. 60. Potential savings. In the short run, for 2012 and 2013, and in case of continuous deterioration of the fiscal situation, the Government could freeze any measure that could increase the salaries and packages for each pensioner beyond their 2011 levels. This freeze could generate up to 0.62 percent of GDP on yearly basis in 2012 (0.15 percent for the last quarter) and 0.83 percent in 2013.36 36 In the absence of detailed figures on the projected increase in the number of retirees for 2012 and 2013, these estimations are only indicative since they were performed with the assumption that the number of retirees would remain unchanged in 2012 and 2013. However, if we assume that the number of retirees would increase by 5 percent 26 Figure 3.9 Pensions and Compensations, 2008–12 (JD mil and % of GDP) 1200 4.6% 1000 4.2% 800 4.0% 4.3% 4.2% 600 400 200 0 2008 2009 2010 2011 Budget-2012 Source: MoF Financial Bulletin; Budget Law; World Bank staff calculations. per year in 2012 and 2013, then the estimated savings would be 0.42 and 0.47 percent respectively. If we project the number of retirees to increase by 10 percent, then the savings would be 0.23 and 0.08 percent. 27 4. Fiscal Implications of the Energy and Power Sectors 61. The energy subsidy bill in Jordan is at the heart of the recent deterioration of public finances. In 2011 the combined deficit of the budget and autonomous agencies reached 18.7 percent of GDP. Of this deficit, 7.7 percent of GDP can be attributed to petroleum subsidies (2.8 percent) and to the electricity company’s deficit (4.9 percent). Large subsidies to the electricity company are relatively new. The sector was in equilibrium until 2009 and deficits started in 2010 and accelerated later. The Government has started taking some consolidation measures (June 2012 electricity tariff increase and cuts on gasoline subsidy) that are expected to reduce energy’s fiscal burden by about 1.4 percent of GDP. 62. More efforts are however needed to put the sector on a sustainable financial path. This chapter takes stock of existing subsidy schemes (section 4.1), and of their recent developments (section 4.2). Sections 4.3 and 4.4 present options for a short-term financial adjustment of the power sector and a fiscal adjustment of fuel subsidies, respectively. Section 4.5 outlines options for longer term reforms. 4.1. Types of Energy Subsidies 63. The two main types of energy subsidies in Jordan are (a) cash subsidies from Government budget to energy companies and (b) noncash subsidies in the form of Government sovereign guaranteed debt and outstanding unpaid financial obligations arising due to the Government’s policy. These two types are elaborated below. 64. Cash subsidies from Government budget to energy companies relate largely to petroleum products. These subsidies cover the gap between selling price and supply cost, especially for household consumption of LPG and kerosene and, more recently, for gasoline and diesel. Between 2008 and end 2010, GoJ phased out cash subsidies on petroleum products, causing a sharp drop in oil subsidies from 2.5 percent of GDP in 2007 to 0.3 percent in 2009. However, starting in December 2010, oil retail prices (except for heavy fuel oil for industrial consumption, power generation and aviation fuels) again were frozen. As a result, petroleum subsidies were reintroduced, reaching 2.8 percent of GDP in 2011. 65. In addition to cash subsidies, GoJ incurs noncash subsidies for energy products. In particular, since 2010 and continuing in 2012, GoJ has provided its sovereign guarantee to the state-owned power company, NEPCO, to facilitate bank borrowings and bond offerings to pay for the cost of back-up liquid fuels for power generation. Furthermore, NEPCO has outstanding fuel bills payable to the oil company, Jordan Petroleum Refinery Company. Supplier payables to JPRC became another source for NEPCO to finance its deficit over the short term. A private company, JPRC is the only refinery and petroleum supplier in Jordan. JPRC has been fronting these bills by taking loans to pay fuel suppliers. JPRC naturally expects to be repaid by NEPCO as soon as possible. The most recent available data from JPRC confirm that NEPCO is actually paying fuel bills to JPRC but with delays of several months. Indeed, from end-2011, the amount of unpaid fuel bills from all customers, including NEPCO, has decreased considerably from JD674 million to JD254 million at end-May 2012. Table 4.1 shows the flow of (a) cash subsidies from the budget and (b) noncash subsidies in the past five fiscal years. 28 Table 4.1 Structure and Flows of Subsidies in Jordanian Energy Sector, 2007–12 Million JD unless otherwise specified 2007 2008 2009 2010 2011 2012 Cash subsidies from the budget Budgetary petroleum subsidies (306) (198) (43) (88) (571) (626) Nominal GDP at market prices 12,131 15,593 16,912 18,762 20,477 22,230 Petroleum subsidies (% of GDP) -2.5 -1.3 -0.3 -0.5 -2.8 -2.8 Government budget deficit including grants (% of GDP) -5.1 -2.2 -8.9 -5.6 -6.7 -6.5 Government budget deficit excluding grants (% of GDP) -7.9 -6.8 -10.9 -7.7 -12.7 -10.7 Noncash subsidies and payables NEPCO surplus/(deficit) (2.0) (40.5) 39.7 (160.1) (1,008) (1,180) (Expressed as % of GDP) 0.0 -0.3 0.2 -0.9 -4.9 -5.3 Source: DoS; Central Bank of Jordan; NEPCO; JPRC. Note: 2012 estimated assuming 50 mil cu ft/day of imported Egyptian natural gas. 4.2. Recent Developments 66. In 2008 Jordan adopted a no-energy-subsidy policy. In February 2008, as part of the national energy strategy, subsidies on most petroleum products were removed. However, in order to shelter lower income households, a smaller subsidy was maintained for household liquefied petroleum gas (LPG), which is used for cooking and space heating. As a result of the subsidy removals, price subsidies on petroleum products declined steeply from 2.5 percent of GDP in 2007 to 0.3 percent in 2009. In March 2008, electricity tariffs also were raised to cost recovery to reflect the higher cost of liquid fuels used in electricity generation—largely heavy fuel oil (HFO) and, to a lesser extent, diesel oil. Table 4.2 Change in Petroleum Subsidies, 2007–12 2007 2008 2009 2010 2011 2012 Budgetary petroleum subsidies (JD mil) -306.0 -197.9 -42.9 -88.2 -571.0 -626.0 Nominal GDP at market prices (JD mil) 12,131 15,593 16,912 18,762 20,477 22,230 Petroleum subsidies (% of GDP) -2.5 -1.3 -0.3 -0.5 -2.8 -2.8 Source: MoF, www.mof.gov.jo; DoS, www.dos.gov.jo; Central Bank of Jordan, www.cbj.gov.jo; Jordan Petroleum Refinery Company (JPRC), www.Jopetrol.com.jo Note: 2012 is estimated 67. From 2005 to 2009, the electricity sector had received no cash subsidies from the budget. Through 2009, overall sector revenues from customers were adequate, supplemented by JD10 million– JD20 million per year of electricity consumption tax intended for rural electrification (approximately 3 percent of electricity revenue). However, from 2010 on, the sector began to incur financial deficits as consumers were insulated from the cost of Egypt’s gas supply disruption. These deficits have been financed not from the GoJ budget but from debt raised and guaranteed by GoJ on behalf of NEPCO. 29 68. Egypt’s gas supply disruption, high oil prices, and political events led Jordan to reintroduce energy subsidies starting in late 2010. Jordan started importing natural gas from Egypt in 2003. This gas has been used exclusively in power generation, complementing the small quantity of locally available natural gas. By 2009 natural gas accounted for 90 percent of power generation fuel, effectively displacing HFO as the main fuel in power generation. Egypt’s gas supply disruption begun in 2010, causing the first drop in supply since 2003. Additional quantities of HFO and diesel were used to make up the gas shortfall, and imports of electricity increased sharply. Egypt’s gas supply disruption worsened in 2011 Figure 4.1 Structure of Jordan’s Power Generation Fuels (000s toe) 100% 106 154 230 332 621 562 881 962 80% 905 782 60% 1,355 3,080 40% 2,396 2,697 2,283 1,381 1,193 20% 849 0% 2004 2005 2007 2008 2009 2010 2011 Diesel Heavy fuel oil Natural gas Source: National Electricity Power Company (NEPCO), www.nepco.com.jo Note: toe = ton of oil equivalent (due to 11 pipeline explosions) and continued into 2012 (a 15th explosion in July). In 2011 Jordan received approximately only 620 million cubic meters (m3) of approximately 2,600 m3, 37 or approximately 24 percent of the base contracted volume from Egypt. Figure 4.1 shows that gas consumption for power generation began to decline from a peak of 3.08 million tons of oil equivalent (approximately 3,300 million m3) in 2009 to 2.28 million tons of oil equivalent (approximately 2,500 million m3 of gas) in 2010. 69. Disruptions in gas supply from Egypt have caused a sizable price shock for Jordan. NEPCO’s role in facilitating the supply of natural gas and other fuels to power plants has placed the company in a difficult financial situation. On an equivalent energy content basis, natural gas prices generally are lower than oil prices. Figure 4.2 shows movement in crude oil and natural gas prices in major markets, expressed in US$ per million British thermal units (Btu), compared to the prices of HFO in Jordan. 37 Equivalent to a supply of 60 mil standard cu ft./da. 30 Egypt’s gas prices to Jordan have been lower than HFO price (and, lower than natural gas prices at major markets). Therefore, Egypt’s gas disruption since 2010 led to a larger price shock. During the same period, this shock was exacerbated by higher oil prices. For instance, in 2011 the cost of HFO in Jordan was between 3 to 4 times higher than the cost of Egyptian gas on a per-Btu basis. Figure 4.2 Average Price of Crude Oil and Natural Gas, 2005–11 (US$/mil BTU) 20 18 16 14 US$ per million BTU 12 10 8 6 4 2 2005 2006 2007 2008 2009 2010 2011 Crude oil, Dubai Natural gas, Europe Natural gas, US Liquefied natural gas, Japan HFO, Jordan Source: World Bank Commodity Price data, www.worldbank.org; JPRC 70. In the absence of commensurate domestic price increases, this has translated into high deficits for NEPCO. Due to the Government-to-government nature of gas contracts and the confidentiality of Egypt’s gas prices, NEPCO has been the entity responsible for gas administration—including payments to gas suppliers on behalf of power plants. For HFO and diesel, some power plants would be compensated by NEPCO for their cost of liquid fuels bought from fuel suppliers. Alternately, especially for new privately owned independent power producers (IPPs), NEPCO is responsible for directly procuring and paying the cost of liquid fuels to the fuel supplier. When there is a mismatch between NEPCO’s selling tariffs (to electric distribution companies and industrial users) and buying tariffs (from power plants and imports), NEPCO incurs the resulting deficits. Imported electricity accounted for 4.4 percent of available electricity in 2010 and rose to 11.0 percent in 2011. That year, the electricity import bill rose from JD54 million (0.3 percent of GDP) to JD195 million (1.0 percent of GDP), further burdening NEPCO’s deficit and Jordan’s Current Account balance. 