Report No. 44350-CV Cape Verde Enhancing Planning to Increase Efficiency of Public Spending Public Expenditure Review (In Two Volumes) Volume II: Background Chapters February 2009 PREM 4 Africa Region Document of the World Bank Currency Equivalent Exchange Rate Effective as of February 2009 Currency Unity = Escudo (CVE) US$1.00 = CVE 70.5 Fiscal Year January 1––December 31 ACRONYMS ADP Water Company of Portugal AGECABO Agency for Promotion of Employment and Local Development ANMCV National Association of Cape Verdean Municipalities AP Public administration APR Annual Progress Report ARE Economic Regulatory Agency ARM Multisectoral Regulatory Agency ASA Empresa Nacional de Aeroportos e Segurança Aérea (parastatal that manages airports) BCA Atlantic Commercial Bank BCN Business Bank of Cape Verde BCV Bank of Cape Verde BIA Inter-Atlantic Bank BT Treasury Bill CECV Caixa Económica de Cabo Verde CFAA Country Financial Accountability Assessment CME Machinery and Equipment Center CPAR Country Procurement Assessment Review COFOG Classification of the Functions of Government of the United Nations CPI Consumer Price Index CVE Cape Verde Escudos DGCI General Directorate for Income Tax DGCP General Directorate for Public Accounting DGO Budget General Directorate DGP General Directorate for Planning DGPE Directorate General for State Assets DGPOC General Directorate for Planning, Budgeting and Management DGT Directorate General of the Treasury DSP Directorate for Payment Services EDP Electricity Company of Portugal ELECTRA Electricity and Water Company EMPROFAC National Company of Pharmaceutical Products ENACOL National Fuel Company of Cape Verde ENAPOR National Port Authority EU European Union FAIMO Labor-Intensive Public Works FBL Framework Budget Law FEF Financial Equilibrium Fund FOB Freight on Board GDP Gross Domestic Product GEP Directorate of Studies and Planning GoCV Government of Cape Verde HC Hidden Cost ICAO International Civil Aeronautics Organization IGF General Inspectorate for Finances ii IMF International Monetary Fund INE National Statistics Institute INFA National Institute of Financial Support to Agribusiness INPS National Institute for Social Protection INTERBASE Company for Sea Products Commercialization INTOSAI International Organization of Supreme Audit Institutions IPSAN International Public Sector Accounting Norms IT Information Technology MDG Millennium Development Goal ME Ministry of Education MFA Ministry of Foreign Affairs MFAP Ministry of Finance and Public Administration MH Ministry of Health MIT Ministry of Infrastructure and Transport MITM Infrastructure, Transport and Maritime Affairs MJIA Ministry of Justice and Internal Administration MFF Municipal Finance Fund MTEF Medium-Term Expenditure Framework NDP National Development Plan NIF Tax Identification Number NOSI Operational Nucleus for Information Systems O&M Operation and Maintenance OECD Organization for Economic Co-operation and Development PETROGAL Petroleum Company of Portugal PIP Public Investment Program PIS Public Investment System PNCP Plano Nacional de Contabilidade Pública (Chart of accounts) PRGF Poverty Reduction and Growth Facility PROMOTORA Venture Capital Stock Corporation PRSC Poverty Reduction Support Credit PSI Policy-Support Instrument QUIBB Questionario Unificado de Indicadores de Bem-Estar (Unified Survey of Well Being Baseline Indicators) RMF Road Maintenance Fund SDR Special Drawing Rights SIGOF Financial Management Integrated System SOE State-owned Enterprise SONANGOL National Fuel Society of Angola STAD Secretariado Técnico de Apoio ao Desenvolvimento (Technical Secretariat for Development Support) STASCO Shell International Trading and Shipping Company Limited TACV Airlines of Cape Verde TdC Tribunal de Contas (General Audit Office) TRANSCOR Private Company of Municipal Transportation US-MCC US Millennium Challenge Account VAT Value-added Tax WTO World Trade Organization Vice President: Obiageli K. Ezekwesili (AFRVP) Country Director: Habib M. Fetini (AFCF1) Sector Director: Sudhir Shetty (AFTPM) Sector Manager: Antonella Bassani (AFTP4) Task Team Leader: Manuela Francisco (AFTP4) iii CONTENTS 1. MACROECONOMIC DEVELOPMENTS ....................................................................................... 1 Robust Growth and Low Inflation.......................................................................................................... 1 External Position .................................................................................................................................. 21 Monetary Policy In Line with the Peg.................................................................................................. 31 Fiscal Consolidation ............................................................................................................................. 51 Improvement in Debt Management...................................................................................................... 51 Recommendations ................................................................................................................................ 61 2. FISCAL PERFORMANCE............................................................................................................... 71 Trends and Composition of Revenue ................................................................................................... 71 Trends and Composition of Public Expenditure................................................................................. 101 Efficiency of Budget Implementation and Role of Public Investment Program ................................ 151 Administrative Expenditure Patterns.................................................................................................. 181 Relationship among GPRSP-1, MTEF, and the Budget..................................................................... 211 3. PUBLIC FINANCE MANAGEMENT........................................................................................... 251 Organizational Structure..................................................................................................................... 251 Budget Process ................................................................................................................................... 261 Revenue Management ........................................................................................................................ 301 Public Expenditure Execution ............................................................................................................ 331 Treasury Management ........................................................................................................................ 341 Payments ............................................................................................................................................ 361 Debt Management .............................................................................................................................. 391 Accounting ......................................................................................................................................... 401 Internal Control .................................................................................................................................. 421 External Control ................................................................................................................................. 441 4. FISCAL DECENTRALIZATION IN CAPE VERDE STATUS AND PERSPECTIVES......... 481 Status of Decentralization................................................................................................................... 481 Responsibilities of the Municipalities ................................................................................................ 491 New Draft law on Decentralization.................................................................................................... 501 Resources of the Municipalities ......................................................................................................... 521 Main Expenditure Items of the mMunicipalities................................................................................ 621 Budget Preparation and Execution ..................................................................................................... 631 Fiscal Balance and Credibility of the Budget..................................................................................... 681 Vertical and Bottom-up Control of Municipalities’ Finances ............................................................ 711 Accountability to Citizens .................................................................................................................. 