Report No. 82013 - AL Albania Public Finance Review Part I: Toward a Sustainable Fiscal Policy for Growth January 2014 Poverty Reduction and Economic Management Unit Europe and Central Asia Region Document of the World Bank CURRENCY AND EQUIVALENT UNITS Currency Unit=Albania LEK LEK 1 = 0.01 US$ US$1 = 104 (As of September 2, 2013) FISCAL YEAR January 1 – December 31 WEIGHTS AND MEASURES Metric System ACRONYMS AND ABBREVIATIONS AMM Albania Market Model IPP Independent Power Producer AMoFTS Albanian Ministry of Finance Treasury IPS Integrated Planning System System ANTP Albanian National Transport Plan KESh Albanian Power Corporation ARA Albanian Road Authority LAC Latin America and the Caribbean BAU Business-as-usual BEEPS Business Environment and Enterprise LGU Local Government Units Performance Survey BOA Bank of Albania LSMS Living Standards Measurement Survey BOOT Build Own Operate and Transfer M&E Monitoring and Evaluation CAD Current Account Deficit MDGs Millennium Development Goals CEM Country Economic Memorandum MoF Ministry of Finance CEZ CEZ Shperndarje, Power Distributor MoH Ministry of Health CFAU Country Fiduciary Assessment Update MoLSA Ministry of Labor and Social Affairs and Equal Opportunities CIT Corporate Income Tax MoPWTT Ministry of Public Works, Transport and Telecommunications CSL Civil Service Law MOTT Ministry of Transport and Telecommunications DAT Directorate of Air Transport MTBP Medium-Term Budget Plan DoPA Department of Public Administration NE Ndihme Ekonomike EAP East Asia and Pacific Region NPV Net Present Value ECA Europe and Central Asia NRW Non-Revenue Water ECB European Central Bank NSDI National Strategy for Development and Integration EIB European Investment Bank NSSED National Strategy for Social and Economic Development ERE Electricity Regulatory Authority OBL Organic Budget Law EU European Union OECD Organisation for Economic Co-operation Development FDI Foreign Direct Investment O&M Operations and Maintenance G&S Goods and Services PEFA Public Expenditure and Financial Accountability GCI Global Competitiveness Index PEIR Public Expenditure and Institutional Review GDCA General Directorate of Civil Aviation PFM Public Financial Management GDP Gross Domestic Product PFR Public Finance Review GDT General Directorate for Taxes Sewage and PIFC Public Internal Financial Control Sanitation GDWSS General Directorate for Water PISA Programme for International Student Assessment GFS Government Finance Statistics PIT Personal Income Tax GNFS Goods and Non-factor Services PPA Power Purchase Agreement GNI Gross National Income PPP Public-Private Partnership GOA Government of Albania PSAP Power Sector Action Plan GP General Practitioners QFD Quasi-Fiscal Deficit GRD General Road Director RPS Retail Public Suppliers (electricity) HH Household SEE South East Europe HII Health Insurance Institute SII Social Insurance Institute HPP Hydropower Plants SUT Sales and Use Tax HSH Albanian Railways TSA Treasury Single Account ICS Investment Climate Survey VAT Value-Added Tax IDA International Development Association WEF World Economic Forum IMF International Monetary Fund WPS Wholesale Public Supplier Vice President: Laura Tuck, ECAVP Country Director: Ellen Goldstein, ECCU4 Acting Sector Director: Roumeen Islam, ECSPE Sector Manager: Satu Kahkonen, ECSPE Task Team Leaders: Abebe Adugna, Doerte Doemeland, ECSPE CONTENTS ACKNOWLEDGEMENTS EXECUTIVE SUMMARY ......................................................................................................................................... I CHAPTER 1. ALBANIA BEFORE AND AFTER THE CRISIS ......................................................................... 1 A. GROWTH AND POVERTY REDUCTION: SIGNIFICANT STRIDES FORWARD .............................................................. 1 B. FISCAL POLICY: UNSUSTAINABLE ........................................................................................................................ 3 C. OBJECTIVES OF THE REPORT.................................................................................................................................. 4 D. HOW THE REPORT IS ORGANIZED .......................................................................................................................... 4 CHAPTER 2. FISCAL SUSTAINABILITY IN THE ABSENCE OF REFORMS ............................................. 7 A. EVOLUTION OF PUBLIC DEBT AND DEBT DYNAMICS............................................................................................. 7 B. STRUCTURE AND COMPOSITION OF ALBANIA’S DEBT ........................................................................................... 9 C. A BASELINE SCENARIO: NO-REFORM ................................................................................................................... 9 D. SENSITIVITY ANALYSIS AND CONTINGENT LIABILITIES ...................................................................................... 13 E. IS ALBANIA’S PUBLIC DEBT TOO HIGH? .............................................................................................................. 15 F. CONCLUSIONS ..................................................................................................................................................... 17 CHAPTER 3. KEY STRUCTURAL REFORMS TO ANCHOR FISCAL SUSTAINABILITY..................... 19 A. ENERGY SECTOR ................................................................................................................................................. 19 Sector Overview .................................................................................................................................................. 19 Sector Performance in 2007-2012 ...................................................................................................................... 20 Prospects and Scenarios for 2013-2020 ............................................................................................................. 22 Policy Options..................................................................................................................................................... 27 B. TAX POLICY AND ADMINISTRATION .................................................................................................................... 29 Measuring Performance: Tax Productivity and Tax Gap ................................................................................... 30 Tax Policy Options.............................................................................................................................................. 31 Tax Administration Issues ................................................................................................................................... 34 Policy Options..................................................................................................................................................... 36 C. PENSION REFORM ................................................................................................................................................ 36 Options for Pension Reform................................................................................................................................ 37 Option 1: Enhance the Contributory Scheme ..................................................................................................... 39 Option 2: Make Alleviating Poverty the Priority ................................................................................................ 41 Policy Options..................................................................................................................................................... 42 D. FISCAL SUSTAINABILITY WITH STRUCTURAL REFORMS ...................................................................................... 44 CHAPTER 4. TRANSPORT AND CAPITAL SPENDING: REBALANCING TOWARD MAINTENANCE TO SUPPORT GROWTH ........................................................................................................................................ 49 A. INTRODUCTION .................................................................................................................................................... 49 B. ROAD NETWORK AND QUALITY .......................................................................................................................... 50 C. EXPENDITURES AND REVENUES .......................................................................................................................... 51 D. FUTURE FINANCING REQUIREMENTS ................................................................................................................... 54 E. POLICY OPTIONS ................................................................................................................................................. 55 CHAPTER 5. CROSS-CUTTING: INSTITUTIONS TO REINFORCE FISCAL POLICY........................... 57 A. REFORM OF PUBLIC FINANCIAL MANAGEMENT .................................................................................................. 57 Budget Planning, Preparation, and Revision ..................................................................................................... 57 Management of Capital Investment .................................................................................................................... 58 Cash Flow Management and Treasury Operations ............................................................................................ 59 Commitment Controls ......................................................................................................................................... 61 Policy Options..................................................................................................................................................... 63 B. RULES TO ANCHOR FISCAL POLICY ..................................................................................................................... 66 Background ......................................................................................................................................................... 66 Numerical Rule or Procedural Targets?............................................................................................................. 66 Implementation ................................................................................................................................................... 69 Conclusions......................................................................................................................................................... 70 CHAPTER 6. CONCLUSIONS AND POLICY OPTIONS ................................................................................ 71 ANNEX 1. KEY FISCAL DATA ...................................................................................................................... 73 ANNEX 1. ENERGY SECTOR: INSTITUTIONAL SETUP AND MARKET MODEL ........................... 76 TABLES Table 2.1: Baseline Projections (in Percent of GDP)................................................................................................... 10 Table 2.2: Albania’s Public Debt Sustainability Framework (Percent of GDP, unless otherwise indicated) .............. 12 Table 3.1: Eastern European Countries with Progressive and Flat PIT rates .............................................................. 34 Table 3.2: Albania: Impact of alternative structural reforms on public debt (Percent of GDP) .................................. 46 Table 4.1: Public Spending on the Transport Sector (nominal € millions) .................................................................. 49 Table 4.2: Albania’s Ranking in the Global Competitiveness Index ........................................................................... 49 Table 4.3: Condition of the National Road Network (percentages)............................................................................. 51 Table 4.4: Budget versus Foreign-Financed Spending on Roads (nominal € millions) ............................................... 52 Table 4.5: Expenditures on the National Road Network (nominal € millions) ............................................................ 52 Table 4.6: Estimated Required Maintenance Expenditure (nominal € millions) ......................................................... 54 Table 4.7: Transport Sector Investment Plan (nominal € millions) ............................................................................. 55 Table 6.1: Summary of Policy Options ....................................................................................................................... 71 Table A1.1: Revenue, Expenditure, and Fiscal Deficit 2005–13 (Percent of GDP) .................................................... 73 Table A1.2: Revenue Structure, Percent of GDP ........................................................................................................ 73 Table A1.3: Current Structure of Taxation of Small and Medium Enterprises in Albania .......................................... 74 Table A1.4: Road User Charges (nominal euro millions) ........................................................................................... 75 FIGURES Figure 1.1: Real Growth and GDP per Capita (US$, right) ...........................................................................................1 Figure 1.2: Credit and Deposits .....................................................................................................................................2 Figure 1.3: Albania: Nonperforming Loans ..................................................................................................................2 Figure 1.4: Export Share by Sector (Percent) ................................................................................................................3 Figure 1.5: Income Convergence between 1998-2012 with EU Per Capita Income (US$ PPP), Percentage Point ......3 Figure 1.6: Real Growth and Fiscal Deficit ...................................................................................................................3 Figure 1.7: Public Debt (Percent of GDP) .....................................................................................................................3 Figure 2.1: Albania’s Debt-to-GDP Ratio over Time and Compared to Other Countries .............................................7 Figure 2.2: Fiscal Balance (Percent of GDP).................................................................................................................8 Figure 2.3: Public Spending (Percent of GDP)..............................................................................................................8 Figure 2.4: Holders of Domestic Debt* .........................................................................................................................9 Figure 2.5: Composition of External Debt*...................................................................................................................9 Figure 2.6: Evolution of Growth and Fiscal Balance................................................................................................... 10 Figure 2.7: Revenue (Percent of GDP)* ...................................................................................................................... 11 Figure 2.8: Debt and Gross Financing Need (Percent) ................................................................................................ 11 Figure 2.9: Albania: Sensitivity Analysis .................................................................................................................... 14 Figure 2.10: Public Debt-to-GDP Ratio with the Contingent Liability in the Energy Sector ...................................... 15 Figure 3.1: Electricity Demand vs Generation, Imports, and Load Shedding, 2007–2012 ......................................... 20 Figure 3.2: System Losses, 2001–2012 ....................................................................................................................... 21 Figure 3.3: Expected Imports to Cover the Supply Gap .............................................................................................. 23 Figure 3.4: Impact of Expected Import Costs on KESh/ WPS, Business-As-Usual Scenario, 2013–2020 ................. 24 Figure 3.5: Expected Financial Gap of the DSO, Business-As-Usual Scenario, 2013–2020 ...................................... 24 Figure 3.6: Expected Average Cost of Delivering Electricity to Final Consumers vs. Current Approved Retail Tariff; Expected Financial Gap of the Electricity Sector, Business-As-Usual Scenario, 2013–2020 ........................... 25 Figure 3.7: Expected Financial Gap of the Electricity Sector, Business-As-Usual, Normal and Aggressive Scenarios ............................................................................................................................................................................ 26 Figure 3.8: Average Cost of Delivering Electricity to Final Consumers under Different Scenarios vs. Current Approved Retail Tariff ........................................................................................................................................ 27 Figure 3.9: Total Tax Revenue, Eastern European Countries (Percent of GDP) ......................................................... 29 Figure 3.10: Albania: Composition of Tax Revenue (Average, 2008–10) .................................................................. 29 Figure 3.11: Corporate Tax Rates, Eastern European Countries (Percent) .................................................................. 33 Figure 3.12: Ratio of Average to Minimum Pensions, OECD Countries and Albania ................................................ 38 Figure 3.13: Minimum Pension and Average Wage as Percent of Average Wage, European and High-Income Countries and Albania......................................................................................................................................... 38 Figure 3.14: Projected Benefits Under Current Policy and Option 1 .......................................................................... 40 Figure 3.15: Projected Pension System Deficits, Baseline and Reform Option 1 ....................................................... 41 Figure 3.16: Projected Pension Deficits, Status Quo and Options 1 and 2 (Percent of GDP) ..................................... 42 Figure 3.17: Debt to GDP ratio under tax scenarios .................................................................................................... 44 Figure 3.18: Public-Debt to GDP ratio under Pension Scenarios ................................................................................ 45 Figure 3.19: Effect of Energy Reform on Contingent Liabilities ................................................................................ 45 Figure 3.20: Public debt trajectory under alternative structural reforms ..................................................................... 46 Figure 4.1: Spending on Roads and Total Capital Expenditure (Percent of GDP) ...................................................... 52 Figure 4.2: Spending on Maintenance (€ /km) ............................................................................................................ 53 Figure 5.1: Budget Versus Actual (Percent) ................................................................................................................ 66 Figure 5.2: Volatility of Economic Indicators ............................................................................................................. 69 Figure A1.1: Revenue as Percent of GDP by Taxes, Eastern European Countries ..................................................... 74 Figure A2.1: The Albanian Market Model .................................................................................................................. 76 BOXES Box 5.1: Numerical Rules............................................................................................................................................ 67 ACKNOWLEDGMENTS This report was prepared by a World Bank team working in Albania during 2013. The principal authors are Abebe Adugna and Doerte Doemeland, who were supported by a thematic team consisting of Munawer Khwaja and Sandeep Bhattacharya (taxes); Yadviga Viktorivna Semikolenova, Gazmend Daci, and Salvador Rivera (energy); Maria Carolina Monsalve and Artan Guxho (capital spending and transport); Anita Schwarz and Lorena Kostallari (pensions); and Rajeev K. Swami, Lewis Hawke, and Jonida Myftiu (public financial management). Erjon Luci contributed significantly to the preparation of the first draft until his appointment as Deputy Minister of Finance in the Albanian government in October 2013. The report was finalized after Mr Luci left the World Bank. Hilda Shijaku supported the team in the final stage of the report. The World Bank team is sincerely grateful to the Albanian Government for its close collaboration during preparation of the report. In particular, thanks are due to senior officials in the Ministries of Finance; Public Works, Transport and Telecommunication; Economy Trade and Energy; Labor and Equal Opportunities; and the Bank of Albania. Special thanks also go to Satu Kahkonen, Yvonne Tsikata, and Roumeen Islam of the World Bank, who provided continuous support, guidance, and technical advice. Useful comments were received in March 2013 at a round table on fiscal policy, jointly organized by the Ministry of Finance, the Bank of Albania, the IMF, and the World Bank; and most recently in December 2013 from the Government of Albania. Finally, the team wishes to thank Ellen Goldstein, country director for Southeast Europe, and Kseniya Lvovsky, former country manager for Albania, for their valuable advice and guidance, and Mismake Galatis for her support in preparation of this report. EXECUTIVE SUMMARY Albania’s rapid growth in the decade up to the 2008 global financial crisis propelled it to middle- income status and helped to reduce poverty. Between 1998 and 2008 annual growth averaged 6 percent in real terms and per capita GDP quintupled to above US$4,000. Although this impressive growth was not matched by a similar surge in job creation, absolute poverty (i.e. the percentage of people living on less than $1.25 a day) dropped from 25.4 percent in 2002 to 12.4 percent by 2008. The global financial crisis in 2008 slammed the brakes on Albania's largely domestic-demand- driven growth. Although Albania was able to avoid a recession, the crisis hit the country’s economy hard. Between 2009 and 2012 GDP growth averaged less than 3 percent. The crisis lowered remittances and other financial inflows and suppressed growth, particularly in the construction sector where the decline has been steep. The Eurozone crisis has further weakened economic recovery because of Albania’s close links to the Greek and Italian economies; exports, remittances, and financial flows have suffered. Tepid growth reversed the fall in poverty, which increased from 12.4 percent in 2008 to 14.3 percent in 2012. As the Eurozone crisis lingers, Albania’s medium-term growth prospects are likely to remain modest, averaging only about 2-3 percent per year. Public debt has surged upward since the crisis. Loose fiscal policy, combined with a depreciation of the lek, brought public debt levels to about 60 percent of GDP in 2011, the legal limit set in 2008. The energy shortage in 2012 and the need for government support to the power generation company in the form of guarantees led the government to ask Parliament to remove the 60 percent debt ceiling from the organic budget law. In December 2012, the Parliament did so, without proposing any other fiscal or debt anchor. Albania’s 2013 budget foresees further increases in public debt. The government has accumulated sizable arrears in payments for public works and VAT refunds. Even as revenues were declining and budgets for public investment were reduced in the wake of the 2008 crisis, many public works continued at the same pace as initially planned as contractors continued to carry out public works, often by drawing on commercial loans. With fiscal space running out, this translated into a buildup of payment arrears – which have adversely affected private sector liquidity and driven up nonperforming loans. While the new government has made good progress in establishing the stock of payment arrears –initial estimates show the amount could be as high as US$670 million or about 5.3 percent of GDP—further efforts are needed to audit and pay the arrears, and institute a commitment control system in the treasury to prevent their recurrence. Why is rising public debt a concern? With Albania’s medium-term growth rate at around 2.0-3.0 percent and the average real interest rate on current public debt of 3.0 percent or higher, the public debt-to-GDP ratio is set to grow rapidly, unless the authorities introduce a serious program of fiscal consolidation. Given the current macroeconomic outlook, the growth rate–interest rate differential implies adverse, and potentially explosive, debt dynamics. In a baseline scenario of no policy reforms, Albania’s public debt-to-GDP ratio is projected to reach 73.5 percent in 2015 and stay above 72 percent over the medium term. By 2018, the central government debt-to-GDP ratio is projected to reach 76.2 percent if the primary deficit holds constant at 2.4 percent throughout the period; 81.5 percent if real GDP growth is 1 percentage point lower than projected throughout; and 84.5 percent if there is a one-time 30 percent real depreciation. The picture looks worse if contingent liabilities in the energy sector are included. i Empirical evidence confirms that high public debt depresses economic activity and significantly increases the probability of default. Although there is no agreement on the optimal level of public debt, empirical analysis of emerging markets has shown that a debt crisis can hit countries even when their debt is surprisingly low: in 55 percent of the debt defaults of middle-income countries, public debt was below the Maastricht benchmark of 60 percent of GDP, and in 35 percent, it was below 40 percent (IMF 2013). For Albania, high and growing debt poses both interest rate and rollover risks. Spending on interest, at about 3.2 percent of GDP, is already much higher than in the other countries of South East Europe (SEE), crowding out more productive spending and representing a major source of vulnerability for the budget. About 57 percent of Albania’s public debt is domestic; about 72 percent of domestic debt has variable interest rates and is affected by interest rate fluctuations. As of 2012, the implied interest rate on domestic debt was 7.7 percent. Furthermore, over 55 percent of domestic debt (almost 30 percent of GDP) should be rolled over within a year, exposing the government to changes in market conditions. Commercial banks are the main holders of Albania’s domestic debt, with Raiffeissen Bank and the Bank of Albania holding nearly half of it. With some European banks recently taking steps to reduce their exposure given the Eurozone sovereign debt crisis and the weak European growth outlook, concerns about the rollover risk have risen. Increased public borrowing also damages growth prospects by crowding out the private sector. In short, fiscal consolidation has now become imperative to reduce the risk of macroeconomic crisis, reduce interest payments, and free up resources for productive investments in education, health, and infrastructure. This report examines closely the opportunities for fiscal consolidation on both the revenue and expenditure sides. What opportunities are there for fiscal consolidation? As Albania’s fiscal deficit is largely structurally driven, reforms in tax policy and administration, pension, and energy could significantly enhance its fiscal sustainability. If Albania were to raise its revenue-to-GDP ratio through tax policy and administration reform from the current 25 percent of GDP to 27 percent in 2014 and keep it there, its public debt-to-GDP ratio would fall below 70 percent by 2016 and reach 62 percent by 2020. If the revenue was further raised to the level of FYR Macedonia (30 percent), Albania’s debt-to-GDP ratio would be down to 44 percent by 2020. Increasing the revenue-to- GDP ratio could be achieved through, among others, raising the CIT rate and/or taxing sales of new residential construction under the VAT. Similarly, if it were to introduce short term pension measures (e.g. temporarily freeze pensions, gradually raise rural contribution rates until they reached the level of urban contributions), it could cut the fiscal deficit by as much as 1.2 percent of GDP by 2016, and reduce its public-debt-to GDP ratio by about 2 percent of GDP. Reforms in the energy sector could help reduce the contingent liabilities to the government budget as promote growth through a reliable energy supply. For example, a distribution loss reduction plan that reduces energy losses by 6 percent by better network management, metering of large tariff customers, customer outreach, and collections efforts could cut the contingent liabilities (of about 2 percent of GDP) by almost half compared to the no reform scenario. Tax policy and administration At about 24 percent of GDP, Albania has one of the lowest tax revenues in Europe. VAT, currently at 20 percent, accounts for about 10 percent of GDP and generates the bulk of revenues. Income taxes (corporate, CIT; and personal, PIT) together contribute just 3.7 percent of GDP, well below the average for countries in Eastern Europe. The PIT is flat at 10 percent with a monthly threshold; the CIT was reduced from 20 to 10 percent in 2009, driven by a region-wide race to improve competitiveness in South East Europe. As a result, Albania now has one of the lowest CIT rates in Eastern Europe—most Eastern European countries have rates of 19 to 21 percent. The low rates limit the scope to bring in more revenues. ii The basic design of Albania’s VAT system is sound. The VAT law has only two positive rates (20 and 10 percent); a zero rate is applied only to exports; and few goods and services are exempted. However, the VAT suffers from deficiencies related to too narrow a tax base and too many exemptions. The productivity of VAT in Albania is 53 percent; the average for Eastern Europe is 66 percent. This means that in Albania, for every 1 percent of the VAT tax rate, the tax administration collects VAT on only 53 percent of total consumption (the base for VAT), leaving 47 percent untaxed. Similarly, an analysis of the VAT gap-- defined as the difference between the amount of taxes the government should collect and the amount of taxes the government actually collects at a given tax rate--suggests that VAT collection in Albania could be improved by 10 to 20 percent. Both the relatively low productivity and the large tax gap suggest that the policy exempts certain sectors or transactions; tax administration is inadequate; corruption and tax evasion may make it impossible to tax the full base; or combinations thereof. One of the main factors contributing to the narrow VAT base in Albania is not taxing new residential construction. Article 20 of the Law on VAT exempts the sale of new construction and instead taxes rentals. In the EU, the practice is the opposite: both sales and rentals of existing real estate are exempt from the VAT but new buildings and improvements are subject to the VAT. Albania should follow the EU model: exempt rental property and tax the sale of new construction, including the land value. If it were to do this, the fiscal impact could be considerable: about 1.5 percent of GDP in additional VAT revenue even under conservative assumptions on taxable base and compliance. Furthermore, recent VAT exemptions on imports of capital goods have further narrowed the base and reduced revenue. In April 2013 the VAT law was amended to exempt, for all businesses, imports of capital goods from the VAT and customs fees. A few years earlier, the government had exempted imports of equipment for investments that cost more than US$500,000 and removed the VAT on steel and cement used for the construction of hydropower plants (HPP). Equipment imported for petroleum exploration and development is also exempt. Besides narrowing the tax base, such exemptions are inconsistent with the EU VAT system. Reversing these exemptions would stop the erosion of the tax base and shift Albania to EU VAT standards. The revenue impact of reversing these exemptions could be a one-off increase of up to 0.5 percent of GDP. Income taxes (PIT and CIT) offer even greater opportunities for raising revenue in Albania. The productivities of PIT (6 percent) and CIT (17 percent) are even lower that of VAT. The PIT productivity rate, for example, means that only 6 percent of the base for income tax (GDP) is actually taxed. A tax gap analysis using information from the Living Standards Measurement Survey (LSMS) shows that the PIT revenue could be four to five times higher than it is currently. Similarly, there is some room to enhance revenue, and more importantly equity, through adjustments to the PIT regime. Albania has a flat PIT with a monthly threshold, but many countries in Europe have progressive rates. Even outside Europe, a vast majority of countries have progressive PIT rates. Making PIT more progressive can improve equity and fairness. The revenue impact of making the PIT progressive is modestly positive or negative, depending on how the rates are structured. Assuming that the current universal deduction of 30,000 LEK would continue; all incomes from 30,000 – 150,000 LEK a month are taxed at 10 percent; and a simple progressive tax of 15 percent is introduced for those earning more than 150,000 LEK, a simple calculation shows that this could yield a very modest additional revenue (of about 0.1 percent of GDP), but the actual impact will clearly depend on how the progressive rates are structured as well as whether the current compliance rate could be maintained. Albania can raise its CIT rate to at least 15 percent without imposing a heavy burden on businesses or harming its regional competitiveness. The race to the bottom with respect to the CIT rates has limited the scope for increasing revenue in Albania. Some countries have already reversed this trend. Serbia, for example, has already raised its CIT rate from 10 to 15 percent. If the CIT were to be raised to iii 15 percent, a potential revenue increase from this measure could be about 0.3 percent of GDP (or 10-15 percent increase) if backed by measures to prevent base erosion and current compliance. Overall, four policy measures can help increase revenue in Albania: expanding the VAT tax base by exempting rental property and taxing the sale of new construction; reversing the recent VAT exemptions; raising the CIT rate from 10 to 15 percent; and making the PIT progressive and introducing risk-based audits. The combined fiscal impact of these policy measures could be as much as 3 percent of GDP assuming Albania’s average compliance. The revenue