31 Table 4.3 Jordan’s Consumption of Petroleum Products, 2008–11 (MT) Total Year LPG Gasoline Jet fuel Kerosene Diesel Fuel oil petroleum 2006 313,072 740,595 299,568 150,073 1,774,362 1,279,228 4,726,564 2007 335,137 839,641 296,747 130,659 1,746,054 1,246,820 4,750,230 2008 321,272 861,177 297,681 99,633 1,508,376 1,096,251 4,352,426 2009 338,553 1,022,515 318,437 110,654 1,613,536 823,043 4,421,713 2010 311,977 1,065,405 350,577 69,355 1,543,479 1,380,905 4,874,155 2011 377,985 1,083,147 353,779 75,166 2,406,709 1,669,826 6,076,074 Source: JPRC. Note: MT = metric ton. 71. The shift in the consumption pattern toward electric heating and cooking, replacing kerosene and LPG, contributed to the burden on the electricity sector. Between 2007 and 2010, diesel/kerosene prices rose 65 percent—from JD0.32/liter (l) to JD0.52/l; and a 12.5-kg LPG tank increased by 53 percent from JD4.25/l to JD6.50/l. These large increases contributed to a general shift toward electric heating and, to a lesser extent, to cooking. Figure 4.3 shows the deep 50 percent decline in kerosene consumption and a slower 20 percent growth in LPG consumption relative to a 40 percent increase in total electricity consumption since 2006. Hence, the price elasticity of kerosene can be estimated at a high -0.77 percent and the elasticity of substitution between electricity and kerosene may be of 0.62, i.e a 1 percent increase in the prices of kerosene leads to a 0.62 percent increase in electricity consumption. Notably, household and domestic electricity consumption increased at a faster rate than total electricity consumption. 38 Consequently, any tariff reform in electricity is likely to reduce electricity consumption and generate lead to substitutions with other – and cheaper – sources of energy. N more general terms, reducing energy and power subsidies would help reduce the relatively high energy intensity of the Jordanian economy. Indeed, in 2009, Jordan’s energy use per US$1000 of GDP was 7 percent higher than the average for Middle- Income countries and 31 percent higher than World’s average.39 Figure 4.3 also shows the sharp increases in diesel and HFO consumption since 2010 due to Egypt’s gas supply disruption. 38 “Domestic� category includes household customers and non-household customers (such as public buildings). Households account for approximately 84 percent of the domestic category. 39 World Bank, World Development Indicators, www.worldbank.org. The GDP is expressed at constant 2005 PPP. 32 Figure 4.3 Trends in Petroleum and Electricity Consumption in Jordan, 2006 Base Year 1.6 1.4 1.2 2006 = 1.00 1 0.8 0.6 0.4 2006 2007 2008 2009 2010 2011 LPG Gasoline Kerosene Diesel Fuel Oil Electricity, total Electricity, domestic Source: NEPCO; JPRC; Electricity Regulatory Commission (ERC), www.erc.gov.jo Figure 4.4 Electricity Consumption by Users, 2008–11 (GWh) 100% 90% 2203 2316 2403 2408 80% 70% 1863 2110 2118 1933 60% 50% 3308 3560 3219 3078 40% 30% 2,641 2,894 3,157 2,400 20% 10% 1,342 1,420 1,493 1,574 0% 2008 2009 2010 2011 Others Public Sector Agricultural Commercial Industrial Household > 300 kWh/mo. Household < 300 kWh/mo. Source: ERC 33 Box 4.1 PPP Structure in Jordan Electricity Sector Electricity Generation In the past 10 years, the electricity generation sector in Jordan went through a series of significant structural reforms. They include the introduction of an IPP scheme, as well as the privatization of a formerly state-owned incumbent generation company, CEGCO in 2007 (51% private ownership). By 2011 only 1 of 4 generation companies—Samra Electric—was fully Government-owned. The first new IPP—Amman East—started full operation in 2009, followed by the second new IPP—Al-Qatrana—in 2011. New IPPs have been granted on an international competitive bidding basis. IPPs’ return on investment is determined through the bidding process and is subject to their meeting their contractual obligations for generation capacity, energy generated, and efficiency. The cost of power generation fuel is structured as a pass-through— based on predetermined efficiency—from power plants to NEPCO to electricity consumers. The Government provides a sovereign guarantee to support NEPCO’s payment obligations to IPPs, including capacity charge, energy charge, and other miscellaneous charges. The Government also provides a sovereign guarantee on project termination charge in case of a contractual default by NEPCO or the Government, or certain force majeure events. Sovereign guarantees have assisted IPPs in obtaining long-term project financing at competitive terms. Electricity Transmission, Single-Buyer, System Operator, and Fuel Facilitator The state-owned NEPCO is responsible for electricity transmission, acts as the single buyer of electricity from power plants and imported power, facilitates procurement of natural gas for all power plants and liquid fuels for some power plants, and responsible for operating Jordan’s electricity network. NEPCO’s selling tariffs (bulk supply tariffs) are regulated by the Electricity Regulatory Commission. Such tariffs have been designed based on a cost recovery approach. Therefore, NEPCO normally records a small profit or loss in a normal operating year. NEPCO has recently experienced large losses due to the large gap between its buying and selling tariffs. Electricity Distribution There are three privately-owned distribution companies in Jordan: JEPCO, EDCO and IDECO. These companies are regulated by the Electricity Regulatory Commission. EDCO and IDECO’ returns on investment are determined based on regulated returns on assets, which is largely a cost plus methodology, subject to efficiency requirements. JEPCO’s returns on investment have been based on a regulated return on equity per the existing concession agreement. Although this is expected to be aligned with the other two companies by end of 2012 when the concession term expires, and a replacement license becomes effective. The distribution companies pay NEPCO the regulated bulk supply tariffs, which are designed to cover their operating cost, regulated investment and returns on investment. Electricity tariffs for end-users are proposed by the ERC and subject to governmental approval. 72. Electricity for agriculture users and water pumping has accounted for approximately 15 percent of total electricity consumption in Jordan over the past five years. The electricity tariff for this group of users has been slightly cross-subsidized by other consumer groups, but not as much as tariffs for households and other domestic users. Figure 4.4 shows that tariffs on agriculture/water pumping had been in line with NEPCO’s average buying tariffs since 2008. This alignment reflects the regulatory effort to have a cost-recovery tariff for this group of consumers. However, in 2011 the tariffs for this group ranged from JD0.054–JD0.060 per kWh, which were below the average cost of NEPCO’s buying tariff by an average of 50 percent–60 percent. 34 73. As a result of these developments, the size of energy subsidies increased substantially in 2011. These subsidies were provided under three categories: a. Cash subsidies to oil company, JPRC. In 2011 JPRC received JD571 million of cash subsidies from GoJ. These subsidies helped cover the gap between the selling prices of petroleum products and their cost of supply in 2011. b. NEPCO borrowing guaranteed by GoJ. The Government guaranteed JD567 million of NEPCO’s long-term debt borrowed in 2011. 40 This guaranteed debt is a direct augmentation of the public sector’s debt by virtue of Jordan’s public debt accounting (also because NEPCO was designed to realize cost-recovery, not large surpluses). Figure 4.5 shows the large gap in 2011 between NEPCO’s average electricity-buying tariff from power plants and import sources, and its average selling tariff to electricity distribution companies and direct consumers. Since end- user electricity tariffs did not fully reflect the cost of back-up fuels, in 2011 NEPCO’s selling tariff was less than half the actual average buying tariff. c. NEPCO’s arrears to private oil company, JPRC. At end-2011, NEPCO had accumulated JD716 million of accounts payable related to power and fuel purchases. Of this amount, JD441 million was payable to JPRC accumulated in 2011 for back-up fuels used in power generation.41 As indicated earlier, JPRC had taken loans to front the payment for fuels on behalf of the power sector. Since NEPCO is a public company, these arrears toward JPRC are another form of contingent liabilities. These arrears amount to several times the capital of JPRC and threaten the financial stability of the company. 42 This situation has important potential implications since one of the shareholders is the National Social Security Corporation, which owns more than 20 percent of the company. In total, energy subsidies amounted to JD1.6 billion in 2011, or 7.7 percent of GDP. Of this total, JD571 million accrued to the budget whereas the remaining amounts took various forms of liabilities. 4.3. Short-Term Options and Recommendations to Address Energy Subsidies in the Electricity Sector 74. Responding to the fiscal crisis caused by high electricity subsidies, the GoJ increased electricity tariffs in certain sectors in June 2012. Banks, telecommunication companies, water pumping, hotels, street lighting, ports, large industries, and mining (phosphate and potash) are among the sectors that would face an estimated 22 percent increase in average unit costs, reaching as high as 150 percent for banks and telecommunications firms. According to projections by the electricity commission, the additional revenues from this increase are estimated at JD118 million for 2012, which would cut electricity sector losses by only 10 percent. The new and old tariff schedules are available in Appendix 3. 40 NEPCO (Public Shareholding Company), Financial Statements, December 31, 2011, May 8, 2012, www.nepco.com.jo 41 Most of these accounts payable were recorded on NEPCO’s books as owed to power generation companies. 42 By the end of 2010, JPRC’s total shareholders’ equity totaled JD74.7 million. 35 Figure 4.5 NEPCO’s Average Selling vs. Buying Tariffs 0.13 0.12 0.11 0.10 JD per kWh SOLD 0.09 0.08 0.07 0.06 0.05 0.04 0.03 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 NEPCO's average selling tariff NEPCO's average buying tariff Source: NEPCO; ERC; World Bank staff estimates. Note: 2012 estimated. 75. Unless additional measures are implemented, the large financial deficit in the power sector is expected to continue. If the lack of Egyptian gas continues, NEPCO deficits are expected to reach JD1.2 billion in 2012 (5.3 percent of GDP) and 900 million in 2013. Due to the large gap between required tariffs and the prevailing average tariffs, a gradual approach is required to smooth the transition for the general public. A number of tariff categories in the May/June 2012 tariff schedule remain below NEPCO’s average buying tariff (Figure 4.6). In fact, if Egyptian gas supply remains at 50 million cubic feet (cu ft) per day for the whole of 2012, and if the subsidy on fuel purchased by NEPCO (estimated at JD222 million) is removed, the company would need approximately JD0.180 per kWh to cover the cost of fuel and nonfuel operating costs and power plants’ operating costs. This range will decline to JD0.145 per kWh if the June 2012 increase produces its full effect and is not absorbed by unexpected increases in fuel prices. 