721 Municipalities’ Capacity .................................................................................................................... 731 Conclusion and Recommendations .................................................................................................... 751 5. INFRASTRUCTURE PUBLIC EXPENDITURE REVIEW ....................................................... 771 Back ground and Motivation .............................................................................................................. 771 Brief Overview of Infrastructure Delivery ......................................................................................... 801 Infrastructure Spending ...................................................................................................................... 891 Investment System ............................................................................................................................. 941 Efficiency of Spending....................................................................................................................... 981 Emerging Messages and Recommendations .................................................................................... 1041 Baseline Scenario ............................................................................................................................. 1141 Alternative and Stress Scenarios ...................................................................................................... 1151 Baseline Scenario ............................................................................................................................. 1151 Alternative Scenarios and Stress Tests............................................................................................. 1161 iv Annexes Annex 1: Joint Bank-Fund Debt Sustainability Analysis ...................................................................... 1091 Annex 2: The Organizational System..................................................................................................... 1251 Annex 3: The Risks of a Payment System ............................................................................................. 1311 Annex 4: Evolution of Formula Based Transfer per Municipality Size................................................. 1341 Annex 5: Revenues for Five Municipalities........................................................................................... 1351 Annex 6: Intergovernmental Arrears––Some data ................................................................................. 1361 Annex 7: Use of PEFA Framework for Municipal Finances ................................................................. 1401 Annex 8: Hidden-Cost Deficit Methodology ......................................................................................... 1441 Annex 9: Accounting for Infrastructure––Methodology........................................................................ 1471 Tables Table 1.1: Real GDP Growth and Inflation, 2000–06 ................................................................................ 1 Table 1.2: GDP: Aggregate Supply 2000–06........................................................................................... 21 Table 1.3: Balance of Payments, 2000–06) .............................................................................................. 31 Table 1.4: Central Government Fiscal Operations, 2000–06 ................................................................... 51 Table 1.5: Key Debt Policy Indicators ..................................................................................................... 61 Table 2.1: Revenues 2002–07 (CVE million)........................................................................................... 91 Table 2.2: Revenues, 2002–07 (% of GDP) ............................................................................................. 91 Table 2.3: Revenues, 2002–07 (% of total revenues)............................................................................... 91 Table 2.4: Revenues, 2002–06 (executed revenues as % of budgeted revenues)................................... 101 Table 2.5: Expenditure, 2002–07 (CVE million) .................................................................................... 121 Table 2.6: Expenditure 2002–07 (% of GDP) ........................................................................................ 141 Table 2.7: Expenditure 2002–07 (% of total expenditure) ..................................................................... 141 Table 2.8: Transfer from Central Government for Autonomous Institutions, 2002–07......................... 151 Table 2.9: Actual Expenditure as % of Budgeted Expenditure, 2002–06 .............................................. 161 Table 2.10: Public Investment Program (PIP) - Execution and funding sources, 2002–07 ..................... 171 Table 2.11: Organic Classification of Executed Expenditures, 2002–07 ................................................. 191 Table 2.12: Organic Classification of Executed Expenditures, 2002–07 ................................................. 191 Table 2.13: Organic Classification of Executed Expenditures, 2002–07 ................................................. 201 Table 2.14: Organic classification of executed expenditures, 2002–07 ................................................... 201 Table 2.15: GPRSP-1 and the Public Investment Program, 2005–06 ...................................................... 221 Table 2.16: GPRSP-1 Benchmarks for Priority Sectors and MTEF Projections, 2004–07...................... 221 Table 2.17: MTEF and Annual Budget Allocations, 2005–07................................................................. 231 Table 4.1: Diversity of Municipalities.................................................................................................... 501 Table 4.2: Source of Revenues ............................................................................................................... 521 Table 4.3: Principal Internal Sources of Revenues................................................................................. 541 Table 4.4: Resources Available as a % of Resources Expected at the Planning Stage .......................... 561 Table 4.5: Comparison Between Share of Formula-based Transfers and Share of Contracts-programs Amounts Received by Each Municipality, 2005–06 ............................ 591 Table 4.6: Principal Budget Items .......................................................................................................... 621 Table 4.7: Budget Tables and Information Annex to be Prepared by Municipalities According to 2005 Local Finances Law................................................................................ 651 Table 4.8: Variation Between Approved Budget and Executed Budget ................................................ 661 Table 4.9: Milestones in Budget Preparation and Control ..................................................................... 671 Table 4.10: Sources of Debt ..................................................................................................................... 691 Table 4.11: Municipal Debts to Public and Para-public Enterprises........................................................ 701 Table 4.12: Number of Staff per Education Level ................................................................................... 741 Table 5.1: Benchmark of Access Indicators ........................................................................................... 781 Table 5.2: Key Infrastructure Assessment for Cape Verde, by Island ................................................... 791 Table 5.3: ELECTRA - Power Statistics ................................................................................................ 811 Table 5.4: Water Supply in Cape Verde................................................................................................. 831 v Table 5.5: Roads Network (2005) .......................................................................................................... 841 Table 5.6: Passengers (arrivals + departures + in transit)....................................................................... 871 Table 5.7: Passengers in-transit .............................................................................................................. 871 Table 5.8: ENAPOR - Traffic Statistics ................................................................................................. 881 Table 5.9: International Benchmarking of Infrastructure Spending Levels, 1998–2004 ....................... 891 Table 5.10: Budget Allocation of Public Expenditure Between Infrastructure and Social Spending - A comparison ........................................................................................... 