impact could be larger if these policies were to be complemented by strengthened tax administration to improve compliance. While tax policy reform is important, it is equally critical that Albania modernizes its tax administration. The large tax gaps calculated for Albania can only be closed with significantly strengthened tax administration and enforcement of existing policies to reduce opportunities for corruption and tax evasion. Revenue performance could be raised by reducing exemptions and tax loopholes and building up the capability of the tax authorities to detect fraud through measures such as data mining and third-party information-matching. Building up the capacity of tax administration and reforming tax policy are not alternatives; they are complements that for best results must be pursued simultaneously. Pensions Although spending on pensions in Albania is lower than the average for Eastern Europe, it is rising. In 2012 Albania spent 5.2 percent of GDP on pensions, much less than the EU countries or the average for Eastern Europe. Albania’s population is still younger than most other European countries and thus has fewer individuals to support through pensions. Pension spending is still low even though the pension program has to finance several generations of elderly, most of whom have full pension rights. Until recently the system has avoided a fiscal crisis, despite the low share of workers participating in it, by compressing benefits; consequently, the average pension in urban areas amounts to 1.1 times the minimum pension. This suggests that contributions and benefits are not closely linked, creating disincentives to full participation and to declaring total earnings. In rural areas, disincentives to contribute are even more perverse because those earning minimum pensions receive additional allowances that raise their total incomes higher than those of pensioners who previously earned more. In past years, the government made sizable increases in pension benefits, particularly for those with lower pensions, so that pension spending increased. Relative to the average wage, minimum pensions are higher in Albania than in most other countries, which makes it difficult to differentiate benefits by earning or the contribution level without raising costs even higher. Spending on pension programs increased by 1 percent of GDP during 2005–12, from 4.9 percent of GDP in 2005 to 5.2 percent in 2012. The lack of a comprehensive approach to the pension reform and sizable ad hoc increases in benefits have exacerbated the fiscal deficit. The pension system, which was in balance in 2005, ran an estimated deficit of about 1.2 percent of GDP in 2012. From 2005 through 2012, total government transfers to the pension system increased from 1.1 percent of GDP to an estimated 1.8 percent. Along with the slowdown in economic growth, the following major factors increased the pension deficit: (a) the plunge in pension contribution rates from 29.9 percent to 21.6 percent in 2009; and (b) sizable increases in pensions well above the inflation rate, particularly for rural dwellers. The fiscal pressure is expected to heighten as the Albanian population ages. The government has two main options for the direction of pension reform: • Keep the current contributory system but tighten the link between contributions and benefits to enhance the replacement rate of incomes individuals earned during their working years. iv • Recognizing that, currently, the pension (or the income earned by the pensioner) is almost the same for all urban pensioners and, although slightly lower, for all rural pensioners, eliminate the contributory scheme entirely and institute a social pension for people above a certain age (which would spotlight the poverty alleviation objective). Those who are unable or unwilling to contribute would then receive at least some basic protection in old age. This kind of a system would eliminate the need to collect contributions and maintain records. Basic benefits could be financed by a rise in income taxes or in the VAT or could continue to be financed by a labor tax—whichever would raise sufficient revenue with the fewest disincentives. The basic social pension could be supplemented by voluntary private savings for those who want to better smooth consumption in old age. The projected impact of these two broad options on fiscal sustainability and pension levels depends on the many sub-options that can be chosen within each option. While illustrative fiscal impact scenarios are discussed in the report for particular sub-options, the full fiscal impact analysis could be made only once the government has chosen a particular direction. As the government defines which path to take and develops the legislation to implement that path, it should consider short-run measures which can provide fiscal savings and support the fiscal consolidation agenda. Short-run fiscal measures can include temporarily freezing pensions, given that increases above the legislated inflation rate have been provided in the recent past, and gradually raising rural contribution rates until they reach the same levels that the urban self-employed pay. These measures can potentially save the Government as much as 0.4 percent of GDP in 2014 alone, but as much as 1.2 percent of GDP in 2016. It is unlikely that the temporary pension freeze could be extended beyond 2016. Energy Albania’s internal electricity generation capacity is about 98 percent dependent on hydropower and thus sensitive to hydrological vulnerability. Dependence on hydrology poses a major risk to the country’s electricity supply and causes enormous variation from year to year in the costs of meeting electricity demand. The costs of covering growing electricity demand are particularly high in poor hydrological years when additional imports are needed to be purchased to offset reductions in domestic hydropower production. If there are not enough imports secured to cover the demand, the system has to resort to load-shedding. The sector has been historically characterized by high transmission and distribution losses, low collections, and high rates of bad debt (accumulated unpaid invoices). While transmission losses were reduced and are now low and comparable with technical losses in developed countries, distribution losses are still high. In March 2009, the Government of Albania privatized the distribution system, OSSh, selling it to CEZ (which bought 76 percent of OSSh shares) to reduce losses and improve sector performance. However, privatization of the distribution system did not work out as expected: acceptance of the level of losses and bad debts by CEZ Sh. came much later than expected; retail tariffs were not adjusted promptly; and collections of the bills of both budgetary and non-budgetary institutions did not improve. By the end of 2012, power losses stood at 45 percent and collection rates at 84 percent. In the fall of 2012, CEZ Sh. stopped importing electricity and KESh, the supplier of last resort, hit its hydro reservoirs hard to cover demand. On January 21, 2013, the energy regulatory agency (ERE)revoked the CEZ Sh. license and appointed a temporary state administrator to run the distribution system. Since the distribution sector operator (DSO) is a public company, distribution losses are now essentially part of the state’s obligations to the sector. As such, the government ultimately bears the financial risks of the energy sector, which constitute significant contingent liabilities on the budget. v While the direct public expenditure or budgetary support to the energy sector remains small, the indirect or contingent liabilities are very large. Direct budgetary support to the sector in recent years included ad hoc liquidity injections into KESh, the generation company, to cover its liquidity gap as well as subsidies to the disabled for electricity bills (80 percent of energy bills of disabled customers is covered by the state; the subsidy was abolished as of May 2012). The Government owes CEZ Sh. over Lek 3 billion (about US$29 million, or about 0.2 percent of GDP) for the subsidy. There is no confirmed estimate of indirect support to KESh by the Government. The indirect support to the sector, however, has been much higher. This includes guaranteed loans of KESh for electricity imports, and significant debt of budgetary and non-budgetary public institutions in electricity payment arrears discussed above. The demand for commercial loans by KESh increased drastically at the end of 2012, when CEZ Sh. stopped importing energy to cover distribution losses. As a result, KESh/ WPS, as the supplier of the last resort, had to increase electricity imports to satisfy demand of tariff customers without resorting to load- shedding. Because of its own liquidity issues related to the non-payments of CEZ Sh., these loans had to be guaranteed by the state. As of spring 2013, the guaranteed overdraft of KESh was 29 billion Lek (US$274 million, or about 2.5 percent of GDP). In addition, contingent liabilities arising from the disputed takeover of CEZ Sh. could amount to another US$ 200-250 million (2.0-2.5 percent of GDP). In a business-as-usual scenario (BAU) –which assumes growth in electricity demand by 3–4 percent a year, normal hydrological conditions, no new generation added to the system and distribution losses gradually reduced to the level of 2010—the sector becomes financially unsustainable. The expected annual financial gap of the sector in this scenario would be about lek 17–20 billion (US$161–189 million), or 1.5 percent of GDP, a lower-bound estimate. In a normal loss reduction scenario, where the distribution operator reduces losses by 6 percent annually through 2016 (the BAU rate would be 3 percent)—through better metering of large customers, clamping down on the illicit use of electricity, and higher collections of 3 percent annually until the collection rate reaches 95 percent, a rate common in the region—the financial gap can be cut by almost half compared with the BAU scenario. The rest can be closed by adjusting end-user tariffs by at least 10 percent. In an aggressive loss reduction scenario, which requires extensive work with large and medium customers, introducing 100 percent consumption-based billing, and eliminating electricity theft, distribution losses could easily be reduced by 10 percent annually for the first three years. Further reducing losses would require additional investments, such as modernizing the network, improving maintenance, and upgrading management. As a minimum these investments would cost about US$50 million over the next three years. Such an aggressive program could cut the financial gap of the sector by over two-thirds compared to the BAU scenario. The rest could be closed through a tariff adjustment. While the above are illustrative scenarios, restoring the financial viability of Albania’s energy sector calls for breaking away from the past by: (a) investing in new generation and diversifying the generation mix; (b) decreasing distribution system losses; (c) adjusting tariffs to reflect the true energy costs; and (d) opening up the Albanian electricity market to regional competition. The market should move toward compliance with EU directives: large consumers could start buying electricity on the market; KESh’s role as supplier of last resort could be done away with; and the DSO could become responsible for a consistent supply of energy to tariff customers. These measures are needed not only to restore the financial viability of the sector and reduce the fiscal risks on the budget but also to effectively support growth through a reliable supply of power. vi What would the structural reforms mean for fiscal sustainability? A combined effect of these structural reforms would reduce Albania’s public debt to GDP ratio singnificantly over the medium term. As an illustrative scenario, we consider a reform scenario whereby Albania (i) raises its revenue-to-GDP ratio from 25 percent of GDP to 27 percent in 2014 through tax policy and administration reform and keeps it at that level thereafter; (ii) puts in place short- run fiscal measures which include temporarily freezing pensions, and gradually raising rural contribution rates until they reach the same levels that the urban self-employed pay; and (iii) implements a (normal) distribution loss reduction plan that reduces energy losses by 4-6 percent through 2014-16 and improves collections by 3 percent annually until collection reaches 95 percent, a common collection rate in the region. The combined effect of tax and pension reforms will be to reduce (relative to the baseline) Albania’s public debt-to-GDP ratio by 2.6 percent of GDP in 2014, 7.7 percent of GDP by 2016 and 17.8 percent by 2020 (see Table 1). The impact of energy sector reform will primarily be to reduce the contingent liabilities arising from the sector. With no reforms, these contingent liabilities are set increase from abour 1.3 percent of GDP in 2014 to about 6.8 percent of GDP in 2020. With the proposed loss reduction reform, the contingent liabilities would be cut by almost half by 2020 compared with the no- reform scenario (Table 1). The energy sector reform would also also enhance the ability of the sector to provide reliable electricity and support investment and growth. Table 1. Albania: Impact of alternative structural reforms on public debt (Percent of GDP) 2013 2014 2015 2016 2017 2018 2019 2020 Public debt under the baseline - no reform 70.5 72.1 73.5 72.7 72.8 72.9 73.0 74.2 Public debt with tax policy and admin reform 70.5 69.6 68.5 65.6 63.7 62.0 60.3 59.6 Public debt with pension reform 70.5 72.2 73.4 72.2 71.6 70.7 69.5 68.4 Public debt with combined tax and pension reforms 70.5 69.7 68.4 65.1 62.5 59.8 56.8 53.8 Contingent liabilities in energy sector – no reform 0.0 1.3 2.5 3.4 4.1 5.1 5.9 6.8 Impact of energy sector reform on contingent liabilities 0.0 0.9 1.6 2.0 2.4 2.9 3.3 3.7 As Albania undertakes these consolidation measures, it could in parallel rebalance its capital spending, particularly in transport, toward maintenance to support growth. Spending on roads— mainly on motorways—rose to 5.7 percent of GDP in 2008 before declining to 2.7 percent in 2010, 2.4 percent in 2011, and 1.9 percent in 2012. Reflecting this investment, Albania’s rankings on the quality of road infrastructure have improved. Still, spending on road maintenance is not sufficient and there is a considerable backlog. Nominal expenditures on maintenance in 2012 were down 25 percent from 2008, while spending on rehabilitation was down 73 percent. Average actual spending on maintenance of national road network during 2008–12 was about €12 million; while it is estimated that Albania would need €40– 45 million to maintain its primary and secondary roads. This means that its road network maintenance allocation will need to triple. While the additional maintenance expenditures could be funded within the current roads allocation, a rebalancing away from investment in new roads toward maintenance and rehabilitation would be critical for preserving the road network and supporting sustainable growth over the medium term. Finally, Albania needs to strengthen its institutions to reinforce financial discipline and strengthen fiscal policy. Institutional reforms are particularly needed with regard to public financial vii management (PFM) and introducing a fiscal rule to anchor policy over the medium term. While Albania’s PFM has improved since 2006, some aspects of planning and execution remain problematic. These relate in particular to over-optimistic revenue projections; and the lack of effective commitment control to prevent payment arrears as evidenced by the build-up of arrears in recent years. Overly optimistic revenue estimates should be made more realistic, for example through partial reliance on independent forecasts, to reduce the scope for sharp adjustments and cut backs mid-year that often hit capital investments, and/ or lead to accumulation of arrears. With regard to commitment control and prevention of future arrears, the recent initiative to register all contracts in the Treasury is a major step forward but there are a number of additional steps the government needs to take to better manage future obligations and instill a system of financial accountability and good financial management throughout government. As for a fiscal rule, having removed the 60 percent-of-GDP ceiling on public debt from the Organic Budget Law, the government is yet to set a new debt ceiling or a fiscal rule. The government should consider putting in place either a procedural or a numerical fiscal rule to anchor fiscal policy over the medium term and send a credible signal to investors and the market that fiscal sustainability will be maintained. viii Summary of policy options Area Policy Options Energy • Expand the generation capacity with equitable risk sharing between the state and private investors. • Reduce fluctuations in the power supply and smooth variation in energy tariffs. • Improve the collection of energy bills from public institutional customers. Settle public institutions' electricity payment arrears and prevent the emergence of further arrears through strengthening of treasury regulations that would mandate payment of utility bills. • Reduce distribution losses. • Increase collections from tariff customers including through improved metering and billing. • Align tariffs more closely to costs, while introducing social assistance that would better protect the most vulnerable, and tighten regulatory compliance. Pensions • Index the average pension to inflation from 2014 onward and avoid costly ad hoc increases; • Gradually raise rural contribution rates to the same levels that the urban self-employed pay. • In the medium-term introduce either 1) an enhanced the Contributory Scheme emphasizing income replacement, or 2) a poverty-focused basic pension. Tax policy • Expand the VAT tax base through exempting rental property and taxing the sale of new construction, including the land value. • Reverse the recent VAT exemptions. • Raise the CIT rate from 10 to 15 percent. • Make the PIT progressive and introduce risk-based audits. Transport • Rebalance spending on roads from capital investments toward maintenance and rehabilitation. • Set up and use a road asset management system—and update it regularly – as a basis for a more efficient and professional management of national road assets. Public Budget Planning, Preparation, and Revision financial • Contain over-optimism in revenue estimates by having independent experts undertake management revenue estimates. Capital Investment Management • Ensure that all major projects are subject to rigorous cost-benefit analysis and prioritization in the context of the budget processes. • Build the investment management capacity (monitoring, reporting, and risk management) for public investment. Cash Management (Treasury) • Establish a direct link between the Treasury and tax collection systems to enable daily and timely cash flow information for managing cash and debt. • Prepare alternative scenarios (baseline and optimistic) for both revenue collections and expenditure outlays; and update them quarterly—to feed into cash and short-term debt management. Commitment Control • Amend the Organic Budget Law to clearly define arrears and clarify the regulations to allow commitments and obligations to be detailed in future budgets as part of the MTBP. • Record commitment of funds for multiyear projects in the treasury system to ensure that claims are taken into account in future budget allocations before budgets are approved. ix CHAPTER 1. ALBANIA BEFORE AND AFTER THE CRISIS A. GROWTH AND POVERTY REDUCTION: SIGNIFICANT STRIDES FORWARD 1.1. Albania’s rapid growth in the decade Figure 1.1: Real Growth and GDP per Capita up to the 2008 global financial crisis propelled (US$, right) it to the middle-income status (Figure 1.1) and helped to reduce poverty. Between 1998 and GDP real growth (%, left) 15 4500 2008 annual growth averaged 6 percent in real 13 terms as per capita GDP increased fivefold, to 10 10 9 9 9 10 4000 7 7 7.7 above US$4,000. Although this impressive growth 6 6 6 6 3500 5 was not accompanied by strong job creation, 5 3 3 3.5 3 3000 absolute poverty 1 dropped from 25.4 percent in 1.6 2500 0 2002 to 12.4 percent by 2008. 2000 -5 1500 1.2. Domestic demand was the main driver GDP per capita (US$, right) 1000 of growth during this period of transition. -10 -10 500 Construction, industry, and services all grew 10 to -15 0 20 percent in real terms between 1998 and 2008, 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 driven in part by remittances and other inflows. Source: INSTAT. Industrial growth relied initially on textile and shoe manufacturing and later on construction materials (mostly cement) and mining and oil exports. Agriculture lost ground as workers transitioned to industry and services, and lack of investment and land fragmentation took their toll. 1.3. The global financial crisis in 2008 slammed the brakes on Albania's largely domestic demand-driven growth. Although it was able to avoid recession, the crisis hit Albania hard. Between 2009 and 2012 GDP growth averaged less than 3 percent, mainly because of basic services and the extraction industry, which sought markets outside the EU. The crisis lowered remittances and other in- flows, which suppressed growth particularly in the construction sector, where the decline has been steep. The Eurozone crisis further weakened economic recovery because of Albania’s close links to the Greek and Italian economies; exports, remittances, and financial flows suffered. Weaker growth reversed the fall in poverty rates, which increased from 12.4 percent in 2008 to 14.3 percent in 2012. As the Eurozone crisis lingers, Albania’s medium-term growth prospects are likely to remain modest, averaging about 2-3 percent. 1.4. Since the crisis the financial sector has remained stable, but continued deterioration of the bank loan quality has restrained credit growth. Albania had seen rapid growth of credit to the private sector, which rose from 4.7 percent of GDP in 2001 to 39 percent in 2011. The banking sector, which is 90 percent foreign-owned, dominates the financial system, accounting for more than 90 percent of total financial assets. Financial stability was preserved despite the difficulties caused by the global crisis. Though between October 2008 and July 2009 deposits dropped by 6 percentage points of GDP, they recovered in 2010 (Figure 1.2). However, nonperforming loans (NPLs) rose from 6.7 percent in 2008 to about 24 percent by end-2013(Figure 1.3). Albania’s NPLs are among the highest in the region, and the difficulties banks confronted in executing on their collateral have kept the rate high. While credit growth 1 This is defined as the percentage of people living with less than $1.25 a day. 1 slowed after the global crisis, Albania’s limited reliance on foreign credit lines prevented a steep drop in loans. The banking system is relatively well-provisioned; since 2007 the ratio of provisions to total loans has quadrupled. By mid-2013, capitalization of the banking sector was 17 percent and deposits, despite having plunged in 2008, were much higher than the pre-crisis levels. Figure 1.2: Credit and Deposits Figure 1.3: Albania: Nonperforming Loans (Percent of total) and Credit Growth (Percent) Credit change (q-o-q, %) Non-performing loans (%) 70 24 60 Total Deposits (% of GDP) 50 19 42 41 38 40 40 36 14 30 30 Total Credit (% of GDP) 22 9 20 o/w FX credit in blue 15 9 4 10 6 7 4 4 4 3 4 4 4 4 5 0 -1 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2004 2005 2006 2007 2008 2009 2010 2011 2012 Source: Bank of Albania. Source: Bank of Albania. 1.5. Net exports have been the main drivers of recovery and growth since the crisis. Exports have tripled since the early 1990s and they now account for more than 30 percent of GDP (Figure 1.4), but at the same time imports have risen by almost as much. Exports of minerals, fuels, and electricity have been growing steadily since 2005, emerging in 2012 as the largest category of exports. Exports of wood and paper products have also grown moderately. After a major expansion in the late 1990s and early 2000s, exports of textiles and footwear have moderated in recent years. Nevertheless, Albania still remains a net importer of traditional consumer goods such as food, despite having potential to be a net exporter in some areas, and of equipment and machinery for domestic production. As a result, Albania’s trade deficit has remained high, though about half has been covered by remittances. Until 2007 foreign direct investment (FDI) had been modest at about 3 percent of GDP, but FDI in recent years has increased, driven by investment in extraction industries (oil and mining) and by the re-capitalization of the banking sector. 1.6. Still, Albania has yet to ensure a durable shift from the old domestic-demand-driven growth model to an export-oriented growth model. For small economies like Albania, a growth model driven by external demand, and largely export growth, is more sustainable. Greater trade and financial integration served as a major income “convergence machine” for earlier entrants into the EU (Figure 1.5). For Albania, accelerating convergence requires a new model of growth that is not dependent on domestic demand. Boosting the export competitiveness of Albania’s economy requires structural reforms in at least three areas: (1) improving the quality of governance, the rule of law, and property rights; (2) strengthening the quality of human capital (education and skills); and (3) overcoming infrastructure bottlenecks, especially in the power and transport sectors (see chapters 3 and 4). However, for Albania to tackle these challenges effectively and free up the resources needed for productive investments in education, health, and infrastructure, it will need to put its fiscal policy on a sustainable path. 2 Figure 1.4: Export Share by Sector (Percent) Figure 1.5: Income Convergence between 1998- 2012 with EU Per Capita Income (US$ PPP), Percentage Point 30 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 100 5 25 23.8 11 5 9 10 90 10 10 5 9 12 13 10 15 16 17 16 19 20 20 21 21.6 80 20 18.7 43 70 31 41 49 43 53 29 60 16.4 47 34 32 67 65 65 65 48 43 15 14.4 65 61 58 55 50 40 10.2 10 9.4 22 25 30 8.3 36 7.5 7.5 7.2 16 10 6 11 6.3 28 30 20 18 20 15 4.8 4.6 19 20 5 3 2 3 3 4 5 8 10 5 16 14 17 16 9 10 9 9 8 8 8 8 7 6 7 6 6 6 0 Food, beverages, tobacco Minerals, fuels, electricity Chemical and plastic products Leather and leather manufactures Wood manufactures and articles of paper Textile and footwear 0 Construction materials and metals Machineries, equipments and spare parts Others RUS SVK POL BGR CZE SVN ALB TUR MNE HUN SRB BiH Cro. MKD Source: Bank of Albania. Source: WEO. B. FISCAL POLICY: UNSUSTAINABLE 1.7. As Albania’s growth prospects have weakened, public debt sustainability has emerged as a key concern. Albania’s high economic growth during 1997- 2007 supported declining fiscal deficits (Figure 1.6) as well as public debt. During 2002-08, public debt declined by about 10 percent of GDP. After 2008, however, as growth slowed, fiscal deficits increased, particularly in the early phase of the crisis. The latter pushed up public debt to about 60 percent of GDP in 2011, the legal limit set in 2008 (Figure 1.7). The energy shortage in 2012 and the need for government support to the power generation company (in the form of guarantees) led the government to ask Parliament to remove the debt ceiling from the organic budget law. In December 2012, Parliament did so, without proposing any other fiscal anchor. Albania’s 2013 budget foresees a further increase in public debt. Figure 1.6: Real Growth and Fiscal Deficit Figure 1.7: Public Debt (Percent of GDP) 15 GDP real growth (%) 70 65 10 62 62 60 60 59 59 60 57 58 59 60 56 56 55 54 53 5 50 40 0 30 -5 -3 -3 -4 -3 -4 -3 -5 -5 20 -6 -7 -7 -10 -8 -8 10 -9 -10 -9 Fiscal balance (% of GDP) -12 -12 0 -15 -14 -13 1997 2000 2003 2006 2009 2012 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 Source: Ministry of Finance, INSTAT. Source: Ministry of Finance, INSTAT. 1.8. Albania’s public debt is among the highest in Eastern Europe. High and growing debt increases vulnerability in terms of both high interest costs and rollover risks. Interest expenditures, at about 3.1 percent of GDP in 2012, are already much higher than in other countries of South East Europe (SEE), crowding out more productive spending and representing a major source of vulnerability for the budget. Over 55 percent of domestic debt (almost 30 percent of GDP) needs to be rolled over within a 3 year, exposing the government to changes in market conditions. Increased public borrowing would also damage growth prospects by crowding out the private sector. 1.9. In addition, the government has accumulated sizable arrears in payments for public works and VAT refunds. Even as revenues were declining and budgets for public investment were reduced in the wake of the 2008 crisis, many public works continued at the same pace as initially planned as contractors continued to carry out public works, often by drawing on commercial loans. With fiscal space running out, this translated into a buildup of payment arrears – which has adversely affected private sector liquidity and driven up nonperforming loans. While the new government has made good progress in establishing the stock of payment arrears –initial estimates show the amount could be as high as US$670 million or about 5.3 percent of GDP—further efforts are needed to audit and pay the arrears and institute a commitment control system in the treasury to prevent their recurrence. 1.10. Fiscal consolidation to reduce public debt over the medium term has become imperative. The growth rate–interest rate differential in the current macroeconomic outlook implies adverse—and potentially explosive—debt dynamics. With Albania’s expected medium-term growth at about 2.0-3.0 percent and the average real interest cost at 3 percent or higher, the public debt-to-GDP ratio is set to grow rapidly unless the authorities take action and pursue fiscal consolidation (see chapters 2). In a baseline scenario of no policy reforms, Albania’s public debt-to-GDP ratio is projected to reach 73 percent in 2016 and stay above 73 percent over the medium term. By 2018, the central government debt- to-GDP ratio is projected to reach 76 percent if the primary deficit holds constant at 2.4 percent throughout the period; 81.5 percent if real GDP growth is 1 percentage point lower than projected throughout; and 85 percent if there is a one-time 30 percent real depreciation. The picture looks worse if contingent liabilities in the energy sector are included. 1.11. The risks of rising public debt—rollover risks, high interest expenditures, and crowding-out of private sector borrowing—have risen to the point that the priority must now shift to fiscal consolidation to reduce public debt over the medium term. Lowering debt is essential to reduce the risk of macroeconomic crisis, reduce interest payments, and free up resources for productive investments in education, health, and infrastructure. C. OBJECTIVES OF THE REPORT 1.12. After experiencing a decade of strong growth, Albania is now embracing a difficult period ahead. Since the 2008 financial crisis economic growth has remained subdued at around 2-3 percent. A low growth outlook is expected to persist for a few years as the eurozone crisis lingers. The slowdown in growth has put more pressure on public finances. With fiscal revenues less buoyant and fiscal space exhausted, the task of sustaining development goals has become more challenging. In particular, the need to ensure fiscal and debt sustainability while supporting growth has become even more critical. 1.13. The objective of this public finance review is two-fold: (i) take stock of the public finance challenges in Albania both on the revenue and expenditure side; and (ii) propose policy options that would help achieve fiscal consolidation and debt reduction over the medium term. D. HOW THE REPORT IS ORGANIZED 1.14. To provide a baseline for the rest of the report, chapter 2 presents a fiscal sustainability analysis that assumes there are no structural policy reforms. It shows that without fiscal consolidation, Albania’s 4 public debt and interest payments will grow over the medium term, and principal debt ratios will shoot up. Public debt becomes unsustainable. 1.15. Chapter 3 discusses options for structural reforms that can anchor long-term fiscal sustainability in three areas: energy, pension reform, and tax policy and administration. • Energy sector reform: Given the current market structure, ultimately the government bears the financial risks of the energy sector, which constitute contingent liabilities on the budget. The energy section of chapter 3 discusses the challenges in the sector and outlines policy options for improving the financial viability of the sector, as well as ensuring a continued, reliable supply of energy – a key factor for attracting private investment and supporting growth. • Pension reform: If allowed to continue in its present form, Albania’s pension system will generate fiscal deficits in the short, medium and long term, and many elderly may not be able to receive pension benefits in the future. The analysis outlines options for reform and implications for pensions and fiscal sustainability. • Tax reform: Albania’s tax revenues are below the average for Eastern Europe, particularly for the corporate income tax (CIT) and personal income tax (PIT). VAT revenue, at about 10 percent of GDP, is above the regional average. The introduction of a flat personal income tax (PIT) of 10 percent and the reduction of the corporate income tax (CIT) in 2007 from 20 to 10 percent have reduced the scope for revenue collection. The tax policy section analyzes options for increasing revenue through tax policy and administration reform. 1.16. The chapter concludes by examining fiscal sustainability under alternative structural policy reform scenarios. It attempts to quantify the fiscal impact of the reform options, and explores the implications for fiscal sustainability of reforms in the energy, pensions, and tax policy and administration areas. 1.17. Chapter 4 examines capital spending, in particular transport spending, from the perspective of fiscal consolidation and growth. Albania has considerably expanded its national road network over the last 10–15 years. The expanded road network has also increased maintenance needs, which are three times higher than what are budgeted today. To overcome the backlog of maintenance and bring the core network to good condition, Albania Road Agency’s maintenance budget would need to be considerably expanded. The chapter makes the case for rebalancing spending toward maintenance and rehabilitation within the same spending envelope. 