76. One swift move could be an immediate increase in electricity tariffs for Governmental customers to reflect the true cost of electricity supply. These customers include national and local entities; the armed forces; and water pumping, street lighting, and other public sector customers. In 2011 these public sector consumers represented 17.6 percent of total power consumption. Increasing their tariffs would help send a credible message to the general public, and raise their awareness of and support for the need for higher tariffs. The new cost-reflective tariffs for governmental customers would raise an additional JD278 million for NEPCO on annual basis in 2012 (JD70 million for the last quarter of the year). Much of this amount would be financed through the removal of the JD222 million (JD56 million for the last quarter) in subsidies that GoJ would pay for NEPCO’s oil purchases. The budget then would 36 directly reflect the cost of actual electricity consumption by Government entities. The budget then also could cross-subsidize other consumers, mainly lower brackets household consumers, that are below the poverty line, i.e, around 15 percent of the population.43 Figure 4.6 NEPCO’s Average Buying vs. End-Consumer Tariff, 2007–12 (JD/ kWh sold) 0.250 0.200 0.150 0.100 0.050 0.000 2007 2008 2009 2010 2011 2012 NEPCO's average buying tariff Domestic - min Agriculture/water pumping Commercial, average Armed forces Large industries - max Source: NEPCO; ERC; World Bank staff estimates 77. After GoJ takes the lead on a cost-reflective electricity tariff, it will become possible to modify the tariffs of other categories of consumers. If adjustment were apportioned according to consumption, the commercial, industrial, and agriculture sectors represent 47 percent of electricity consumption and therefore should be responsible of 47.3 percent of electricity losses, that is, JD660 million. Households, who represent 35.1 percent of electricity consumption, should bear JD490 million of overall electricity losses. This goal would require increasing the cost of electricity per household by 100 percent–150 percent. Instead of moving towards such a massive increase in tariffs, the GoJ may decide to fulfill the target of JD280 to 300 million savings from NEPCO agreed under the SBA with the IMF by generating more revenues from household consumers while fully compensating the first two spending deciles that are below poverty line. Hence, only 66 percent of the losses (JD320 million) could be recovered and part of the additional savings (JD15 to 20 million) can be used to compensate the first two deciles (see chapter 5 for details on household tariff reforms). The Government may decide to bring NEPCO to cost recovery which implies an increase in tariffs on industrial and commercial consumers without putting additional burden on households. The remaining JD170 million losses from households could then be cross- subsidized either by other types of consumers (public sector and businesses) or through cash transfers. Of course, these estimates are done on an annual basis. Thus, obtaining the order of magnitude for the likely 43 World Bank staff calculations based on actual figures for 2011 provided by ERC, www.erc.gov.jo. 37 impact on the last quarter requires dividing the figures by 4. Table 4.4 in the below shows for 2012 a scenario of full cost recovery with an estimation by large categories of consumers of (i) the share in total consumption, (ii) the imputed losses, (iii) the imputed additional contribution to reach cost recovery and, (iv) the estimated cross subsidy. Table 4.4 Estimation of Losses and Cost Recovery by Large Categories of Consumers for 2012 Estimation Additional Imputed Losses Cross Subsidies Category Share Total Tariff Contribution if Cost (JD 000s) (JD 000s) Paid (JD 000s) Recovery (JD 000s) Public Sector 17.6% 175,443 246,254 278,422 32,168 Economic Sectors 47.3% 559,081 663,143 802,766 139,623 and Other Households 35.1% 415,303 492,603 320,812 (171,792) Grand Total 100% 1,149,826 1,402,000 1,402,000 - Source: World Bank staff calculation, based on data from ERC, www.erc.gov.jo, Note: 2012 estimated assuming 50 mil cu ft/day of imported Egyptian natural gas and 7.07 percent increase in consumption. 78. Beyond the need to eliminate operational deficits, the large liabilities that have built up in the system require specific Government action. The accumulated debts of the power sector reached approximately JD1.2 billion by end-2011 and are increasing. By June 2012, most of these deficits were funded directly through NEPCO’s indebtedness, but a residual remains on JPRC’s books. Based on GoJ’s own share of electricity consumption, one option would be for the Government to pay upfront its share of the incremental power generation fuel cost from 2010 and 2011. GoJ first could decide how much of this amount could be covered from the budget. The amount could be equal to the combined public sector and armed forces’ share in electricity consumption of 17.6 percent in 2011. Assuming that 17.6 percent, or JD211 million, could be covered by the budget, the balance of JD1.0 billion could be collected retroactively from certain groups of electricity users over future years. 79. Once it has paid its own share of consumption, the Government can proceed with the retroactive collection of other electricity deficits accumulated in 2010 and 2011. All consumers except the lifeline block of household consumers could be levied with retroactive fuel surcharges. Table 4.5 shows a calculation of such a retroactive collection scheme with a deduction of JD211 million cash from GoJ (representing 17.6 percent of electricity consumption). In addition, Table 4.5 shows two scenarios of cross-subsidies. The first excludes the first 300 kWh of household consumption (11.6 percent of total electricity consumption, and a share of JD140 million in accumulated losses). The second excludes the first 200 kWh (1.6 percent of total consumption and a share of JD19 million in losses). 80. A public relations campaign could stress the fact that these retroactive surcharges are specific to recover the debt accumulated by NEPCO. In the first, scenario, it is estimated that the remaining JD1.0 billion could be collected over 3 months in 2012 plus 5 years at a rate of JD0.0158 per kWh; or over 3 months in 2012 plus 8 years at a rate of JD0.0093 per kWh. The first scenario assumes an average electricity demand growth of 5.5 percent per year over 2013–20. As an example, the most recent electricity tariffs effective in June 2012––electricity tariffs for household and domestic users––range from JD0.086 per kWh to JD0.235 per kWh for consumption above 300 kWh blocks. Hence, a retroactive fuel surcharge of JD0.0158 per kWh would add approximately between 7 and 18 percent to the latest June tariffs for domestic users. The Government should stress that retroactive surcharges aim to recover the 38 debt accumulated in 2010 and 2011 and will be removed in the future. In the second scenario, the rate for the 5-year period would be JD0.0139; and for the 8-year period, JD0.0081. The JD0.0.139 surcharge corresponds to an addition ranging from 6 percent–16 percent over the last tariff adjustment. 4.4. Short-Term Options and Recommendations to Address Energy Subsidies in the Petroleum Sector 81. Jordan discontinued monthly petroleum price adjustments in December 2010 when crude oil prices were approximately US$90 per barrel. In recent months, oil prices have started to decline so should provide opportune timing for GoJ to resume periodic petroleum product price adjustments. Starting May 2012, the GoJ has raised the price of octane-95 gasoline to JD1.0 per liter (US$1.40 per liter, or $5.30 per US gallon) and octane-90 gasoline to JD0.70 per liter (US$1.00 per liter or $3.80 per US gallon), which is a move in the right direction. 82. However, a more broad-based petroleum price adjustment is needed because gasoline represents approximately 55 percent of transport sector sales by volume. In fact, in 2011 oil consumption in the transport sector totaled approximately 1.9 million tons. Of this quantity, 10 percent was octane-95 gasoline, and the remaining was split approximately equally between octane-90 gasoline and diesel. JPRC has observed a gradual shift from the higher priced octane-95 toward the lower priced octane-90. This trend suggests that future oil pricing strategy would have to focus on octane-90 gasoline and diesel to recoup paid subsidies and reduce future subsidies. In fact, the high substitutability between oil products should incite the government to raise prices simultaneously in order to maintain relative 39 Figure 4.7 Retroactive Fuel Surcharge from 2010 and 2011 for Electricity Generation, 2012–18 Scenario 1. Excludes HH electricity consumption below 300kWh; GoJ pays for public sector users (100 percent of poor plus part of the non-poor) 2012 2013 2014 2015 2016 2017 2018 2019 2020 Mil kWh sold to DISCOs and large customers 16,703 17,622 18,591 19,613 20,692 21,830 23,031 24,298 25,634 Subtract: transmission and distribution losses (2,171) (2,290) (2,416) (2,549) (2,690) (2,838) (2,994) (3,159) (3,332) Mil kWh sold to end-users 14,532 15,331 16,174 17,064 18,002 18,993 20,037 21,139 22,303 Subtract: first 300kWh of household (4,243) (4,477) (4,723) (4,983) (5,257) (5,546) (5,851) (6,173) (6,512) consumption & prepaid public sector Mil kWh eligible for 2010/2011 retroactive fuel 10,289 10,854 11,451 12,081 12,746 13,447 14,186 14,967 15,790 surcharge Fuel surcharge (JD mill) @ JD 0.018 per kWh 162.6 171.5 180.9 190.9 201.4 212.5 224.1 236.5 249.5 Fuel surcharge (JD mill) @ JD 0.015 per kWh 132.7 140 147.7 155.8 164.4 173.5 183 193.1 203.7 Fuel surcharge (JD mill) @ JD 0.011 per kWh 95.7 100.9 106.5 112.4 118.5 125.1 131.9 139.2 146.8 Scenario 2. Excludes HH electricity consumption below 200kWh (75 percent of the poor) 2012 2013 2014 2015 2016 2017 2018 2019 2020 Mil kWh sold to DISCOs and large customers 16,703 17,622 18,591 19,614 20,693 21,831 23,031 24,298 25,634 Subtract: transmission and distribution losses (2,171) (2,291) (2,417) (2,550) (2,690) (2,838) (2,994) (3,159) (3,332) Mil kWh sold to end-users 14,532 15,331 16,174 17,064 18,003 18,993 20,037 21,139 22,302 Subtract: first 200kWh of household (2,790) (2,944) (3,105) (3,276) (3,456) (3,647) (3,847) (4,059) (4,282) consumption and prepaid public sector Mil kWh eligible for 2010/2011 retroactive fuel 11,742 12,388 13,069 13,788 14,546 15,346 16,190 17,081 18,020 surcharge Fuel surcharge (JD mill) @ JD 0.016 per kWh 163.2 172.2 181.7 191.6 202.2 213.3 225.0 237.4 250.5 Fuel surcharge (JD mill) @ JD 0.013 per kWh 132.7 140.0 147.7 155.8 164.4 173.4 182.9 193.0 203.6 Fuel surcharge (JD mill) @ JD 0.009 per kWh 95.1 100.3 105.9 111.7 117.8 124.3 131.1 138.4 146.0 Source: World Bank staff calculation, based on data from ERC, www.erc.gov.jo, and NEPCO, www.nepco.com.jo Note: DISCOs = Distribution companies 40 prices reduce overall consumption instead of shifting consumption between products. In 2011 the JD571 million cash subsidies for petroleum products corresponded to an average subsidy of JD140 per metric ton of subsidized fuels (approximately JD0.10 per liter). The new prices of gasoline introduced in June are likely to reduce the 2012 estimated cash subsidies by approximately JD145 million. Last, the actual cash subsidies could be lower if Egypt’s gas supply situation improves; oil prices continue to decline during the remainder of 2012; and/or other petroleum product prices are increased. 44 Figure 4.8 Crude Oil Movements, 2010 to Present (US$/barrel) 130 120 110 100 90 80 70 60 2010M10 2010M11 2010M12 2011M01 2011M02 2011M03 2011M04 2011M05 2011M06 2011M07 2011M08 2011M09 2011M10 2011M11 2011M12 2012M01 2012M02 2012M03 2012M04 2012M05 2012M06 Crude oil, Brent Crude oil, Dubai Crude oil, WTI Source: World Bank Commodity price data 83. Regarding cooking gas and kerosene, the Government may resort to targeted subsidies for selected consumers. LPG is used widely by Jordanian households for cooking and, to a lesser extent, for space heating. Predominantly poor households also use kerosene for space heating. At the country level, LPG demand has increased by approximately 20 percent over the past five years, reaching 377,985 tons in 2011. In contrast, demand for kerosene declined by approximately 50 percent over the same period, reaching 75,166 tons in 2011. To safeguard against adversely affecting the social welfare of low-income 44 On November 13, 2012, the Government has removed fuel subsidies and reduced LPG subsidy and is currently implementing a system of cash compensation benefitting to between 80 to 90 percent of resident Jordanians. According to preliminary estimations, the Government expects budgetary savings to reach JD500 million (about 2.1 percent of GDP), based on an estimated subsidy bill of JD800 million and on an estimated cost for cash compensation at around JD300-350 million. However, World Bank staff preliminary estimation shows that about half of the savings might actually be related to the removal of the subsidy on petroleum products consumed by NEPCO (see estimation of this subsidy in paragraph 75, page 41). Thus, since NEPCO deficit might increase by the same amount as the decline in budgetary subsidies on NEPCO’s fuel purchases (other than gas), the saving on overall public sector deficit would be around 0.12 percent for the last month and a half of 2012 and around 1 percent of GDP for the year 2013, which would remain in any case a substantial contribution to overall fiscal adjustment. 