921 Table 5.11: On- and off-budget Infrastructure Expenditures ................................................................... 921 Table 5.12: Public Expenditures by Infrastructure Sector........................................................................ 931 Table 5.13: Comparison of Official Development Assistance for infrastructure ..................................... 941 Table 5.14: Sources of Public Spending for Infrastructure ...................................................................... 941 Table 5.15: Comparison GPRSP-1, MTEF, Annual Budget .................................................................... 951 Table 5.16: Transports: Comparison GPRSP-1, MTEF, Annual Budget................................................. 961 Table 5.17: On-budget Investment (Execution) by Sector ....................................................................... 971 Table 5.18: Financial Indicators for SOEs ............................................................................................... 991 Table 5.19: Cost coverage Annual Average, 2001–06 ........................................................................... 1011 Table 5.20: Revenues and (operative) costs allocation, 2004–06........................................................... 1011 Table 5.21: Cape Verde’s Hidden Cost in Electricity and Water Distribution, 2001–06....................... 1041 Figures Figure 2.1: Consumer Price Index, 2000–06 ............................................................................................. 21 Figure 2.2: Reserve Coverage (Months of Prospective Imports of Goods and Services), 2000–06.......... 41 Figure 2.3: Evolution of Emigrants’ Deposits (Stock) and Remittances (flows), 2000–07....................... 41 Figure 3.1: Government revenue collection system................................................................................. 311 Figure 3.2: Disbursements - Payment orders ........................................................................................... 381 Figure 3.3: System Fact and Accounting Registration ............................................................................ 421 Figure 4.1: Resource Composition, 2005 (2004 for Praia) ...................................................................... 531 Figure 4.2: Transfers or Land Sales (CVE), 2005 (2004 for Praia)......................................................... 541 Figure 4.3: Share of Capital and Recurrent Expenditures in 2005 Budget .............................................. 631 Figure 4.4: Share of Capital and Recurrent Expenditures (staff and non staff) in 2005 budget .............. 631 Figure 4.5: Comparison, Resources, and Expenditures: Selected Municipalities, 2004-06 ................... 681 Figure 5.1: Population Density ................................................................................................................ 771 Figure 5.2: Effective Electricity and Water Tariff................................................................................... 821 Figure 5.3: Losses Due to Power Outages - A comparison ..................................................................... 821 Figure 5.4: Sanitation Services ................................................................................................................ 841 Figure 5.5: TACV - Performance Indicators, 1997–2005 ....................................................................... 861 Figure 5.6: ASA - Distribution of Aircrafts and Passengers, 2006.......................................................... 871 Figure 5.7: Institutional Responsibilities for Infrastructure Delivery...................................................... 911 Figure 5.9: Hidden cost in Electricity and Water Distribution .............................................................. 1031 Boxes Box 2.1: Medium-Term Expenditure Framework................................................................................... 241 Box 3.1: Summary of the Draft Budget Framework Law ....................................................................... 281 Box 4.1: “Local Power� in the Cape Verdean Constitution.................................................................... 491 Box 4.2: Municipality of São Vicente - Impact of Introducing IT.......................................................... 751 Box 5.1: Chile’s Model of Investment Planning ..................................................................................... 971 vi Macroeconomic Developments ROBUST GROWTH AND LOW INFLATION The Government’s development strategy for 2004–07 was set out in the Growth and Poverty Reduction Strategy Paper (GPRSP-1). The GPRSP-1 was built on the Interim Poverty Reduction Strategy Paper, which in turn was underpinned by the National Development Plan. GPRSP-1 was part of a broad strategy underway in Cape Verde since the beginning of the decade to promote growth and reduce poverty. The GPRSP-1 focused on five key pillars: (i) promoting good governance; (ii) improving competitiveness and private-sector-led growth, (iii) fostering human and capital development; (iv) strengthening social security and solidarity; and (v) improving infrastructure and land-use management. The authorities concluded in May 2008 the second Growth and Poverty Reduction Strategy Paper (GPRSP-2), which covers the period 2008-11 and maintains the thrust of the GPRSP-1, being broadly based on the same five pillars. Cape Verde has continued to experience robust growth over the last years. GDP growth averaged 6.3 percent over 2002–06, with real GDP growth in 2006 reaching 10.8 percent. Estimates for 2007 indicated that GDP growth rate was likely to be approximately 7 percent. With the exception of 2004, real GDP growth exceeded GPRSP-1 estimations.1 Reflecting this solid performance, Cape Verde graduated to middle income status on January 1, 2008. Prospects for the medium term also are optimistic. According to the most recent projections, GDP growth is expected to stand at 7.5 percent on average over 2008–10. Even though it is still high, unemployment fell to 18.6 percent in 2006, down 6.1 percentage points from 2005. Table 0.1: Real GDP Growth and Inflation, 2000–06 (%) Indicator 2002 2003 2004 2005 20061 20072 Real GDP growth (annual percentage change) 4.9 4.7 4.3 6.5 10.8 6.9 Real GDP per capita growth (annual percentage change) 3.0 2.3 2.5 4.6 8.8 5.0 CPI annual average (annual percentage change) 1.9 1.2 -1.9 0.4 5.4 4.9 Source: Ministry of Finance and Public Administration, IMF, and staff estimates. Notes: 1 Preliminary estimates. 2 Projections. The strong economic growth is driven largely by tourism-related activities. The fast economic growth in Cape Verde is mostly explained by public and private investment directed principally toward infrastructure development and tourism. In 2006 tourism receipts amounted to approximately 19 percent of GDP, whereas in 2002 they reached only 10 percent. Tourism also is driving growth in other industries, most notably construction, transportation, telecommunications, and financial services. Tourism-related investments are expected to continue to evolve positively, with approximately US$1 billion worth of FDI in tourism in the pipeline. On the other hand, agriculture, fisheries, and the manufacturing industry capture only a 1 PRSP projections for real GDP growth rate were: 5.5%, 6.0%, 6.5%, and 7.0% for 2004, 2005, 2006, and 2007, respectively. For more details, see annex 1. 