1.18. Chapter 5 discusses the institutional dimensions of sustainable fiscal policy: the need to reform public financial management (PFM) to strengthen financial discipline; and the need to put in place a fiscal rule as a strategic anchor to guide fiscal policy over the medium term. Recent Public Expenditure and Financial Accountability assessments (PEFAs) have shown that Albania’s PFM improved between 2006 and 2011, in particular with respect to implementation of the treasury system, budget predictability, and control of budget execution. However, challenges remain, especially with regard to planning and budgeting and control of arrears. On the execution side, the recent build-up of arrears has revealed serious loopholes in the commitment control system. The analysis gives options for improving PFM and discusses alternative fiscal rules that can be considered by the Albanian authorities. 1.19. Chapter 6 concludes and summarizes the policy options for the government 5 . CHAPTER 2. FISCAL SUSTAINABILITY IN THE ABSENCE OF REFORMS A. EVOLUTION OF PUBLIC DEBT AND DEBT DYNAMICS 2.1. Albania’s public debt-to-GDP ratio has steadily increased since 2007, reaching 61.5 percent in 2012. Between the mid-1990s and 2007, it gradually climbed to 53.4 percent. Since then, it has continued to trend upward. In 2008 and 2009 large primary deficits of 2.7 and 3.8 percent of GDP and a depreciation of the lek in 2009 pushed up debt. In 2011 and 2012 a slowdown in growth and unforeseen lending to KESh, the national electricity company further raised the debt burden (Figure 2.1a). At the end of 2012, the public debt-to-GDP ratio reached 62.4 percent—above the Maastricht threshold of 60 percent–with the debt ratio further increasing to 70.5 percent in 2013. Albania has now the highest debt to GDP ratio among the Western Balkan countries (b). Figure 2.1: Albania’s Debt-to-GDP Ratio over Time and Compared to Other Countries a. Debt Dynamics 2003–12 b. Gross Debt-to-GDP in 2011, 2012, 2013 10 64 8 80.0 62 2011 2012 2013 (proj.) 6 60 70.0 4 2 58 60.0 0 -2 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 56 50.0 54 -4 40.0 -6 52 -8 30.0 50 -10 -12 48 20.0 Primary deficit Real GDP growth 10.0 Other automatic debt dynamics Other debt creating flows Residual Change in public debt 0.0 Public debt to GDP ratio ALB BiH KOS MKD MNE SRB Source: IMF. Source: IMF World Economic Outlook 2013 (April). 2.2. Between 2012 and 2013, public debt rapidly increased by about 8 percent of GDP, principally due to a GDP revision, and electoral spending. About 5.8 percent of GDP increase in public debt was due to a GDP revision that resulted from a methodological change in the treatment of implicit subsidies to the energy sector. The remainder was due to the large fiscal deficit of 6.2 percent of GDP in 2013 (versus the original planned target of 3.5 percent deficit) and recognition of the stock of payment arrears. 2.3. With close to 80 percent of Albania’s exports destined for the EU and 60 percent for Italy alone, the euro crisis has, with a lag, spilled over to Albania. In 2008 the Albanian economy boomed as GDP grew to 7.5 percent. Though it slowed to about 3.3 percent between 2009 and 2011, it was still strong compared to neighboring countries as oil production came on stream. The increase in electricity, mineral, and oil exports compensated for the contraction of traditional exports, such as textiles, and of remittances, two-thirds of which have traditionally come from Greece and Italy. In 2012, growth fell to 1.3 percent as a severe drought hit in the first quarter of the year and as domestic consumption decelerated. 7 2.4. Albania’s fiscal deficit averaged 3.4 percent between 2005 and 2012, except in 2008 and 2009, when it deteriorated sharply (Figure 2.2). Revenues as a share of GDP climbed from 24.4 percent in 2005 to 26.7 percent in 2008 but have declined since 2010. In 2012, they reached 24.9 percent—close to the 2005 level. As revenues fell, so did spending (Figure 2.3). Personnel expenditures declined from 6.1 percent in 2008 to 5.2 percent in 2012. Capital expenditures also contracted. Social insurance spending as a share of GDP has risen steadily since 2000, reaching 7.0 percent in 2008 and then 9.0 percent in 2012. Figure 2.2: Fiscal Balance (Percent of GDP) Figure 2.3: Public Spending (Percent of GDP) 40% 0.0% 50% 30% -2.0% 20% -4.0% 10% -6.0% 0% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 0% -8.0% Capital Expenditures 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Other Primary expenditure (right axis) Social Insurance incl. Health Revenue and grants (right axis) Operations & Maintenance Interest Source: MoF. Source: MoF. 2.5. By 2012, capital expenditures had fallen to about 4.6 percent of GDP, the lowest level in the new millennium; spending on operations and maintenance also declined. Since 2008 about half of capital expenditures have been for transport, of which over 95 percent went to roads, leaving less than 5 percent for the ports and less than 0.5 percent for rail. Expenditures on roads rose in 2008 to 5.7 percent of GDP—reflecting large investments in motorways—before declining to 2.7 percent in 2010, 2.4 percent in 2011, and 1.9 percent in 2012. Spending on operations and maintenance averaged 2.6 percent of GDP between 2005 and 2012 but has been slackening since then. There is some evidence that Albania spends too little on maintaining its road network and that some reallocation from capital expenditures to operations and maintenance would make public spending more efficient (see chapter 3). 2.6. Social insurance absorbs nearly one-third of Albania’s total public spending, close to 9 percent of GDP. About 30 percent of these outlays are for the health sector and another significant part for pensions. Albania has a very young population: only 9.7 percent of its people are 65 and older, well below the EU average. In 2010, the country spent about 6.2 percent of GDP on pensions, significantly less than the average EU country, but the pension system has become increasingly burdensome; pension deficits have widened due to reductions in contribution rates and ad hoc pension increases for the rural residents. In 2010 the total fiscal deficit of the pension system was nearly 2.8 percent of GDP, including government contributions for rural residents. 2.7. Interest payments as a share of GDP are high at more than 3 percent of GDP but have declined slightly, from 3.4 percent in 2010 to 3.1 percent in 2012. Though debt increased from 58.5 8 percent of GDP in 2011 to 62.4 percent in 2012, interest payments fell from 3.4 percent of GDP to 3.1 percent as the average nominal interest rate on domestic debt dropped from 8.6 to 7.7 percent. 2 B. STRUCTURE AND COMPOSITION OF ALBANIA’S DEBT 2.8. As of end 2012, about 57 percent of Albania’s debt was domestic. More than 50 percent of domestic debt is short-term, mostly 12-month zero coupon bills. In recent years, the government has extended the average maturity of its domestic debt from almost 350 days at the beginning of 2012 to almost 450 days by April 2013. 3 The longer-term domestic debt portfolio is dominated by two- and five- year coupon-bearing bonds. Five-year bonds have floating rates, which are reset annually. Rates on all other bonds are fixed. Most of Albania’s domestic debt is in leks, but the government has been issuing a few short-term euro-denominated T-bills in the domestic market (1.34 percent of total domestic debt as of end 2012). Banks hold about 80 percent of Albania’s domestic debt, half of which is held by the Bank of Albania (BoA) and Raiffeissen Bank (Figure 2.5). Figure 2.4: Holders of Domestic Debt* Figure 2.5: Composition of External Debt* Domestic 32% short-term Multilateral 41% 24% Domestic long- term Eurobond Domestic 9% 55% Other private Guaranteed creditors 12% Bilateral External 1% 26% Source: MoF. Source: MoF. *Data as of end 2012. *Data as of end 2012. 2.9. Albania’s external public debt consists of multilateral and bilateral official loans, syndicated bank borrowing, and a Eurobond. By the end of 2012, 55 percent of Albania’s external public debt was owed to multilateral creditors, reflecting the fact that Albania reached middle-income status only in 2009. The share of non-concessional external borrowing has risen steadily, from less than 20 percent in 2006 to nearly 50 percent in 2012. Albania has raised two syndicated bank loans, one in 2007 for €200 million and another in 2009 for €192 million, which was paid back with the Eurobond issued in 2010. In 2010, the government issued its first Eurobond, which had a five-year maturity and raised €300 million. 4 Consequently, the implied interest rate on external debt climbed from 1.4 to 2.2 percent and external interest payments as a share of GDP from 0.2 to 0.6 percent between 2006 and 2012. C. A BASELINE SCENARIO: NO-REFORM 2.10. The no-reform scenario assumes that there will be no major structural reforms. No-reform projections expect real GDP growth to recover slowly over the medium term, rising from 1.6 percent in 2 The average nominal interest rate is calculated as interest payments in year (t) as a share of the debt stock as of end year (t-1). 3 This did not raise interest costs: during this period, the average yield declined from 8 percent to 7.7 percent. 4 Albania has a B1 credit rating with a stable outlook from Moody’s and B+ from Standard and Poor’s. 9 2012 to 0.7 percent in 2013 and 2.0-2.5 percent in 2014-15 and 3.0-3.5 percent in 2016-17 (Table 2.1, Figure 2.6a). This rate, though below Albania’s 10-year average of 4.6 percent, is close to its five-year average of 3.7 percent. Domestic demand will remain weak over the medium term due largely to depressed private consumption, credit and reduced government spending (due to the planned fiscal consolidation). The clearance of public payment arrears between 2014-2016 is expected to support growth through improved liquidity to the private sector. Extractive industries, agriculture, and services will continue to support growth over the medium term, but with weak domestic demand and only a gradual recovery in the Eurozone, growth is likely to remain modest at 2-3 percent in 2014-16, increasing to only 3.5 percent by 2017. Table 2.1: Baseline Projections (in Percent of GDP) 2013 2014 2015 2016 2017 2018 Revenue 24.0 24.5 24.6 24.6 24.6 24.6 Expenditure 30.1 31.5 31.0 29.0 29.0 29.0 Fiscal Balance -6.2 -7.0 -6.4 -4.4 -4.4 -4.4 Public Debt 70.5 72.1 73.5 72.7 72.8 72.9 Memorandum Item: 2.5 3 3.5 Real GDP growth 0.7 2.1 3.5 Source: MoF, Bank Staff Calculations. 2.11. Revenues are expected to remain at 24.6 percent over the medium term (Table 2.1, Figure 2.6b). The Albanian government implemented several tax reforms in recent years, such as the introduction of a flat personal income tax and a low corporate income tax in 2007–08, and most recently in 2012 and 2013 several tax reductions, e.g., a VAT reduction for machinery and a personal income tax reduction on small salaries. The law for national tax was amended in August 2011 removing the customs/import tax on used cars as well as sales tax within the country. Albania’s revenues, equal to 25.3 percent in terms of GDP in 2012, are the lowest among the South East European countries (Figure 2.7). The revenue ratio is projected to decline further to 23.7 percent in 2013 due to the changes made in tax policy. In the medium term, the ratio has the potential to increase to 24.6 percent as the government makes efforts to strengthen tax administration. Figure 2.6: Evolution of Growth and Fiscal Balance a. Real GDP Growth (Percent) b. Fiscal Balance (Percent of GDP) 10% 40% 0.0% Projection Projections 8% 30% -2.0% 6% 20% -4.0% 4% 10% -6.0% 2% 0% -8.0% 2002 2004 2006 2008 2010 2012 2014 2016 2018 0% 2002 2004 2006 2008 2010 2012 2014 2016 2018 Primary expenditure (right axis) Revenue and grants (right axis) Real GDP growth 10 year historical average Fiscal balance (left axis) Source: World Bank projection. Source: World Bank projection. 10 2.12. Non-interest expenditures are expected to drop to the pre-2008 level of about 25 percent in the medium term. In 2014, they are projected to peak at 27.8 percent of GDP as the government clears a part of the arrears estimated at, 5.3 percent of GDP. 2.13. Capital expenditures, including maintenance, are assumed to stay at about 5 percent of GDP until 2019. This is significantly below the 10-year historical average of 6 percent but still higher than the figures in the latest medium-term budget plan. 5 Maintaining capital spending at this level will be essential if Albania is to realize its medium-term growth goal of 3.5 percent. Furthermore, considering the current under-investment in maintenance and a significant backlog in road maintenance work, holding to this level of capital spending will need to be accompanied by reallocation within capital expenditure from new investment to operations and maintenance—a difficult but essential step if Albania is to realize its growth potential. 2.14. This scenario assumes the Albanian pension system will continue to be in deficit. The baseline assumes that the status-quo of the pension system prevails. Pensions are assumed to be price- index from 2014 onwards in line with current legislation and rural contributions are kept at their 2013 level. Under these assumptions, the pension deficit is projected to reach 2.9 percent of GDP in 2025. When government contributions to rural pensioners are included, the 2025 deficit reaches 3.3 percent. As a result, social assistance spending would reach 7.75 percent by 2018. The baseline projections ignore the fact that the government will probably need to provide assistance in some form to the elderly who have no pension rights. Additional social programs in support of the elderly would further increase the debt burden. The baseline projections also do not foresee any ad hoc pension increases. If pensions would grow in line with wages as has been the case in recent years, the public deficit would widen significantly. Figure 2.7: Revenue (Percent of GDP)* Figure 2.8: Debt and Gross Financing Need (Percent) 50 80% 30% Gross Financing Need (in % of Debt-to-GDP ratio 40 25% 60% 20% 30 40% 15% 20 10% GDP) 10 20% 5% 0 0% 0% 2002 2004 2006 2008 2010 2012 2014 2016 2018 External Debt (left axis) Domestic Debt (left axis) Gross Financing Need (right axis) Source: IMF WEO. Source: MoF, World Bank Staff calculations. *Refers to general government revenue. 5 The 2013 medium-term budget plan has capital spending at 3.9 percent in 2014, 3.5 percent in 2015, and 3.6 percent in 2016. The projections do not include any significant investment in railways. 11 Table 2.2: Albania’s Public Debt Sustainability Framework (Percent of GDP, unless otherwise indicated) Projection 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 Debt- stabilizi ng I. Baseline Projections primary balance 10/ Public sector debt 1/ 54.7 59.5 58.5 60.3 62.4 70.5 72.1 73.5 72.7 72.8 72.9 -0.4 o/w foreign-currency denominated 15.0 23.3 25.2 26.0 27.0 26.8 31.0 31.6 31.3 31.3 31.3 Change in public sector debt 2.1 4.8 -0.9 1.7 2.2 8.0 1.6 1.4 -0.7 0.1 0.1 Identified debt-creating flows (4+7+12) 1.7 4.6 0.1 2.6 3.0 4.5 4.3 2.9 0.3 0.0 0.0 Primary deficit 2.7 4.4 0.4 0.5 0.3 2.9 3.3 2.4 0.4 0.4 0.4 Revenue and grants 26.7 26.0 26.1 25.8 24.9 24.0 24.5 24.6 24.6 24.6 24.6 Primary (noninterest) expenditure 29.4 30.4 26.5 26.3 25.2 26.9 27.8 27.0 25.0 25.0 25.0 Automatic debt dynamics 2/ -1.3 1.6 -0.2 0.6 2.1 2.8 1.0 0.4 -0.1 -0.4 -0.4 Contribution from interest rate/growth differential 3/ -2.3 0.4 -0.2 0.5 2.1 1.8 0.8 0.2 -0.1 -0.4 -0.4 Of which contribution from real interest rate 1.4 2.1 1.9 1.2 1.9 2.2 2.2 2.2 2.0 2.0 2.0 Of which contribution from real GDP growth -3.7 -1.7 -2.1 -0.7 0.2 -0.4 -1.4 -2.0 -2.1 -2.4 -2.4 Contribution from exchange rate depreciation 4/ 1.0 1.2 0.0 0.0 0.0 1.0 0.2 0.3 0.0 0.0 0.0 Other identified debt-creating flows 0.3 -1.4 -0.1 1.5 0.5 -1.2 0.0 0.0 0.0 0.0 0.0 Residual, including asset changes (2-3) 0.4 0.2 -1.0 -0.8 -0.8 3.5 -2.7 -1.5 -1.0 0.1 0.1 Gross financing need 5/ 26.5 26.6 20.5 21.2 20.8 23.6 26.8 23.2 23.9 23.3 25.8 in billions of U.S. dollars 3.4 3.2 2.4 2.6 2.7 3.0 3.5 3.2 3.4 3.6 4.2 Projected Key Macroeconomic and Fiscal Assumptions Average Real GDP growth (in percent) 7.7 3.3 3.8 1.2 -0.3 0.7 2.1 3.0 3.0 3.5 3.5 3.0 Average nominal interest rate on public debt (in percent) 6/ 6.1 6.1 6.1 5.7 5.4 5.3 5.6 5.8 5.8 5.8 5.8 5.8 Average real interest rate (nominal rate minus change in GDP deflator, in percent) 3.1 4.1 3.5 2.2 3.3 3.6 3.3 3.3 3.0 3.0 3.0 3.0 Inflation rate (GDP deflator, in percent) 3.0 2.0 2.6 3.5 2.1 1.7 2.3 2.5 2.8 2.8 2.8 2.6 1/ Central government gross debt. 2/ Derived as [(r - p(1+g) - g + ae(1+r)]/(1+g+p+gp)) times previous period debt ratio, with r = interest rate; p = growth rate of GDP deflator; g = real GDP growth rate; a = share of foreign-currency 3/ The real interest rate contribution is derived from the denominator in footnote 2/ as r - π (1+g) and the real growth contribution as -g.4/ The exchange rate contribution is derived from the numerator in footnote 2/ as ae(1+r). 5/ Defined as public sector deficit, plus amortization of medium and long-term public sector debt, plus short-term debt at end of previous period. 6/ Derived as nominal interest expenditure divided by previous period debt stock. 7/ The key variables include real GDP growth; real interest rate; and primary balance in percent of GDP. 8/ The implied change in other key variables under this scenario is discussed in the text. 9/ Real depreciation is defined as nominal depreciation (measured by percentage fall in dollar value of local currency) minus domestic inflation (based on GDP deflator). 2.15. Under these assumptions, Albania’s public debt-to-GDP ratio is projected to reach 73.5 percent in 2015 and stay above 72 percent in the medium term (Figure 2.8 and Table 2.1). Its gross financing need will plateau at 26.8 percent in 2014 and 23.2 percent in 2015, up from 23.6 percent in 12 2013. The increase in the 2014 -2016 gross financing need is driven by the assumed arrears clearance of Lek 72 billion. 6 In 2015, financing needs remain high due to the bullet payment on the Eurobond, which it is assumed will be rolled over with external debt. Overall, foreign currency debt is projected to continue at about 43 percent of total central government debt, in line with the government’s current Medium-term Debt Management Strategy. Reliance on short-term domestic debt is expected to decline. The average interest rate on domestic debt is expected to stay at the 2012 level of 7.7 percent and on external debt to increase slightly over the medium term as Albania contracts new external, nonconcessional loans. D. SENSITIVITY ANALYSIS AND CONTINGENT LIABILITIES Sensitivity to shocks 2.16. Albania’s public debt burden indicators will significantly deteriorate if there are adverse shocks to primary deficit, growth, exchange rate and interest rate. By 2018, the central government debt-to-GDP ratio is projected to reach 76 percent if the primary deficit remains constant at 2.4 percent throughout the projection period; 81.5 percent if real GDP growth is 1 percentage point lower than projected throughout the period; and 85 percent if there is a one-time 30 percent real depreciation (Figure 2.9). 6 As of September 2013, Albania’s government arrears have been estimated at lek 72 billion, 5.3 percent of GDP. The estimate includes arrears in transport, water and sanitation, health, and education as well as VAT refunds, CIT returns, and electricity bills. 13 Figure 2.9: Albania: Sensitivity Analysis 85 Baseline and historical scenarios 28 90 Interest rate shock (in percent) Baseline 85 i-rate 80 shock 73 23 80 74 75 75 70 18 70 73 Historical67 Baseline 65 13 65 60 60 8 55 55 50 50 3 2008 2010 2012 2014 2016 2018 2008 2010 2012 2014 2016 2018 90 Growth shock (in % per year) 90 Growth PB shock 85 shock 85 83 81 80 80 76 75 73 75 70 Baseline 70 73 Baseline 65 65 60 60 55 55 50 50 2008 2010 2012 2014 2016 2018 2008 2010 2012 2014 2016 2018 90 Combined shock 3/ 95 85 Combine 90 30 % d shock 77 85 depreciat 85 80 80 ion 75 75 73 70 Baseline 73 70 65 Baseline 65 60 60 55 55 50 50 2008 2010 2012 2014 2016 2018 2008 2010 2012 2014 2016 2018 Sources: WB staff estimations based on MOF data. 1/ Shaded areas represent actual data. Individual shocks are permanent one-half standard deviation Figures in the boxes represent average projections for the respective variables in the baseline and scenario being presents ten-year historical average for variable is also shown. 2/For historical scenarios the historical averages are calculated over the ten-year period and the information is used to project debt dynamics five years ahead. 3/Permanent ¼ standard deviation shocks applied to real interest rate, growth rate and primary balance. 4/One-time depreciation of 30 percent and 10 percent of GDP shock to contingent liabilities occur in 2014, with real depreciation defined as nominal depreciation (measured by percentage fall in dollar value of local currency) minus domestic inflation (based on GDP deflator). 14 Contingent liabilities in the energy sector 2.17. The Albanian government supports the Albanian Power Corporation (KESh) with loan guarantees and ad hoc subsidies. Albania’s electricity supply has not been able to keep up with demand for the last decade, and distribution losses are high. If not enough electricity is generated to meet demand, the wholesale public supplier KESh/WPS must import electricity at a high cost, since the entire region is hydro-dependent. 7 Approved tariffs cover only the agreed, not the actual, level of losses. Costs of CEZ Sh.’s imports not covered by the tariff were lek 6.4 billion in 2011 (US$61 million) and lek 13.3 billion (US$126 million) in 2012. Imports are financed by commercial loans, which are often extended with government guarantees. As of spring 2013, government guarantees to KESh amounted to lek 29 billion (around 2.2 percent of GDP). KESh has also benefitted in the past from ad hoc liquidity injections in the form of direct loans to KESh. 2.18. The electricity sector faces a financial gap—a contingent liability-- Figure 2.10: Public Debt-to-GDP Ratio with the which over time could add Contingent Liability in the Energy Sector 85 82.0 significantly to Albania’s public debt. 79.0 80.7 82.7 As electricity demand continues to 80 73.4 grow, if no new generation is added, the 75 electricity supply gap would increase. 70 70.5 72.7 72.9 74.5 72.1 Electricity imports are projected to rise 65 from about 200 GWh in 2014 to almost 60 2000 GWh in 2020. At an expected 2013 2014 2015 2016 2017 2018 2019 2020 2021 import price of €60 /MWh, the costs for Baseline without contingent liabilities in the KESh/WPS to cover these imports energy sector would be about lek 17–20 billion (1.3- Baseline with contingent liabilities in the energy 1.5 percent of GDP) annually if retail sector tariffs are not adjusted. As a result, the Source: Bank Staff Calculations. public debt-to-GDP ratio would reach 81.4 percent in 2018, compared to 76.3 percent without the energy gap (Figure 2.10). E. IS ALBANIA’S PUBLIC DEBT TOO HIGH? 2.19. History abounds with public debt crises. Many of today’s high-income countries went through debt defaults or rescheduling during their emerging market stage of development, some as long ago as the 13th century (Reinhart and Rogoff 2010). Contrary to widespread perceptions, the probabilities of a debt crisis seem to be determined by total public debt rather than by external debt alone. History also shows that the economic and social costs of these crises can be high in that countries in debt distress face higher financing costs and lower GDP and trade. A common estimate puts default “costs” at 2 percent of GDP growth (Sturzenegger 2002). De Paoli, Hoggarth and Saporta (2006) report much higher output losses— on the order of 7 percent per year for their median country. Tomz (2007) finds that output is 1.4 percent below trend during periods of default compared to 0.2 percent above trend when the borrower is in good standing. 2.20. While empirical evidence confirms that high public debt depresses economic activity and significantly increases the probability of default, there is no agreement on what level of public debt is optimal because of the uncertainty surrounding the costs of taxation. It can be argued that zero debt 7 If not enough imports can be secured, as happened in 2007, the system has to revert to load-shedding, which have high economic and social costs. 15 may be desirable if taxes are distortionary. Moreover, a given debt ratio could be sustainable in one country but unsustainable in another. There exists empirical evidence that the higher a country’s debt level, the lower the quality of its policies and institutions; and the higher the probability of shocks, the more likely it will fall into debt distress (Kray and Nehru, 2004). While the risk of debt distress clearly tends to rise with debt, what exact level is acceptable and compatible with fiscal sustainability is to some extent a matter of judgment. Still, empirical analysis of emerging markets has shown that countries can enter debt crisis even when debt is surprisingly low: in 55 percent of debt defaults of middle-income countries, public debt was below the Maastricht benchmark of 60 percent of GDP, and in 35 percent it was less than 40 percent (IMF 2013). 2.21. High public debt carries the risk that adverse shocks will push up borrowing costs and trigger sustainability concerns. Global economic factors, such as commodity prices and country interest rates, have been found to play a major role in precipitating sovereign debt crises (Reinhart and Rogoff 2008). Risks that could rapidly lead to a buildup of debt, depending on the structure of the debt portfolio, are exchange rate and interest shocks, rollover risks, and unforeseen drops in revenues and GDP. 2.22. If much of the public debt is denominated in foreign currency, depreciation could immediately and significantly raise the debt burden, especially since currency depreciation often takes place in connection with other adverse shocks to output, exchange rates, or market confidence, which worsens debt dynamics (Eichengreen, Hausmann, and Panizza 2003). Albania’s exchange rate is an independent float that has been rather stable since 2009. 8 Still, Albania has significantly increased its share of foreign currency–denominated debt in recent years: from 32.7 percent in 2008 to 42.3 percent by 2013. A difficult external environment could reduce Albania’s export share and dry weather could raise energy import costs putting pressure on the exchange rate. 2.23. Albania is highly exposed to interest rate risks. Because about 72 percent of its domestic debt (including t-bills) has variable interest rates, it is affected by interest rate fluctuations. Since Albania only recently graduated from low-income country status, the share of external debt with fixed interest rates is still high at 73.6 percent but has been declining gradually since 2010. Exposure to the interest rate risk is also a concern because there is a high probability that global rates will rise from current historically low levels. 2.24. Albania’s current debt portfolio is also exposed to the domestic rollover risk, especially given the current uncertain external environment. About 32 percent of Albania’s central government gross debt—18.6 percent of GDP—consists of short-term domestic instruments in local currency. Commercial banks, the main holders of domestic debt, are already reducing their exposure. The Albanian banking sector is dominated by subsidiaries of foreign European banks. Demand for government debt thus depends on decisions of parent banks, which could decide (as they have in the past) to offload government paper. Banks may also become reluctant to buy Albania government debt if its risk of debt distress heightens. 2.25. High debt also reduces the scope for countercyclical policies. When a government can no longer finance its deficits, it is forced to contract spending or raise revenues. This often happens when fiscal policy must help stabilize or stimulate the economy. Inability to implement countercyclical policy can entail large economic and welfare costs. Any country would thus be well-advised to maintain some fiscal buffers, especially under the current uncertain economic circumstances. 8 In 2008, a nominal 20 percent depreciation of the lek against the euro contributed significantly to the increase in Albania’s debt burden, but up to 2011 repercussions of the euro crisis were less severe in Albania than in other countries. 16 2.26. While the threat of debt distress does not seem imminent, slow and sustained fiscal consolidation is essential to prevent the debt burden from rising further and to build up some fiscal buffers. The Albanian economy seems to be less volatile than the economies of its neighbors (see chapter 5). Still, the domestic rollover risk and exposure to interest rate and exchange rate risks are significant. F. CONCLUSIONS 2.27. With no reforms, Albania’s public debt-to-GDP ratio would stay above 72 percent over the medium term. If there are adverse shocks, debt burden indicators would deteriorate significantly. By 2018 Albania’s central government debt-to-GDP ratio would reach 76 percent if there was a constant primary deficit of 2.4 percent throughout the projection period; 81.5 percent if real GDP growth is 1 percentage point lower than projected; and 85 percent if there was a one-time 30 percent real depreciation. If the contingent liabilities in the energy sector are included, the ratio would reach 78 percent by 2018 which compares to 72.9 percent under the baseline without contingent liabilities. 2.28. While there are no established guidelines for the optimal amount of public debt, Albania seems to have eroded its fiscal buffers. It is thus important that the Albanian government adopt a medium-term program of fiscal consolidation that will decisively reduce its public debt and ensure fiscal sustainability. The next chapter discusses the structural reform areas that offer opportunities for fiscal consolidation over the medium and long term. 17 CHAPTER 3. KEY STRUCTURAL REFORMS TO ANCHOR FISCAL SUSTAINABILITY 3.1. Chapter 2 showed that, without fiscal consolidation, Albania’s public debt becomes unsustainable. It also pointed out that sustained fiscal consolidation must be underpinned by structural reforms. This chapter discusses options for structural reforms in three areas—energy, pensions, and tax policy and administration--that can help improve fiscal sustainability over the medium- to long-term. Toward the end, the chapter quantifies the fiscal impact of the reform options, and explores the implications for fiscal sustainability of alternative structural reforms. A. ENERGY SECTOR Sector Overview 9 3.2. Albania’s power sector is operated by the Albania Power Corporation, KESh (generation), Transmission System Operator, OST (transmission) and CEZ Shperndarje, CEZ (distribution). The sector is regulated by the Electricity Regulatory Authority (ERE). Both KESh and OST are publicly owned while 76 percent of CEZ Sh. is owned by a Czech public company, CEZ. In August 2004, all non- household customers were granted the right to become eligible consumers and choose their own suppliers. By the end of 2012, Albania had 7 eligible customers with annual consumption of 600 GWh per year (14 percent of all electricity sales). In January 2013 CEZ Sh’s licence was revoked by ERE and its control and operation placed under interim administration. 3.3. Albania’s internal electricity generation capacity is about 98 percent dependent on hydropower and thus sensitive to hydrological vulnerability. Dependence on hydrology poses a major risk to the country’s electricity supply and causes enormous variation from year to year in the costs of meeting electricity demand. The costs of covering growing electricity demand are particularly high in poor hydrological years when additional imports are needed to be purchased to offset reductions in domestic hydropower production. If there are not enough imports secured to cover the demand, the system has to resort to load-shedding. 3.4. KESh is responsible for the overall energy supply in Albania. KESh operates as a power generation company as well as a Wholesale Public Supplier, WPS. The WPS/ KESh buys the electricity produced by KESh at a regulated price, as well as electricity produced by Independent Power Producers (IPPs) 10 to the extent needed to serve tariff customers, at a higher prices agreed in Power Purchase Agreements (PPAs). KESh may sell any electricity not taken by the WPS, on the market. Under the Energy Law WPS/ KESh is also the “supplier of the last resort”: WPS has to buy additional electricity on the market to meet the demand of domestic tariff customers. WPS is obliged to purchase power from IPPs first (because of the obligations under the PPAs), after purchasing power generated by KESh, and then import electricity to the extent needed to serve tariff customers. 9 See Annex 2 for more detail. 10 IPPs are hydro power plants which have entered into power purchase agreement (PPA) with the WPS. 19 3.5. Historically the sector has been characterized by high levels of transmission and distribution losses, low levels of revenue collection, and high levels of bad debt. 11 To address these challenges, the Government of Albania privatized the distribution system, OSSh, in March, 2009, with the signature of a Share Purchase Agreement (SPA). CEZ, the Czech energy corporation, paid €102 million for 76 percent of OSSh shares, with a pre-agreed Regulatory Framework (See Appendix 2 for more detail on Regulatory Statement), constituting part of the SPA ratified by the Albanian Parliament. Transmission losses however have been gradually reduced and are now comparable with technical losses in developed countries. 3.6. The privatization of the distribution system did not work out as expected. Acceptance of the level of losses and bad debts happened later than expected. Retail tariffs were not timely adjusted. Bill collections from budgetary and non-budgetary institutions did not improve. Power losses increased to 51.1 percent by the end of 2012; collection rates decreased from 85 percent in 2008 to 73 percent in 2011 and to 64 percent in 2012. In the fall of 2012, CEZ Sh. had exhausted its resources, and without the support of its holding company, stopped electricity imports to cover its net losses. KESh, as the supplier of last resort, overused its hydro reservoirs to cover demand. On January 21, 2013, ERE revoked CEZ Sh. license due to noncompliance of its obligations under its license and appointed a temporary state administrator to run the distribution system. 3.7. Consequently, the energy sector in Albania remains in crisis. It is technically unreliable, a drain on fiscal sources, commercially inefficient, unable to meet growing regulated consumer demand from its own facilities, and highly vulnerable to volatile weather patterns. Sector Performance in 2007-2012 3.8. The supply gap grew during the Figure 3.1: Electricity Demand vs Generation, Imports, last 10 years increasing the need for and Load Shedding, 2007–2012 electricity imports or load-shedding. 