41 households in Jordan, the Government may need to provide targeted subsidies for LPG and kerosene for selected consumers, replacing a blanket subsidy for all users. Based on 2011 LPG demand, a targeted subsidy scheme for LPG covering 15 percent–35 of demand cost approximately JD4 million–JD10 million to administer per years (at an average subsidy of about JD70 per ton of LPG); versus JD47 million for a full-fledged subsidy (see Table 4.6). Table 4.5 Scenarios for LPG Subsidies Reform (JD mil) Estimated overall LPG subsidies (JD mil) Subsidy per ton of LPG in JD 70 210 350 Total for 15% of demand 4.0 12.1 1 20.1 2 Total for 35% of demand 9.4 9 28.2 2 4 47.0 Source: World Bank staff estimates. 4.5. Long-Term Options to Address Energy Subsidies and Diversify Sources of Energy 84. It would be advisable for Jordan to consider returning to a no-energy subsidy policy. Jordan’s no-energy-subsidy policy was sound and encouraged efficient use of both energy and national economic resources. In addition, Jordan’s energy pricing arrangement for all energy products is already in place and could be reactivated promptly. 85. Subject to proper design, the authorities may consider establishing an emergency energy fund to safeguard against future energy price shocks in Jordan. The experience of Jordan’s rural electrification fund (Rural Fils Fund)—for which JD0.001 is collected for every kWh of electricity sold— could be applicable. Inflow to the emergency fund would come from petroleum and electricity sales. The proceeds of the fund could be provided to energy operators during a price shock and be reimbursed afterward. For example, total petroleum sales in 2010 were JD2.38 billion; and electricity wholesales were JD0.67 billion, totaling JD3.05 billion. 45 A 1.0 percent emergency fund collected from combined petroleum and electricity sales would generate JD30.5 million. International experience shows that the design of the legal and operational framework of these funds is critical to their success. The authorities also could consider using this fund to “park� outstanding liabilities currently on NEPCO’s balance sheet (unpaid bills to JPRC, bank loans, and bonds for backup fuels), which arose due to Government’s decision on electricity tariffs. These liabilities then could be paid off gradually through Government budget allocation and through retroactive fuel surcharges (section 4.3). By removing these outstanding liabilities from NEPCO, company’s balance sheet and financial accounts could be quickly returned to equilibrium, thus and enabling NEPCO to proceed with its business and new investment on its own as in the past. 86. Specific to the electricity sector, it might be worthwhile to consider separating NEPCO’s electricity transmission/power system operator business from its electricity and fuel trading business. A transparent transmission margin methodology could be introduced to govern the transmission/power 45 JPRC sales for petroleum and NEPCO sales for electricity. 42 system operator business. This arrangement would shield the electricity transmission and system operator business from negative financial spillover from the electricity/fuel trading business, which happened in recent years. The resulting new electricity/fuel trading entity would operate based on Jordan’s sound electricity tariff methodologies in place between the actors and stakeholders of the sector. They comprise power generators, the single-buyer of electricity (NEPCO), distribution companies, and end-consumers. When there is a justifiable need for subsidies, a transparent subsidy arrangement could be put in place without affecting the transmission/power system operator business. 87. For the long term, Jordan should be in a better position to rebalance its energy supply. Jordan imports around 95 percent of its primary energy each year and has always placed high emphasis on energy supply security. The National Energy Strategy; updated in 2007; had already identified several options for further enhancing energy security. These include: (i) diversifying natural gas sources through diversification of energy imports, (ii) strengthening regional trade of gas and electricity, (iii) developing renewable energy, (iv) promoting energy efficiency, (v) developing domestic energy resources, including oil shale, and (vi) developing nuclear energy. If implemented, these measures would gradually reduce the burden on Jordan’s oil import and help prevent further energy shock due to over-reliance on only a few sources of energy. The introduction of Egypt gas import in 2003 was in fact a key milestone in reducing Jordan’s dependence on oil for electricity generation, but the Egypt gas import disruption since 2010 had again highlighted the risks of depending on one source and one supplier of energy. 88. Over the medium term, diversifications of sources of fuel should be the main alternative to Egyptian gas. In the next 3 to 5 years, import of liquefied natural gas from the international market and import of electricity from neighboring countries should be the main alternatives to Egypt gas import and continuation of oil product import. Import of natural gas via pipeline from other countries would take longer to implement. Renewable energy (such as wind and solar) would not achieve a large scale that is adequate for meeting energy demand during this same timeframe. The envisaged plan for oil shale and nuclear energy would also take longer to implement. 43 5. Distributional Impact of Electricity Tariff and Consumer Subsidy Reforms 89. This chapter analyzes the potential impacts on households of electricity tariff reforms and the removal of food subsidies. Chapter 4 looked at the fiscal implications of an increase in electricity tariffs. Chapter 5 focuses on the implications of a differentiated tariff increase for households (HH), specifically the impacts on poverty and welfare. If electricity consumption brackets are modified, full cost recovery from households with a cross-subsidy to the lowest bracket would generate annual savings of 1.4 percent of GDP. These savings would cover 65 percent of electricity losses related to household consumption and 23 percent of overall electricity losses. The remaining JD170 million annual losses (JD43 million for the last quarter of 2012) from lower-bracket consumption would need to be covered by the Government or be cross-subsidized by industrial and commercial consumers. This chapter also shows that eliminating food subsidies and redirecting 25 percent of the savings to support households in the 3 lower deciles would generate an annual net savings of 0.75 percent of GDP (0.16 percent for 2012 if the measure is implemented in the last quarter of the year). Section 5.1 examines different scenarios of electricity tariff increases for households. Section 5.2 simulates a removal of food subsidies. 5.1. Scenarios for Cost Recovery from Residential Users 90. The tariff increases of June 2012 included limited changes in the tariffs on residential consumers (households). Only the tariffs on residential consumers in the bracket of 600 kWh and above were increased. Residential consumers make up approximately 80 percent of all consumers in Jordan and consume approximately 35 percent of total electricity. It is well documented that existing universal price subsidies not only place a large fiscal burden on the Government but also disproportionately benefit the relatively better-off households. 46 Given the high cost and the inefficiencies of electricity subsidies, GoJ could consider a more proactive change in the nature of tariffs on residential customers. Indeed, the impact of the recent tariff increase is limited since it affects less than 10 percent of households and, at most, would generate additional annual revenues of JD3.5 million from this category of consumers. 91. This section examines the cost recovery capacity of increases in tariffs on residential consumers under various scenarios with focus on sparing the lower income households. The tariff structure for residential consumers in Jordan is divided in six cost brackets based on their monthly consumption levels. In such a setup, tariff revenues can be raised either by increasing tariffs on the existing cost brackets or by increasing the number of cost brackets and setting tariffs on each. Simulations of cost recovery are developed below based on these alternative scenarios using 2010 Household Expenditure and Income Survey (HEIS) data. 47 HEIS has collected electricity and overall consumption information from a sample of 11,223 households across Jordan. 48 The extent to which revenues increase 46 IMF, “Welfare Impact of Price Subsidy Reform in Jordan� (Washington, DC, September 2011); World Bank, Electricity Subsidy Note,� 2011. 47 Jordan Department of Statistics: Household Expenditures and Income Survey 2010 (HEIS) (GoJ, Amman), http://www.dos.gov.jo/dos_home_e/main/index.htm 48 Results are based on simulations done in World Bank, Electricity Subsidy Note, 2011. 44 would depend not only on the tariffs that are set but also on how households adjust their demand in response to increased tariffs. 92. Tariff increases would be capable to raise 0.2 percent–1.4 percent of GDP per year in extra revenues, without hurting the poor. The simulation results suggest that, depending on the underlying assumptions, increases would help cover JD4.2 million–JD27 million of the overall monthly electricity subsidies. Adjusting the bracket structure appears more effective in raising revenues than increasing tariff rates without modifying the brackets. However, while residential users can be an important source of cost recovery, long-term cost recovery will necessitate taking a mix of measures, including exploring alternative energy sources. 5.1.1 Residential Electricity Usage Patterns under Existing Tariffs 93. Electricity subsidies reflect the differences in the costs of producing, transmitting, and distributing electricity above the cost-recovery tariff from users. In Jordan, all of these excess costs are borne by the publicly owned NEPCO. The latter currently runs at an average monthly deficit of approximately JD100 million. 49 Effectively, NEPCO’s losses amount to the total electricity subsidies. Estimates based on data from HEIS 2010 show that households were spending an estimated JD23 million per month on electricity in 2011. Since households accounted for approximately 35 percent of all electricity consumption, Jordan’s households can be inferred to receive approximately JD35 million per month in subsidies and are responsible for 2.2 percent of GDP’s worth of NEPCO’s losses. Households are paying less than half the price needed for full cost recovery from them. In a simple sense, if households were not subsidized and were to meet the true costs of their electricity use, electricity revenues from households would increase by approximately 150 percent. Undoubtedly, this (theoretical) increase would be very large, underscoring the extent of financial imbalances in Jordan’s electricity sector. 94. In considering how revenues can be raised from their present levels, it is first important to understand the nature of household electricity expenditures under the current tariff structure. Following the revisions in June 2012, households face 7 tariff brackets that increase with monthly consumption levels. Tariffs range from JD0.033/kWh for the lowest bracket (1–160 kWh/month) to JD0.235/kWh for the highest bracket (1000+ kWh/month) (Table 5.1). The vast majority of households–– an estimated 85 percent––consumes 161 kWh–500 kWh per month. Only 2.0 percent pay the lowest bracket, and an even smaller 0.4 percent pays the highest bracket. The former brackets and tariffs prior to the most recent tariff increase are available in appendix 3. 