1 small share of the economy (Cape Verde does not have natural resources and labor; electricity and water costs are high). Table 0.2: GDP: Aggregate Supply 2000–06 (% of GDP) Component 2002 2003 2004 2005 20061 Supply Primary 6.8 6.5 7.1 6.6 6.0 Secondary 20.0 20.3 23.1 23.7 22.3 Tertiary 72.9 73.2 69.8 69.7 71.0 Source: Ministry of Finance and Public Administration, IMF, and staff estimates. Note: 1 Preliminary estimates. Inflationary pressures are moderate and supply driven. Bolstered by a firm monetary policy and an exchange rate pegged to the Euro, consumer inflation has been subdued in the past few years (-1.9 and 0.4 percent respectively in 2004 and 2005). However, in 2006 inflation reached approximately 5.4 percent, largely reflecting supply shocks (poor rainfall temporarily drove up domestically produced food prices). In 2007, inflation stood at around 4.4 percent, once again driven by food prices (domestically produced). The recent increase in international food prices has not yet had an impact on domestic prices (GPRSP-2), owing to the large stocks of cereals. As these stocks are reduced, prices may increase, but this is not likely to threaten fiscal or external stability in the short term given the fiscal space and comfortable level of international reserves. Inflation over the medium term is expected to stabilize at around 2 percent consistent with the currency peg. Figure 0.1: Consumer Price Index, 2000–06 195.00 190.00 185.00 180.00 175.00 170.00 200 2001 200 200 200 200 200 CPI CPl excluding f ood, energy and wat er Source: National Institute of Statistics, IMF, and staff estimates. Note: Index level 1989 = 100. EXTERNAL POSITION The exchange rate has served Cape Verde well as an anchor for price and financial stability. The anchor currency, the Euro, also is the currency of its main trading partners (76 2 percent of exports, 80 percent of imports, and more than 70 percent of total remittances).2 The appreciation of the Euro (and hence the CVE) against the dollar does not appear to have affected Cape Verde’s external competitiveness, reflecting the country’s relatively stronger links to Europe with respect to trade, tourism, and remittance flows. In addition, the sharp increase in oil prices could have had a much larger impact if it had not been partially offset by the appreciation of the Euro against the US dollar. According to the IMF CGER-type analysis, the REER is in line with fundamentals and policies. Furthermore, the strong performance of FDI suggests improved competitiveness. The external current account deficit has narrowed over recent years, falling from over 14.4 percent of GDP in 2004 to approximately 5 percent of GDP in 2006. The improvements in the external position reflect primarily increasing revenues from tourism, effective demand management, and a more diversified import source base. Exports are estimated to have grown by 35 percent in 2006, compared to 24 percent in 2005. In contrast, imports grew by 23.4 percent in 2006 compared to 0.5 percent in 2005.3 The increase in imports has been largely driven by FDI- related imports (investment goods). Table 0.3: Balance of Payments, 2000–06 (US$ million) Component 2002 2003 2004 2005 2006 Trade balance -236.2 -307.7 -378.8 -347.1 -391.2 Imports -278.1 -360.4 -436.1 -435.5 -472.4 Exports 41.9 52.7 57.3 88.4 81.2 Current account balance Excluding official transfers -106.5 -126.0 -114.8 -54.2 -65.4 Including official transfers -71.0 -77.2 -62.1 -33.8 -50.6 Capital and financial account (net) 96.7 63.2 94.9 78.8 104.6 Overall balance 26.9 -7.1 34.9 56.1 47.7 Source: Central Bank of Cape Verde and IMF. MONETARY POLICY IN LINE WITH THE PEG Monetary policy was consistent with the goal of strengthening the sustainability of the exchange rate pegged to the Euro by steadily building up reserves during the entire analysis period. In the context of the fixed exchange rate regime, monetary policy is subordinated to the target level of international reserves deemed appropriate to support the peg. The Central Bank of Cape Verde (BCV) has successfully fulfilled its objective: reserve coverage has increased steadily over the last years (figure 1.4). However, in the first semester of 2008, the accumulation of reserves declined due to the reduction of current transfers and private debt re- payments. The decrease of remittances from emigrants is due to the depreciation of the dollar against the Euro and the global economic slowdown. Furthermore, the global financial crisis, which is affecting some of the countries that invest in Cape Verde, has a repercussion in the degree of the foreign investment. 2 Central of Bank of Cape Verde, Annual Report (2007). 3 Measured in CVE. 3 To prevent capital inflows from leading to inflationary pressures, the BCV has intervened since 2006 by selling bank securities. Emigrant deposits (stock) continue to grow but at a decreasing rate, which is largely explained by the narrowing spread between Euro and domestic deposit rates. BCV is closely monitoring interest rate differentials with the Euro area and the US and is targeting external interest rates differentials to prevent the outflow of emigrant deposits (which represent 40 percent of the total bank deposits). Furthermore, remittances flows fell in 2004 and 2007. In 2007 they dropped by CVE 179 million, which may be related to the depreciation of the dollar and to the raised living standards in Cape Verde. Even so, remittances in 2006 represented 10 percent of the GDP. Figure 0.2: Reserve Coverage (Months of Prospective Imports of Goods and Services), 2000–06 4 Source: Central Bank of Cape Verde. 3.5 3 2.5 2 1.5 1 0.5 0 2000 2001 2002 2003 2004 2005 2006 Figure 0.3: Evolution of Emigrants’ Deposits (Stock) and Remittances (flows), 2000–07 30000 25 25000 20 20000 15 15000 10 10000 5 5000 0 0 Dec-02 Dec-03 Dec-04 Dec-05 Dec-06 Oct-07 Stock of emigrant deposits (CVE billion) Remittances (annual flows) Growth rate of stock deposits Source: Central Bank of Cape Verde. 4 FISCAL CONSOLIDATION Fiscal policy is consistent with macroeconomic stability and debt sustainability. Over the past few years, Cape Verde has demonstrated prudent fiscal policy. As a result of improved tax collection and expenditure control, the fiscal deficit, including grants, averaged 3.8 percent of GDP during 2002–06 (4.4 excluding 2004),4 In 2007 the deficit was expected to stand at 4.6 percent of GDP, reflecting a continuing strong revenue performance due to economic growth, improvements in tax administration, and restraint on expenditures. These factors have reduced the government’s resorting to domestic credit. The 2008 Budget law suggests that the fiscal policy stance will be firm. Recurrent spending will decline as a share of GDP due to a reduction (as a share of GDP) of wages and salaries, goods and services, and subsidies, thus freeing resources for capital expenditures. Following the elimination of petroleum product and utility tariff subsidies, the budget does not, in principle, allow for oil-related subsidies. Furthermore, hiring and promotions have been frozen until the revision of the Career and Salary System Plan, expected in 2008. Table 0.4: Central Government Fiscal Operations, 2000–06 (CVE million) 2002 2003 2004 2005 20061 20072 Total revenues (CVE million) 23,491 22,445 26,334 28,178 30,826 35,301 Total expenditure (CVE million) 27,019 25,641 26,853 31,911 35,400 40,154 Interest rate payments 2,162 1,994 2,056 1,927 1,920 1,883 GDP at current prices (CVE million) 72,758 79,527 82,116 88,733 101,551 105,178 Overall balance, incl. grants (CVE million) -3,528 -3,196 -519 -3,733 -4,574 -4,853 Overall primary balance, incl. grants (CVE -1,366 -1,202 1,537 -1,806 -2,654 -2,970 million) Overall primary balance, incl. grants (% -1.88 -1.51 1.87 -2.04 -2.61 -2.82 GDP) Overall balance, incl. grants (%GDP) -4.85 -4.02 -0.63 -4.21 -4.50 -4.61 Source: Ministry of Finance and Public Administration, IMF, and staff estimates. Notes: 1 Preliminary accounts. 2 Approved budget. IMPROVEMENT IN DEBT MANAGEMENT Significant progress has been achieved with debt management. The total central government net debt and guarantees began a declining trend in 2005 and have been declining sharply since then. They fell from 87 percent in 2005 to 77 percent of GDP in 2006 owing to the rapid GDP growth and prudent fiscal stance. The domestic debt-to-GDP ratio was reduced from 33 percent in 2005 to 29 percent in 2006, and was expected to reach 23 percent by the end of 2007. The Policy-Support Instrument (PSI) program aimed to lower the domestic debt to close to 20 percent of GDP by 2009. However, net domestic debt is now projected to decline to below 20 percent by the end of 2008, one year ahead of the IMF program-Policy Support Instrument (PSI) schedule. An analysis of the sustainability of Cape Verde’s public debt carried out in December 2007 4 This is in part explained by the large amount of grants (9% of GDP), the highest in the analysis period. 5 concluded that, in spite of the likely gradual reduction in access to concessional loans, Cape Verde’s debt stock and flow indicators will remain below their policy-dependent thresholds throughout the projection period (annex 1). Table 0.5: Key Debt Policy Indicators (% of GDP) 2002 2003 2004 2005 2006 20071 Total nominal gov debt 87.3 85.3 89.0 87.1 77.3 66.4 External gov debt 58.1 56.2 54.0 53.8 48.0 43.5 Domestic gov debt (net of dep.) 29.2 29.2 35.0 33.3 29.3 22.9 External debt service (% of exp.) 15.6 10.6 11.3 8.6 5.7 5.0 Notes: 1 Projections. Source: Ministry of Finance and Public Administration, IMF, and staff estimates In July 2006, the IMF approved the request by the government for a three-year Policy Support Instrument. The PSI is designed to support the government’s economic objectives and policy framework for 2006–09. This program focuses on measures to reduce macroeconomic risks, provide a margin of safety against exogenous shocks, and address the prospects of a longer-term decline in highly concessional external support. The third review was completed in December 2007. RECOMMENDATIONS5 Monetary policy should continue being oriented toward further accumulation of foreign exchange reserves to consolidate the credibility of the peg. To that end, monetary authorities should continue managing liquidity actively through the appropriate issuance of short-term bills by the Central Bank. In addition, the Central Bank should continue closely monitoring interest rate differentials with the Euro area and USA to prevent declines in remittances. Fiscal policy should continue to be dominated by fiscal prudence, which will help create space to absorb potential shocks and preserve the low risk of debt distress. Authorities should continue building the budget on conservative estimates of projected revenues, which has helped prevent over-spending. Furthermore, the authorities should continue implementing measures to avoid the accumulation of arrears both at the levels of central government and municipalities. 