8,000 Despite increasing demand, there has been no load-shedding since 2010 (Figure 3.1), 7,000 which has been a major achievement, 6,000 particularly in view of the rising demand (-) (-) 5,000 (-) and increasing supply gap. 4,000 3.9. The cost of imports to cover 3,000 demand has been growing in recent Load Shedding (GWh) 2,000 Net Imports (-)/ Exports (+) (GWh) years. Since the whole region is hydro- dependent, the price of imports grows 1,000 Net Generation (GWh) during dry seasons. The Albanian 0 electricity retail tariffs are designed to 2007 2008 2009 2010 2011 2012 pass through the total cost of service to the retail consumers which include the Source: The World Bank. purchase of energy at regulated price from the WPS, the transmission tariff, the distribution tariff, and a small margin to cover the Retail Public Supplier’s (RPS) operating costs. WPS’s regulated price is determined by the weighted average of the different prices paid by WPS for energy (i.e., KESh internal generation, IPPs generation, and imports). 11 Bad debt represents the amount owed to the distribution operator for sales that have not been paid and deemed irrecoverable. It is therefore written off as a loss/ and classified as an expense/ because the debt cannot be collected after all reasonable efforts to collect it have been exhausted. 20 However, WPS tariffs are usually not adjusted to reflect increased costs of imported electricity, and the burden of additional import costs during dry years has been borne by WPS/ KESh as the supplier of last resort. 3.10. Distribution losses impose a huge burden on the distribution operator (Figure 3.2). Most of the system losses occur in distribution networks. Approximately 15 percent of the total losses are technical (6-8 percent above the international average), while the balance is due to illicit use of energy. Figure 3.2: System Losses, 2001–2012 a. Distribution Losses b. Billing vs. Collections 8,000 50% 50 100.0 44 44 40% 40 36 80.0 6,000 32 36 34 30 30 30% 30 25 26 27 27 60.0 24 25 Bln Lek 23 4,000 20 20 20 21 GWh 20% 20 16 18 40.0 13 15 10 2,000 10 20.0 10% 0 0.0 - 0% 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 Energy Received by Distribution Billed Distribution Losses Source: The World Bank. Source: The World Bank. 3.11. The distribution operator is obligated to import electricity to cover the system losses. Following the privatization of the distribution sector, and as per the 2009 Regulatory Statement agreed at the time of privatization, the approved tariff only covered the agreed, not actual, level of losses. Since losses were not decreased as planned in 2009-2012 but rather increased, electricity imports became a huge financial drain on the resources and net worth of the privatized distribution company. In 2011 and 2012, the cost to CEZ Sh of imports not covered by tariff was 6.4 and 13.3 billion Lek, respectively (US$61 million and US$126 million). 3.12. Collections from tariff customers have remained low. Collections from regulated customers have fallen since 2007, reaching an all-time low in 2012 (64 percent, see below). The budgetary (government) and non-budgetary (water supply, hospitals, schools, etc.) public institutions have been consistently among those non-paying customers. At the end of 2012, the debt of public institutions was 10.7 billion Lek (US$110 million). 3.13. Faced with the high losses and poor collection rates, the distribution operator CEZ Sh. exhausted its resources and by end-2012 was insolvent and unable to pay its bills to other companies in the sector. Because of its poor financial state, CEZ Sh. essentially stopped importing electricity to cover its losses at the end of 2012. By 2013, CEZ Sh. owed about 36 billion Lek (US$343 million) to KESh, 5.5 billion Lek (US$ 52 million) to OST, 15.3 billion Lek (US$144.5 million) to traders and 2.8 billion Lek (US$26.5 million) in VAT penalties and 7.4 billion Lek (US$ 75 million) in loans to CEZ a.s. 12 CEZ Sh’s debts are booked as receivables by KESh and OST; and, with the help of a revaluation of assets in 2011, on paper they remain solvent. However, in reality both KESh and OST are financially dependent on CEZ Sh. and consequently are not viable due to lack of liquidity. 12 Including Lek 4. 8 billion assumed by CEZ a.s. upon the subrogation of the IFC and EBRD loans in 2012 21 3.14. While the direct public expenditure or budgetary support to the energy sector remains small, the indirect or contingent liabilities are very large. • Direct budgetary support to the sector in recent years included ad hoc liquidity injections into KESh, the generation company, to cover its liquidity gap as well as subsidies to the disabled for electricity bills (80 percent of energy bills of disabled customers is covered by the state; the subsidy was abolished as of May 2012). The Government owes CEZ Sh. over Lek 3 billion (about US$29 million) for the subsidy. There is no confirmed estimate of indirect support to KESh by the Government. • The indirect support to the sector, however, has been much higher. This includes guaranteed loans of KESh for electricity imports, and significant debt of budgetary and non- budgetary public institutions in electricity payment arrears discussed above. The need in commercial loans for KESh increased drastically at the end of 2012, when CEZ Sh. stopped importing energy to cover distribution losses. As a result, KESh/ WPS, as the supplier of the last resort, had to increase electricity imports to satisfy demand of tariff customers without resorting to load-shedding. Because of its own liquidity issues related to the non-payments of CEZ Sh., these loans had to be guaranteed by the state. As of spring 2013, the guaranteed overdraft of KESh was 29 billion Lek (US$274 million, or about 2.5 percent of GDP). • Contingent liabilities arising from the disputed takeover of CEZ Sh. could amount to another US$ 200-250 million plus the costs of international arbitration. This comprises the repayment of an existing loan to CEZ a.s. of Lek 7.4 billion and the reimbursement of the initial price of OSSH of €102 million. 3.15. Distribution losses above the normal technical losses (of ±8 percent), inadequate WPS tariffs, and too broadly defined block/lifeline tariffs are an implicit subsidy to consumers. • In 2012, the financial cost of this subsidy is estimated to be around Lek 15.5 billion (US$165 million), or about 1.5 percent of GDP. Before privatization of the distribution system, the subsidy was borne by the Government. After the privatization, the subsidy was supposed to be gradually eliminated, as distribution losses were reduced. It was expected that if the losses were not decreased according to the schedule outlined in the Regulatory Statement, the subsidy would be borne by CEZ Sh. However, after CEZ Sh. stopped importing to cover its losses and was taken over by ERE, the cost of this subsidy will again be borne by the government and de facto by all the companies in the sector. • The current block tariffs are not sufficiently focused and they allow for a significant number of consumers to benefit from the lower tariffs intended for the more vulnerable groups. As such this constitutes an additional subsidy. Prospects and Scenarios for 2013-2020 3.16. Absent major reforms and financial support from the government, the prospects for the sector look bleak. By most reasonable measures both KESh and CEZ Sh. are insolvent and can only continue to operate with the support of the government. The extent of this incremental support is illustrated in the following three scenarios, which comprise “Business as Usual”, and a “Break-away scenario”, incorporating normal to aggressive loss reduction programs. The level of this support assumes, however, that the three state-owned or controlled companies are at least able to meet their debt service commitments as they fall due, are successful in reaching a reasonable settlement on the dispute with CEZ a.s., and are able to fund their capital requirements from internal resources. 22 Business-As-Usual Scenario 13 3.17. The business-as-usual scenario assumes that: • Albania’s electricity demand continues to grow at the same rate as over the past 10 years, i.e. 3-4 percent per year; • normal hydrological conditions; • no new generation added to the system; and, • distribution losses gradually being reduced to the level of 2010. 14 3.18. Under this scenario, demand is expected to reach 10,000 GWh by 2020 while the expected electricity supply gap would reach 5,000 GWh (or 50 percent of total expected demand) by 2020 (Figure 3.3). Given the current structure of the electricity sector, the projected supply gap will be covered by purchasing additional imports. The risk of covering the expected gap with imports is not equally shared between the government, investors, and customers, but is entirely borne by the state. The growing supply gap would therefore put additional financial pressure on KESh/ WPS as the supplier of last resort and on publically administered distribution operator, DSO. KESh/WPS imports are expected to increase from a little over 200 GWh in 2014 to almost 2,000 GWh in 2020; DSO’s imports to cover losses are expected to stay between 2400 and 2900 GWh per year. 3.19. With a projected import price of Figure 3.3: Expected Imports to Cover the Supply Gap €60 /MWh (8.4 Lek/kWh), the costs of KESh/WPS to cover those imports would increase from 2 billion Lek (0.15 percent of GDP) in 2014 to 16 billion Lek (1.2 percent of GDP) in 2020; the costs of distribution operator to cover losses would be around 25 billion Lek (US$236 million) annually. As the result of the increased import costs, there would be upward pressures on the wholesale tariffs. Assuming that the wholesale tariffs are not adjusted and remain at the current level, there would be a need to directly support KESh and/ or guarantee their short-term commercial loans to buy electricity. The direct support to KESh Source: The World Bank estimations. could reach 4 billion Lek (0.3 percent of GDP) in 2020 from zero currently (Figure 3.4, a and b). 13 This is a conservative scenario, since it excludes capital expenditure, debt service and interest on bridging finance (funding gap). 14 If the state administrator of the distribution system decreases losses by 4 percent per year (similar losses reduction rate as outlined in the current Regulatory Statement), then the distribution system would bounce back to 30 percent distribution losses (the system’s status quo level of losses) by the end of 2016. 23 Figure 3.4: Impact of Expected Import Costs on KESh/ WPS, Business-As-Usual Scenario, 2013–2020 a. Estimated Average Cost of WPS* vs. Current b. Expected KESh Financial Gap Approved Wholesale Tariff 5.0 4 3.8 4.1 4.2 4.0 3.5 3.2 3.7 3.8 4.0 3.2 3 2.7 3.0 2.7 2.5 2.2 Lek/kWh Billion Lek 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2.2 2 1.7 1.9 2.0 1.5 1 0.7 1.0 0.5 0.1 0 0.0 0 2013 2014 2015 2016 2017 2018 2019 2020 2013 2014 2015 2016 2017 2018 2019 2020 Average estimated cost of WPS Current approved WPS wholesale tariff Source: The World Bank estimations. * Estimated Average Cost of WPS: average cost WPS bears from purchasing electricity from KESh, IPPs and traders (imports). 3.20. Similarly, assuming that the distribution losses are reduced to the level of 2010 (30 percent) by 2016, but the loss coverage allowed in the tariff stays at the same level as defined in the Regulatory Statement, i.e. 17 percent after 2014, the DSO would remain a loss-making company: the expected financial gap would be around 9-10 billion Lek (US$85-95 million) and by 2020 only 52 percent of import costs would be covered by the tariff. Since the DSO is a public sector company, those losses are essentially a part of public obligations to the sector (Figure 3.5). The expected financial gap will decrease by 2020 compared to 2013 due to reduced distribution losses and increased sales. Figure 3.5: Expected Financial Gap of the DSO, Business-As-Usual Scenario, 2013–2020 a. Estimated Average Cost of the DSO vs. Current b. Expected Financial Gap of the DSO Approved Distribution Tariff 10 9.6 45 7.00 40 40 41 8.6 39 38 39 9 40 36 38 6.00 8 7.5 35 31 32 6.5 6.5 6.5 6.5 6.5 29 30 5.00 7 30 28 25 TWh Lek/kWh Billion Lek 6 5 5 5 5 5 5 5 5 23 4.00 25 21 5 20 3.00 4 15 3 2.00 10 2 5 1.00 1 0 0 0.00 2013 2014 2015 2016 2017 2018 2019 2020 2013 2014 2015 2016 2017 2018 2019 2020 Total cost of DSO, including operating costs Approved DSO revenue Current approved DSO tariff Average cost of DSO Expected sales, TWh Source: The World Bank estimations. 3.21. The retail tariffs are supposed to pass through the cost of buying electricity from the WPS plus the transmission tariff plus the distribution tariff plus a small margin to cover the RPS operating costs. Under the business-as-usual scenario, where the retail tariffs are not adjusted properly to reflect the full costs of electricity and collection rates do not improve but stay at 83 percent as expected for 2013, the 24 expected financial gap of the sector would be in the order of 17-20 billion Lek (US$161-189 million) annually (Figure 3.6). This is a lower bound estimate of the financial gap. Figure 3.6: Expected Average Cost of Delivering Electricity to Final Consumers vs. Current Approved Retail Tariff; Expected Financial Gap of the Electricity Sector, Business-As-Usual Scenario, 2013–2020 a. Average Cost of Delivering Electricity to Final b. Expected Financial Gap of the Electricity Sector Consumers vs. Current Approved Retail Tariff 14.0 12.5 25 12.0 22 11.4 10.9 11.1 11.2 11.3 11.4 21 12.0 20 20 19 18 18 9.5 9.5 9.5 9.5 9.5 9.5 9.5 9.5 19 17 10.0 Billion Lek 15 Lek/kWh 8.0 6.0 10 4.0 5 2.0 0 - 2013 2014 2015 2016 2017 2018 2019 2020 2013 2014 2015 2016 2017 2018 2019 2020 Financial gap of the sector, unchanged collection rate of Average cost of delivering energy to final consumers 83% Approved average retail tariff Source: The World Bank estimations. 3.22. The business-as-usual scenario referred to above is clearly unsustainable. Given that the state bears all the system supply risk, it poses a threat that it may become more difficult for the state to secure enough imports to cover Albania’s growing supply gap. If WPS and distribution costs are not passed on to the final consumers, the system may be forced to resort to load shedding, which would have damaging consequences to the economy and growth. The sector will have no funds for either operation and maintenance or investments to sustain its operation and improve its efficiency. This could potentially lead to devastating consequences and the sector could simply collapse. In addition the sector needs investments to diversify generation, re-commission the Vlore thermal power station to mitigate, in part, the climatic volatility, and reduce distribution system losses. If these capital needs are incorporated into the calculus, then the estimated financial gap of the sector would be much higher. A Break-Away Scenario: Restoring Energy Sector Sustainability 3.23. The sector is on the verge of banruptcy. The annual financial losses and debt service requirement of the magnitude projected is simply not sustainable without improving financial viability and sufficient funds to support an aggressive investment program to improve the condition of the distribution system as well as diversify and expand generation to cover growing supply gap. A break- away scenario assumes: • investment in new generation and diversifying the generation mix; • decreasing the distribution system losses through the design and implementation of a comprehensive loss reduction program; • enforcing tariff adjustments to reflect the full cost of electricity supply; and • opening up the Albanian electricity market to regional competition. 3.24. Investing in internal generation and diversifying electricity generation mix would decrease the electricity supply gap and reduce the need for imports; hence it would decrease the risk of Albania 25 not being able to secure its electricity supply. It is estimated that generation investment needs of Albania are at least US$780 million to cover the expected supply gap. Given that the generation investment needs are enormous, the immediate step should be to ensure full operation of Vlore thermal power plant. When Vlore becomes operational, WPS would not be as exposed to climatic conditions and its need to import would be reduced by at least 500 GW annually. It would also improve security of supply and reliability of the system and reduce the risk of load shedding during dry periods. 3.25. It is also important that private investors are attracted to develop additional hydro capacity in the country. However, the government needs to be cautious in contracting new liabilities. Currently, the PPAs with private medium and small HPP concessions are such that KESh/ WPS is obliged to buy their electricity at a 10 percent above import price. As a result, the estimated financial liability of KESh/ WPS towards IPPs is estimated to be over 30 percent of the total WPS costs in 2013 despite the fact that IPPs electricity would cover only 4 percent of expected demand. Given the increasing needs in new generation and imports to satisfy a growing demand, it is important to ensure that there is proper risk- sharing, so that new investments’ risks are not borne by the state alone. 3.26. In order to increase the financial viability of the sector, the DSO needs to introduce a constructive, realistic and affordable loss reduction program. The business-as-usual scenario assumed a loss reduction plan to reach the level of 30 percent losses by 2016. This however is very slow, not very practical and is not a sustainable solution for the sector. Albania currently has a high level of illicit use of electricity and poor proper metering and consumption based billing system (about 20 percent of tariff consumption is not metered, including large tariff consumers) and poor information management system. Other countries that faced similar problems have managed to decrease their distribution losses much faster and without much additional investments. 3.27. Under a normal loss Figure 3.7: Expected Financial Gap of the Electricity Sector, reduction scenario, the DSO Business-As-Usual, Normal and Aggressive Scenarios could achieve a much larger loss 30 reduction of 4-6 percent per year till 2016 through improving 21 22 1 19 18 20 metering of large tariff customers, 20 18 17 Billion Lek improving network management 15 13 11 11 12 and decreasing the illicit use of 11 10 10 10 7 8 77 electricity. Combined with 7 5 6 5 improved customer outreach and improved collections, including 0 from budgetary and non-budgetary 2013 2014 2015 2016 2017 2018 2019 2020 intuitions, by 3 percent annually Financial gap of the sector, Business-As-Usual until collection reaches 95 percent Financial gap of the sector, normal loss reduction program (a common collection rate in the region), these measures would cut Source: The World Bank estimations. the financial gap of the sector by almost half compared with the business-as-usual scenario (Figure 3.7). 3.28. Under an aggressive loss reduction scenario that would include extensive work with large and medium customers, introducing a 100 percent consumption-based billing and eliminating electricity theft, the distribution losses could easily be reduced by 10 percent annually for the first three years of the program. In order to reduce losses further, additional investments would be required. Those would include modernizing the network, improving its maintenance and upgrading its management. These investments could be in the order of minimum US$50 million over the three years. The result of such an aggressive program could be cutting the financial gap of the sector by over 2/3 compared to the business- as-usual scenario (Figure 3.7). 26 3.29. To completely eliminate the financial gap in the sector, the end-user tariffs would need to be adjusted to reflect the true cost of electricity supply delivery to the final consumers. Figure 3.8. depicts how the final cost under the three scenarios discussed above (BAU, Normal, and Aggressive) compare to the current approved tariffs. Figure 3.8: Average Cost of Delivering Electricity to Final Consumers under Different Scenarios vs. Current Approved Retail Tariff 15.0 12.5 12.0 11.6 11.0 11.4 11.1 11.2 11.3 11.4 10.5 10.8 10.9 10.7 10.8 10.9 11.0 9.9 10.2 10.1 10.0 10.2 9.9 9.9 10.0 10.0 9.5 9.5 9.5 9.5 9.5 9.5 9.5 9.5 Lek/kWh 5.0 0.0 2013 2014 2015 2016 2017 2018 2019 2020 Average cost of delivering energy to final consumers, Business-As-Usual Average cost of delivering energy to final consumers, normal Average cost of delivering energy to final consumers, aggressive Source: The World Bank estimations. 3.30. The initial estimated increase in end-user tariffs would be in the order of 10-30 percent, depending on the program implemented by the DSO. Such an increase needs to be accompanied by the reform of the social assistance to improve its targeting to make sure that the most vulnerable are protected from the negative consequences of the reform. Tariff reform also should be accompanied by a public awareness campaign to explain to the population the rationale of the tariff reform and planned mitigation measures to cushion its impact. 3.31. In order to share the risks between the government, investors and consumers, the market needs to open up: all the large consumers need to acquire their electricity in the market rather than from the DSO (i.e. de facto becoming eligible); the function of the supplier of the last resort should be removed from KESh and eliminating WPS; the responsibility to supply tariff customers should be fully given to the DSO. Such a market structure is consistent with EU Directives and adopting it would bring Albania closer to the regional integration. Moreover, implementation of existing market rules should be strictly enforced. In current Albanian electricity market, the existing market rules are poorly implemented and there is poor payment discipline in the sector. Lack of implementation of market rules and a related incentive/penalty structure prevents a competitive marketplace for energy from developing to send proper price signals for efficient consumption and investment decisions. Policy Options 3.32. The sector is in crisis. The annual financial losses and debt service requirement of the magnitude projected is simply not sustainable. Incremental reforms need to be successfully implemented and supported by well-functioning institutional arrangements and capacities. This will entail a number interconnected measures by Government, KESh, OST, DSO and ERE : 3.33. Government: • Settle intercompany arrears and establish a mechanism to keep all payments current thereafter. Verification and set-off of arrears comprising the purchase by MOF of CEZ’s MoH, VAT, and public institutions receivables at face value amounting to about ±US$ 160 27 million, which in turn would be used to reduce arrears to both KESh and OST in a like amount taking into account the rights of other creditors (including CEZ a.s.). 15 This will regularize ERE’s decision on bad debts, 16 strengthen the companies’ balance sheets, improve their domestic borrowing capacity and put them on a path to recovery. Thereafter the government should actively support DSO’s efforts to collect bills from strategic institutional consumers. • Increase investments in additional generation capacity in order to reduce reliance on imports to cover the projected supply gap. This should include re-commissioning the Vlore thermal plant as soon as possible and attracting private capital to develop small and medium concessions. • Ensure that, for new investments, risks are better shared between the state and private investors. This could include renegotiations of recently concluded PPAs to eliminate the market premium borne by KESh and also share the risks more equally between the state and private investors. • Explore ways to reduce fluctuations in power supply and smooth the costs of energy between dry and wet periods (e.g., through efficiently operating stabilization fund or hedging/insuring against dry seasons). 3.34. Distribution System Operator: • Improve the operation and corporate governance of the distribution company. Initially this could be done by identifying and recruiting an efficient and cost effective operator to take control of the company and implement the agreed turn-around plan under a performance based management contract. • Introduce a more aggressive loss reduction program: Engage with large and medium customers, improve network management, increase in consumption-based billing and decrease in illicit use/theft of electricity should all be a part of the plan. • Increase collection rates from tariff customers, including better customer outreach and support thorugh a joint public awareness campaign. 3.35. Energy Regulatory Agency: • Reform and fully enforce the regulatory framework. One of the key elements in protecting the public interest is the avoidance of any residual regulation which effectively prevents the achievement of financial viability by the operating electricity entities. The public is not well served by a regulatory framework that undermines the electricity utilities ability to raise capital for modernization and expansion. Despite the expected public opposition, electricity tariffs must be further rebalanced and realigned with underlying costs (including those stranded by previous decisions). Such rebalancing is a precondition for encouraging efficient entry into the generation segment of the market—and consequently for fulfilling the electricity sector’s economic potential. 15 These official arrears can then, with the support of the donor group, be settled internally by the government and its agencies. 16 ERE’s decision to include a bad debt provision of only 16.85 percent in the retail tariff for 2012-2014 is predicated on the basis that the arrears that budgetary and non-budgetary institutions should not be regarded as bad debts by the bad debt study undertaken by Deloitte pursuant to the CEZ Share Purchase Agreement 28 • Regionalizing regulatory policy. In a region that includes small countries like Albania, regional policy coordination and regulatory cooperation might represent a pragmatic approach to dealing with the problem of limited domestic regulatory capacity and the high fixed costs of regulation. The regionalization of regulatory policy could advance domestic electricity reform and enhance regulatory credibility. • Reform of social assistance. This would include appropriate changes to the block tariffs to improve targeting to exclude consumers with the ability to pay, and ensure the protection of only the most vulnerable. 3.36. Finally, the Albanian electricity market model should be amended to comply with the requirements of the EU Directives. This means that all commercial consumers could start buying electricity in the market, KESh’s role as a supplier of last resort could be removed (de-facto eliminating WPS); and the DSO could become responsible for a consistent supply of energy to the tariff customers. Governance of the sector could be improved, market rules should be enforced, and the transparency of the sector increased. B. TAX POLICY AND ADMINISTRATION 3.37. At about 24 percent of GDP, Albania has one of the lowest tax revenue intakes in Eastern Europe (Figure 3.9). Both its Personal Income Tax (PIT) and Corporate Income Tax (CIT) revenues are flat at 10 percent and below the average for Eastern Europe, though VAT revenue, at about 10 percent of GDP, is above the regional average (Figure 3.10). The flat PIT rate at 10 percent and the reduction of the CIT rate from 20 percent to 10 percent in 2007 have limited the scope for bringing in more revenues within the current tax regime. Figure 3.9: Total Tax Revenue, Eastern Figure 3.10: Albania: Composition of Tax Revenue European Countries (Percent of GDP) (Average, 2008–10) Property Small Other taxes and local business Customs 6% taxes tax duties 3% 1% 3% Profit tax 6% VAT Personal 40% income tax 10% Excise tax 13% Social insurance contributio ns 18% Source: World Bank; MoF. Source: World Bank; MoF. 3.38. Tax policy in Albania is based on a mix of indirect and direct taxes. For indirect taxes there has been a significant move from distorting trade taxes to the VAT and excise taxes (see Annex 1). The VAT, currently at 20 percent, generates most indirect tax revenues. The share of direct taxes is relatively small: CIT accounts for only 6 percent and PIT for 10 percent of total revenue, including social security contributions. 29 3.39. This section reviews Albania’s tax policy with a view to identifying policy and administrative options for making Albania’s public finances more sustainable. The slowdown in growth since 2008 has increased the pressure on public finances, especially revenues. One goal is to see which aspects of the present tax structure might usefully be altered in ways that are congruent with bringing in more revenue while also reducing adverse tax impacts on economic growth and efficiency. Measuring Performance: Tax Productivity and Tax Gap 3.40. A country’s tax revenue performance can be assessed using a variety of methods. One method is to examine the productivity of its tax revenues. Another method is to measure the tax gap for the different categories of taxes. In the following section, we assesses the tax productivity and tax gaps in Albania for the principal taxes— the VAT, CIT, PIT and excise taxes Tax Productivity 3.41. Tax productivity is defined as tax collected as a percent of the tax base (GDP or consumption) divided by the standard tax rate. For VAT productivity the base is total consumption. For PIT and CIT, the base is GDP. In that sense, productivity measures how much of the tax base is taxed and how much is left untaxed, either due to tax policy choices such as exemptions, loopholes and tax holidays that narrow the tax base, or due to weaknesses in administration, corruption and collusion. In an ideal world, productivity would be 100 percent if the entire base is taxed, there are no exemptions and there is no leakage due to tax evasion or collusion. 3.42. The productivity of VAT in Albania is 53 percent; the average for Eastern Europe is 66 percent. This means that in Albania, for every 1 percent of the VAT tax rate, the tax administration collects VAT on only 53 percent of total consumption (the base for VAT), leaving 47 percent untaxed. This may be because the policy is to exempt certain sectors or transactions; administration of the tax is inadequate; or corruption and tax evasion make it impossible to tax the full base. 3.43. The productivity of PIT at 6 percent and CIT at 17 percent is even lower. For PIT the productivity rate means that only 6 percent of the base (GDP) is actually taxed. The productivity of CIT is only slightly better (the averages for Eastern European countries are 7 percent for PIT and 16 percent for CIT). Clearly, in both there is scope for better revenue performance through, e.g., reducing exemptions and tax loopholes and building up the capability of the tax authorities to detect tax fraud through measures such as data mining and third-party information-matching. Also, the opportunities for better revenue performance are much higher for income taxes, especially PIT, than for VAT. Tax Gap 3.44. A tax gap is defined as the difference between the amount of taxes the government should collect and the amount of taxes the government actually collects for a given tax regime. It points to areas of revenue where room exists to expand collection based on the existing tax regime. As part of this review, a detailed analysis of the tax gap was conducted for Albania using both the top-down and bottom- up approaches for the main taxes, that is, VAT, PIT, CIT and excise taxes. 17 The results of this analysis 17 For the tax gap analysis the following models/approaches were used: (i) in the bottom-up approach, the audit information from the tax department was used to estimate the tax gaps in VAT, CIT and excises.17. In the top-down approach, the consumption side model applying the 2009 Supply-Use Table (SUT) and aggregated data from VAT returns was used to estimate the VAT tax gap. For the tax gap in CIT, data from the enterprise survey were used. For the tax gap in PIT, in the absence of income distribution data, expenditure distribution by deciles of income was used together with data from LSMS of 2008. For excise taxes on tobacco and alcohol, the consumption data from 30 were as follows: 3.45. The VAT gap: The analysis indicates that it may be possible to improve VAT collection in Albania by 10 to 20 percent. Based on analysis of the sales and use tax (SUT), the tax gap was found to be about 10 percent; based on audit information, the gap was estimated to be 16 to 20 percent. 3.46. PIT and CIT tax gaps: Analyzing the PIT tax gap using monthly household expenditure from the 2008 Living Standards Measurement Survey (LSMS) shows that the PIT revenue can be considerably enhanced from the current level of lek 28 billion. The PIT revenues could be at least doubled with more efficient policies and more effective tax administration. Audit information also suggests that the CIT tax gap is about 38 percent. If no provision is made for non-random audits, the evasion potential reaches almost 66 percent. 18 This shows that there is much higher potential for raising more revenue from CIT than from VAT. This is not surprising; there are far more opportunities for underreporting profits or overstating expenses with regard to CIT. Finally, although the tax gaps could not be estimated due to data constraints, anecdotal evidence suggests that there is some scope for increasing revenue collection on tobacco, alcohol, and fuel. Tax Policy Options This section discusses tax policy options for making Albania’s public finances more sustainable. These policy options relate to: (i) the broadening of the VAT base; and (ii) increasing the revenue from income taxes, both CIT and PIT, through rate increases. Broadening the VAT Base 3.47. The basic VAT structure in Albania is generally sound. The VAT law has only two positive rates (20 and 10 percent); a zero rate is applied only to exports; and few goods and services are exempted. The VAT is payable for: (a) all taxable supplies of goods and services realized against the payment by a taxable person, as part of economic activity in the Republic of Albania; and (b) all imports of goods into Albanian territory. The VAT threshold for registration is lek 5 million of annual business turnover. All legal and physical persons involved in exporting or importing must register to pay the VAT. 3.48. However, there are serious deficiencies with the narrowing of the VAT tax base for real property transactions and capital goods imports. 19 Broadening the VAT base in these areas would make it possible to mobilize more revenue from the VAT. VAT and Real Estate 3.49. One of the biggest sources of potential revenue loss for Albania is not taxing new residential construction. Albania’s practice differs from that of what most other countries do. Article 20 of the Law on VAT exempts sale of new construction and instead taxes rentals. In the EU, although both sales and rentals of existing real estate are exempt, new buildings and improvements are subject to VAT. Applying the tax to new buildings in effect amounts to charging a prepaid VAT on future services (for use or subsequent sale) at the time of purchase. Increases in the value of property are typically not taxed. LSMS of 2008 was used. 18 The analysis of the tax gap based on the enterprise survey did not produce a robust result. It may be that the enterprise survey data does not reflect total profit in the economy because enterprises are underreporting profits. 