49 Deficits are due predominantly to disruptions to the gas pipeline that has transported natural gas from Egypt to Jordan since early 2011. In the absence of this gas supply, NEPCO has been forced to import costlier fuel oil for electricity consumption. For a thorough analysis of NEPCO’s deficit and debt, refer to chapter 4. 45 Table 5.1 Tariff Brackets and Household Consumption of Electricity Mean monthly electricity Bracket kWh Tariff by bracket (JD) consumption (JD) Share of HH (%) 1 1–160 0.033 4.5 2.00 2 161–300 0.072 11.0 40.20 3 301–500 0.086 21.8 45.10 4 501–600 0.114 37.5 7.30 5 601–750 0.141 51.1 3.46 6 751–1000 0.168 73.2 1.60 7 > 1000 0.235 128.0 0.40 Source: Government decision of June 2012; World Bank, “Electricity Subsidy and Household Welfare in Jordan� using Household Expenditures and Income Survey 2010 (HEIS) (Washington, DC, December 2011); www.dos.gov.jo; NEPCO data www.nepco.com.jo. 95. Electricity expenditures are an important item for Jordanian households, in particular, for the poor. On average, electricity accounts for 2.6 percent of Jordanian households’ expenditures (Table 5.2). Poorer households dedicate a larger share of their expenditures (4.6 percent) to electricity. Not surprisingly, wealthier households consume much more electricity than poorer households, and they spend much more on electricity. The top decile of households spends an estimated JD4.8 million per month, approximately 4 times higher than the bottom decile (JD1.3 million). Table 5.2 Per Capita Expenditures by Decile Expenditure Mean total Mean electricity Electricity expenditure decile expenditure (JD) expenditure (JD) as share of budget (%) 1 252 11.7 4.6 2 378 13.8 3.7 3 461 15.2 3.3 4 535 16.4 3.1 5 610 16.9 2.8 6 696 18.8 2.7 7 803 20.7 2.6 8 954 23.0 2.4 9 1188 26.5 2.2 10 2022 39.9 2.0 All 802 20.5 2.6 Source: World Bank, Electricity Subsidy Note, 2011; www.dos.gov.jo 46 5.1.2 Residential Electricity Tariffs - Policy Reforms for Cost Recovery 96. There are 2 broad approaches of generating more revenues from households: (a) increasing tariff rates within the existing 6 tariff brackets that existed prior to June 2012 and (b) reforming the existing tariff brackets as well as adjusting tariff levels taking into consideration the decision of June 2012. 50 The first set of simulations is on expenditures quintiles. Using quintiles does not enable identifying within the upper quintile the upper user’s decile that was subject to the June 2012 decision. Hence, the June 2012 increase can be taken into consideration only in the last set of simulations that are conducted on users’ deciles. In any case, as noted earlier, the impact of the June 2012 increase remains very marginal since it increases revenues from households by only JD0.292 million per month. Increase tariffs within existing structure prior to June 2012 decision 97. If Jordanian households were to pay their burden of subsidies, tariffs would have to more than double. Full cost recovery of all electricity subsidies from residential users alone would imply increasing tariff rates even more––by 200 percent to 500 percent (depending on the assumptions). Increases of such magnitude would place enormous burdens on households and are unrealistic from a practical standpoint. Thus, this section explores the implications of more moderate, but hardly inconsequential, increases in tariffs of 30 percent–60 percent. 98. In simulating tariff increases, several assumptions need to be made regarding the effects on inflation and on the behavioral patterns of households, that is, how households would react to price increases (by reducing electricity usage and/or by switching to alternative sources of energy). Price increase is also needed to reduce the energy intensity of the economy. First, with higher tariffs, the increased electricity bill would directly impact household welfare. Moreover, to the degree that other prices rise as a result of the increase in households’ electricity prices, households would be indirectly impacted. However, the World Bank 51 estimated the indirect impact of electricity price increases in Jordan to be small. 52 Hence, in the ensuing analysis, only the direct inflation impact is considered. 99. Second, to account for how households might react to increased tariffs, assumptions need to be made about the demand elasticity of electricity. Under the assumption of zero elasticity, household demand would remain unchanged if tariffs were imposed, and households would continue to consume the same amount of electricity as before the increase. This scenario is somewhat remote, although for an energy item such as electricity, demand elasticity arguably would be low. Elasticity estimates are difficult to compute because the data requirements are demanding. However, elasticity estimates from other studies can provide some guidance as to the likely magnitude in Jordan. For instance, recent studies 50 Methodology and results here are based on the analysis carried out in World Bank, Electricity Subsidy Note, 2011. Additional details are available in that study. 51 World Bank, Electricity Subsidy Note, 2011. 52 Simulations based on an input-output approach suggest that a 30 percent increase in tariff for all brackets would increase production prices by 0.62 percent on average, whereas a 60 percent increase would increase production prices by 1.24 percent on average 47 suggest that demand elasticity for electricity in the United States ranges from -0.2 to -0.3, 53 and in India from -0.3 to -0.4. 54 In light of all of these findings, the present analysis estimates the impacts of tariff increases first by assuming elasticity to be zero and then by assuming elasticity to be -0.3 or -0.6. Scenario 1. Increase in tariffs with an assumption of zero price elasticity 100. Table 5.3 presents the results of increasing tariffs by 30 percent–60 percent for all tariffs brackets except the lowest (0–160 kWh/month), which originally was designed to protect the poor from paying high electricity prices. The simulations suggests that a 30 percent increase in tariffs under the assumption of zero elasticity (that is, no reduction in electricity use) and no adjustments for inflation would increase household electricity expenditures by less than 30 percent because the lowest bracket is untouched. The increases are moderately higher for the relatively wealthier households. Revenues are estimated to increase by JD5.2 million per month, approximately 5 percent of NEPCO’s monthly deficit. With the same assumptions, if tariffs were to increase by 60 percent, revenues would increase by approximately JD10.4 million, approximately 10 percent of NEPCO’s monthly deficit. Table 5.3 Increases in Total Revenues (with zero elasticity) Expenditure quintile Increase tariffs 30% Increase tariffs 60% 1 0.62 1.25 2 0.77 1.54 3 0.91 1.84 4 1.13 2.25 5 1.74 3.52 All 5.19 10.4 Source: World Bank, Electricity Subsidy Note, 2011. Scenario 2. Increase in tariffs with different price elasticity assumptions 101. Assuming that households make behavioral adjustments to electricity usage following tariff increases would change the outlook. As households respond to higher prices by consuming less, the revenues from higher tariffs decline compared to the case of zero price elasticity. If elasticity is assumed to be -0.3, a 30 percent tariff increase is projected to increase monthly revenues by approximately JD4.2 million. Of course, the savings in subsidies would be different from these revenue figures since there also would be savings from reduced electricity usage. 102. The scenarios below suggest that the deficit would be reduced, although not by enough to cover the households’ shares of the subsidy burden. Although households increase their electricity expenditures considerably under any of the above tariff increases, poverty is not projected to increase by 53 M.A. Bernstein and J. Griffin, “Regional Differences in the Price-Elasticity of Demand for Energy,� National Renewable Energy Laboratory, Subcontract Report NREL/SR-920-39512, RAND Corporation (Santa Monica, CA, February 2006). 54 S. Pachauri and M. Filippini, “Elasticities of Electricity Demand in Urban Indian Households.� Energy Policy 32 (2004): 429–36. 48 much. For example, with an elasticity of -0.3, if tariffs increase by 30 percent and 60 percent, the poverty rate is estimated to increase by 0.32 percent and 1.45 percent, respectively. With an elasticity of -0.6, if tariffs were to increase by 30 percent and 60 percent, poverty would increase by 1.45 percent and 4.05 percent, respectively. Table 5.4 Impact on Revenues of Increases with an Elasticity of -0.3 Expenditure quintile Increase tariffs 30% Increase tariffs 60% 1 0.30 0.51 2 0.48 0.81 3 0.68 1.17 4 0.97 1.65 5 1.81 3.08 All 4.24 7.22 Source: World Bank, Electricity Subsidy Note, 2011. Scenarios with Reformed Tariff Structure after June 2012 Decision 103. An alternative to increasing tariffs on the existing brackets is to increase the number of brackets. Prices then can be set to better mirror peoples’ actual willingness to pay for electricity. Although, in practice, increasing the number of brackets would reduce the level of “consumer surplus,� the added brackets would increase electricity revenues more substantially. For a simulation based on such a design, the number of tariff brackets is increased from 6 to 10. These 10 brackets can be selected such that there are an equal number of individuals in each bracket (based on electricity usage statistics). The tariffs for the first two are kept unchanged. Except for the first two brackets, tariffs are increased according to progressive rates. The objective of additional savings is set in accordance with the JD280 to 300 million savings agreed between the Government and the IMF. To reach this objective, the progressive rates are set in a range from 23 to 41 percent and are increased by increments of 3 percentage points. To account for the increase of June 2012, JD0.337 million were added to the base case scenario. Simulations then were conducted under the assumption of zero elasticity and no inflation. 104. These simulations provide an upper bound of the possible revenue generation under such a reform. Simulation results are presented in Table 5.5. The consumption for 2012 is projected to increase at the average rate observed between 2008 and 2011: 8 percent per year. The population is now distributed evenly across brackets, and the total tariffs paid by households roughly doubles from JD25.2 million to JD51.8 million per month, generating additional savings of JD318.5 million. With the increased tariffs, the marginal and average prices paid by all users, including the poor, increase. Indeed, while deciles include the same number of individuals, they have a different number of households since household size is larger in the poorest deciles and smaller for the richest deciles. Hence, due to their large size, some poor households in the lowest two deciles would consume in the upper tariff brackets and would be hit by the tariff increase despite the exemption of the two first brackets. The Government may provide cash transfer to households in the two poorest deciles and totally shelter them from the impact of tariff increases. This cash transfer can be granted through the electricity bill. The cost would not exceed JD13.5 million per year, hence leaving a net saving of JD305 million (1.4 percent of GDP) for NEPCO. At the end, the larger users of electricity pay a disproportionately larger share of the increase than the 49 smaller users, making this design attractive on equity grounds. This illustration is a possible scenario for reforming the tariff structure. While its feasibility remains tributary of political considerations, the design offers an attractive framework for thinking about reform. Table 5.5 Increases in Revenues from Reformed Tariff Structure Current Tariff Characteristics Structure Reformed Tariff Structure Total Bracket electricity Total electricity Total electricity Revenue Decile thresholds consumption expenditures expenditures increase (kWh) (kWh) (mil kWh) (mil JD) (mil JD) (mil JD) 1 160 26 1.25 1.69 0.44 2 241 29 1.49 2.16 0.68 3 271 31 1.64 2.35 0.71 4 300 34 1.80 2.69 0.89 5 326 33 1.94 3.28 1.34 6 356 38 2.13 3.77 1.64 7 393 42 2.46 4.55 2.09 8 447 47 2.87 5.76 2.89 9 536 58 3.52 8.00 4.48 10 Maximum 84 6.15 17.54 11.38 All 421 25.24 51.78 26.55 Source: 2012 projected based on World Bank, Electricity Subsidy Note, 2011. 