5 All the recommendations (from all the chapters) are presented in Annex 1, Volume 1 6 Fiscal Performance TRENDS AND COMPOSITION OF REVENUE The share of total revenue to gross domestic product (GDP) for 2002–06 shows a stable trend, with some declines in 2003 and 2006.6 Total revenues as a percentage of GDP has decreased slightly from 32 percent in 2002 to approximately 30 percent in 2006 (table 2.2). It averaged 31 percent over the whole period. Despite the stable trend throughout the period, it declined approximately 4 percentage points in 2003 and 1.4 percentage points in 2006. Deterioration in revenue performance during these two years was due largely to a decrease in foreign aid and, in 2003 due also to a slowdown of fiscal revenue growth. Foreign aid represents, on average, 25 percent of total revenue, with minimums of 22 percent in 2003 and 19 percent in 2006. A slight increase in total revenue, expressed as a percentage of GDP, was anticipated for 2007 (from 30 percent in 2006 to almost 34 percent in 2007). Supported by two comprehensive reforms in indirect taxation introduced in early 2004 (tariff reforms and introduction of the VAT), tax revenue performance has been strong. Total fiscal revenues declined in 2003 (stood at 18.8 percentage of GDP) and increased steadily, averaging 21 percent after 2004. The steady revenue performance has been due largely to the introduction of VAT, which has substantially compensated the loss of the import tax and helped improve Cape Verde’s fiscal administration. Import tax revenues represented an average of 8.4 percent of GDP in 2002 and 2003 and decreased to 4.8 percent from 2004 through 2006. For 2007, they were expected to represent 5.2 percent of GDP. On the other hand, VAT represented, on average, 7.5 percent of GDP for 2004-06, whereas consumption taxes represented approximately 3 percent of GDP for 2002–03. Import tax revenues are expected to decline in the future due to revised custom taxes. In early 2004 the government implemented a radical fiscal reform that eliminated most of the taxes and contributions administered by Customs. The number of tariff brackets was reduced from 64 to 7, and the maximum tariff was leveled from 250 percent to at 50 percent. However, Cape Verde’s tariffs are still high, and in the context of WTO membership, tariffs were expected to be revised. On December 18, 2007, the WTO’s General Council approved a package of agreements that spells out the terms of Cape Verde’s accession. In this process, all remaining (tariff and nontariff) measures to protect local industry were abandoned. Furthermore, a new Customs Code was prepared and is expected to be approved by the National Assembly in 2008. This code will provide progressive liberalization of import rules and procedures. Collection of fiscal revenues improved and is expected to improve further as a result of additional reforms in the tax administration. The Enactment of Decree-Law No. 35/2003 brought about an improved recovery of fiscal arrears, since it endowed the General Directorate of Taxes (Direcção Geral de Contribuição e Impostos, or DGCI) with the power to affect compulsory collection through confiscation of assets and bank accounts. Additional 6 The data used in this analysis are classified as official information of the Ministry of Finance and Public Administration. The information corresponds to data concerning the definitive state accounts for 2002–05; temporary accounts for 2006; and the budget approved for 2007. 7 improvements are expected through the introduction of the Fiscal Identification Number (FIN), which came into being in 2006 and is expected to render better management of taxpayer data. A project to automate revenue management and taxpayer current account management is being implemented as an instrument to increase the control and efficiency of tax collection. Reforms in tax administration also include the end of the agreement with Banco Comercial do Atlântico (BCA) as the sole tax collector and revision of legal texts (new IUR on corporate and individual tax, and stamps). Until 2006, BCA was the only tax collector and retained 1 percent of the taxes collected. During the first semester of 2006, agreements were signed with other institutions whereby a flat rate of 200–300 ECV per collection was defined as the remuneration for its services. BCA signed a similar agreement in May 2007, with retroactive effects dated July 2006. This has provided taxpayers with more payment options and is expected to reduce revenue forgone. Direct taxes represented, on average, 6.8 percent of GDP from 2002 to 2006 but are below their potential owing to the extended and unfounded system of fiscal exemptions and large fiscal evasion. Preliminary data for 2006 suggest that revenues lost to fiscal exemptions correspond to approximately 12 percent of fiscal revenues. There also is the belief that fiscal evasion is massive. The extended system of exemptions and the wide-ranging evasion contributes to the narrow and skewed nature of the taxpayer base. Approximately 15 corporations account for 66 percent of total tax revenues; one of them accounts for one-third of the total (corporations as a whole represent 44 percent of direct tax revenues). Therefore, to reduce the dependence on large taxpayers as a source of revenue, rationalization of exemptions and fiscal incentives is very important. The law is being drafted (there is already an inventory of the existing laws) and is expected to be submitted to Parliament in 2008. Recurrent revenues have shown an ascending trend, whereas capital revenues (consisting mainly of foreign aid) have shown annual fluctuations, highlighting the ever-present uncertainty attributable to this source of income. Foreign aid has undergone ups and downs from as low as 6 percent of GDP in 2003 and 2006, to 9 percent in 2002 and 2004. With Cape Verde’s graduation to the status of middle-income country, its eligibility for foreign aid and concessional funds is expected to decrease over time, making it even more important for the country to create fiscal space to respond to future pressures. Overall, total revenues projections have been close to revenues outturns. However, this closeness hides significant individual deviations. Nonfiscal revenues and foreign aid present great overestimation. (table 2.1). For example, in 2003 and 2005, the execution rate of foreign aid was approximately only 63 percent. This is not surprising. In Cape Verde, as in many other countries, aid projections, particularly for project financing, tend to be overly optimistic. On the other hand, revenues from VAT and international transactions have been often underestimated. The difficulties in making projections emerge from the lack of an adequate model. In this regard, staff from the Ministry of Finance and Public Administration are being trained to build capacity in financial programming. 