19 Agriculture is also virtually zero-rated. Any changes in taxation in this sector require deeper fiscal-economic and social analysis that is beyond the scope of this report. 31 3.50. Albania could, consistent with the EU model, tax the sale of new construction, including the land value, to generate additional VAT. Those who do not sell new construction (self-consumption or government construction or other arrangements that avoid selling) will absorb input taxes fully. Businesses that are not exempt will get full input credit. Households and government will pay the tax if they purchase fully built units. 3.51. What would be the revenue impact of this? Assuming that the ratio of land value to gross fixed capital formation in construction is a very conservative 1:1, the total potential taxable value is around LEK 700,000 million. If total exempt production (including for threshold reasons, purchase of goods and taxed services, construction attributed to government and other adjustments) was seventy percent of this figure that would still leave LEK 210,000 million that could be taxed annually. Of this, LEK 10,000 million is output of new residential construction and the rest is land value. Assuming a 50 percent compliance rate, Albania can still collect LEK 10,000 million from this sector at the maximum VAT rate of 20 percent – which amounts to about 1.5 percent of GDP (or a 15 percent increase in VAT collection from the current level). VAT Exemption on Capital Goods Imports 3.52. Recent VAT exemptions on imports of capital goods have not only reduced revenue but are also inconsistent with the EU VAT system. In April 2013 the VAT law was amended to exempt imports of capital goods by all businesses from the VAT and customs fees. A few years earlier, the government had exempted imports of equipment for investments that cost more than US$500,000 and removed VAT on steel and cement used for the construction of hydropower plants (HPP). Equipment imported for the petroleum exploration and development is also exempt. Besides narrowing the tax base, these exemptions created huge opportunities for misreporting the end use: it would be impossible for any tax authority to control the end use if steel or cement, commonly used in construction, is diverted for uses other than a HPP. 3.53. It is important that the government reverses these VAT exemptions, both to stop erosion of the tax base and to shift back to the EU-standard VAT. Many countries allow deferred VAT payments for imported machinery to help businesses with liquidity and investment cash. In that case, the VAT collection is deferred for a fixed period, say, a year or until the plant starts production. When sales commence, the firm notionally pays the VAT on capital equipment and then can claim input credit. In effect, after this adjustment, the VAT is collected on the firm’s sales transactions. Albania has in recent years introduced, in addition to this deferral system, a general exemption of capital equipment from VAT payments. The revenue impact of reversing these exemptions could be a one-off increase of as much as 0.5 percent of GDP. Strengthening Income Taxes 3.54. Income taxes together contribute just 3.7 percent of GDP in Albania, which is very low by regional standards. The lower productivities of PIT and CIT compared to that of VAT suggest that these taxes offer even greater opportunities for raising revenue in Albania. Below we discuss those opportunities – and what might be their revenue impact. 32 Figure 3.11: Corporate Tax Rates, Eastern European Countries (Percent) Source: Tax at a Glance. Corporate Income Tax 3.55. Driven by a region-wide race in the Balkans and several other European countries to reduce CIT, Albania cut its rate from 20 percent to 10 percent in 2009, as did Bulgaria, Bosnia and Herzegovina, Kosovo, FYR Macedonia, Montenegro, and Serbia. The logic was that this would make Albania a favored destination for foreign investments and reduce informality and corruption. Recent surveys, however, have shown that informality has not been significantly reduced. With regard to FDI, since all countries in the immediate neighborhood reduced their rates, the effect seems to have been neutralized, with investment inflows depending more on the general business environment (physical and social infrastructure, healthy and skilled labor). With the reduction in the CIT rate, Albania now has one of the lowest CIT rates in Eastern Europe (Figure 3.11). Most Eastern European countries have rates of 19 to 21 percent. As a result of the low CIT rate, this tax contributes only 6 percent of total tax revenues. 3.56. Albania could raise its CIT rate to at least 15 percent to increase revenue. The race to the bottom caused revenues to decline. Serbia, like some other countries, has since raised its CIT rate from 10 to 15 percent. The revenue loss can be bridged in part by raising the CIT rate to 15 percent without imposing a heavy burden on businesses. The potential revenue increase from this measure would be up to 0.3 percent of GDP (or a 10-15 percent increase from the current level, from around 10 billion LEK to about 15 billion LEK) if backed by measures that would ensure higher average compliance. Personal Income Tax 3.57. Similarly, there is some room to enhance revenue, and more importantly equity, through adjustments to the PIT regime. All individuals resident in Albania are subject to taxes for all incomes earned anywhere in the world, though nonresidents are subject only to taxes on incomes earned within Albanian territory. 3.58. Albania has a flat PIT with a monthly threshold, but many countries in Europe have progressive rates (Table 3.1). Even outside Europe, a vast majority of countries have progressive PIT rates. Making PIT more progressive can improve equity and fairness. The revenue impact of making the PIT progressive is modestly positive or negative, depending on how the rates are structured. Assuming that the current universal deduction of 30,000 LEK would continue; all incomes from 30,000 – 150,000 LEK a month are taxed at 10 percent; and a simple progressive tax of 15 percent is introduced for those earning more than 150,000 LEK, a simple calculation shows that this could yield a very modest additional revenue (of about 0.1 percent of GDP), but the actual impact will clearly depend on how the progressive rates are structured as well as the whether the current compliance rate could be maintained. 33 Table 3.1: Eastern European Countries with Progressive and Flat PIT rates Progressive Rates Flat Rates Albania, Belarus, Bosnia Herzegovina, Armenia, Azerbaijan, Bulgaria, Croatia, Kosovo, FYR Macedonia, Romania, Moldova, Montenegro, Poland, Russia, Serbia, Turkey, Ukraine Source: Tax at a Glance. 3.59. A more progressive PIT rate should be complemented by audits of PIT returns. At present there is audit only of firms with respect to the PIT withheld from employees. Given the large PIT tax gap, audits might yield substantial revenue. The introduction of universal filing for PIT two years ago was a move in the right direction but it would need to be complemented by careful audits in order to maintain or even increase revenue collection from this tax. Small Business Taxation 3.60. Firms in Albania with annual turnover of no more than lek 8 million are considered small businesses whose tax accounts are mostly administered at the local level. Small businesses with turnover of 2 to 8 million lek prepay a fixed tax to local authorities, but they must submit an annual return to the General Directorate of Taxes (GDT), supported by a simplified balance sheet. The small business tax paid to the local authorities is credited to the Annual Return file of the central tax administration. The system generally works well for small businesses. Tax Administration Issues 3.61. An important factor in any tax reform is the capacity of the tax administration and the general institutional context within which it operates. Weaknesses in the tax administration to detect and combat tax evasion may result in significant gaps in tax collection. As recent IMF and World Bank reports have pointed out, modernizing and improving tax administration remains an important challenge in Albania. 3.62. Countries that have a large informal sector typically have a high tax gap and poor tax administration. According to a recent study on the size of the shadow economy 20, the shadow economy in Albania was estimated at 35.1 percent of GDP, close to the average for Eastern Europe. Collusions between tax officials and tax evaders may aggravate the size of the shadow economy. In Albania, according to the Transparency International Corruption Barometer, respondents ranked corruption as a serious problem 21. A high share of the shadow economy, corruption and inadequate and inefficient legislative and accounting institutions are related to high levels of tax evasion and poor tax administration. 22 3.63. A recent World Bank report 23 discussed several tax administration measures that the General Directorate of Taxes (GDT) can consider to improve tax compliance. Two of the measures are worth highlighting: 20 Schneider, Friedrich and Andreas Buehn (2007). “Shadow Economies and Corruption All Over the World: Revised Estimates for 120 Countries”, Linz, Germany. 21 Transparency International Corruption Barometer 2013.Transparency International. 22 McGee, Robert W. and Tatyana B. Maranjyan (2006) “Tax Evasion in Armenia: Causes, Cures and Policy Implications”. Andreas School of Business Working Paper, Barry University 23 Albania Tax Mission Aide Memoire, April 10-19, 2011, World Bank. 34 • Establishing an Active Tax Payer List (ATL): A high percentage of the GDT taxpayers do not file their returns in Albania. In 2011, for example, approximately 18 percent of VAT payers, 41 percent of Profit Tax payers, 23 percent of PIT payers, and 30 percent of Social Insurance payers did not file their returns. This problem can be addressed by the GDT in multiple ways. First, by cleaning up the tax register which contains a large number of inactive taxpayers, which were registered by the Central Registry without adequate verification that they would qualify for taxes. The inactive taxpayers could be removed from the register after a verification that they are no longer in business or otherwise liable to taxation. Second, by introducing an Active Taxpayer List (ATL)--a list of those taxpayers who are complying regularly with their filing requirements. The idea is that GDT will produce an initial list containing all compliant taxpayers; after that, if a taxpayer fails to file their returns for a given period (say, 3 periods for VAT or 2 periods for annual income tax for example) they will be removed from the ATL. Once the taxpayer files the missing returns, the system will re-introduce them in the ATL immediately. The GDT portal will provide a TIN look-up facility that will allow anyone to verify if a taxpayer is on the ATL. Only taxpayers appearing in the ATL may issue valid VAT invoices and carry out imports and exports. The GDT systems will identify the misuse of VAT invoices issued by non-ATL taxpayers and indicate this to the GDT taxpayers automatically when they e-file their tax sales and purchases. The program to help taxpayers prepare their VAT returns will be modified to check the supplier’s TIN in the ATL (via the GDT portal). If the supplier is not in the list, the program will warn the taxpayer that his supplier is not authorized to issue VAT invoices. Customs will not process imports or exports for non-ATL traders. The ATL is a variation of the EU VAT Information Exchange System (VIES) system, which was set up with the introduction of the single market on 1 January 1993. • The ATL chould be implemented gradually; first, for VAT filers and then for profit tax filers, and finally for other taxes. At the beginning, the system would only provide warnings to taxpayers who use invoices from non-ATL taxpayers. After a certain period (e.g. 3-6 months from implementation date), the credit for invoices issued by non-ATL taxpayers should be denied. Income deductions for purchases from non-ATL taxpayers should be disallowed as well. Also, imports and exports from non-ATL taxpayers should not be released by Customs unless they regularize their situation with GDT. • Reducing the limit for cash payment from Lek 300,000 to Lek 150,000: The implementation of threshold for cash payments was a very important tool to fight informality and to facilitate audit. As of end-2013, the threshold established in the Tax Procedure Code is Lek 300,000. This means that all payments for transactions above this threshold are made by check or go through banks, which would leave a documentary trail for VAT credits and payments. This threshold could further be reduced to Lek 150,000 to reduce cash transactions and minimize opportunities for the nonpayment of VAT in the informal economy. 24 Representatives from the private sector agree that reducing the threshold is a reasonable next step to further reduce the informal economy. 3.64. Last but not least, the government can consider using the tax gaps calculated in this report as a baseline to establish targets for reducing the tax gaps in VAT, CIT, and PIT and define the associated enforcement initiatives. The success of such enforcement initiatives could then be evaluated based upon their effect on reducing the gaps. 24 From January 1, 2014, the government has reduced the cash payment threshold to 150,000 Lek. 35 Policy Options 3.65. The analyses above indicate that Albania can raise its revenue performance by reducing exemptions and tax loopholes, expanding the tax base in particular for the VAT, increasing the rates for CIT and PIT, and building up the capability of the tax authorities to detect fraud, reduce tax evasion, and improve compliance. The government can consider the following policy options to increase revenue: • Expand the VAT tax base: One of the biggest sources of potential VAT loss for Albania is not taxing new residential construction. Article 20 of the Law on VAT exempts sale of new construction and instead taxes rentals. In the EU, both sales and rentals of existing real estate are exempt from the VAT but not new buildings and improvements. Albania could follow the EU model and exempt rental property and tax the sale of new construction, including the land value. • Reverse the recent VAT exemptions: Recent VAT exemptions on imports of capital goods have not only reduced revenue but are also inconsistent with the EU VAT system. Reversing these exemptions could both stop the erosion of the tax base and shift Albania to EU- consistent VAT standard. • Raise the CIT rate. Albania can raise its CIT rate to at least 15 or even 20 percent without imposing a heavy burden on businesses or losing any regional competitiveness. • Make the PIT progressive and introduce risk-based audits: Albania could move toward a progressive PIT, complemented by introduction risk-based individual audits, to increase revenue. 3.66. Finally, the large tax gaps calculated for Albania can only be closed with significantly strengthened tax administration and enforcement of existing policies to reduce opportunities for corruption and tax evasion. Building up the capability of the tax administration to detect fraud and reduce opportunities for evasion—for example through establishing an active tax payer list or reducing the threshold for cash transactions—is thus critical. Still, strengthened tax administration and reforming tax policy are not alternatives; they are complements that for best results must be pursued simultaneously. C. PENSION REFORM 3.67. The Albanian pension system, established in 1947, provides old age, disability, and survivor coverage to contributors. Almost all the elderly are covered, mainly because almost all workers participated in the pension system during the socialist era, but the number of workers now making contributions is low, so that its coverage of the elderly will undoubtedly decline. 3.68. Although spending on pensions in Albania is lower than the average for Eastern Europe, it is rising. In 2012 Albania spent 5.2 percent of GDP on pensions, much less than the EU countries or the average for Eastern Europe. Albania is still younger than most of the other countries in Eastern Europe and thus has fewer individuals to support through pensions. Pension spending is still low even though the pension program has to finance several generations of elderly, most of whom have full pension rights. Until recently the system has avoided a fiscal crisis, despite the low share of workers participating in it, by compressing benefits; consequently, the average pension in urban areas amounts to 1.1 times the minimum pension. This suggests that contributions and benefits are not closely linked, creating disincentives to full participation and to declaring total earnings. In rural areas, disincentives to contribute are even more perverse because those earning minimum pensions receive additional allowances that raise their total incomes higher than those of pensioners who previously earned more. In past years, the government made sizable increases in pension benefits, particularly for those with lower pensions, so that 36 pension spending increased. Relative to the average wage, minimum pensions are higher in Albania than in most other countries, which makes it difficult to differentiate benefits by earning or contribution level without sending costs even higher. Spending on pension programs increased by 1 percent of GDP over 2005–12, from 4.9 percent of GDP in 2005 to 5.2 percent in 2012. 3.69. The lack of a comprehensive approach to pension reform and sizable ad hoc increases in benefits has exacerbated the fiscal deficit. The pension system, which was in balance in 2005, ran an estimated deficit of about 1.2 percent of GDP in 2012. The total burden of the pension system on the government budget was 50 percent higher than that amount when government contributions for rural contributors are included. From 2005 through 2012, total government transfers to the pension system increased from 1.1 percent of GDP to an estimated 1.8 percent. Along with the possible impact of the slowdown in economic growth, the following major factors helped push up the deficit: (a) the plunge in pension contribution rates from 29.9 percent to 21.6 percent in 2009; and (b) sizable increases in pensions well above the inflation rate, particularly for rural dwellers. The fiscal pressure is expected to heighten as the Albanian population ages. 3.70. In 2006 the World Bank and the government conducted a joint social insurance review to look at the main problems in the pension system and identify broad options for reform. After the review, the government chose to concentrate on improving pension administration rather than undertaking a comprehensive pension system reform. The current (rising) pension system charge on the fiscal balance has now prompted the government to put reform back on the agenda. While the 2006 review is useful as a starting point, with the changes in the economic context as well as pension parameters (e.g., lower contribution rates, sizable increases in pensions, particularly rural), some options have since changed. Options for Pension Reform 3.71. While the objective of the changes Albania has introduced in recent years was to improve the pension system, the changes were ad hoc undermining fiscal sustainability, fairness and incentives. The government was endeavoring to provide better benefits to pensioners and improve their income security, but the changes were made without a comprehensive approach to all the challenges in the system. The result: a pension system that is overly redistributive and fiscally unsustainable. In urban areas, there is significant redistribution from high- and middle-income earners to low-income earners. As a result, average and minimum pensions are now virtually equivalent. The benefit structure appears to be flat, although to be eligible for a pension when they retire, people must have paid in amounts indexed to their incomes, which naturally differ. As a result, the urban scheme has significant disincentives to contribute for longer periods and to declare full earnings. The scheme for rural residents requires transfers from the budget both to supplement contributions and to finance the pension system deficit. Consequently, rural pensions have become a de facto social assistance scheme, although one administered as if it were contributory. Lower contribution rates and sizable increases in pensions have affected system fiscal sustainability. Although the government has to finance an increasing pension system deficit, the extensive informality in the labor market means that a significant share of the elderly will not have access to a pension benefit in future. 3.72. Pension systems around the world have two objectives, alleviation of poverty in old age, which involves redistribution to those at risk of poverty, and replacement of working age earnings, which implies higher pensions for those who earned and contributed more. There are trade-offs between the two objectives, since basing pensions strictly on contributions and earnings will not prevent poverty among the lifetime poor or those with incomplete earning histories. At the same time, focusing on poverty alleviation alone will leave middle- and higher-income contributors dissatisfied and with little incentive to contribute. Some countries, like Ireland, have chosen to give priority to alleviating poverty; others such as 37 Germany have been more concerned to replace earnings, leaving social assistance to deal with relieving old age poverty. Most countries have chosen a compromise of some poverty alleviation and some income replacement. The Albanian system is currently very close to an almost exclusive focus on poverty alleviation (see Figure 3.12 and 3.13). Figure 3.12: Ratio of Average to Minimum Pensions, OECD Countries and Albania 7 6 5 4 3 2 1 0 Turkey USA Sweden Luxembourg New Zealand Australia Iceland Portugal Slovak Republic Canada Austria Hungary Poland Israel Italy Greece Korea Mexico Belgium Norway Japan Switzerland Netherlands Finland Czech Republic Denmark Ireland UK Slovenia France Spain Estonia Germany Chile Albania Source: OECD Pensions at a Glance 2011 and Social Insurance Institute (SII). Figure 3.13: Minimum Pension and Average Wage as Percent of Average Wage, European and High-Income Countries and Albania 60 Minimum Pension as % of 50 40 Average Wage 30 20 10 0 Iceland Sweden Luxembourg Hungary USA Australia Austria Albania rural Korea Canada Poland Slovak Republic Portugal Greece Turkey Czech Republic Italy Israel New Zealand Denmark Finland Spain Belgium Netherlands Estonia Japan France Switzerland Mexico Norway UK Ireland Slovenia Albania urban Chile Germany Source: OECD Pensions at a Glance 2011 and SII. 3.73. The government has two main options for the direction of pension reform: • Keep the current contributory system but tighten the link between contributions and benefits to enhance the ability of the pension to replace the incomes individuals earned during their working years. • Recognizing that currently the pension (or the income earned by the pensioner) is almost the same for all urban pensioners and, although slightly lower, for all rural pensioners, eliminate the contributory scheme entirely and institute a social pension for people above a certain age (which would reflect the poverty alleviation objective). Those who are unable or unwilling to contribute would then receive at least some basic protection in old age and would eliminate 38 the need to collect contributions and maintain records without differentiating much between pensioners. Basic benefits could be financed by a rise in income taxes or in the VAT or could continue to be financed by a labor tax—whichever would raise sufficient revenue with the fewest disincentives. The basic benefit could be supplemented by voluntary private savings for those who want to better smooth consumption in old age. 3.74. The projected impact of these two broad options on fiscal sustainability and pension levels are analyzed below. Within each of the options there are many sub options; what has been modeled is only one of those sub options. Once the government chooses a particular direction, the options can be modified and the impact projected. In all cases, the reforms are projected to begin in 2014. Option 1: Enhance the Contributory Scheme 3.75. The intent here would be to reinforce the contributory system by providing a social pension for all those who have not contributed enough to eventually collect a pension. The particular elements of the option were chosen with the objectives of maintaining the contributory system but also achieving fiscal consolidation. Should fiscal constraints ease, elements of the reform package could be altered to allow for greater spending. 3.76. This option would gradually eliminate the redistributive aspects of the pension system. The reform package shown is for illustrative purposes only, to show what a reform reinstating the contributory system might look like given fiscal consolidation. The actual elements of a reform package might require different eligibility conditions, provide different benefits, and introduce changes faster or slower. The package has the following elements: • A gradual increase in the maximum pension, indexing it to 150 percent of nominal wage growth from 2014 onward • Gradual elimination of the basic benefit, holding it constant in nominal terms from 2014 on, and keeping the annual accrual rate to 1 percent per year of service • Restoration of a gap between average and minimum pensions by holding the minimum pension constant in nominal terms until 2020 and thereafter indexing it to inflation • Indexation of the average pension to inflation from 2014 on. By 2012 the combination of this and the previous element will produce an average pension equal to twice the minimum, which would continue thereafter. • A rise in the retirement age for women to age 65 between 2014 and 2020, followed by an increase in the retirement age for both men and women to age 70 by 2080, in line with increases in life expectancy • Institution of a social pension equal to the 2014 minimum pension for rural workers and all who do not qualify for the urban pension, but only available at age 70. The rural pension system with its subsidized contribution rates will be eliminated, with farmers no longer required to pay contributions but in turn receiving only the social pension. Current rural contributors who are willing to make the full urban contribution without subsidy may continue contributing and upon retirement will receive the same pension as an urban worker. Current rural pensioners will receive pensions indexed to inflation. 3.77. This option clearly raises urban benefit levels (Figure 3.14). While the current policy results in a drop in future benefits from the current 33 percent of the average wage to 10 percent, the reform option limits the drop to 31 percent. The initial slightly higher benefit rate comes from the gradual lifting of the 39 cap on benefits beginning in 2014, with higher benefit rates in later years also dependent on linking benefits to contributions paid, without the repressive impact of the current cap. Figure 3.14: Projected Benefits Under Current Policy and Option 1 Projected Benefits as Percent of Average Wage 40.0% 35.0% 30.0% 25.0% 20.0% 15.0% 10.0% 5.0% 0.0% 2010 2013 2016 2019 2022 2025 2028 2031 2034 2037 2040 2043 2046 2049 2052 2055 2058 2061 2064 2067 2070 2073 2076 2079 Base Case Urban Inflation Indexation Option 1 Urban Source: World Bank PROST projections. 3.78. Higher benefits coupled with greater coverage of the elderly will in the long run result in higher deficits in the pension system, but the fiscal consolidation measures in the reform package imply smaller deficits over the short and medium term, with savings of as much as 1.4 percent of GDP by 2020. The left-hand panel of Figure 3.15 shows that the fiscal consolidation measures described for Option 1 would reduce deficits immediately. Furthermore, the better rates of return for higher- and middle-income individuals are likely to result in higher income declarations and more revenue being collected by the Social Insurance Institute (SII) and the tax authorities, which will further narrow the deficit. 3.79. The reform option provides better retirement outcomes for both contributors and those working in the informal sector, but in the long run these benefits will raise costs above the current system—but only if the current system continues to be indexed to inflation. The right-hand panel of Figure 3.15 shows government outlays for the whole pension system, including the deficits of the urban and rural systems and the subsidies for rural contributions in the baseline, compared to the deficits of the contributory urban system and the cost of the noncontributory social pension in the reform option. Savings are substantial, again about 1.4 percent of GDP by 2020. The current system only becomes cheaper than the reform option in 2052, but the reform option provides pension income to the 56 percent of the elderly who in the current system would have no access to pensions over the long run. It is unlikely to be socially sustainable to leave 56 percent of the population with no support during old age; under the current system many people will require social assistance in the future. Once the cost of that social assistance is added to the costs of the current system, there may be little difference, even in the long run, between the current system and the reform option. 40 Figure 3.15: Projected Pension System Deficits, Baseline and Reform Option 1 Projected Pension Deficits in Urban Projected Pension Deficits, 2010-2080 System, 2010-2080 (Percent of GDP) (Percent of GDP) 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2080 2010 2015 2020 2025 2030 2035 2040 2045 2050 2055 2060 2065 2070 2075 2080 (.5%) (.5%) (1.0%) (1.0%) (1.5%) (1.5%) (2.0%) (2.0%) (2.5%) (2.5%) (3.0%) (3.5%) (3.0%) (4.0%) (3.5%) (4.5%) (4.0%) (5.0%) Base Case :Total (Urban+Rural+Rural Govt Cont ) Urban Option 1 urban Option 1 Total (urban deficit +spending on social pensions) Source: World Bank PROST projections. Option 2: Make Alleviating Poverty the Priority 3.80. Since the Albanian system already is close to a uniform pension, an alternative would be to eliminate the contributory system altogether and implement a flat pension for all at the retirement age. Figures 3.12 and 3.13illustrate how close the Albanian system already is to providing flat benefits. A number of countries, among them Ireland, Australia, New Zealand, Kosovo, Kazakhstan, and Georgia, provide similar flat benefits to all citizens above a certain age, and in all but Ireland, the benefits are financed from general revenues. The difficulty in moving in this direction immediately is that all those who have contributed to the pension system would rightfully expect to get something more than the social pension received by those who have never contributed. If the government pays people the value of the rights they have already accumulated in the pension system without collecting contribution revenue, it will run even larger deficits than currently. One solution to this dilemma would be to abolish the contribution requirement immediately, terminate future accrual of pension rights, but agree to pay out on rights that have already accrued. These rights could be financed by a new labor tax, which initially could be of the same magnitude as the current contribution rate to social insurance. With benefits paid out of general revenue, the new labor tax would add to general revenue. This approach would eventually add revenue to the Treasury as pension obligations decline over time and the labor tax continues to be imposed. At this point, or when the Treasury determines it can pay the remaining expenditures out of other revenue, the labor tax can be gradually reduced. This is the approach taken by Georgia, which initially converted its contribution into a social tax and then merged the social tax with the income tax so that individuals were paying only a single higher income tax that included the previous social tax. 3.81. This option would cover many of the same reforms as the previous option: • Holding the minimum pension constant in nominal terms until 2020 and then indexing it to inflation; 41 • Indexing the average pension to inflation from 2014 onwards; • A rise in the retirement age for women to age 65 between 2014 and 2020, followed by an increase in the retirement age for both men and women to age 70 by 2080, in line with increases in life expectancy; • Institution of a social pension equal to the 2014 minimum pension, for all workers but not available until age 70. The rural pension system and its subsidized contribution rates would be eliminated, with farmers no longer required to pay contributions but in turn receiving only the social pension; • No contributions required from either rural or urban Albanians, with past urban contributors receiving either the social pension or what they have accrued in the former pension system, whichever is higher, and rural contributors receiving only the social pension, which is equal to the minimum rural pension they would have receive in any case; • Collecting a new labor tax in the amount of 21.6 percent of wages initially, to be decreased over time as fiscal space opens up. 3.82. Figure 3.16 shows the net projected impact of Figure 3.16: Projected Pension Deficits, Status Quo and Options 1 this option on general and 2 (Percent of GDP) revenue: the labor tax would 1.0% bring in revenue, and the Treasury would pay pension 2014 2019 2024 2029 2034 2039 2044 2049 2054 2059 2064 2069 2074 2079 costs to the SII for (1.0%) distribution to pensioners. (2.0%) When the balance is negative, other Treasury (3.0%) resources will have to cover (4.0%) pension spending; when the balance is positive, labor tax (5.0%) Current system Option 1 Option 2 revenue exceeds pension payments, and the Treasury Source: World Bank PROST projections. earns additional revenue. Since there will be large fiscal savings over the current system, which reach 1.8 percent of GDP by 2020 and continue on up, some could be retained as fiscal savings and some used to reduce the labor tax. 3.83. While Option 1 pays much higher benefits for those who contribute than Option 2, the latter has the advantage of replacing pension contribution rates with a gradually diminishing labor tax, which might help curb informality and generate growth. However, the lower long-run benefits under Option 2 are not likely to be sufficient for middle- and higher-income individuals. Providing supporting infrastructure for voluntary automatic enrollment or even mandatory savings schemes will become critically important to give middle- and higher-income individuals the opportunity to replace part of the income they earned during their working years. From the standpoint of the individual, capacity for saving will increase as pension contributions are eliminated and the successor labor tax is gradually reduced. Policy Options 3.84. The Albanian pension system is fiscally unsustainable; it is projected to ultimately provide very low benefits; it gives higher- and middle-income individuals incentives to under declare 42 earnings and hours or avoid the system entirely; and it subsidizes the rural self-employed at the expense of the rest of the system. These problems are intertwined: The system is costly because the government reduced contribution rates in an effort to reduce informality; it raised pensions above the rate of inflation particularly for lower-income pensioners; and it provides increasing subsidies for rural contributors. Ultimately, it is expected to provide low benefits: to control costs the government restricted the maximum pension, but the maximum pension gives higher- or middle-income earners little incentive to contribute or report on full earnings and lower contributions further raise the costs and subsidies required to keep the system running. 