105. This analysis should be considered a framework for thinking about cost recovery in the electricity sector. The discussion suggests that increasing tariffs could raise revenues to reduce the deficit for which the households are responsible but cannot fully make up for it. Indeed, under the increased number of brackets scenario, JD26.6 million of additional monthly revenues would be generated, leaving a deficit related to household consumption of JD14.2 million. However, reforming the tariff structure by increasing the number of brackets as well as raising tariffs offers more promise for cost recovery. At the same time, it also is a more equitable approach to raising revenues. However, cost recovery through increasing tariffs can be only a part of the broader cost recovery strategy. Options of Reform of Electricity Tariffs for Household Consumers 106. The Government of Jordan may wish to consider scenarios for increasing tariffs of electricity to households along the following lines: a. Privilege the redesign of consumption brackets according to users’ deciles b. Set tariffs so that only consumers in the lower five deciles continue to be subsidized c. Ensure that 65 percent (JD320 million in 2012 on an annual basis) of losses imputed to households are recovered through the new tariff structure d. Subsidize the remaining 35 percent (JD170 million in 2012 on an annual basis) through other users or through the budget (chapter 4). 50 5.2. Food Subsidies 107. Food subsidies are estimated to have contributed significantly to Jordan’s 2011 deficit. In 2010 the country spent an estimated JD103 million on food subsidies. In 2011 that amount was estimated to have increased to JD218 million. For 2012, food subsidies are expected to reach JD185 million, equivalent to 0.8 percent of GDP and 7.8 percent of Government deficit. Jordan’s food subsidies included those on the imported wheat and barley used as animal feed. The wheat subsidy is absorbed by the Ministry of Industry and Trade, which sells imported wheat to mills at subsidized prices. GoJ also sets prices for certain types of flour and bread so that customers receive the subsidy. Subsidies on imported barley used as animal feed are provided to farmers registered with the Ministry of Agriculture. While this barley is not consumed directly by households, its removal could increase the price of domestically produced meat. 55 108. While existing food subsidies are important to the well-being of poor households, it turns that more than 80 percent of the total subsidies go to the non-poor households.. Estimates from the Household Expenditure and Income Survey of 2010 (HEIS) suggest that, poorest households in the bottom decile, spend 3.6 percent of their expenditures on subsidized food items, whereas for the richest top 3 deciles, the share is below 1.0 percent. But the poorest two deciles receive only 14 percent of total food subsidy. Hence, untargeted food subsidies are costly and inefficient. Table 5.6 presents simulations of the effects of subsidy cuts across Jordanian households. These simulations are based on the assumption that changes in subsidies would be transmitted fully to households. The analysis also assumes no change in household consumption patterns as a result of subsidy removal, in other words, that demand elasticity for the subsidized food items would be zero. Eliminating these subsidies completely, with no compensating measures, would increase the poverty headcount by approximately 2 percentage points and severely affect the poorest 2 decile households. 55 World Bank, Jordan Subsidies Note, 2011. 51 Figure 5.1 Household Expenditures on Subsidized Food Items and Household Receipts of Subsidies as a Share (%) of Their Budgets (by per Capita Expenditure Decile) 6 Expenditure share (%) Subsidy incidence (%) 5 4 3 2 1 0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 All Source: Calculations based on HEIS 2010 109. Substantial savings will be generated by reducing or eliminating subsidies and moving to targeted subsidies that protect the poorer segments of the population. Simulations suggest that, to have recovered 40 percent of the cost of the food subsidies in 2011, existing prices of subsidized items would have had to increase by approximately 50 percent. For full cost recovery, households would have to face approximately 130 percent higher prices. The estimated annual cost recovery from such increases would range from JD71 million to the total amount of subsidies projected for 2012, that is, JD185 million, or 0.8 percent of GDP (4.1). While such price increases are considerable, approximately 75 percent of cost recovery would be coming from the top 7 expenditure deciles. Thus, an amount of approximately 0.2 percent of GDP of savings could be diverted to protect the poorer segments of the population. 52 Table 5.6 Simulations of Cost Recovery from Subsidy Cuts Cuts in subsidies (JD/per year) Proportion of Increase Increase prices Increase prices impact (%) prices by 50% by 100% by 130% Decile 1 5.1 10.2 13.2 7.1 Decile 2 5.9 11.8 15.3 8.3 Decile 3 6.6 13.2 17.1 9.2 Decile 4 7.0 14.0 18.1 9.8 Decile 5 7.5 15.0 19.4 10.5 Decile 6 7.7 15.4 20.0 10.7 Decile 7 8.0 16.0 20.7 11.2 Decile 8 8.2 16.4 21.2 11.5 Decile 9 8.1 16.2 21.0 11.3 Decile 10 7.3 14.6 18.9 10.2 Population 71.4 142.8 185 100.0 Source: Calculations from HEIS 2010. Options for Reform of Food Subsidies 110. The Government of Jordan may wish to consider to immediately moving away from subsidizing food item. Therefore, the Government could: a. Remove 75 percent of the food subsidy, across the board, expected at JD185 million for 2012. This action would save the budget JD139 million and 0.6 percent of GDP annually (JD35 million and 0.16 percent of GDP for the last quarter), b. Reallocate the remaining 25 percent of the subsidy (JD46 million and 0.2 percent of GDP annually in 2012) to support the poorest segments of the population. The reallocation could occur along the lines of an efficient targeting system using the databases of a reformed NAF or other institutions. 53 6. Reducing the Deficits of Own Budget Agencies: the Case of the Water Sector 111. The deficit of autonomous public companies and entities has widened, rising to 5.9 percent of GDP. Of this amount, the water sector deficit exceeded 1.2 percent. In this chapter, we tackle the inefficiencies and fiscal waste from the water sector. The chapter proposes options to enhance revenues and cut the losses in the sector. These options would generate savings of 0.2 percent of GDP for the budget in 2012 on a yearly basis (0.05 percent in the last quarter) and would close the 1.3 percent of GDP deficit of autonomous water companies (or reduce it by 0.35 percent in the last quarter). In the following, we will first summarize the characteristics of the sector in Section 6.1. Section 6.2 provides a detailed analysis of the main sources of waste in the sector and Section 6.3 presents options of reforms and potential savings. 6.1. Characteristics of the Water Sector in Jordan 112. The water sector is indeed one of the major contributors to the overall deficit of the public sector in Jordan. Spending on water is high, fragmented, and inefficient. Jordan is one of the most water-poor countries in the world, which has used almost all of its available water resources and is increasingly dependent on non-conventional water resources that tend to be expensive. Access to improved water supply and sanitation services 56 is high: 98 percent of the population having access to improved water supply and 56 percent to piped sewerage facilities. Yet, this performance comes at a very high fiscal cost. The net cost of Jordan’s water sector in the 2012 budget is expected to amount to JD333 million, equivalent to 1.5 percent of GDP and to 8.5 percent of the overall deficit of the public sector.57 Expenditures are expected at JD527 million, while revenues stand at JD195 million. 58 113. Several public entities govern and operate the sector. These are the Water Authority of Jordan (WAJ), Jordan Water Company (Miyahuna), Aqaba Water Company (AWC), Yarmuk Water Company (YWC), and the Ministry of Water and Irrigation (MoWI), which includes the Jordan Valley Authority (JVA). WAJ is an own-budget agency that manages bulk water supply and retail distribution for locations in which commercialization of distribution services has not occurred. Miyhauna and AWC are own- budget agencies that provide retail distribution and water and wastewater treatment. YWC services Jarash, Ajloun, Mafraq, and Irbid. JVA comes under the budget and depends directly on MoWI. JVA provides essential water for irrigation in Jordan Valley. 114. The sector benefits from large subsidy transfers. A large part of these subsidies are used to fund the capital investment program in the water sector. In addition, underpricing of water services, 56 Ninety-seven percent of the population had improved water access (World Bank, World Development Indicators 2012) in 2010 and compares to a world average of 88.3 percent. 57 World Bank staff calculations based on 2012 Budget Law and own-budget agencies budgets, available on the website of the General Budget Directorate, www.gbd.gov.jo. Of this overall number, approximately JD44 million (0.2 percent of GDP) would appear directly in the budget as the sum of the deficit (JD29.3 million) of JVA, which is directly under MoWI, plus GoJ transfers to own-budget water authorities (JD14.7 million). 58 These revenues do not include any donor grants, but only refers to revenues generated from user charges. 54 inefficiencies in revenue collection and non-revenue water losses provide consumers with additional (implicit) subsidies. As a result, critical maintenance and investment are deferred, with current taxpayers and water users implicitly “borrowing� from the future. It is important to unbundle the source of inefficiencies to determine which ones are the most costly and hence make the most sense to tackle. 6.2. Sources of Waste and Deficits in the Water Sector 115. The main drivers behind the sector deficit are inefficiencies in revenue collection, high operation and maintenance costs, affordability concerns and underpricing of water services. Closing the financing gap in the sector is likely to be a longer-term process in which the appropriate mix of instruments will change over time. 116. Low collection rates. Collection efficiencies vary widely between the different sector agencies but they can be significantly improved upon. In the case of JVA, for instance, the complete separation of expenses and revenues leaves the agency with few incentives to collect revenues, and only about 75 percent of the revenues were actually collected in 2010. The high dependency on subsidies may provide little incentive to water providers to collect revenues, while it does not provide users with strong incentives to pay their bills on time. 117. Inefficient service provision translating in high operational costs. Staff productivity in utilities and JVA is low compared to benchmarks used globally. Non-revenue water losses are also a source of inefficiency. 