8 Table 0.1: Revenues 2002–07 (CVE million) 2002 2003 2004 2005 20061 20072 Total recurrent revenues 16,951.48 16,971.69 18,741.48 21,419.73 24,626.45 26,731.19 Tax revenues 14,948.92 14,935.53 16,636.67 18,539.44 22,609.73 24,178.20 Direct taxes 5,505.61 5,146.60 5,394.58 5,815.33 6,952.41 7,496.62 Income tax 5,505.61 5,146.60 5,394.58 5,815.33 6,952.41 7,496.62 Indirect taxes 9,443.31 9,788.93 11,242.09 12,724.11 15,657.32 16,681.58 Consumption/VAT tax 2,420.60 2,160.25 5,591.80 6,551.89 8,438.91 8,610.09 Tax on international transactions 6,126.63 6,755.33 3,976.80 4,231.32 4,888.71 5,458.47 Nontax revenues 2,002.56 2,036.16 2,104.82 2,880.29 2,016.72 2,553.00 Net lending 220.13 537.58 231.43 146.98 301.07 200,00 Other national sources - 39.98 - - - - External grants 6,319.16 4,896.02 7,360.75 6,611.55 5,898.73 8,369.75 Total revenues 23,490.77 22,445.27 26,333.66 28,178.25 30,826.25 35,300.94 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. Table 0.2: Revenues, 2002–07 (% of GDP) 2002 2003 2004 2005 20061 20072 Total recurrent revenues 23.3 21.3 22.8 24.1 24.3 25.4 Tax revenues 20.5 18.8 20.3 20.9 22.3 23.0 Direct taxes 7.6 6.5 6.6 6.6 6.8 7.1 Income tax 7.6 6.5 6.6 6.6 6.8 7.1 Indirect taxes 13.0 12.3 13.7 14.3 15.4 15.9 Consumption/VAT tax 3.3 2.7 6.8 7.4 8.3 8.2 Tax on international transactions 8.4 8.5 4.8 4.8 4.8 5.2 Nontax revenues 2.8 2.6 2.6 3.2 2.0 2.4 Net lending 0.3 0.7 0.3 0.2 0.3 0.2 Other national sources 0.0 0.0 0.0 0.0 0.0 0.0 External grants 8.7 6.2 9.0 7.5 5.8 8.0 Total revenues 32.3 28.2 32.1 31.8 30.4 33.6 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. Table 0.3: Revenues, 2002–07 (% of total revenues) 2002 2003 2004 2005 20061 20072 Total recurrent revenues 72.16 75.61 71.17 76.02 79.89 75.72 Tax revenues 63.64 66.54 63.18 65.79 73.35 68.49 Direct taxes 23.44 22.93 20.49 20.64 22.55 21.24 Income tax 23.44 22.93 20.49 20.64 22.55 21.24 Indirect taxes 40.20 43.61 42.69 45.16 50.79 47.26 Consumption/VAT tax 10.30 9.62 21.23 23.25 27.38 24.39 Tax on international transactions 26.08 30.10 15.10 15.02 15.86 15.46 Nontax revenues 8.52 9.07 7.99 10.22 6.54 7.23 Net lending 0.94 2.40 0.88 0.52 0.98 0.57 Other national sources - 0.18 - - - - External grants 26.90 21.81 27.95 23.45 19.14 23.71 Total revenues 100.00 100.00 100.00 100.00 100.00 100.00 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. 9 Table 0.4: Revenues, 2002–06 (executed revenues as % of budgeted revenues) 2002 2003 2004 2005 20061 Total recurrent revenues 97.3 91.6 99.9 107.8 100.9 Income tax 109.1 87.5 90.1 94.2 100.0 Consumption/VAT tax 107.7 68.7 129.5 122.3 114.6 Tax on international transactions 104.6 137.0 118.6 114.0 108.6 Nontax revenues 59.4 62.1 75.0 97.5 59.8 External grants 101.5 63.0 103.9 64.3 83.8 Total revenues 96.8 82.0 99.0 96.0 88.2 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. Recommendations • Finalize the draft law on tax exemptions. Reduction of tax exemptions is important to increase VAT effectiveness, and to enlarge the narrow tax base, thus reducing the current dependency on a few companies. It also is recommended to unify all exemptions in the single law covering all fiscal incentives. • Develop a legal and strategic action framework to ensure operability of tax inspections and its link to other competent institutions involved in criminal research. • Develop and implement IT supporting projects, and programs for tax management inspection. • Reinforce inspections related to high risk tax-payers and critical areas. TRENDS AND COMPOSITION OF PUBLIC EXPENDITURE The economic classification of expenditure shows a stable pattern of expenditure. Throughout the last 5 years, on average, 64 percent of resources financed the recurrent expenditures, while the remaining resources (on average, 36 percent) paid for the public investment program (PIP) (table 2.6). This broad distribution is affected by annual variations that are explained by the unpredictable execution of the current PIP and, to a lesser degree, by the privatization costs of some public enterprises that spilled over from the late 1990s to the first half of the current decade. In 2002, 2003, and 2004, the cost of privatizations reached 0.3 percent, 0.8 percent, and 0.3 percent, respectively, of total public expenditures. The expenditures pattern for 2002–06 highlights the continuous dominance of nondiscretionary expenditure. The three categories (salaries and benefits, transfers and subsidies, and payment of interest rates) represented on average 83 percent of total recurrent expenditures and 53 percent of total expenditures. Continued increases of nondiscretionary expenditures will restrict the fiscal space available to the government to react to external shocks. In the medium-long run, the implementation of some reforms will decrease their weight in the budget. However, it is important to emphasize that, in the short run in a constrained budget, if the government does not wish to increase the overall deficit, the weight of any potential budgetary 10 cuts resulting from a negative shock would fall onto discretionary expenditures such as goods and services or the investment budget. The share of salaries and benefits to GDP increased from 9.6 percent in 2002 to 12.7 percent in 2005, an increase of 30 percent. In absolute terms, this share increased by 84 percent from 2002 to 2007, or approximately 10 percent a year. This increase reflects the impact of various salary increases instituted during the first half of the decade to improve the standard of living of civil servants. It also reflects the hiring of teachers and health care professionals, and reinforcements for the police force for public security. For 2007, the expectation was that salaries and benefits would represent approximately 12.3 percent of GDP. To contain the escalation in wages and benefits, the government has decided to freeze promotions and recruitments7 until the Plan Positions, Careers and Salaries (PCCS) is revised. Furthermore, the validation of the civil service database, which began in August 2006, has been completed and integrated with the payroll. The validation brought about some savings by eliminating some “ghost� civil servants from the payroll. Given the budget constraints, the government should pursue measures to restrict the growth of the wage bill. Furthermore, in the context of the revision of PCCS stricter conditions should be defined for promotion and for eligibility to the “quadro privativo�8 status. The recently approved decree (Decreto-Lei 21/2006) on pensions contributes to an increase in salaries and benefits’ share of the budget (already reflected in the 2007 budget). With the new law, the responsibilities of the state as an employer increased significantly. While continuing to pay pensions to all civil servants who joined the public sector before December 31, 2005, the government also must transfer contributions to the Instituto Nacional da Previdência Social (INPS). The monthly transfers to INPS include 23 percent of the wage bill of the new civil servants, plus 8 percent of the current active civil servants and pensioners (for the health program). Even though, in consolidated terms, there is a gain in moving civil servants to the INPS system, because it provides fewer benefits, the integration worsens the unsustainability of the INPS (a contingent liability).9 The implementation of parametric reforms––which should consider either a raise in the contribution rates or a substantial cut in benefits10––has been under consideration. Estimates of the impact of a number of reforms were undertaken in the 2006 PER. Within the group labeled “social transfers and subsidies,� the government was successful in controlling the cost of scholarships. This category covers transfers to other government levels (autonomous institutes and municipalities) and subsidies for petroleum products, as well as transfers for embassies and for students abroad. As in 2002, the authorities decided to reduce funds for financing public scholarships and went on to encourage participation in costs by students and families eligible for these scholarships. It is estimated that the expenditure on scholarships was reduced by almost 44 percent between 2002 and 2006 (from 555 million CVE in 2002 to 314 million in 2006). 7 Hiring and promotions have been frozen until the revision of the CSSP (with the exception of the Health, Education, Social Protection, and Security sectors). 8 Quadros Privativos are technicians with higher salaries and special benefits. 9 Preliminary estimates indicate that the INPS system will be in deficit around 2040, as the system matures and the demographic situation improves. 10 Cape Verde’s social security system is very generous by international standards (2006 PER). 11 The amount of oil products subsidies registered in one year’s accounts often relates to previous years. Therefore, the account readings are very misleading. Because of Cape Verde’s cash basis approach to the budget,11 one year’s subsidies usually are carried forward to the next years’ budget. This practice seriously understates the current fiscal position. The budget does not provide adequate support for planning and monitoring current-year spending priorities against current-year resources. Furthermore, oil subsidies often are not recorded as subsidies, but rather as “extraordinary expenditures.� Other times, they are not recorded at all in the budget. As the payment of subsidies relates to previous years’ budget, often times the subsidies are recorded as “extraordinary expenditures.