3.85. With government policy at a crossroads, one alternative would be to re-establish a tight link between contributions and benefits, separating the income replacement function of pensions from the poverty alleviation function. After numerous policy changes, the system is close to providing a uniform pension for most people, with income from the average pension almost equal to income from the minimum pension. Either the government needs to better link contributions and benefits and decompress current benefits or it needs to eliminate the contributory benefit and provide a poverty-alleviating flat benefit to all elderly and disabled. With the first choice, the currently high minimum pension relative to current average wage means that decompressing pension benefits will raise costs unless there is fiscal consolidation of benefits. However, individuals will get higher pensions in the longer run and have greater incentive to contribute. There is political pressure to raise the current rural pension, which is de facto social assistance, to the level of the urban minimum pension despite minimal rural contributions; instead it should be replaced with a social pension for non-contributors that would also cover those in urban areas who do not work or contribute. The advantage for workers is that their contributions directly support worker benefits and are no longer channeled to the social assistance component. From the government’s perspective, the political pressure to equalize rural with urban pensions should disappear and the government should be able to provide benefits more affordably. 3.86. The alternative would be to direct the public system exclusively to poverty alleviation, leaving income replacement to savings instruments, voluntary or mandatory. Such a pure social pension has been implemented in Kazakhstan, Georgia, and Kosovo, where a flat, poverty-alleviating benefit is provided to all above a certain age. The advantage for workers is that they would no longer need to make contributions and the labor would gradually decline, with positive implications for labor costs and growth. The advantage for government is that the difference between informal and formal workers would disappear as would the cost of collecting contributions and maintaining records since everyone would get same pension upon retirement. In moving from a contributory to a noncontributory system, one approach modeled here would be to provide benefits based on rights already accrued up to the date of the reform and provide a benefit equal to the rural minimum to everyone else, indexed for inflation. This approach will need to be complemented by vigorous promotion of voluntary savings to give higher- and middle-income individuals the opportunity to replace more of their working-age income during retirement. 3.87. Given the current fiscal situation, the Government needs to adopt in short-run measures which can provide fiscal savings while it defines the long-run reform path and develops the legislation to implement that path. Short-run fiscal measures could include temporarily freezing pensions, given that increases above the legislated inflation rate have been provided in the recent past, and gradually raising rural contribution rates until they reach the same levels that the urban self-employed pay. These measures can potentially save the Government as much as 0.4 percent of GDP in 2014 alone, but as much as 1.2 percent of GDP in 2016. It is unlikely that the temporary pension freeze could be extended beyond 2016. 43 D. FISCAL SUSTAINABILITY WITH STRUCTURAL REFORMS 3.88. This section assesses the implications for fiscal sustainability of the structural policy reform scenarios discussed above. The fiscal sustainability analysis presented in chapter 2 assumed a policy of no structural policy changes. This chapter assesses fiscal sustainability under scenarios for the tax policy and pension reform. It also explores the impact of the energy sector reform that will reduce the contingent liabilities in the sector. A short description of each policy reform scenario is followed by an assessment of its implications for fiscal sustainability. Tax Reform 3.89. As discussed above, four policy measures can help increase revenue in Albania: expanding the VAT tax base through exempting rental property and taxing the sale of new construction; reversing the recent VAT exemptions; raising the CIT rate from 10 to 15 percent; and making the PIT progressive and introducing risk-based audits. The combined fiscal impact of these policy measures could be as much as 3 percent of GDP assuming Albania’s average compliance. The revenue impact could be larger if these policies were to be complemented by strengthened tax administration to improve compliance. 3.90. Improvements in Figure 3.17: Debt to GDP ratio under tax scenarios revenue ratios over the medium term would significantly reduce debt burden indicators. If Albania were to raise its revenue-to- GDP ratio from the current level of about 24 percent of GDP to 27 percent in 2014 and keep it there, by 2020 the public debt-to-GDP ratio would fall below 70 percent by Source: WB Simulations. 2016 and reach 65 percent (Figure 3.17); and if the revenue-to-GDP ratio went up to 30 percent, the level of FYR Macedonia, and there were no severe shocks, the debt-to-GDP ratio would reach 57 percent. Increasing the revenue-to- GDP ratio to 27 percent could, for example, be achieved by raising the CIT rate from 10 to 15 percent or taxing sales of new residential construction under the VAT. Pension Reform 3.91. Given the current fiscal situation, the Government needs to engage in short-run measures which reduce the pension deficit while it defines its long-run pension reform path. Short-run fiscal measures could include temporarily freezing pensions, given that increases above the legislated inflation rate have been provided in the recent past, and gradually raising rural contribution rates until they reach the same levels that the urban self-employed pay. These measures can potentially save the Government as much as 0.4 percent of GDP in 2014 alone, but as much as 1.2 percent of GDP in 2016. It is unlikely that the temporary pension freeze could be extended beyond 2016. Still, by 2016, the public-debt-to GDP ratio would decline by about 2 percent of GDP if pensions were frozen and contributions for rural pensioners increased in line with past government announcements (Figure 3.18). 44 Figure 3.18: Public-Debt to GDP ratio under Pension Scenarios 110.0% 100.0% 90.0% 80.0% 70.0% 60.0% 50.0% 40.0% 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 Baseline Option 1 Universal social pension without labor tax Universal social pension with labor tax Source: World Bank Calculation. Energy Sector Reform 3.92. The energy sector needs an affordable program to reduce distribution losses. DSO could reduce losses by 6 percent by better network management, metering of large tariff customers, customer outreach, and collections efforts (Figure 3.12). These measures could cut the sector financial gap (i.e. sector contingent liabilities) by almost half compared to the business-as-usual scenario. Modernizing the network, maintaining it well, and upgrading its management in an aggressive reform scenario could cut the financial gap by more than two-thirds. This scenario would, however, require investments of about €30 million over the next five years. 3.93. The normal loss reduction program would significantly reduce the contingent liabilities caused by the energy sector (Figure 3.19). Figure 3.19: Effect of Energy Reform on Contingent Liabilities This scenario would 85% bring the public debt 80% Pu blic Debt-to-GDP ratio ratio down from 82.7 75% to 77.9 percent by 70% 2020. The aggressive reform scenario would 65% not improve the ratio 60% over the medium term 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 because additional Baseline with energy gap collections would need Normal energy scenario to be invested in the Aggressive scenario network. Source: Bank Staff Calculations. Implications for Fiscal Sustainability 3.94. Fiscal consolidation through reforms to tax policy tax administration and pension reform would significantly enhance Albania’s fiscal sustainability; and the reforms in the energy sector would not only help reduce fiscal risks arising from the sector but also enhance the ability of the sector to provide reliable 45 electricity and support investment and growth. Albania can thus reduce its public debt and assure continued macroeconomic stability, but also support sustainable growth. Table 3.2: Albania: Impact of alternative structural reforms on public debt (Percent of GDP) 2013 2014 2015 2016 2017 2018 2019 2020 Public debt under the baseline - no reform 70.5 72.1 73.5 72.7 72.8 72.9 73.0 74.2 Public debt with tax policy and admin reform 70.5 69.6 68.5 65.6 63.7 62.0 60.3 59.6 Public debt with pension reform 70.5 72.2 73.4 72.2 71.6 70.7 69.5 68.4 Public debt with combined tax and pension reforms 70.5 69.7 68.4 65.1 62.5 59.8 56.8 53.8 Contingent liabilities in energy sector – no reform 0.0 1.3 2.5 3.4 4.1 5.1 5.9 6.8 Impact of energy sector reform on contingent liabilities 0.0 0.9 1.6 2.0 2.4 2.9 3.3 3.7 Source: Bank Staff Calculations. Figure 3.20: Public debt trajectory under alternative structural reforms 85 81.5 80 79.1 76.0 77.0 75 73.2 73.4 72.2 72.2 71.6 70 70.5 69.7 70.7 68.4 65 65.1 62.5 60 59.8 55 50 2013 2014 2015 2016 2017 2018 Baseline - no reform Tax policy and admin reform Public debt with pension reform Combined tax and pension reforms Low growth no reform Source: World Bank staff calculations Note: Tax policy reform: increase the CIT from 10% to 15%, reverse VAT exemptions Pension reform: Freeze pension, gradually raise rural contributions 3.95. A combined effect of these structural reforms would reduce Albania’s public debt to GDP ratio singnificantly over the medium term. As an illustrative scenario, we consider a reform scenario whereby Albania (i) raises its revenue-to-GDP ratio from 25 percent of GDP to 27 percent in 2014 through tax policy and administration reform and keeps it at that level thereafter; and, (ii) puts in place 46 short-run fiscal measures which include temporarily freezing pensions, and gradually raising rural contribution rates until they reach the same levels that the urban self-employed pay; and (iii) implements a (normal) distribution loss reduction plan that reduces energy losses by 4-6 percent through 2014-16 and improves collections by 3 percent annually until collection reaches 95 percent, a common collection rate in the region. The combined effect of tax and pension reforms will be to reduce (relative to the baseline) Albania’s public debt-to-GDP ratio by 2.4 percent of GDP in 2014, 7.6 percent of GDP by 2016 and 10.4 percent by 2020 (see Table 3.2 and Figure 3.20). The impact of energy sector reform will primarily be to reduce the contingent liabilities arising from the sector. With no reforms, these contingent liabilities are set increase from abour 1.3 percent of GDP in 2014 to about 6.8 percent of GDP in 2020. With the proposed loss reduction reform, the contingent liabilities would be cut by almost half by 2020 compared with the no-reform scenario (Table 1). The energy sector reform would also also enhance the ability of the sector to provide reliable electricity and support investment and growth. 47 CHAPTER 4. TRANSPORT AND CAPITAL SPENDING: REBALANCING TOWARD MAINTENANCE TO SUPPORT GROWTH A. INTRODUCTION 4.1. In recent years most of Albania’s capital spending has gone to the transport sector, and within transport, primarily to roads. For 2008–12, 95 percent of transport expenditures were for the national road network, rising as high as 99 percent in 2008 (Table 4.1). Less than 5 percent has gone to ports and less than 0.5 percent to rail. Maritime and civil aviation have increasingly received both donor and private financing. In light of this, this chapter will be mainly concerned with the road sector. Table 4.1: Public Spending on the Transport Sector (nominal € millions) 2008 2009 2010 2011 2012 Roads (national network) 505.2 487.1 241.8 210.1 156.9 Ports 3.4 3.1 5.0 10.0 4.5 Railways 1.6 0.2 0.4 0.9 0.4 Air Transport 0.1 0.1 0.1 0.1 0.1 Total 510.4 490.6 247.2 221.1 161.8 Source: Ministry of Finance. Note: The data are for what the budget terms capital expenditure and exclude subsidies. 4.2. Albania has invested heavily in motorways in recent years. October 2010 saw the opening of a US$535 million modern motorway, which has a 5.8 km two-bore tunnel. The 61 km, four-lane highway stretches from the town of Rrëshen, in the heart of the country, up to the northeastern village of Kalimash. It is the central leg of a 171 km highway traversing the country from the Adriatic Sea to the northeastern border with Kosovo, at a cost of about €950 million. The new roadway cuts travel time along the route from six to two hours, boosting both coastal trade and northeast tourism. It also provides a vital connection not just within Albania but across the region, linking markets to the Adriatic port of Durres and stimulating economic growth as Albania prepares for EU accession. Other motorways are close to completion; some of the more important are in the North–South Corridor from the Montenegro border to Greece and the West-East TEN Corridor from Western Coastal Albania to FYR Macedonia and Greece to the east. 4.3. Businesses consider the quality of Albania’s transport Table 4.2: Albania’s Ranking in the Global Competitiveness infrastructure, particularly roads, Index to have improved in recent years. 2007–08 2012–13 For more than 30 years the annual Overall ranking 108 89 competitiveness reports of the World Number of countries ranked 134 144 Economic Forum (WEF) have Roads 114 59 examined numerous factors that Railroad infrastructure 109 119 enable national economies to achieve Ports 124 96 sustained economic growth and long- Air transport 76 66 term prosperity. Since 2005, the WEF Source: WEF. Note: Lower ranking indicates better competitiveness. has based its analysis on the Global Competitiveness Index (GCI), which captures 12 pillars of competitiveness, including infrastructure. A 49 comparison of Albania’s GCI rankings (Table 4.1) for 2007–08 through 2012–13 indicate that there is a perception of significant improvement in the quality of road infrastructure and smaller improvements in ports and airports. For instance, whereas in 2007–08 railways were perceived as being in better condition than roads, the opposite is true today (Table 4.1 and Table 4.2). Similarly, whereas in 2007–08 infrastructure was considered the second most problematic factor for doing business in Albania, by 2012– 13 it was only the ninth most problematic. 4.4. Still, Albania faces challenges in the road sector. Sustainable conservation of the new road infrastructure requires rebalancing spending toward maintenance and rehabilitation. The country also needs a more disciplined approach to the planning and phasing of priority investments and a more professional approach to management of road assets to ensure that they are adequately maintained. B. ROAD NETWORK AND QUALITY 4.5. According to the Transport Strategy for 2008–13, the road network in Albania is about 18,000 km long. It consists of (a) some 3,719 km of national roads—about 1,300 km of primary roads (the main corridors) and about 2,500 of secondary roads; (b) about 10,500 km of interurban roads; and (c) another 4,000 km under the jurisdiction of autonomous units, enterprises, and companies. However, there is some uncertainty about the actual extent of the road network, even the national roads managed by the Albanian Road Authority (ARA). According to Albanian National Transport Plan 2 (ANTP2), the network consists of 1,300 km of national roads of primary importance and 1,000–1,500 km of secondary importance, while about 1,000–1,500 km are regional (secondary) or local (tertiary), with many of the later only partially paved or unpaved. ARA figures put its total network at 3,740 km. 4.6. These discrepancies point up a possible need to reclassify roads in terms of their functional requirements and the jurisdiction responsible for them. 25 This recommendation was put forth in the 2006 public expenditure review as critical for making spending on roads more efficient (World Bank 2006). It was based on a 2005 study that identified about 1,000 km of secondary roads for which GRD was then responsible that could be reclassified as local and transferred to local government units. Downgrading some roads from national to secondary and from local to private could create significant savings through lower maintenance standards and therefore lower costs. That study also found that of some 4,411 km of district roads, 3,500 km provided access to private or communal forestry and mines, with little justification for the public sector to maintain them. 4.7. According to ANTP2 only 42 percent of the national road network is in a good or very good condition (Table 4.3). While the primary network is well-defined and information on road condition is available, the same cannot be said for the secondary network, the condition of over 45 percent of which is not known (Table 4.4). Moreover, only 16 percent of the secondary network is in good or very good condition. Nevertheless, these figures suggest improvement since 2006, when the World Bank (2006) found that only 16.5 percent of the national or main network was in good condition.26 The tertiary network, however, is in worse condition. Little of it is paved, but in recent years. Because of Albania Development Fund projects there have been significant improvements in the condition of rural roads. 25 Functional classification is the process by which roads and highways are grouped according to the type of service they provide. Individual roads and streets do not serve travel independently but as part of a network through which traffic moves. Because functional classification describes the importance of a particular road or network of roads to the overall system, it is critical in assigning priorities to projects and establishing road design standards that meet the needs of the traffic served. 26 However, because the ANTP2 dates from before 2010, it is likely that due to insufficient maintenance overall road conditions have worsened from the 42 percent quoted. 50 Table 4.3: Condition of the National Road Network (percentages) Network National Primary Secondary Very good 32 55 5 Good 10 12 11 Medium 26 30 30 Poor 5 1 9 Under construction 1 2 5 Unknown 25 0 45 Source: ANTP2. 4.8. Albania’s emphasis on new roads has not been always supported by quality construction work. Among the signs of poor quality are premature failure of new pavement and some structures (the major landslide along the new Milote-Morine Highway in 2010 comes to mind); use of different design standards on adjacent sections; poor and contradictory signage and other failures to ensure safety; and non-repair of electrical fixtures and fittings. There is a perception that lack of maintenance has resulted in poor value for money from some capital expenditures. There is also concern about how some projects, primarily those supported from public funds, are planned, implemented, monitored, and maintained. 4.9. Even though the rate of fatalities is dropping, in Albania road safety is a major social and public health issue. Annual fatalities have fallen from a peak of 384 in 2007 to 322 in 2011; the fatality rate has nearly halved over the last decade, from 15 casualties per 10,000 vehicles to 7.8 in 2011. However, this still compares unfavorably with neighboring countries and is about six times higher than in the best-performing EU countries. Over the same period the number of casualties and injured has quadrupled, from 616 in 2000 to 2,472 in 2011. The condition of the road network, rising traffic, driver behavior, and poor enforcement together have prevented a convergence to the standards of better- performing neighbors. Estimates in 2010 suggested that the economic cost of road accidents in Albania was about €130 million annually, 1.5 percent of GDP (Sweroad Consultancy Report 2010). The government has focused on enforcing road safety laws and regulations rather than giving attention to safety audits and black-spot improvements, incorporating safety standards into road design and construction, increasing public awareness, and enforcing the laws. Road safety is unlikely to improve unless it is given comprehensive and robust multi sectoral attention. C. EXPENDITURES AND REVENUES 4.10. Spending on roads (mainly motorways) rose to 5.7 percent of GDP in 2008 before dropping to 2.7 percent in 2010, 2.4 percent in 2011, and 1.9 percent in 2012. These figures exclude accumulated arrears, which in the roads sector alone are estimated at about €50 million.27A significant amount of construction, rehabilitation, and maintenance is donor-financed: donor financing averaged 37 percent for 2008–2012 and reached 64 percent in 2012 (Table 4.4). Although maintenance has risen as a percentage of total spending on roads (Figure 4.1), that does not reflect an increase in actual expenditures but simply the fact that total spending on roads has gone down and maintenance now makes up a larger proportion than it did than five years earlier. 27 The construction association claimed a year ago that unpaid bills for completed road works amounted to more than €200 million, but the World Bank believes that this amount is now down to €50 million. The Bank has not received official data on outstanding road arrears, and the tables in this chapter do not include accumulated arrears. 51 Figure 4.1: Spending on Roads and Total Capital Expenditure (Percent of GDP) Roads Tot.Investments 10 8 6 4 2 0 2005 2007 2009 2011p Source: MoF, IMF, World Bank. Table 4.4: Budget versus Foreign-Financed Spending on Roads (nominal € millions) 2008 2009 2010 2011 2012 Capital investments for construction, rehabilitation, reconstruction 499 485 242 210 181 Budget 223 431 191 139 64 Foreign financed 276 54 51 71 117 Recurrent expenditure for road maintenance 11 11 9 10 9 Percent of total 2.15 2.30 3.75 4.62 4.60 Total 510 496 251 220 190 Source: Ministry of Transport. 4.11. Albania is not spending enough on maintaining the national road network; the backlog is considerable. As a start to addressing this problem, real annual budgetary allocations to operations and maintenance of the national road network increased between 2005 and 2008. However, they have since dropped considerably—nominal expenditures on maintenance in 2012 are down 25 percent from 2008, and spending on rehabilitation is down 73 percent. 28 Overall spending on the national road network has dropped steeply in the last four years (Table 4.5), reflecting reduced capital investments on motorways. Table 4.5: Expenditures on the National Road Network (nominal € millions) 2008 2009 2010 2011 2012 Maintenance 14.6 12.7 10.1 12.0 10.9 Rehabilitation 22.7 37.5 23.3 9.3 6.2 Investment 259.0 496.7 251.1 276.4 152.4 TOTAL 296.4 546.9 284.5 297.7 169.4 Source: Albanian Road Authority. Note: Includes budget and donor funding. 4.12. Per km expenditures on maintaining the national network are low compared to regional comparators (Figure 4.2). Albania spent an average of €2,914/km on its network in 2012—very low 28 Using Albanian lek figures, in real terms the decline was 24.5 percent for maintenance and 72.6 percent for rehabilitation. 52 even compared to regional comparators, which also do not spend enough on maintenance. For example, in 2012 Serbia spent nearly three times as much on maintenance as Albania did (Figure 4.2). Clearly, the approach Albania has so far taken is not sustainable. Figure 4.2: Spending on Maintenance (€ /km) Romania (2011) 9,303 Serbia (2012) 9,039 Kosovo (2011) 3,742 Albania (2012) 2,914 - 1,000 2,000 3,000 4,000 5,000 6,000 7,000 8,000 9,000 10,000 Sources; Road agencies and World Bank estimates. 4.13. Under spending on road maintenance has significant costs. Depending on the amount of traffic, adequate maintenance can sustain road pavement beyond the original design life, reducing the need for reconstruction. Inadequate maintenance exacerbates the rate of deterioration, hastening failure of the pavement and engendering a need for reconstruction that could have been avoided. Heggie and Vickers (1998) report that in current terms reconstructing a paved road is three to five times more expensive than maintaining it, and in net present value terms about 35 percent more. The result is the accumulation of a maintenance backlog; a postponement of spending that translates into much higher spending in the future. Failure to maintain a paved road has also been estimated to increase user costs by a factor of three in terms of additional time, fuel, and vehicular wear and tear, which have a direct impact on the price of goods and the expenses consumers pay for goods and services. 4.14. Road user charges reached €257 million in 2012, up 10 percent from five years earlier.29 There have been significant changes to road user charges, including revisions to the fuel excise tax rates, introduction in 2010 of the circulation fuel tax, and removal of the registration tax and taxes on imported vehicles—the annual tax on transport vehicles replaced the axle load and registration tax; new vehicles are not subject to the tax for the first three years. In 2012, actual spending on the national road network was less than the charges collected (Table 4. 4). However, the taxes listed in Table 4.6 accrue to the state, not directly to the Albanian Road Authority, which is dependent on budget and donor funds. While calculation of a cost recovery ratio would require allocating a defined percentage of road user charges to the national road network, on the basis of traffic count data—which is complicated by the lack of such data for local roads—it is likely that the bulk of revenues are generated by users of national roads, which suggests that the cost recovery ratio is likely to be close to one if all spending on the road sector is included. From a welfare economic perspective, road user charges are meant to cover the short-term marginal costs of operating roads in order to ensure optimal use of infrastructure; the evidence suggests that revenues from road users in Albania are significantly above those costs. 29 In real terms the rise was 11.9 percent, compared to 10.4 percent in nominal euro terms. 53 D. FUTURE FINANCING REQUIREMENTS 4.15. As part of ANTP2, comprehensive estimates were made of the maintenance expenditures the national network required. Using the HDM-4 model, maintenance costs were calculated under four scenarios for a 20-year period for 1,198 km of the primary network, where information on traffic levels and road condition is most complete (Table 4.4). The analysis suggests that the annual expenditure needed to maintain primary roads is about €24 million. With less than that, vehicle operating costs and travel times will rise with road roughness. For the secondary road network, there is less information and considerable uncertainty about the condition of 45 percent of it but the maintenance requirements have been estimated at €21 million, bringing total annual maintenance needs to €40–45 million annually. 4.16. The average maintenance spending on the national road network for 2008–12 was about €12 million, less than one-third of what seems to be needed (Table 4.6). Of major concern is how the government will be able to finance the additional work—or at a minimum, raise spending on road maintenance from the current levels. Table 4.6: Estimated Required Maintenance Expenditure (nominal € millions) 30 Total Cost Annual Cost Primary network No budget constraint 473 24 Annual expenditure capped at €32 million 473 24 Annual expenditure capped at €22 million 440 22 Annual expenditure capped at €17 million 375 19 Secondary network Consultant estimate 427 21 TOTAL 40 to 45 Source: ANTP2. 4.17. One effective way to address road maintenance needs is to rebalance spending to preserve past road investments. The ANTP2 Investment Plan (Table 4.7) foresees €1,292 million of road development projects for 2010–15, rising to €1,731 million for 2016–20, or €302 million annually for the 10-year period. While these projects are to be financed from a combination of budgetary sources, donor funds, and perhaps private investors, there is a need to ensure that maintenance of existing road assets can be financed within the budget constraints imposed by Albania’s fiscal position, both its annual fiscal deficit targets and its total indebtedness. If plans exceed what the medium-term fiscal framework allows, network preservation needs to take precedence over road expansion. The Transport Investment Plan for 2010–30 shifts the balance toward preservation of the network to the second period, 2016–30 there is a solid case for rebalancing spending toward maintenance even before 2016. 30 These estimates cover the 20-year period 2010–2030. 54 Table 4.7: Transport Sector Investment Plan (nominal € millions) 2010–15 2016–30 Total Annual Road development projects national network 1,292 1,731 3,023 151 Network preservation for national network 248 680 928 46 Primary 101 297 398 20 Secondary 117 309 426 21 Bridges 11 16 27 1 Road safety 28 11 39 2 Secondary and local road program (200km annually) 66 499 566 28 Urban transport 230 390 620 31 TOTAL 1,873 3,329 5,202 260 Source: ANTP2. 4.18. Earmarking revenues from user charges by setting up a road fund is not the best option. ANTP1 recommended consideration of creating a stand-alone road fund exclusively for maintenance. Though this has not yet been done, ANTP2 maintains that it is a good option for financing maintenance.31 While it would guarantee a certain level of maintenance funding, from a macroeconomic perspective earmarking road user revenues reduces fiscal flexibility by restricting the government’s ability to spend money on its own priorities and is thus often opposed vigorously by finance ministries. Building and consolidating institutional capacity to beef up both budgetary processes and planning, together with reassessing the balance between spending on new capital investments and maintenance and rehabilitation of existing assets would help meet maintenance financing needs. 4.19. Measures to improve efficiency and private participation are essential; for instance, building up the ARA is a priority for efficient use of resources. Road projects, particularly those funded from domestic sources, do not seem to be prioritized on solid grounds. Returning responsibility for geotechnical investigations from the construction to the design phase would markedly improve quality. Another with considerable potential for improvement is the inadequacy of supervision when a locally funded contract is constructed. Modernizing maintenance practices by extending output and performance- based maintenance contracting could also offer significant efficiency gains. Infrastructure PPPs have benefits that need to be explored. E. POLICY OPTIONS 4.20. The key recommendation is that to support economic growth, spending on roads needs to be rebalanced from capital investments to maintenance and rehabilitation. Capital investments could also be carefully phased so that they do not come at the expense of maintenance and rehabilitation. Modernizing maintenance practices by extending output and performance-based contracts could offer significant efficiency gains. 4.21. Other steps that the government could consider to improve management and efficiency in roads investment are: • In light of the recent significant accumulation of payment arrears in public works, it is important to ensure that future spending is aligned with revenues to prevent arrears to 31 European Commission, Europe Aid/127468/C/SER/AL. First Five-year Review of Albanian National Transport Plan. Draft Final Report, Part III, p.10. Projected funded by the EU prepared and implemented by Louis Berger. 55 contractors from accumulating. This requires a commitment to not start rehabilitation and construction work unless adequate funding has already been committed to it. • There is a need to review functional classification of the road network and reclassify roads in line with their functional requirements and the jurisdiction responsible for their preservation. The ARA needs to set up and use a road asset management system—and update it regularly to ensure that it is useful as a planning tool. Such a system is a prerequisite for more efficient and professional management of national road assets. • A number of measures that have been proposed to improve the design, construction, and supervision of Albanian roads would ensure better-quality road works and more efficient use of public monies, particularly for projects financed by domestic resources. • In future the road concession program should focus on operations and maintenance—brown field rather than green field projects. Concessions for highway projects involving large capital expenditures are unlikely to materialize without significant minimum traffic guarantees or availability payments; these need to be carefully reviewed for value-for-money and any potential contingent liabilities. 