59 USAID estimated that the 2010 revenue loss was JD170 million (0.9 percent of GDP). The benefits from NRW reduction are obvious, but such programs require strong institutional capacity and substantial financial resources. Reducing NRW is not only a technical issue. If water tariffs are too low, the costs of reducing NRW may exceed the benefits of saving water, as any water “saved� can only be sold below costs. 118. Inefficiencies in the sector’s capital investment programs. An assessment of the capital budget shows that the total budget of the water sector includes a very high level of current expenditure. In 2010, 50 percent of the total water budget was made up of current expenditure, increasing to 53 percent in the 2012 budget. The high level of recurrent costs suggests that investment is essentially crowded out by the high cost of operating and maintaining the current systems. 119. Inefficient and inequitable subsidies. To overcome the concerns of the affordability of water services, it is important that subsidies are properly targeted. A recent World Bank report highlights the inefficiencies of the water tariff system in Jordan from a distributional perspective. 60 Utilizing data from the Household Expenditure and Income Survey (HEIS) of 2008, the study shows that rich households receive almost twice the amount of subsidies as do poor households. For example, the top decile receives approximately 21 percent of the total water subsidies to households, whereas the poorest decile only 6 percent (Figure 2.7). While individuals in the poorest decile receive an estimated average transfer of 59 USAID, Jordan Public Expenditure Perspectives (PEP), USAID-Funded Fiscal Reform II Project (Amman, December 2011), www.frp2.org 60 World Bank, “Hashemite Kingdom of Jordan: An Analysis of Consumption Subsidies� (Washington, DC, July 2011). Draft delivered to the Ministry of Planning and International Cooperation. 55 JD21 per year, those in the richest decile receive more than three times as much, or approximately JD69.4. 61 Inefficiencies also extend to irrigation tariffs, where prices are set far below operation and maintenance costs, and are combined with quota allocations62 that tend to favor crops that require higher water consumption. Figure 6.1 Water Subsidies and Transfers to Households (%) Distribution of water subsidies by deciles of per capita Average annual per capita transfer value (JD) of water consumption subsidies by deciles of per capita consumption Series1, 80 D10, 69.4 20.6% 60 44.2 38.3 40 32.5 33.7 30.1 26.7 27.7 Series1, 22.7 24.7 21 D1, 6.2% 20 0 D1 D2 D3 D4 D5 D6 D7 D8 D9 D10 All Source: World Bank staff calculations; World Bank 2011. 120. Water tariffs are set below O&M cost recovery levels. Water tariffs are in general set below the level of operation and maintenance costs. In addition, tariffs are not adjusted on a regular basis. Irrigation water rates have not been adjusted since 1994, corresponding to 50 percent erosion in real prices according to the dynamic of the GDP deflator. In the case of water supply and sewerage services, rate increases have also been rather infrequent. Prices need to keep up with inflation to avoid against erosion of the revenue base of a water agency, but also to ensure that consumers do not receive any perverse incentive to consume more water. The contractual agreement between the water distribution companies and public companies ensure that these distribution companies are profitable. However, only the Aqaba Water Company generates an operating surplus, whereas the other two utilities still generate operating deficits that require government transfers. WAJ which is managing the remainder of the utilities also undertakes investments and accumulates debt. In 2010, with the execution of the Disi water project, the debt of WAJ alone reached JD700 million 63 (3.7 percent of GDP) and is estimated to have reached JD924 61 Domestic water consumption is priced on an increasing block tariff (IBT) basis, in which the first block is set at a low unit price to subsidize the poor. However, consumption is based on the monthly consumption per meter, not per individual or household. Hence, the correlation between water use and income remains weak. In fact, an average of 20 percent of households in Jordan use the same meter, and this average could go up to 39 percent in governorates such as Zarqa. Poor households, who tend be larger in size and share meters among different households, consume larger amounts of water. Hence, they end up in the upper tariff blocks. Thus, the subsidy will be benefitting small households with higher incomes and individual meters (USAID, Jordan PEP, 2011). 62 The current annual quotas correspond to 3,600 cum/ha for vegetables, 7,650 cum per ha for citrus, and 12,550 cum per ha for bananas. 63 USAID, Jordan PEP, 2011. 56 million (4.2 percent of GDP) in 2011. This debt raises concerns over the Government’s liabilities.64 GoJ is guaranteeing this debt and, in the absence of any tariff adjustment, ultimately would have to assume this debt. 6.3. Options for Reform and Potential Savings 121. The following measures can be adopted immediately to move the sector to higher levels of cost recovery that ensure the long-term financial sustainability of the water services. Although tariff increases are necessary, improved sector governance, better management of public funds and efficiency improvements in the different water agencies will need to be undertaken too. a. Reducing inefficiencies to lower the cost of water by determining which inefficiencies can be tackled most effectively to reduce the cost of service. b. Improving investment efficiency. Investment efficiency can be improved by (i) improving sector and investment planning, while screening investments to ensure that their benefits outweigh costs; (ii) improve the capacity to procure, disburse, audit and monitor sector resources; and (iii) maintain a sharper focus on providing incentives in the allocation of investment funds. c. Increase the revenues of the utilities: x Before embarking on tariff increases revenue-generating agencies such as JVA and the water utilities need to increase their billing and collection efficiencies, which will decrease the need for tariff increases; x Ensure that water tariffs are adjusted for inflation on an annual basis to ensure that water tariffs are not systematically eroded, and tariffs do not provide consumers with additional incentives to increase consumption in an already very water-constraint environment; x Review the tariff structure for household consumption to eliminate subsidies for middle- income and rich consumers. The increasing dependence on bottled water for drinking can be taken as a proxy for ability to pay for water services65 with almost 32 percent of the population in 2009 using bottled water for drinking water purposes; x Administering water subsidies to the poor outside the tariffs through existing or new targeting mechanisms; x Ensure that non-residential water users pay tariffs at cost recovery levels. d. Increase the revenues of JVA 64 Per its accounting regulations, GoJ records all guarantees provided for state-owned enterprises (SOEs) as explicit debt. However, not all WAJ debt is government guaranteed. SOE can independently raise debt from domestic private creditors, mostly commercial banks, against its balance sheet. Yet, all foreign debt is borrowed directly by the Government and transferred to WAJ. Risks arise if WAJ goes bankrupt or its deficit increases to a level at which it can no longer access commercial loans to sustain its operations. In these cases, the Government will have to intervene by providing either direct cash transfers or sovereign debt guarantees, as it did recently to the electricity utility, NEPCO. 65 According to the findings of USAID, while Miyahuna’s average tariff was JD0.85 per m3, water sold via private water trucks was priced 4.5 times higher at JD3.8 per m3. The higher price can be taken as a proxy for the willingness to pay. Source: USAID, Jordan PEP, 2011. 57 x Raising irrigation water tariffs for at least medium- and large-quantity users 66 x Rationalize the quota allocations; x JVA also provides water to industries and the Water Authority of Jordan: x In the case of WAJ, the tariff should at least cover basic operation and maintenance costs for that service x Industrial water users should at least pay the agreed water rates. These measures would incentivize more efficient water use in the agriculture sector, enhance its economic returns, and reduce waste. The measures would eliminate the 0.2 percent of GDP fiscal burden of the sector that is supported by the budget and eliminate the 1.3 percent deficit of autonomous water companies. The immediate fiscal impact and the longer term efficiency outcome could be enhanced through (a) creating an incentive mechanism for farmers who reduce or reallocate water quotas, and (b) stopping allocations of land rights to people with illegal wells and increasing penalties for offenders. 66 Here again, some figures available from USAID show that the willingness to pay may exist: whereas JVA sells water to small farmers at JD0.012 per m3, these farmers pay 17 times more, at JD0.21 per m3, for the water they desalinate. Source: USAID, Jordan PEP, 2011. 58 7. Conclusion 122. The success of any reform plan hinges on the Government’s ability to maximize the political, social, and economic returns from its chosen measures. This maxim is particularly true in the case of Jordan. The country is in a region rocked by political turmoil and instability and is itself going through a difficult and hesitant political transition. In such a context, politically sensitive measures of fiscal consolidation can become even more delicate to implement. However, not all measures are of the same degree of sensitivity. In addition, political sensitivity must be assessed against the tradeoff between appeasing certain segments of the population and risking a major fiscal and financial crisis that could generate pain and suffering for the majority of Jordanians. Box 7.1 Six proposed Criteria for the Selection of Short-Term Policy Measures Popular demands and political pressure in times of economic downturn are not always conducive to reforms, especially those related to fiscal adjustment. Therefore, the Government of Jordan needs to prioritize its options for fiscal savings. This note proposes six criteria to evaluate and select short-term fiscal policy adjustment measures. These criteria are intended to cover a range of economic, political, and fiscal concerns. The six criteria and their functions are described below: 1. Magnitude: Examines whether the measure on spending or foregone revenue generates substantial savings to the budget, 2. Efficiency: Examines whether the measure on the spending or foregone revenue item improves the use of resources and reduces the crowding out resources from other vital alternatives, or the distortive use of natural, human, or financial resources (moral hazard). 3. Distributional Impact: Examines the beneficiaries of a particular item and whether the measure aiming to reduce spending or the increase revenues is progressive or regressive, 4. Dynamic: Observes the rate to which expenditure or tax exemption/arrear has expanded in recent past (5 years for example), 5. Poverty and social impact: Evaluates whether the measure on spending or forgone revenue has a neutral or positive poverty and social impact or not, 6. Impact on future growth: Examines whether containing or reducing spending/tax exemptions and arrears has a neutral or positive impact on growth. The framework then calculates a weighted rating for each measure and ranks these reforms by highest returns. 