� Other times, the subsidies payments are not registered at all in the budget, because payments due to the oil companies are converted in protocols, a common practice in Cape Verde.12 As a result, actual payments of oil subsidies are much higher than presented in the budget.13 In April 2006, subsidies on oil products were eliminated to protect the budget from open- ended commitments. Nevertheless, given the lack of adjustment of prices by the Economic Regulatory Agency (ARE), subsidies continue to exist. As lags in the adjustments of the prices have been consistent, government covers the difference (between the import price and the price established by ARE) to the oil companies.14 For instance, referring to subsidies covering May- December 2007, the amount due to Shell, (after subsidies had been eliminated) was converted into a protocolo: CVE 123 million. Furthermore, because the elimination of the subsidies was not reflected in an adequate adjustment of electricity and water tariffs, tariff deficit was accumulated toward ELECTRA (the water and electricity company): in May 2006-February 2007, CVE 550 million. Table 0.5: Expenditure, 2002–07 (CVE million) 2002 2003 2004 2005 20061 20072 Total recurrent expenditure 17,248 17,245 17,834 19,627 22,087 23,602 Primary Recurrent expenses 14,861 15,251 15,779 17,699 20,167 21,720 Goods and Services 454 1,426 1,287 1,606 1,307 2,127 Salaries and pensions 7,015 9,886 10,219 11,230 11,553 12,973 Transfers and subsidies 4,999 2,502 2,631 2,718 5,136 4,094 Subsidies on oil prices 966 137 - 450 - - Other recurrent expenditure 2,394 796 1,416 2,145 2,171 2,526 Interest Rate Payments 2,162 1,994 2,056 1,927 1,920 1,883 Interest rate payments ext. debt 725 517 550 543 522 530 Interest rate payments dom. debt 1,437 1,478 1,506 1,384 1,398 1,352 Extraordinary expenditure 225 640 225 - - - Restructuring costs 225 640 225 - - - Capital expenditure 0,052 0,229 0,173 0,231 0,102 0,211 Public investment program 9,719 8,167 8,845 12,053 13,211 16,340 11 As a result of the practice of accounting for expenditures only when payments are made (payment of the subsidies usually is made against taxes). The accrual approach will be in place once the new chart of accounts is implemented (expected for 2008). 12 Protocols state an agreement whereby the government recognizes the debt and agrees to pay at a certain moment with no interest rate. This has been a common recourse in Cape Verde that bypasses the budget. 13 The 2006 PER estimates that in 2005 the amount of subsidy to ELECTRA (to cover the difference between the regulated price and the ELECTRA price of CVE 37.9/liter) reached CVE 382 million. The subsidy to the oil companies in 2005 due to the lag in the adjustment of prices was CVE 1,347 million. 14 This is discussed in more detail in chapter 5. 12 Total expenditure 27,019 25,641 26,853 31,911 35,400 40,154 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts; 2 Approved budget 13 Table 0.6: Expenditure 2002–07 (% of GDP) 2002 2003 2004 2005 20061 20072 Total recurrent expenditure 23.7 21.7 21.7 22.1 21.7 22.4 Primary recurrent expenses 20.4 19.2 19.2 19.9 19.9 20.7 Goods and services 0.6 1.8 1.6 1.8 1.3 2.0 Salaries and pensions 9.6 12.4 12.4 12.7 11.4 12.3 Transfers and subsidies 6.9 3.1 3.2 3.1 5.1 3.9 Subsidies on oil prices 1.3 0.2 0.0 0.5 0.0 0.0 Other recurrent expenditure 3.3 1.0 1.0 2.4 2.1 2.4 Interest rate payments 3.0 2.5 2.5 2.2 1.9 1.8 Interest rate payments ext. debt 1.0 0.6 0.6 0.6 0.5 0.5 Interest rate payments dom. debt 2.0 1.9 1.8 1.6 1.4 1.3 Extraordinary expenditure 0.3 0.8 0.3 - - - Restructuring costs 0.3 0.8 0.3 - - - Capital expenditure 0.1 0.3 0.2 0.3 0.1 0.2 Public investment program 13.4 10.3 10.8 13.6 13.0 15.5 Total expenditure 37.1 32.2 32.7 36.0 34.9 38.2 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts; 2 Approved budget. Interest rate payments have been decreasing steadily during the analysis period. Table 2.7 shows that interest rate payments decreased from 8 percent of total expenditure in 2002 to 5.4 percent in 2006. During 2003–06, these payments reached an average of 6.6 percent of total expenditures (2 percent of GDP). It is expected that, as a result of the decline in borrowing (as percentage of GDP), interest payments will continue lessening. Table 0.7: Expenditure 2002–07 (% of total expenditure) 2002 2003 2004 2005 20061 20072 Total recurrent expenditure 63.8 67.3 66.4 61.5 62.4 58.8 Primary recurrent expenses 55.0 59.5 58.8 55.5 57.0 54.1 Goods and services 1.7 5.6 4.8 5.0 3.7 5.3 Salaries and pensions 26.0 38.6 38.1 35.2 32.6 32.3 Transfers and subsidies 18.5 9.8 9.8 8.5 14.5 10.2 Subsidies on oil prices 3.6 0.5 0.0 1.4 0.0 0.0 Other recurrent expenditure 8.9 3.1 5.3 6.7 6.1 6.3 Interest rate 8.0 7.8 7.7 6.0 5.4 4.7 payments Interest rate payments ext. debt 2.7 2.0 2.0 1.7 1.5 1.3 Interest rate payments dom. debt 5.3 5.8 5.6 4.3 3.9 3.4 Extraordinary expenditure 0.8 2.5 0.8 - - - Restructuring costs 0.8 2.5 0.8 - - - Capital expenditure 0.2 0.9 0.6 0.7 0.3 0.5 Public investment program 36.0 31.9 32.9 37.8 37.3 40.7 Total expenditure 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. There are pending questions of risk assumption by the government concerning state-owned enterprises (SOEs), autonomous institutions, funds, and simple services liabilities. Autonomous institutions have their own limited resources but also are recipients of large budget 14 transfers (table 2.9). Government data indicate annual transfers ranging from 2.4 billion CVE– 3.3 billion CVE. In the long term, these large transfers appear to question the autonomous nature of these institutions, as well as the soundness of such an arrangement. An in-depth analysis should be conducted, provided solid data are available. Regarding the SOEs, a study of the liabilities of TACV (Cabo Verde Air Lines), ENAPOR (port authority), EMPROFAC (distributor of pharmaceutical products), INTERBASE (frozen fish company), and Empresa Nacional de Aeroportos e Segurança Aérea (ASA) (airports) could shed light on implicit risks for the central budget. The Ministry of Finance and Public Administration has set up a unit to follow up on state participations Table 0.8: Transfer from Central Government for Autonomous Institutions, 2002–07 (CVE million) 2002 2003 2004 2005 2006–est.1 2007–bud.2 Total 2,897 3,391 3,498 4,257 1,573 3,739 Of which: Own resources of FSA 416 531 373 933 205 529 Of which: Transfer from central 2,481 2,860 3,125 3,324 1,367 3,210 government Expenditure 2,739 3,220 3,338 3,681 2,818 3,739 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts/ 2 Approved budget. Recommendations Better control of nondiscretionary expenditure and open-ended commitments is needed to guarantee adequate fiscal space and ability to cope with potential shocks. Furthermore, to help plan and monitor current-year expenditures against current-year resources, it is important to record all expenditures and revenues. The following measures should be taken into account: • Continue restraining the wage bill. • Take greater control of transfer effectiveness (for example, payment of scholarships). • Reduce open-ended commitments. • Record all expenditures and gross revenues in the appropriate category in the budget, in the year they are relative to. • Apply the automatic mechanisms for oil petroleum products and utility tariffs, in line with the legislation. EFFICIENCY OF BUDGET IMPLEMENTATION AND ROLE OF PUBLIC INVESTMENT PROGRAM Patterns of Budgetary Execution: Programmed, Revised, and Executed In Cape Verde, as in many developing countries, the execution rate of recurrent expenditures is much higher than that of capital expenditures. Recurrent expenditure execution rates typically are more than 90 percent of the originally approved budget. However, the best implementation rate for capital expenditure during 2002–06 was 81 percent in 2006. This difference stems from the already discussed characteristics of the budgetary execution process: (i) nondiscretionary expenditures determine the implementation of recurrent 15 expenditures; and (ii) the fluctuating flow of foreign aid determines the implementation of the investment program. Table 0.9: Actual Expenditure as % of Budgeted Expenditure,1 2002–06 2002 2003 2004 2005 20062 Total recurrent expenditure 107.9 93.7 97.1 97.7 91.4 Goods and services 69.4 85.1 76.3 91.9 69.8 Salaries and pensions 96.0 92.9 91.6 98.1 90.5 Transfers and subsidies 92.6 90.0 93.4 79.4 78.2 Other recurrent expenditure 517.8 151.0 158.7 113.5 189.9 Interest rate payments 100.0 92.1 158.7 121.3 106.5 Public investment program 63.8 59.0 63.6 72.9 81.1 Source: Ministry of Finance and Public Administration. Notes: 1 During the period of analysis, there were two budgetary amendments (in 2002 and 2005). 