56 CHAPTER 5. CROSS-CUTTING: INSTITUTIONS TO REINFORCE FISCAL POLICY A. REFORM OF PUBLIC FINANCIAL MANAGEMENT 5.1. Albania has participated in a number of detailed reviews of its public financial management (PFM), among them two PEFA assessments (2006 and 2011), a Public Expenditure Review (2006), annual EU-SIGMA reviews, and more specific analysis by the World Bank, the IMF, the EC, and other organizations. The various reviews have plotted the significant progress Albania has made in improving PFM. They have also identified areas where change has been difficult and where more needs to be done. 5.2. This chapter looks at four areas where immediate action is needed to improve PFM discipline: budget planning, preparation, and revision; capital investment management; cash management and financial systems; and expenditure commitments and arrears. It briefly outlines the current situation and presents options for improvement, drawing where appropriate on good international practices. It concludes with specific recommendations for action. Budget Planning, Preparation, and Revision 5.3. In 2008–10 there were significant differences between Albania’s budgeted and actual revenue (2011 PEFA assessment). While spending has been well-controlled, the accuracy of revenue forecasts has been disappointing. The true picture of revenue and expenditure planning and management is more complex than PEFA scores reveal. On the revenue side, the period covered by the PEFA assessment was characterized by substantial uncertainty because of the international financial crisis, which had a considerable effect on growth and tax revenues. Throughout the period of uncertainty, overly optimistic revenue estimates persisted, in part because the depth and the length of the international financial crisis were underestimated. The Ministry of Finance sought to limit the impact of overly optimistic revenue estimates by taking a more conservative approach and seeking expanded contingencies on the expenditure side of the budget. However, these proved insufficient. 5.4. As a consequence of the optimistic revenue projections, liberal spending plans have been approved during budget preparation, which have had to be cut back during the budget year when revenue collections fell short. This has had a heavy impact on discretionary expenditure, especially capital investments, and on the accumulation of arrears, because commitments made and approved before the cuts could not be cancelled or deferred; all administrators could do was postpone payment of obligations until sufficient funds became available. Not surprisingly, this resulted in some perverse behavior by businesses, such as withholding invoices until cash was available so that they could postpone the obligation to pay VAT until they could expect payment. It is difficult for businesses to adjust the pace of work to take into account the potential for late payment because contracts generally require that work be performed expeditiously. These practices also made it impossible for the Ministry of Finance (MoF) to know the full extent of government liabilities. They may also have affected the willingness of local firms to bid for contracts or build into tender bids premiums to allow for the cost of delayed payments. 5.5. Complex decision-making arrangements also limit the effectiveness of measures to contain expenditure commitments. For example, decisions on major road investments are made by the Strategic Planning Committee attached to the Prime Minister’s Office but decisions on capital investment projects are made by the Ministry of Transport. Priorities for the use of project funds are set by a supervisory board and the spending is managed by the Albanian Roads Authority (ARA). Maintenance expenditure is capped at 16 percent of the total budget allocation with no regard to project characteristics. In the health 57 sector the cost of pharmaceuticals and the number of items to add to the list of subsidized products are decided by the Council of Ministers; the Ministry of Health sets capital investment priorities; the Health Minister appoints hospital directors; the Health Insurance Fund manages primary and secondary health care programs and payments for pharmaceuticals; and local governments manage application of regional health policies. Negotiations with pharmaceuticals suppliers occur during the budget year, after Parliament has approved budget allocations, and the list of approved drugs may not be complete until well into the fiscal year. It can also be changed when new drugs are added, requiring an additional allocation to meet the extra costs. If pharmaceuticals estimates approved in the annual budget are above the actual cost, there will be a saving, but often the estimates are lower than the eventual cost and actual commitments exceed the budget. The shortfall is managed by allowing unpaid liabilities to accumulate toward the end of the year until funds become available in the new budget year. 5.6. This process is so fragmented that it is difficult to ensure that each set of decision makers take all relevant considerations into account. Each has a specific perspective derived from its mandate and mode of operations but operates separately from the other decision making and management agencies. This can lead to such disconnects as approval of more activities and expenditures than the funds available can accommodate, altering the balance between new construction and service delivery or operations, or diverting funds from other priorities without full awareness of the consequences. 5.7. The framework for preparation of the Medium-Term Budget Plan (MTBP) is established by the generic budget law 32 approved in 2008, supported by detailed instructions 33 from the MoF. The law requires that the Minister of Finance, in January of each year, “prepare for revision and approval of the Council of Ministers a report on macroeconomic assessment and forecast for (a) the last two years, (b) the budgetary year, and (c) forecasts for the following three budget years.” 34 Preparation of the MTBP is therefore a rolling process. 5.8. The onset of the international financial crisis, which heightened the difficulties of preparing annual let alone medium-term budget estimates, has undermined the significance of the MTBP as a planning and management tool. The focus has shifted to estimates for annual or even for shorter periods as budgets are adjusted to meet rapidly changing conditions. These changes have had consequences for multi-year expenditures but the implied adjustments to future obligations have been allowed to accumulate without full recognition of how they affect the budget process and the line ministry management plans. 5.9. Another source of pressure on the spending side of the budget comes from policy decisions and changes to laws during the budget year that were not anticipated when the budget was drawn up. When the changes have financial implications, they add to budget demand, generally without offsets in other spending or new sources of revenue to cover new spending requirements. Where there is already demand for more funds than are available because revenue estimates were overly optimistic, pressure from unfunded new policies and laws makes it very difficult for the government to meet its obligations. Management of Capital Investment 5.10. As discussed in chapter 4, spending on public investments has expanded significantly in recent years, particularly in the transport sector—where spending on maintenance has declined. This 32 Law No.9936 Date 26.06.2008. On Management of Budgetary System In The Republic of Albania 33 “Budget Institution Internal Rules for Preparation of MTBP Submissions and for Monitoring the Delivery of Outputs.” 34 2011 Albania PEFA Assessment. 58 highlights a serious weakness in capital investment management because the full cost of investments is not considered when they are approved. 5.11. Little has changed since the 2010 World Bank study which concluded that, despite closer links between capital investment and integrated planning arrangements, there are still weaknesses in the way projects are developed and considered in the budget process. Some foreign grants are still omitted from the budget and a significant number of unplanned projects are introduced late in the budget process without a comprehensive readiness assessment. 5.12. Major investment projects are discussed by the Strategic Planning Committee attached to the Prime Minister’s Office with input from the MoF. However, because the processes for preparing multiyear plans, screening and approving investment, and preparing annual budgets are still not fully integrated, the links are fragile. 5.13. Albania has a sound planning system in the National Strategy for Development and Integration (NSDI), a medium-term budget framework, and a system to integrate all sector and ministry planning documents with the budget and national development plans. It is therefore surprising that planning for public investments has not been organized on a more orderly basis. This has resulted in excessive investment in roads at the expense of other sectors, and inadequate provision for maintenance, repairs, and restoration to keep assets efficient and sustainable. The version of the NSDI now in preparation offers a timely opportunity to reconsider how capital investment priorities could be incorporated and updated more effectively. There is an opportunity to build in procedures for adapting the plan to significant changes in underlying assumptions, and more flexibility in response to changing circumstances and new information. 5.14. The government could take a variety of actions to make management of capital investment more efficient and effective –including a cost-benefit analysis, better prioritization, and monitoring and evaluation. The difficulty of managing public finances during the international financial crisis may also have limited the extent to which a broad PFM reform agenda could be pursued, given the limited resources available to the MoF. A shrinking revenue and the need for substantial downward adjustment of spending during a budget year has meant that most of the adjustment burden was borne by capital expenditure—one of the few discretionary items in the budget. Cash Flow Management and Treasury Operations 5.15. Over the past few years the government has continued to reinforce Albania’s core PFM system— notably through the adoption of the Treasury Single Account (TSA) and the development of the Albania MoF Treasury System (AMoFTS). A number of enhancements to the latter are planned, in particular adding program modules to cover budget preparation, external assistance, and medium-term budget planning. This will give the MoF a fully functional financial management information system. The system will also be expanded from the present 36 Treasury District Offices to all participating entities (budget institutions and a limited number of municipalities) to enable direct expenditure management and registration of pre-commitments and contracts (commitments can already be registered and entered in the system). 5.16. However, these planned enhancements will still leave some significant cash management and commitment control gaps. Further technical assistance is needed to build up and modernize these core functions. Better cash management data and more accurate commitment information will reinforce the MTBP as future-year projections account for future liabilities and commitments the government has already entered into but that multiyear budgets and plans do not yet recognize. 59 5.17. As budgets are being formulated, each budgetary institution submits to the Treasury with the draft annual budget forecasts of expected monthly cash flow for the entire year. The cash flow forecasts have three components: spending on personnel; other recurrent expenditures; and investment outlays. The Treasury may revise the original forecasts if it considers them unreliable or unrealistic. The Treasury plan can be updated either because of a justified request from a budget institution or by the General Directorate of Treasury if there is a liquidity shortage. 5.18. In recent years, the annual cash flow projections have been made and cash management plans developed well after the new fiscal period had begun. Ideally, annual cash management plans should be finalized by the end of the previous fiscal year, not several months into the new one. There are a number of areas of current practice where quick remedies would allow for more timely and accurate information to be collected and monitored (see below). That in turn will enhance the quality and accuracy of projections of government cash flow needs, cash availability, and need for debt financing instruments. 5.19. Adoption of the TSA is a major accomplishment in the evolution of PFM in Albania. It established the building blocks for a consolidated approach to cash management, improved cash control, and reduced the need for short-term borrowing. The General Directorate of Treasury manages all aspects of the TSA, and therefore all expenditures and revenues of the central government, which consists of ministries, departments, agencies, regional and local government units, and special funds. 5.20. The Treasury reconciles all of its cash balances with the TSA subaccounts in the Central Bank of Albania (BoA) every day. A few donor-financed projects are financed by foreign currency accounts, also maintained in the Central Bank, that are not part of the TSA cash balance.35 However, they are reconciled monthly and reported to the General Directorate of the Treasury. Public enterprises maintain their bank accounts outside the TSA, in the commercial banking system. Since 2012 public enterprise data have been included monthly in preparation of budget reports and annual financial statements. This has given the government a more accurate picture of how much liquidity is available. 5.21. On the revenue side, the Treasury system is not currently directly connected with the tax collection system, though it is recognized that an improved interface is needed to better integrate tax collections into overall cash management. The Treasury receives daily information on total revenues collected through the banking system, but the disaggregated revenues are available only monthly, which complicates cash flow analysis and the government’s ability to effectively manage its cash position. 5.22. Current cash flow management practice, introduced in 2012 to try to mitigate the decline in actual revenue collection, is mostly an administrative rather than a management or analytical task. Cash flow management essentially consists of small adjustments to the budget plan, since it is not built on realistic revenue inflows and cash outflows. Cash management and cash forecasting are considered separate functions, which also makes it hard for the government to finance spending effectively. Because cash forecasting is still weak, actual cash management is more a passive function than an active approach to the reality of low or declining liquidity. 5.23. The Treasury system has the technical capacity to provide the data necessary for cash and liquidity management purposes, but neither the full range of commitments nor the full value of contracts is entered into the system promptly. This limits the effectiveness of any reform to modernize cash flow management. 35 MoF Instruction: On Working Procedures of the Treasury System and Relations with Budgetary Institutions, No 3486/2005. 60 5.24. While information on revenues collected is available in the Treasury Bank of Albania account, it does not seem to be incorporated into cash flow projections and cash management plans. The revenue forecasts of the Macroeconomic and Budget Department are not sufficient for preparing reliable cash- flow forecasts. This is a hurdle for more efficient and effective cash management and results in both unrealized revenues and additional financing costs. 5.25. Current practices do not meet the standards for preparing realistic cash-flow plans: • The cash flow plans produced do not reflect expected in- and outflows; currently they are merely seasonally adjusted versions of the budget. • Though such plans are considered obligatory administrative functions, expectations have not been delineated. It should be made clear that cash-flow plans should reflect what is realistically likely so that they can support the decisions that have to be made. • The plans usually lack analysis; when there is analysis, the results are not built into subsequent cash-flow plans. • Although line ministries prepare cash-flow plans, they do not regularly revise or adjust them. Commitment Controls 5.26. Some commitment controls for both salary and non-salary financial transactions are part of the current control system. In 2013, the Treasury began requiring that contracts be submitted with payment requests. This has made it possible for the Treasury to construct a schedule for remaining payments and update the government’s cash management plan. However, some contracts, particularly those for infrastructure, have multiyear payment schedules, and the Treasury records future-year liabilities and commitments only as a lump sum of residual obligations for future years, even though more precise details of future annual liabilities are available. This practice undermines the medium-term planning and future cash flow projections that line ministries prepare for MoF and Treasury approval. 5.27. The accumulation of arrears is a major concern for the government not only due to the fiscal risk it poses but also because it raises the prices vendors and suppliers charge, because arrears cause liquidity problems for private firms doing business with government. As a result, some of their loans and lines of credit become nonperforming, with a negative impact on the banking system and the financial sector. 5.28. Arrears accumulate in Albania for a combination of reasons. As noted, the effectiveness and reliability of budget planning and management and cash management are areas where reform and performance improvements are needed. The lack of current and timely information about commitments and obligations and the failure to record future obligations accurately are also causing arrears to accumulate. The government is now moving to document and clear current arrears, but there are other steps it can take to ensure that the problem does not recur. 5.29. Expansion of the AMoFTS budget execution system will certainly help in registering transactions promptly and getting the information directly to the Treasury. One of the simplest steps the government can take is to properly define arrears (i.e., payable obligations that are overdue by more than 30 days) in the Organic Budget Law, which currently does not recognize them. 5.30. Another problem has been that acknowledging that there are arrears has been challenging. A common international practice is to measure the time between receipt of an invoice and the date the invoice is actually paid. In Albania, since the Treasury executes all payments on behalf of line ministries (the AMoFTS has not yet been extended to all primary budget institutions), the government considers 61 only the time when the Treasury itself receives the invoice—not when the line ministry received it. Moreover, it is very difficult to determine when an invoice was first received, because the date of the invoice is not consistently recorded in the system. Often invoices are not sent to the Treasury for payment until it is clear that the Treasury has enough liquidity to fulfill pending obligations. 5.31. Treasury instructions on the recording, processing, and reporting of transactions are clear but not always followed. However, the accumulation of arrears, particularly in the transport, education, and health sectors, has stimulated efforts to understand the full scope of the problem. Making the necessary assessment will put the government in a stronger position to clear its arrears. 5.32. At present, commitment controls are effective only in terms of the annual appropriation; that is, no matter what monthly cash flow forecasts individual budget institutions prepare, they are likely to enter commitments into the system only up to the total uncommitted funds in their appropriation. In the current system, contractual obligations for the entire year are entered when the commitments are signed, and the Treasury has taken measures to enforce this new procedure, but multiyear commitments are not currently captured effectively. Monthly payments are validated against planned commitments; should deviations appear, adjustments are negotiated with Treasury. The Treasury complements the appropriation-level commitment controls with periodic cash rationing to ensure that payments do not exceed the cash available. 5.33. The government’s public procurement system, which manages all tenders and processes, has a pre-commitment module available. This module can capture information that will allow the Treasury to more accurately analyze the government cash position and debt financing options and take steps to prevent line ministries from entering into commitments that may threaten the annual deficit target. While the Treasury does receive procurement plans from line ministries, they are submitted on paper, not registered or analyzed electronically. Consequently, they are not currently factored into cash management analysis. It is vital that the government link the Treasury and procurement systems to allow for a daily, real-time flow of information to support cash management and commitment control. 5.34. Daily payments are controlled very tightly by the Central Treasury and its 36 Treasury District Offices. Payments are made only when cash is available. The Central Treasury imposes limits on total daily payments, payment by different Treasury District Offices, and specific types of expenditure (salaries and operational). It can further control which invoices are paid by setting up processing priorities when individual invoices are registered in the system. Daily reports on payments pending for approved expenditures and payments in the authorization pipeline are extracted from the system and distributed to Central and District Office personnel. 5.35. Ceilings for commitments are based solely on the limits set by the budget appropriation. The budget law says that “Authorizing officers of general government units shall maintain information on financial commitments, and shall not allow undertaking of any new commitment if that exceeds the limit of the budget appropriation.” 5.36. In practice there is much less predictability, due to the financial crisis and shortfalls in revenue collection. The MoF in the early stages of budget execution can contain expenditure by freezing parts or all of appropriations and by adjusting budget institution Treasury Plans. Reductions imposed by the MOF are regularized in the mid-year Supplementary Budget and in the final Normative Budget document. 5.37. As noted in the 2011 PEFA Report, Albania has made substantial progress in adopting internal controls to complement controls embedded in the electronic system. The Organic Budget Law passed in 2008 separates financial policy functions (MoF responsibility) from financial administration (line ministry responsibilities). It was complemented by a Financial Management and Control Law, a Law on 62 Internal Audit, the Public Internal Financial Control (PIFC) Policy and Implementation Plan Law, and the Law on Financial Inspection. Numerous Council of Ministers decisions were approved, as were two manuals, on financial controls and on internal audits. 36 The government issued its comprehensive Financial Management and Control Manual in 2010. 37 The manual provides clear direction on all aspects of the PIFC system being implemented. While the legal and structural bases for PIFC are in place, there is considerable work to be done in fully reflecting this fundamental change in internal control throughout management. Many line managers do not understand their responsibilities for operation of their internal financial control system; nor do they understand the major change in the role of the internal auditor from one of inspection to one of advisor on controlling and mitigating risks—including control over expenditure commitments and on-going liabilities entered into by line ministries. Policy Options Albania would need to strengthen its institutions to reinforce financial discipline. While Albania’s public financial management (PFM) has improved since 2006, some aspects of planning and execution remain problematic. The government could consider the following measures: 5.38. Budget Planning, Preparation, and Revision • Contain over-optimism in revenue estimates. Steps that can help improve the realism of revenue forecasts include: (i) setting up an advisory council of independent experts, ideally attached to Parliament or to the High Audit, and/or (ii) references to consensus forecasts and the forecasts of credible financial institutions; (iii) publication of the full assumptions underlying the revenue projections together with the draft MTBP and the annual budget document. • Limit the accumulation of excessive spending obligations within the budget process using two measures successfully applied in other countries through amendment to the Law on Management of the Budget System and budget instructions. o First, ensure that decisions made during the fiscal year that have spending implications recognize government and Parliament-approved ceilings. Where the decisions result in spending obligations that could not be known when budgets were prepared or are outside the control of the spending authorities, build a reasonable contingency allocation into budget plans or identify possible savings from other spending items in advance to keep the net impact on the budget neutral. o Second, have the Ministry of Finance (MoF) carefully vet new policies and changes in legislation before they are finalized to identify the likely fiscal impact. Where there is a financial impact that has not been incorporated into the budget, give the MoF authority, in cooperation with the ministry responsible for the measure, to identify offsetting savings or additional revenue to cover the full cost of the proposed changes. Where full offsetting changes cannot be agreed for the budget year, have the MoF draft a plan for meeting the costs that honors government fiscal policy. • Over the medium term, a move towards a more robust medium-term budget framework specifiying the following aspect would be critical: (i) Spending limits, through which the government makes a commitment to a future expenditure level; (ii) Multi-year controls that 36 Ibid. 37 Minister of Finance, Manual on Financial Management and Control, Order no.89780/ 2010. 63 would allow the government to restrict expenditure pressures arising beyond the annual budget horizon; (iii) Credible baseline estimates which demonstrate the future costs of the current policies, a planning/contingency reserve to absorb the impact of unexpected events; and, (iv) A reconciliation that decomposes deviations of outturns from forecasts for each ministry, policy area, or program into various factors (macroeconomic determinants, operational parameters, accounting changes and policy measures) to increase the framework’s transparency and improve accountability. Capital Investment Management 5.39. It is crucial that reforms to capital investment take account of the public expenditure operating environment and the political considerations driving public investment decisions. The most beneficial actions Albania could take to improve the management of capital investments are as follows: • Ensure that all major projects are subject to rigorous cost-benefit analysis and prioritization in the context of the NSDI and budget processes, as the MoF Public Investment Management Procedures require as part of a graduated appraisal framework. There should be no exceptions. Major projects put forward for consideration outside the normal budget procedures should at most be given in-principle support and then subjected to comprehensive cost-benefit and readiness assessments to ensure that they are affordable, of higher priority than other investments, and are ready to proceed so that they do not tie up funds that could be used more immediately or more effectively. Give projects final approval only within the annual budget process and if they comply with fiscal policy settings, such as sector mid-term spending ceilings. • Ensure that design and preparation are completed before calls for proposals or initiation of tender procedures. There should be a clear explanation of what is to be purchased and the precise specifications to avoid any need for later renegotiation or delays once work begins. • Require explicit project plans before final approval of major projects that incorporate monitoring, reporting, and risk management arrangements. The plans should cover the entire lifecycle of the project, including repairs and maintenance and recurrent operational costs. If assets need to be transferred or disposed of at the end of the project life, the project should specify how that will be done. • Evaluate the outcomes of each major project on completion or implementation. The evaluation should assess how well the project has achieved objectives set out in the approval phase and identify lessons learned for future investments. • Identify ways to build the investment management capacity of the MoF Department for Public Investment and of line ministries. The 2010 World Bank Public Investment Management report noted that Albania had limited capacities for preparing and assessing public investments. Ideally, additional resources should be directed to the investment preparation function and considered in allocation of investment funds so that organizations with major public investment responsibilities, such as the MoF review function, have staff resources commensurate with those responsibilities. The need for additional practical guidance and training on analytical techniques should also be considered, as the World Bank report recommended. 64 Cash Management 5.40. To reform cash management practices, the government could consider the following options: • Specify requirements for preparation of cash-flow plans: Plans should reflect expected real in- and outflows (the closer to the due date the closer o the real liquidity situation is likely to be). o They are to be regularly updated, on a rollover basis. o The Treasury is to define preliminary content; reports might differ by line ministry, depending on the regularity, volume, and specifics of the expected cash flow. o Plans are to be regularly analyzed by the Treasury and MoF departments (e.g., Debt Management for analysis of short-term financing needs and financing instruments). • Define roles and responsibilities for preparing forecasts and analyzing data: o Cash-flow plans are to be prepared by the line ministries that have the best insight into expected cash-flow needs, i.e., units that have concise information on expected flows. o The Treasury should prepare its cash-flow plan on the basis of the information line ministries provide, taking into account commitments that have already been made but are not yet factored (though the responsibility of Treasury is broader than this). o The Treasury cash-flow plan should not be a mechanical summary of the plans received, but should be based on Treasury analysis that identifies trends (especially with regard to revenue collections) and draws conclusions. o As part of its work on the annual cash management plan, the Treasury should prepare different scenarios (pessimistic, optimistic, and most realistic) for both revenue collections and expenditure outlays; ideally these should be updated at least quarterly. This differentiated analysis will allow for a much more powerful approach to cash and [short-term] debt management. • The government should strengthen the link between the Treasury and tax collection systems to enable daily and timely flow of information, which will in turn reinforce the basis for cash flow analysis and practices for managing cash and debt. Commitment Control 5.41. Commitments are controlled through extension of the AMoFTS, and the recent initiative to register all contracts in the Treasury system is having some positive effect. However, there are still steps the government needs to take to reform commitment control to better manage future obligations and instill financial accountability and good financial management practice throughout the government. • Amend the Organic Budget Law to carefully define arrears and clarify the regulations to allow commitments and obligations to be detailed in future budgets as part of the MTBP. • Record commitment of funds for multiyear projects in financial statements and in budget preparation documents to ensure that claims are taken into account in future budget allocations before budgets are approved. • To reinforce the government’s ability to manage commitments in general and to facilitate more accurate cash flow management, link the AMoFTS to the general procurement system. That will give the MoF and the Treasury more timely information on present and future commitments so they can more effectively manage cash availability and short-term debt financing. 65 B. RULES TO ANCHOR FISCAL POLICY Background 5.42. In December 2012 the Albanian parliament removed the 60 percent-of-GDP ceiling on public debt, including the debt of sub-national authorities and guarantees, from the Organic Budget Law. 38 The government has not yet set a new debt ceiling or a fiscal rule. Albania’s only binding fiscal commitment is that the contingency fund (an unallocated appropriation to cover risks related to revenues and expenditures as budgets are implemented) and the reserve fund (an appropriation to finance unusual expenditures that cannot be forecast); together these funds cannot exceed 3 percent of total spending. Art. 12 of the Organic Budget Law states that new borrowing should not exceed capital spending. While this reduces government discretion with regard to recurrent spending, it cannot ensure convergence to a specific debt level. 5.43. Quantitative medium-term targets currently guide Albania’s fiscal policy. The government mentions in its medium-term fiscal framework that fiscal consolidation will be the main priority. 39 The framework also sets a ceiling on total net borrowing of government as a share of GDP over the medium term. 5.44. Albania’s previous GDP Figure 5.1: Budget Versus Actual (Percent) and revenue forecasts have been 2 optimistic. Between 2010 and 2012, tax revenues as a share of GDP were 0 on average 1.7 percentage point lower 2010 2011 2012 -2 than budgeted and nominal GDP 2.3 percent lower than forecast (Figure -4 5.1). As revenues tended to be lower than expected by mid-year, the Tax revenue (in % of GDP) authorities curtailed capital spending GDP in billions of leks to hold the fiscal deficit steady. The government also accumulated arrears Overall balance (in % of GDP) on the invoices of road contractors. Source: IMF 2013. Numerical Rule or Procedural Targets? 5.45. Fiscal rules are generally defined as constraints on fiscal policy expressed as summary indicators of fiscal performance. Indicators might be the government budget deficit, borrowing, debt, a major component thereof, or a combination of these. Ideally, a fiscal rule should be clearly defined, transparent, simple, flexible, and adequate to promote the final goal, enforceable, consistent, and supported by sound policies (Kopts and Symansky 1998). Some fiscal rules emphasize procedures; others set forth permanent numerical targets. 