123. This note proposes a framework to assist the Government in prioritizing policy options to achieve the deficit target. The framework offers a set of six criteria to select the priority short-term expenditures and revenue measures. These criteria take into account the (a) magnitude of savings generated by the measure, (b) distortions in the use of resources created by the item under scrutiny, (c) growth pattern of the item and whether it has been steadily increasing or decreasing in the past few years; and (d) extent to which proposed measures affect beneficiaries regressively or a progressively (box 7.1). Additionally, the framework calls for taking into account the impact of the planned cuts (e) on poverty and social equity, and (f) on future growth prospects. Proposed cuts would be ranked according to these 59 criteria, and the most relevant cuts would be executed first. The weighting of these criteria is not binding and could be modified by decision-makers based on informed judgment (Table 7.1). 124. Beyond the immediate concerns of fiscal adjustments, to restore long-term fiscal sustainability, Jordan has to explore all possible policy options. While short-term fiscal consolidation has to be implemented at the level of the budget, fiscal sustainability has to be restored at the level of the public sector in general and for the long term. This restoration requires implementing medium- to long-term measures to deal with the deep structural causes of Jordan’s fiscal crisis. This report details various options for short- to medium- term fiscal adjustments, including measures that GoJ already is implementing. The report also explores possible options for medium- to long-term fiscal consolidation. All such short- and medium-term reforms are listed in the Matrix of Policy Objectives and Actions accompanied by policy objectives, actions to reach these objectives, and time horizons (Table 7.2). The purpose of this matrix is to serve as a tool for reflection and support for decision-making. The matrix thus can be modified as needed to fulfill these tasks. 60 Table 7.1 Framework for the Selection of Policy Actions Savings - 2013 Criteria** Additional Measures Suggested for the Budget Ratio to Poverty and Impact on Rating over Amount JD Magnitude Efficiency Distributional Dynamic GDP Social Impact Future Growth 100 Freezing Nominal Wage (Civil 60 0.25% 0 1 0 1 1 1 40 Servants and Military)* Freezing Hiring (Civil Servants 66 0.27% 0 1 0 1 1 1 40 and Military)* Wage Bill Cuts on 2011 Civil Servants 156 0.65% 1 1 1 1 1 1 100 Allowances Freeze land acquisition * 36 0.15% 0 0 0 0 1 0 10 Freeze investment spending Capex 29 0.12% 0 1 1 1 1 1 70 except "economic affairs"* Freeze public pensions 200 0.83% 1 0 0 1 0 1 50 Pension Losses of JVA and Transfers to Autonomous Water Companies 44 0.18% 0 1 1 0 1 1 60 Eliminated Adjusting Water Tariffs Removing 75% of Subsidy and 169 0.70% 1 1 0 1 0 1 60 Enhancing Targeted Transfers Food Subsidy Making Sure that no New Arrear Accumulation Occur During the 200 0.83% 1 1 1 1 1 1 100 Year the Year Zero Tax Arrears in 61 Savings - 2013 Criteria** Additional Measures Suggested for the Budget Ratio to Poverty and Impact on Rating over Amount JD Magnitude Efficiency Distributional Dynamic GDP Social Impact Future Growth 100 Recovering Accumulated Arrears 213 0.89% 1 0 1 0 1 1 80 Over a Period of 8 Years Arrears Recovering Removing the GST and Customs Exemptions and Cuts Granted in 83 0.34% 1 1 1 0 1 1 90 2008* Exemptions Removing GST Total 1,255 5.23% Weights 30% 10% 30% 10% 10% 10% GDP 23,998 Note: * Measures that are being partially or totally implemented by the Government, we are suggesting here to reinforce them. ** We put 1 if the criteria is satisfied, 0 otherwise, for example, preventing new tax arrears accumulations would generate a lot of saving, improve the efficiency of the use of resources (through reduced financing of raising deficits), burden the few (those who practice tax evasion), reverse an upward dynamic witnessed in the previous years, has a rather neutral or positive poverty and social impact, and a neutral or positive impact on future growth. The measure's rating is 100. Hence, this measure is politically, fiscally, and economically a best candidate and could be implemented with no delays. 62 Table 7.2 Matrix of Policy Objectives and Actions Time Reforms Policy objective Action horizon I. Public Spending Wage bill Contain expansion of public sector wage bill to generate Freeze nominal wages (civil servants and military) ST* expenditure savings Freeze hiring (civil servants and military) ST Cut allowances ST Capital spending Improve quality of public investment spending under Freeze land acquisitions** ST fiscal austerity conditions Freeze capital spending except on “economic affairs�** ST Water sector Remove net cost of sector on budget through: Review tariff structure for household consumption MT* (a) Increasing revenues of water utilities to at least cost recovery Administer water subsidies for poor through targeting, not through tariffs MT Review pricing of nonresidential consumption such as public sector consumption MT Review O&M spending as to reduce losses from physical capital deterioration MT (b) Reduce nonrevenue water Increase wholesale water prices to incentivize public and private distribution companies to reduce leakages and improve collections MT Raise agriculture water tariffs for medium and large users MT (c) Reduce agriculture sector subsidies (explicit and Introducing a tariff premium for farmers with high water-consumption implicit) crop quotas such as bananas and citrus MT Social assistance Generate efficiency gains from social assistance Use better targeting mechanism (for example, proxy means-testing) ST-MT programs and enhance transparency of these accounts Eliminate cross-transfers among social programs and from social programs to other institutions ST-MT Public pensions Contain expansion of public sector pension to generate Freeze public pensions ST expenditure savings 63 Time Reforms Policy objective Action horizon II. Energy and Power Sectors Electricity sector Eliminate ongoing deficit of NEPCO and recover Gradually increase electricity tariffs to cost-recovery levels for ST-MT accumulated losses from 2010 and 2011 Governmental customers first and then for households and other sectors** Introduce a retroactive fuel surcharge from 2010 and 2011––first on ST-MT public sector, then on households and other consumers––that is proportional to their share of consumption Shield electricity transmission and system operator Consider separating NEPCO’s electricity transmission/power system MT business from negative spillover from electricity/fuel operator business from its electricity and fuel trading business trading business Energy sector Reduce energy subsidy Resume application of monthly formula for gasoline and diesel products ST and consider applying it on a reduced time basis such as weekly** Replace blanket subsidy on LPG and Kerosene for all users with a targeted subsidy scheme covering 15 to 35 percent of demand coming ST from more vulnerable groups Establish an emergency fund to safeguard for future energy price shocks Return to no-subsidy policy in sector to encourage LT efficient use of energy III. Household Electricity and Food Subsidies Household Generate additional revenues from Household Increase tariffs within existing brackets for residential consumers** ST-MT electricity subsidy electricity consumption Reform electricity tariffs structure by increasing number of brackets ST-MT Food subsidies Generate savings by reducing or eliminating food Adopt targeted food subsidy mechanism for wheat and barley ST-MT subsidies and moving toward targeted system that protects poor segments of population IV. Domestic Revenues Increase taxes on Increase domestic revenues over very short term Increase resident fees for non-Jordanians** ST selected goods and services Increase in money transfer fees** ST Increase tax on cars** ST 64 Time Reforms Policy objective Action horizon Introduce sales fee on mobile phones and air-conditioners** ST Remove exemptions on agricultural usage items** ST Increase tax on air tickets** ST Increase tax on alcohol and tobacco products** ST Stop granting any type of tax exemptions ST Suspend all types of tax credits ST Raise mine-exploring fees ST-MT Remove all exemptions from sales tax except on very limited and specific ST-MT cases based on international best practice** Structural tax Engage in comprehensive tax reform to boost domestic Stop accumulating tax revenue arrears and recover large existing stock ST-MT reforms revenues sustainably through restoring sufficiency and efficiency of existing system Revisit income tax law mainly to (a) reduce household exemption threshold, (b) increase income tax rates, (c) introduce progressivity and ST-MT tax brackets for individuals and businesses, and (d) increase taxes on interest revenues Stop forgone revenues generated by investment promotion law and review all legislation granting tax exemptions with objective of amending them MT Review and consider capital gains, property and real-estate, and interest incomes taxation Broaden tax base through taxing new resources that MT could enhance feeling of equity and fairness Source: World Bank staff. Note: * = Short-Term (ST) = up to 1 year; Medium-Term (MT) = up to 3 years. ** = Indicates actions that have been totally or partially executed or on which formal Government decisions have been taken. Refer to report for details on status of action. 65 Appendices Appendix 1. Revenues, Expenditures, and Deficits in Jordan, 2011–13 (JD mil)) 2011 2012 2013 JD mil Domestic revenues 4,199 4,663 5,264 Budget expenditures 6,797 7,047 7,505 Budget deficit -2,598 -2,384 -2,241 NEPCO deficit -1,008 -1,180 -900 Other autonomous agencies’ combined deficit -208 -380 -329 Overall public sector deficit -3,814 -3,944 -3,470 Interest payments 429 572 732 Primary budgetary balance -2,169 -1,812 -1,509 Primary public sector balance -3,385 -3,372 -2,738 Debt 14,483 17,603 19,917 GDP 20,477 22,230 23,998 Ratios to GDP (%) Budget deficit -12.7 -10.7 -9.3 NEPCO deficit -4.9 -5.3 -3.8 Other own budget agencies’ combined deficit -1.0 -1.7 -1.4 Overall public sector deficit -18.6 -17.7 -14.5 Interest payments 2.1 2.6 3.1 Debt to GDP 70.7 79.2 83.0 Primary budgetary balance -10.6 -8.2 -6.3 Primary public sector balance -16.5 -15.2 -11.4 Average cost of debt 3.9 4.2 Nominal GDP growth 8.6 8.0 Debt stabilizing primary balance -3.0 -2.8 Gap between debt stabilizing primary balance and projected primary balance -5.1 -3.5 JD mil Government measures 740 955 Budget deficit, excluding measures -3,124 -3,196 Overall public sector deficit, excluding measures -4,684 -4,425 Primary budgetary balance, excluding measures -2,552 -2,464 Primary public sector balance , excluding measures -4,112 -3,693 Debt, excluding measures 18,343 20,872 Ratios to GDP (%) Government measures 3.3 4.0 Budget deficit, excluding measures -14.1 -13.3 Overall public sector deficit, excluding measures -21.1 -18.4 Primary budgetary balance, excluding measures -11.5 -10.3 Primary public sector balance , excluding measures -18.5 -15.4 Debt, excluding measures 82.5 87.0 Gap between debt stabilizing primary balance and projected primary balance excluding -8.5 -7.5 measures Source: MoF; IMF-SBA documents; World Bank staff scenarios and calculations 66 Appendix 2. Flowchart for Implications of Exogenous Shocks in Jordan Positive Price Sbock I I I I I I Iuh.",ption, Ca sh r .. m l'.,.., Sub>i, F""" ula, u .. d, :\I«:hani,." , a doptod I Nega tive Pr ice Sb ock I s< ..".no, fOf Food ""rn Qu. stion: Would tho N~ Shock b< Suffri.." to R . Slo« Fi>ca l Sum,inabilil:a I An",,-.., Unlihly, :\1. u.""At. Not AUT 0\ L\ TIC!ILL Y RI \'I RSIIIU I Response Options I I hpendi_ Opbooo: Lim11uMiohe< ~ mooyNl • :~ • Cool! T""" r.n (liOcolly difficul, ., """'" • Tariff R. f""" oJ>ein!; ~N) : ~:::."'.':", - I I CuduioA Im pwiui SUfnCII l\CY I:»: U'm IQUITABLI Tu a tionloB ypu RI n: RSi BILIT Y I SSU[ S J»J1R'5IOf' SUST AIl'iABILIT Y 67 68 Appendix 3. Tariffs Structure in Jordan I l I l111 1"1 ,1 1 J l J J llll!!IJ!! !!!! 11 . II I'll " 11 I ~ "~J 111111111111111111.1 !!!!!lllll!11 .:j .. illlllill ililil! .iI Jil , illj!!!~J!!!!ll!!!!lll!!!!!!!! .,00$'" ~Jla .=.!!~ .=8,li' ,- · - 1 · ·