5.46. The government could reinforce procedural rules to ensure that fiscal consolidation remains on track. To enhance transparency, accountability, and prudent fiscal management, procedural rules define the main features of the budget process. They tend to be based on numerical fiscal objectives that each new government must submit after election. These objectives cover at least the length of a 38 Article 58 of the Law on the Management of the Budgetary Systems. 39 The medium-term fiscal framework is prepared jointly by the MoF General Directorate of Budget (GDB) and the Macroeconomic Directorate. After approval by the Council of Ministers, it is published. It includes estimates of growth, revenues, spending, and other key variables. 66 parliament and the need to be consistent with legally binding principles for fiscal policy. The rule include (a) principles for fiscal policymaking; (b) a requirement that the government to set a target for one or more fiscal indicators; (c) the fiscal strategy statement in which the targets are based; and (iv) arrangements for reporting performance against the targets. Australia, New Zealand, the United Kingdom, and other countries place great emphasis on procedural rules. 5.47. The government could also again consider a numerical rule (see box 5.1). Numerical rules establish a permanent, legally binding target for one or more fiscal aggregates (revenues, spending, etc.) that permanently constrain fiscal policy. Setting such a rule could help enhance budgetary discipline and, depending on the rule, improve policy coordination between different levels of government. Box 5.1: Numerical Rules A fiscal rule can target one or more variables, such as the budget balance, the public debt ratio or expenditures and revenues. Each target serves different purposes and has unique own advantages and drawbacks. a) The objective of budget balance rules is fiscal sustainability. The budget balance is determined by variables that are largely under the control of the government, except for interest payments. They can provide clear operational guidance. Budget balance rules can be specified as overall balance, structural or cyclically adjusted balance, and balance “over the cycle.” While the first type of rule does not have economic stabilization features, the other three explicitly account for economic shocks. However, estimating the adjustment, typically through the output gap, makes the rule more difficult to communicate and monitor. A balance “over the cycle rule” has the added disadvantage that remedial measures may be put off to the end of the cycle. A “golden rule,” which targets the overall balance net of capital expenditure, is less linked to debt. b) Debt ceilings also serve the objective of fiscal sustainability. They are easy to communicate and the most effective in terms of ensuring convergence to a debt target. As Albania has experienced in the past, it can require ad hoc fiscal adjustments, which mostly take place at the expense of capital expenditures. The public debt ratio can also be significantly affected by variables outside the control of the government, such as the exchange rate, interest rates, or the realization of contingent liabilities. c) Expenditure rules, such as ceilings on primary or current spending, expenditure growth, or percent of GDP, can help limit the budget deficit but they cannot ensure fiscal sustainability unless combined with a balance, debt, or revenue rule. They can, however, provide operational guidance for the required fiscal consolidation when they are accompanied by debt or budget balance rules. Similarly, revenue rules, which are used rarely, mostly serve other purposes, such as improving revenue collection or limiting tax increases. Given the trade-offs, many countries combine two or more fiscal rules. Using a combination can help address the gaps of each target. For example, a debt ceiling in combination with an expenditure rule would provide not only a link to debt sustainability but also short- to medium-term operational guidance and allow for some counter cyclicality. A combination of a debt and cyclically adjusted budget balance rule could do the same. Lithuania and Poland, for example, complemented their debt rules over time with expenditure or budget balance rules. Countries choose different rules depending on their stage of development. In low-income countries, debt rules are predominant. In advanced countries, cyclically adjusted budget balance rules dominate; even for emerging market economies, this type of numerical target is difficult because it requires calibration of the output gap, which is particularly challenging for countries going through structural transformation. As a result, emerging economies with cyclically adjusted budget balance rules tend to use thresholds of actual economic activity rather than an output gap. 5.48. A numerical rule can send a powerful signal of commitment and enhance credibility, as long as it is supported by strong political commitment. It could help lower Albania’s public debt burden while heightening the credibility of the consolidation path. However, the rule can only be successful with 67 clear political commitment and institutions to monitor and enforce it. Otherwise, it may create problems of its own: 1. Some rules, such as debt ceilings or deficit rules can be procyclical, especially in bad times. Because rules that overcome this limitation tend to be complex, they are more difficult to communicate, monitor and implement. 2. Some rules, such as debt ceilings or deficit rules, can lead to ad hoc reductions in capital spending, which is easier to cut than current spending but may have higher economic and social returns. Albania has experienced ad hoc capital spending adjustments in the past. 3. Any rule can be revoked as the 2012 removal of the debt ceiling from the organic budget law illustrates. 4. Rules can also motivate off-budget operations or reduce transparency, as demonstrated by the accumulation of arrears for road construction.40 5.49. Establishing another fiscal rule would be in line with what is happening throughout Europe. As part of the 2011 Enhanced Stability and Growth Pact, EU member states are advised to translate EU fiscal rules into national legislation that has “a sufficiently strong binding and durable nature.” Also, the Fiscal Compact that the European Council signed in March 2012 introduces several elements for national fiscal rules that member states are expected to adopt by 2014. For example, the compact requires countries to adopt legally binding rules that limit the annual structural deficit to maximum of 0.5 percent for countries with debt above 60 percent of GDP. The rules include a commitment to continuously reduce the public-debt-to GDP ratio to 60 percent at an annual pace of no less than 1/20th the distance between the observed level and the target. To ensure effective implementation of the fiscal rules, the European Commission set out recommendations to make MTBFs more binding, prepare budgets in a more top- down sequence, report more often and comprehensively on fiscal developments and risks, and give independent counsels a larger role. 5.50. One potential disadvantage of numerical rules is that if they have to consider cyclical fluctuations, they soon become very complex, which may undermine their transparency and make them too complicated to be managed with the available capacity. Cyclical adjustments of fiscal balance rules, for example, require calculation of output gaps or trend nominal growth rates, which can be difficult to calibrate. This can undermine credibility unless there is at least one independent institution that can provide credible economic statistics. A procedural rule circumvents this problem because it does not require a fixed numerical value for fiscal targets. 5.51. Albania’s economy is less volatile than those of neighboring countries (Figure 5.2); its economic growth, revenues, expenditures, and deflators do not fluctuate as much. Its inflation targeting regime has been effective in overcoming shocks and keeping inflation low and stable. Low economic volatility makes it easier to apply a simple and transparent numerical rule. However, adoption of a numerical rule should be accompanied by reinforced budgetary procedures. In fact, most countries have legislated elements of both procedural rules and numerical targets (Kumar and Ter-Minassian 2007). 40 Practices to circumvent numerical rules have included reclassifying expenditure from current to capital items in the context of current balance budget rules; using debt instruments not covered by debt ceilings or reclassifying expenditures as acquisitions of financial assets to circumvent net lending rules (Kumar and Ter-Minassian 2007). 68 Figure 5.2: Volatility of Economic Indicators Source: Data from IMF WEO 2013 (Spring). Implementation 5.52. Procedural or numerical, a fiscal rule should be transparent and easy to communicate. The key objective of a rule would be to maintain fiscal sustainability, reduce debt in the medium term, and keep it at a more prudent level over the longer term. The rule should be clearly defined and easy to communicate to ensure that compliance with it is easy to monitor. Given Albania’s history of inaccurate forecasting, the rule should not rely on future variables. Targets embedded in the rule should be revised rarely, e.g., as part of the electoral cycle, and they should capture a large share of public finances and cover the consolidated government. 5.53. Since rules are not a substitute for sound fiscal management, Albania needs to reinforce its public fiscal management to enhance the credibility of its policies. Clear areas of weakness are its unreliable forecasts of the main fiscal variables, the commitment control failures that allowed arrears to accumulate, and the inadequate medium-term budgetary framework. It would be thus important to: (1) have forecasts scrutinized by independent experts, for example, an advisory group of academic and economic professionals from the private sector or a standing fiscal advisory council under the authority of Parliament rather than the administration41; (2) ensure that spending decisions during a fiscal year take into account government- and parliament-approved ceilings; and (3) improve medium-term fiscal planning. 5.54. In current circumstances a procedural fiscal rule might be more beneficial for Albania. A procedural rule would have three advantages: (1) the country would not be tied to a permanent numerical 41 Since 2008, Ireland, Portugal, Romania, Serbia, and the United Kingdom, among other countries have established or are establishing fiscal councils (see, e.g., Calmfors 2012; and Debrun, Gerard, and Harris 2011). 69 target, which would be desirable given the volatile external context and the fact that Albania is still in a state of economic transition. (2) Each government will have to set its own target, which could increase political commitment to meet it. (3) Since the debt ceiling has been removed and the PFM process would first need to be strengthened to ensure successful application of a numerical target, a procedural fiscal target could help lay the foundation for an eventual credible permanent fiscal target. 5.55. A procedural rule would require amendments to Albania’s laws. The law would have to cover (1) permanent principles for fiscal policy; (2) a requirement that the government set fiscal targets to be submitted to Parliament in a Statement of Fiscal Policy after each election; (3) what must be in the Statement of Fiscal Policy; and (4) reporting arrangements. Art. 4 of Albania’s Budget Law currently sets out as a principle “fiscal discipline consistent with macroeconomic stability, and stable economic and social development,” but it does not say anything about fiscal sustainability, such as “maintaining prudent levels of public debt so as to maintain fiscal sustainability.” 5.56. A procedural rule would require that nominal expenditure ceilings for each ministry be binding. The rule would require that the MoF each year present to Parliament a fiscal strategy laying out the target for the central government; binding nominal expenditure ceilings for each ministry; and a summary of the policy measures required to implement the strategy. Under Art. 23 of the Organic Budget Law, the Minister of Finance already submits to Parliament a macro fiscal assessment of the previous two years and the next three years as part of the annual budget, which is passed with a simple majority. Parliamentary discussion of this assessment tends to be limited, and although the MoF assessment and forecasts are sent as information to Parliament, they are not translated into binding ceilings on expenditure for the out-years. Likewise, the sector ceilings communicated to the line ministries are merely indicative; they are not binding. The government could consider a system of mandatory medium-term budget ceilings at the start of the annual budget preparation cycle. This would encourage a more robust and effective approach to planning (see previous section on PFM). 5.57. Successful implementation of a fiscal rule, whether procedural or numerical, requires sufficient budget controls and incentives for policymakers to abide by the rule. The rule can only be monitored and remedial actions taken promptly if measurement of budget aggregates is reliable and timely. The government of Albania already publishes fiscal data every month. Transparency, for example through regular publication of outcomes and external reviews, is vital for encouraging compliance. Building up Albania’s audit systems, both the high state audit and the internal audit system, would support budget execution discipline. Conclusions 5.58. Adoption of a legally binding procedural fiscal rule would support Albania’s fiscal consolidation. Procedural fiscal rules require the government to set legally binding targets. Together with the MTBP they can improve the budget process and lend credibility to fiscal policy over a medium term. Given that the debt ceiling was removed in 2012 and even while the ceiling was in place it was at times undermined by lack of political commitment and inadequate PFM, a procedural rule might be preferable to a permanent numerical fiscal target in the Albanian context. 70 CHAPTER 6. CONCLUSIONS AND POLICY OPTIONS 6. 1. With rising public debt high and a lower growth outlook, fiscal consolidation is a priority for Albania. For this the authorities should move immediately to enhance revenue, reduce spending and contain the fiscal risks, particularly in the energy sector. Rebalancing transport spending away from capital toward maintenance would increase the economic return on infrastructure spending and help support growth as would reforms in the energy sector. Better PFM will be critical to ensure that fiscal resources are used efficiently and transparently; and a credible fiscal policy anchored by a fiscal rule would help provide positive signal to the market and build confidence of investors, both external and domestic. In the medium term the government also needs to define a pension reform agenda; meanwhile, it will need to rein in the pension fiscal deficit through limiting pension increases to inflation rather than approving costly ad hoc increases. The main policy recommendations of this report are summarized in Table 6.1. 6. 2. A second phase of this public finance review will focus on how to improve the quality of spending and efficiency in education, health, and social protection. It will also look into public administration and staffing issues as a crosscutting theme for improving public spending efficiency and service delivery. Table 6.1: Summary of Policy Options Area Policy Options Energy • Expand generation capacity with equitable risk sharing between the state and private investors. • Reduce fluctuations in the power supply and smooth variation in energy tariffs. • Improve the collection of energy bills from public institutional customers. • Reduce distribution losses. • Increase collections from tariff customers including through improved metering and billing. • Settle public institutions' electricity payment arrears and prevent the emergence of further arrears. • Align tariffs more closely with costs, while introducing social assistance that would better protect the most vulnerable, and tighten regulatory compliance. Pensions • Index the average pension to inflation from 2014 onward and avoid costly ad hoc increases; • Gradually raise rural contribution rates to the same levels that the urban self- employed pay. • In the medium-term introduce either 1) an enhanced the Contributory Scheme emphasizing income replacement, or 2) a poverty-focused basic pension. Tax policy • Expand the VAT tax base through exempting rental property and taxing the sale of new construction, including the land value. • Reverse the recent VAT exemptions. • Raise the CIT rate from 10 to 15 percent. • Make the PIT progressive and introduce risk-based audits. 71 Transport • Rebalance spending on roads from capital investments toward maintenance and rehabilitation. • Set up and use a road asset management system—and update it regularly – as a basis for a more efficient and professional management of national road assets. Budget Planning, Preparation, and Revision Public financial management • Contain over-optimism in revenue estimates by having independent experts undertake revenue estimates. Capital Investment Management • Ensure that all major projects are subject to rigorous cost-benefit analysis and prioritization in the context of the budget processes. • Build the investment management capacity (monitoring, reporting, and risk management) for public investment. Cash Management (Treasury) • Establish a direct link between the Treasury and tax collection systems to enable daily and timely cash flow information for managing cash and debt. • Prepare alternative scenarios (baseline and optimistic) for both revenue collections and expenditure outlays; and update them quarterly—to feed into cash and short-term debt management. Commitment Control • Amend the Organic Budget Law to clearly define arrears and clarify the regulations to allow commitments and obligations to be detailed in future budgets as part of the MTBP. • Record commitment of funds for multiyear projects in the treasury system to ensure that claims are taken into account in future budget allocations before budgets are approved. 72 ANNEX 1. KEY FISCAL DATA Table A1.1: Revenue, Expenditure, and Fiscal Deficit 2005–13 (Percent of GDP) Actual 2005 2006 2007 2008 2009 2010 2011 2012 Revenues and grants 24.5 25.3 25.5 26.7 26.0 26.5 25.3 25.0 Tax revenue 22.0 23.0 23.4 24.3 23.6 23.5 23.2 22.9 Expenditures 28.1 28.6 29.2 32.3 33.1 29.7 28.7 28.5 Overall balance (including grants) -3.6 -3.2 -3.7 -5.5 -7.0 -3.1 -3.4 -3.5 Primary balance (including grants) -0.5 -0.4 -1.1 -2.7 -3.9 0.3 -0.3 -0.5 Public debt 56.9 55.9 53.7 55.5 59.4 58.5 59.4 60.8 Source: World Bank staff, IMF, and Albanian Ministry of Finance estimates Table A1.2: Revenue Structure, Percent of GDP 2005 2006 2007 2008 2009 2010 2011 2012 Total revenues 24.4 25.5 25.6 26.8 26.2 26.5 25.0 25.0 Tax revenue 22.0 22.8 23.2 24.3 23.6 23.5 23.1 22.9 National taxes 16.2 17.2 18.0 18.9 18.2 18.2 17.9 17.8 Value-added tax 7.7 8.3 8.9 9.8 9.6 9.3 9.1 9.0 Corporate income tax 2.3 2.5 2.1 1.7 1.5 1.4 1.5 1.2 Excise tax 2.2 2.6 2.9 3.0 2.9 3.2 3.1 2.9 Personal income tax 0.9 1.0 1.5 2.3 2.3 2.2 2.1 2.0 Other national taxes 1.5 1.4 1.5 1.3 1.2 1.5 1.6 2.2 Customs duties 1.6 1.6 1.0 0.8 0.7 0.6 0.5 0.4 Local taxes 1.4 1.2 1.0 1.0 1.1 1.0 0.9 0.8 Property tax 0.1 0.1 0.1 0.1 0.1 0.2 0.1 0.2 Small business tax 0.4 0.3 0.2 0.2 0.2 0.2 0.2 0.1 Social insurance contributions 4.3 4.4 4.3 4.4 4.3 4.4 4.3 4.3 Nontax revenue 1.7 1.8 2.3 2.1 2.1 2.6 1.7 1.8 Source: Ministry of Finance 73 Figure A1.1: Revenue as Percent of GDP by Taxes, Eastern European Countries a: VAT Revenue b: CIT Revenue c: PIT Revenue Table A1.3: Current Structure of Taxation of Small and Medium Enterprises in Albania Category Income Tax VAT Administration Micro Simple patent (fixed amount N/A (unless freelancer) Local (municipal authority) < lek 2 million based on type of business and location) Small (no VAT) Simple Patent and income tax N/A (unless freelancer) Local (municipal authority), lek 2 - 5 million prepayment based on turnover. Annual balance sheet to GDT Final tax 10 percent of the net Final income tax paid to GDT and profit based on annual balance- transferred to local authority sheet. Small (with VAT) Simple patent and income tax Regular VAT, monthly Local (municipal authority), lek 5 - 8 million prepayment based on annual filing/payment Annual balance sheet to GDT turnover. Final income tax at 10 Final income tax paid to GDT and percent on net profit based on transferred to local authority annual balance-sheet. VAT administered by GDT Medium Corporate Income Tax Regular VAT General Directorate of Taxes ≥ lek 8 million (Central government) 74 Table A1.4: Road User Charges (nominal euro millions) 2008 2009 2010 2011 2012 Excise on domestic and imported fuel 172.1 160.2 158.1 163.2 160.1 Registration tax 5.3 5.3 5.1 5.1 0.0 Axle load tax 0.0 0.0 0.0 0.0 0.0 Carbon tax (environmental tax) 4.7 3.9 7.7 12.6 17.1 Circulation tax of foreign vehicles 2.0 1.4 0.0 0.0 0.0 Tax over imported used vehicles 22.2 21.3 17.5 20.3 40.9 Circulation fuel tax 0.0 0.0 6.9 27.0 31.8 Annual tax over used transport vehicles 0.0 0.0 0.0 1.4 8.0 Other 0.1 0.0 0.0 0.0 0.0 Total 206.3 192.0 195.4 229.6 227.8 Source: Ministry of Finance 75 ANNEX 1. ENERGY SECTOR: INSTITUTIONAL SETUP AND MARKET MODEL Figure A2.1 Structure of the Albanian Power Sector Generation Transmission Distribution METE CEZ KESh 24% 76% Tariff Hydro 90% Customers Thermal 1% OST OSSh/ Other Hydro CEZ 8% SPPs/ IPPs Eligible 1% Customers Figure A2.1: The Albanian Market Model SPP: small power producer connected to the electricity distribution system. Qualified suppliers: market participants licensed to supply eligible customers. Source: USAID. The government approved the Albanian Market Model (AMM) in early 2008, which made KESh responsible for the security of overall supply within Albania. Under the AMM structure, KESh operates as both a power generation company and Wholesale Public Supplier (WPS). The WPS/ KESh buys at a regulated price the electricity produced by KESh and electricity produced by independent power 76 producers (IPPs) 42 at a price agreed in Power Purchase Agreements (PPAs), to the extent needed to serve tariff customers. Any electricity not taken by the WPS, KESh may sell on the market. The WPS/ KESh is also the supplier of the last resort: it has to buy any additional electricity needed, to meet the demands of final tariff customers. WPS is obliged to purchase power from IPPs first because of the PPA obligations after purchasing power generated by KESh and then import electricity as needed to serve tariff customers. 42 IPPs: hydro power plants that have entered into power purchase agreement (PPAs) with WPS. 77 REFERENCES Ahmad, Ehtisham. 2010. “Design of a VAT for the GCC Common Market.” In Fiscal Reforms in the Middle East: VAT for the GCC, edited by Ehtisham Ahmad and Abdelrazak Al-Faris. (Cheltenham, UK: Edward Elgar). Alexeev, Michael, and William Pyle. 2003. “A Note on Measuring the Unofficial Economy in the Former Soviet Republics.” Economics of Transition 11 (1): 1–23. Andreoni, James, Brian Erard, and Jonathan Feinstein. 1998. “Tax Compliance.” Journal of Economic Literature 36: 818 –60. Andritzky, Jochen. 2011. “Evaluating Designs for a Fiscal Rule in Bulgaria.” IMF Working Paper WP/11/272, Washington, DC: International Monetary Fund. Bagachwa, M.S.D., and A. Naho. 1995. “Estimating the Second Economy in Tanzania.” World Development 23 (8): 1387–99. Bajada, Christopher. 2002. Australia’s Cash Economy: A Troubling Issue for Policy Makers. Aldershot, UK: Ashgate Publishing Company. ——, and Friedrich Schneider. 2003. “The Size and Development of the Shadow Economies in the Asia- Pacific.” Discussion Paper, Department of Economics, University of Linz, Austria. Bird, Richard M., and Oliver Oldman, eds. 1990. Taxation in Developing Countries (Fourth Edition). Baltimore and London: Johns Hopkins University Press. Boadway, Robin, Maurice Marchand, and Pierre Pestieau .1994. “Toward a Theory of the Direct-indirect Tax Mix.” Journal of Public Economics 55: 71–88. Brou, Daniel, and Kirk A. Collins. 2001. “Winning at Hide and Seek: The Tax Mix and the Informal Economy.” Canadian Tax Journal 49 (6): 1539–61. Cebula, Richard J. 1997. “An Empirical Analysis of the Impact of Government Tax and Auditing Policies on the Size of the Underground Economy: The Case of the United States, 1993-94.” American Journal of Economics and Sociology 56 (2): 173–85. Chatterjee, S., K. Chaudhury, and Friedrich Schneider. 2003. “The Size and Development of the Indian Shadow Economy and a Comparison with 18 Other Asian Countries: An Empirical Investigation.” Discussion Paper, Department of Economics, University of Linz, Austria. Debrun, Marc Gérard and Jason Harris. 2011. “Fiscal Policies in Crisis Mode: Has the Time for Fiscal Councils Come at Last?”. Background Paper No.2. International Monetary Fund, Fiscal Affairs Department Das-Gupta, Arindam. 1994. “A Theory of Hard-to-Tax Groups.” Public Finance, 49 (Supp.): 28–39. Del’Anno, Roberto. 2003. “Estimating the Shadow Economy in Italy: A Structural Equation Approach.” Discussion Paper, Department of Economics and Statistics, University of Salerno, Italy 79 De Paoli, Bianca, Glenn Hoggarth and Victoria Saporta. 2006. “Cost of Sovereign Default”, Financial Stability Paper No. 1, Bank of England. Engelschalk, Michael. 2004. “Creating a Favorable Tax Environment for Small Business Development in Transition Countries.”, in: Alm, James, Jorge Martinez-Vazques and Sally Wallace (eds.). “Taxing the Hard-to-Tax: Lessons from Theory and Practice”, Elsevier. Eichengreen, Barry, Ricardo Hausmann and Ugo Panizza. 2003. “Currency Mismatches, Debt Intolerance and Original Sin: Why they are not the Same and Why it Matters,” NBER Working Paper 10034, National Bureau of Economic Research, Cambridge, MA. Erbas, S. Nuri. 1993. “Presumptive Taxation: Revenue and Automatic Stabilizer Aspects.” IMF Working Paper WP /93/69. Washington, DC: International Monetary Fund. Feige, Edgar L. 1994. “The Underground Economy and the Currency Enigma.” Public Finance/Finances Publiques, 49 (Suppl): 119–36. ——. 1996. “Overseas Holdings of U.S. Currency and the Underground Economy.” In Exploring the Underground Economy, edited by Susan Pozo. (Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. Feinstein, Jonathan S. 1999. “Approaches for Estimating Noncompliance: Examples from Federal Taxation in the United States.” Economic Journal, 109 (456): 360 –69. Fleming, M.H., J. Roman, and G. Farrel. 2000. “The Shadow Economy.” Journal of International Affairs 53 (2): 64–89. Friedman, E., S. Johnson, D. Kaufmann, and P. Zoido-Labton. 2000. “Dodging the Grabbing Hand: The Determinants of Unofficial Activity in 69 Countries.” Journal of Public Economics 76 (4): 459–93. Giles, David E. A.; Tedds, Lindsay M. 2002. “Taxes and the Canadian Underground Economy.” Canadian Tax Paper No. 106, Canadian Tax Foundation, Toronto, Ontario. Hartzenburg, G.M., and A. Leimann. 1992. “The Informal Economy and its Growth Potential.” In Economic Growth in South Africa, edited by E. Adebian and B. Standish. Oxford, UK: Oxford University Press, 187–214. Heggie, Ian, & Piers Vickers. 1998. Commercial Management and Financing of Roads.” World Bank Technical paper No. 409. Washington DC Hill, Roderick, and Muhammed Kabir. 1996. “Tax Rates, the Tax Mix, and the Growth of the Underground Economy in Canada: What Can We Infer?” Canadian Tax Journal/Revue Fiscale Canadienne 44 (6): 1552–83. International Monetary Fund. 2013. “Albania – 2012 Article IV Consultation.” Washington, DC: IMF. Johnson, Simon, and Daniel Kaufmann. 2001. “Institutions and the Underground Economy.” In A Decade of Transition: Achievements and Challenges, edited by O. Havrylyshyn and S. Nsouli. Washington, DC: International Monetary Fund. Johnson, Simon, Daniel Kaufmann, and Andrei Shleifer. 1997. “The Unofficial Economy in Transition.” Brookings Papers on Economic Activity, Fall. Washington, DC: Brookings Institution. 80 Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobatón. 1998a. “Regulatory Discretion and the Unofficial Economy.” American Economic Review 88 (2): 387–92. Johnson, Simon, Daniel Kaufmann, and Pablo Zoido-Lobatón. 1998b. “Corruption, Public Finances and the Unofficial Economy.” World Bank Discussion Paper, Washington, DC. Jung, Young H., Arthur Snow, and Gregory Trandel. 1994. “Tax Evasion and the Size of the Underground Economy.” Journal of Public Economics 54: 391–402. Kaufmann, Daniel, and Aleksander Kaliberda. 1996. “Integrating the Unofficial Economy into the Dynamics of Post-socialist Economies: A Framework of Analyses and Evidence.” In Economic Transition in Russia and the New States of Eurasia, edited by B. Kaminski. London, UK: M. E. Sharpe. Kraay, Aart and Vikram Nehru. 2004. “When is external debt sustainable?”, World Bank Policy Research Working Paper No. 3200, Washington, DC. Kumar, Manmohan, and Teresa Ter-Minassian. 2007. “Promoting Fiscal Discipline”. Washington, DC. Kopts, George and Steven A. Symansky. 1998. What are fiscal policy rules? Series: Occasional Paper No. 162. Washington, DC. International Monetary Fund. Lackó, Mária. 1998. “The Hidden Economies of Visegrad Countries in International Comparison: A Household Electricity Approach.” In Hungary: Towards a Market Economy , edited by L. Halpern and Ch. Wyplosz, (Cambridge, MA: Cambridge University Press, 128–52. ——. 2000. “Hidden Economy – An Unknown Quantity: Comparative Analysis of Hidden Economics in Transition Countries 1989– 95.” Economics of Transition 8 (1): 117–49. Levy Yeyati, Eduardo and Ugo Panizza. 2006. “The Elusive Costs of Sovereign Defaults.” CIF Working Paper No. 11/2006. Lippert, Owen, and Michael Walker, eds. 1997. The Underground Economy: Global Evidences of its Size and Impact. Vancouver, BC: Fraser Institute. Loayza, Norman V. 1996. “The Economics of the Informal Sector: A Simple Model and Some Empirical Evidence from Latin America.” Carnegie-Rochester Conference Series on Public Policy 45: 129–62. Mogensen, Gunnar V., Hans K. Kvist, Eszter Körmendi, and Soren Pedersen. 1995. “The Shadow Economy in Denmark 1994: Measurement and Results.” Study No. 3, The Rockwool Foundation Research Unit, Copenhagen, Denmark. Morris, B. 1993. “Editorial Statement.” International Statistical Yearbook. Budapest, Hungary. Murphy, Richard. 2012 “Closing the European Tax Gap.” A report for Group of the Progressive Alliance of Socialists & Democrats in the European Parliament. Tax Research LLP. OECD. 2011. “Pensions at a Glance 2011: Retirement-Income Systems in OECD and G20 Countries”. Paris, France. 81 Palda, Filip. 1998. “Evasive Ability and the Efficiency Cost of the Underground Economy.” Canadian Journal of Economics 31 (5): 1118–38. Portes, Alejandro. 1996. “The Informal Economy.” In Exploring the Underground Economy, edited by Susan Pozo. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. Pozo, Susan, ed. 1996. Exploring the Underground Economy. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. Republic of Albania – Ministry of Finance. 2013. Debt Indicators as of March 31, 2013. Tirana: General Directorate of Public Debt Management. ——. 2013. Economic and Fiscal Programme 2013–2015, January 2013.Tirana: MoF. ——. 2013. Public Debt Management Strategy 2013–2015. Tirana: General Directorate of Public Debt Management. Reinhart, Carmen, and Kenneth Rogoff. 2008. “This Time is Different: A Panoramic View of Eight Centuries of Financial Crises.” NBER Working Paper 13882. Cambridge, MA: National Bureau of Economic Research. Rider, Mark, and Roubina Ather Ahmed. 2008. Pakistan’s Tax Gap: Estimates by Tax Calculation and Methodology.” ISPWP 0811 working paper. World Bank, Washington, DC, and Georgia State University. Rogoff, Kenneth. 1998. “Blessing or Curse? Foreign and Underground Demand for Euro Notes.” Economic Policy: The European Forum 26: 261–304. Schneider, Friedrich. 1997. “The Shadow Economies of Western Europe.” Journal of the Institute of Economic Affairs 17 (3): 42–48. ——. 1998. “Further Empirical Results of the Size of the Shadow Economy of 17 OECD Countries over Time.” Discussion Paper, Department of Economics, University of Linz, Linz, Austria. ——..2000. “The Increase of the Size of the Shadow Economy of 18 OECD Countries: Some Preliminary Explanations.” Paper presented at the Annual Public Choice Meeting, March 10–12, Charleston, SC. ——. 2002. “The Value Added of Underground Activities: Size and Measurement of the Shadow Economies of 110 Countries All Over the World,” Centre for Tax System Integrity. ——. 2003a. “The Size and Development of the Shadow Economy around the World and the Relation to the Hard-to-Tax.” Working Paper, Johannes Kepler University, Linz, Austria. ——. 2003b. “The Shadow Economy.” In Encyclopedia of Public Choice, edited by Charles K. Rowley and Friedrich Schneider. Dordrecht, Netherlands: Kluwer Academic Publishers. ——, and Dominik H. Enste. 2000. “Shadow Economies: Size, Causes, and Consequences.” Journal of Economic Literature 38 (1): 77–114. ——, and Dominik H. Enste. 2002. The Shadow Economy—An International Survey. Cambridge, MA: Cambridge University Press. 82 Spiro, Peter S. 1993. “Evidence of a Post-GST Increase in the Underground Economy.” Canadian Tax Journal/Revue Fiscale Canadienne 41 (2): 247–58. Sturzenegger, Federico. 2002. “Default Episodes in the 90s: Factbook and Preliminary Lessons.” Manuscript, Universidad Torcuato Di Tella, Buenos Aires, Argentina. Terkper, Seth. 2003. “Managing Small and Medium-Size Taxpayers in Developing Economies.” Tax Notes International, 13 January: 211–234. Thomas, Jim J. 1992. Informal Economic Activity. New York and London: Harvester Wheatsheaf. ——. (1999). “Quantifying the Black Economy: ‘Measurement without Theory’ Yet Again?” Economic Journal 109 (456): 381–389. Tomz, Michael. 2007. Reputation and International Cooperation: Sovereign Debt Across Three Centuries. Princeton, NJ: Princeton University Press. Witte, A.D. (1987). “The Nature and Extent of Unreported Activity: A Survey Concentrating on Recent U.S. Research.” In The Unofficial Economy: Consequences and Perspectives in Different Economic Systems, edited by S. Alessandrini and B. Dallago (Aldershot: Gower). World Bank. 2003. “Ukraine: Tax Policy and Tax Administration.” Report No. 26221-UA, World Bank, Kyiv. ——. 2006, “Albania—Restructuring Public Expenditure to Sustain Growth: a Public Expenditure and Institutional Review.” Poverty Reduction and Economic Management Unit, Europe and Central Asia Region, Report No. 36453-AL, Washington, DC. ——.. 2011. “Albania – Debt Management Performance Assessment (DeMPA)." Washington, DC ——. 2013. “Albania – Pension Policy Note: Key Challenges and Directions for Reform.” Washington, DC. Yoo, Tiho, and Jin K. Hyun. 1998. “International Comparison of the Black Economy: Empirical Evidence Using Micro-level Data.” Paper presented at the 1998 Congress of the International Institute of Public Finance, Cordoba, Argentina. 83