2 Preliminary accounts. In the category of recurrent expenditures, the implementation rate of nondiscretionary expenditures (salaries and benefits, transfers and subsidies, and interest payments) is typically high. For example, in the subcategory salaries and benefits expenditure, the outlay is usually over 90 percent of that originally programmed. In the subcategory interest payments, this rate exceeded 100 percent in 2004, 2005, and 2006. The subcategory transfers and subsidies had an execution rate above 90 percent in 2002–06, and in 2005 and 2006. It dropped to approximately 80 percent (as noted above, the payment of some subsidies is not included in the budget). This high implementation rate of nondiscretionary expenditure often is obtained at the expense of discretionary expenditure such as goods and services. In the category of goods and services, the expenditure execution rate versus planned budget varied from 69 percent in 2002 to a peak of 92 percent in 2005, and then down to 70 percent in 2006. Programming and implementation of the Public Investment Program (PIP) PIP programming has been over optimistic (table 2.10). The highest execution rate during the period of analysis was 81 percent in 2006, and it had fallen as low as 59 percent (2003). The average execution rate during the period was 69 percent. This variable performance is due primarily to the fact that PIP implementation depends to a large extent on foreign aid materialization. Table 2.11 shows that, during the 5-year analysis, more than 80 percent of the investment was funded by external resources. In 2004 this rate climbed to 95.2 percent. Whenever the disbursement of external funds is delayed or does not materialize, implementation may be suspended or even cancelled. Some projects, mostly the ones with important social dimensions, have been implemented, even when foreign aid does not materialize or is insufficient. A wide range of secondary investments and small maintenance tasks have been implemented through domestic funding. Similarly, when a project to be funded with external resources has priority but lacks significant endowment, the Treasury mobilizes funds to bridge the financing gap over a short period, that is, until foreign aid arrives. In this manner, it is possible to continue project implementation. However, in the case of very large investments such as infrastructure projects, implementation is delayed or remains incomplete. 16 Table 0.10: Public Investment Program (PIP) - Execution and funding sources, 2002–07 2002 2003 2004 2005 20061 20072 PIP (CVE million) 9,719 8,167 8,845 12,053 13,211 16,340 of which (in % of total PIP) Internal funding 13.1 14.1 4.8 15.7 16.8 17.5 External funding 86.9 85.9 95.2 84.3 83.2 82.5 PIP (% nominal of GDP) of which (in % of total PIP) Internal funding 1.8 1.4 0.5 2.1 2.2 2.7 External funding 11.6 8.8 10.3 11.4 10.8 12.8 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts. 2 Approved budget. Variability in foreign aid funds is rooted in two main factors: (i) the nature of aid (project versus budgetary aid); and (ii) delays related to donor or recipient political or policy implementation processes. In the first place, as is typical in many developing countries, most foreign aid to Cape Verde is channeled toward project financing. This type of aid is prone to delays whenever the bureaucracies of the donor and the recipient differ with respect to public procurement, financial management, or other requirements related to domestic regulation. Furthermore, this type of aid generates project coordination units that often operate outside national budgetary processes, thus making the budgetary unification process difficult. Second, foreign donors frequently must abide by their own national objectives or policies, which are not always aligned with the objectives of the recipient country. It is expected, however, that aid predictability will improve, as budget support has gained prominence during the last years, thanks to the creation of the Budget Support Group. Created in 2005, the BSG included the World Bank, the European Union, and the Dutch cooperation.15 The coordinated approach used by the BSG to align and harmonize their support for poverty reduction around the GPRSP-1 has played a catalytic role in bringing in new partners. As a result, BSG was expanded in 2007 to a total of 6 partners when the African Development Bank (AfDB) and Austria, and Spain cooperations joined the group. The BSG expanded further when the Portuguese cooperation announced at the GAT meeting (Grupo de Apoio à Transição) in December 2007 that it will join too. During the last 3 years, approximately 25 percent of foreign aid was in the form of budget support. Deficiencies in monitoring project implementation for projects implemented through direct aid makes difficult in-depth and efficient analysis of investment expenditure execution. The deficiencies in monitoring project implementation financed with external sources are due, to some extent, to lack of capacity. Capacity building in monitoring is a priority. Furthermore, it also is important that development partners collaborate by making data available at the appropriate times. 15 The creation of the BSG had an important role in reducing government transaction costs (missions are conducted jointly) as well as in harmonizing donor policies (donors have a joint policy matrix). 17 Recommendations Given that recourse to large amounts of credit is not sustainable, it is recommended that the Cape Verde authorities implement a set of actions to guarantee a more reliable inflow of foreign aid: • Prepare a pluri-annual framework with the donors as part of the MTEF process. • Encourage partners to make directly available updated financial data of funded projects to establish transparency in the presentation of quarterly and annual state accounts. ADMINISTRATIVE EXPENDITURE PATTERNS The Ministry of Finance and Public Administration is clearly dominant in expenditure, representing on average 43 percent of the total. A significant part of these expenditures are related to restructuring costs from privatizations, payment of oil subsidies, payment of arrears, and interest rate payments. From 2002–06, the Ministries of Education and Health accounted for, on average, 24 percent and 8 percent, respectively, of total recurrent expenditures. These rates support the assertion that growth and poverty reduction had been national priorities even before the adoption of GPRSP-1-I in 2004. However, most of these recurrent expenditures consist of salaries and Public Administration benefits. An evaluation covering several years showed that more than 60 percent of the total annual budget for each sector is devoted to salaries and other benefits. The expenditure level of the Ministry of Internal Administration, estimated at 6.4 percent of total expenditure for 2007, has increased over the last few years. The increase results from new recruitments for security. Concerns with security related to narcotics trafficking have risen in the past few years. To prevent itself from being transformed into a golden highway for drug routes and other types of traffic, Cape Verde has hired more security staff to strengthen the control of the maritime and air frontiers.16 16 One-quarter of all cocaine consumed in Western Europe is trafficked through West Africa. 18 Table 0.11: Organic Classification of Executed Expenditures, 2002–07 (% of total of recurrent expenditure) 2002 2003 2004 2005 2006 20071 Presidency of the Republic 0.49 0.52 0.56 0.57 0.46 0.45 National Assembly 2.31 2.69 2.58 2.66 2.55 2.47 Head of Government 1.86 2.31 1.92 1.78 1.64 2.00 Supreme Court 0.11 0.15 0.13 0.11 0.11 0.18 General Attorney 0.08 0.06 0.09 0.10 0.11 0.12 Accounts Court 0.19 0.17 0.21 0.20 0.21 0.24 Ministry of Foreign Affairs 4.08 3.93 3.60 3.52 3.15 3.19 Ministry of Finance and Public Administration 48.24 39.40 40.84 40.64 45.10 42.20 Ministry of Justice 2.35 2.61 2.57 2.54 2.35 2.79 Ministry of Internal Administration 4.20 4.63 4.92 5.39 5.33 6.47 Ministry of Defense 2.87 3.24 3.19 3.09 2.75 2.69 Ministry of Agriculture and Fisheries 2.20 2.83 2.52 2.34 1.01 2.13 Ministry of Education and Higher Learning 20.57 25.59 25.12 24.85 24.46 23.72 Ministry of Health 7.51 8.09 7.62 8.48 7.46 7.71 Ministry of Infrastructure and Transports 0.98 1.61 1.45 1.46 1.45 1.34 Ministry of Economy, Growth and 0.40 0.43 0.65 0.60 0.53 0.78 Competitiveness Ministry of Culture 0.38 0.40 0.81 0.52 0.50 0.58 Ministry of State Reform and Public 0.24 0.27 0.21 0.16 0.14 0.13 Administration Ministry of Labor and Solidarity 0.93 1.05 1.02 0.99 0.68 0.80 Total Expenditures 100.0 100.0 100.0 100.0 100.0 100.0 Source: Ministry of Finance and Public Administration. Note: 1 Approved budget. Table 0.12: Organic Classification of Executed Expenditures, 2002–07 (% of initial budgetary allocations) 2002 2003 2004 2005 20061 Presidency of the Republic 99.5 97.9 102.1 112.0 101.1 National Assembly 94.0 94.0 86.6 97.8 100.0 Head of Government 79.9 94.5 72.6 124.1 81.9 Supreme Court 100.0 174.0 131.7 121.0 100.0 General Attorney 100.0 69.6 89.9 122.7 100.0 Accounts Court 106.2 85.8 100.4 9.6 100.0 Ministry of Foreign Affairs 92.1 87.7 84.8 91.3 91.9 Ministry of Finance and Public Administration 133.0 95.9 111.2 113.3 86.9 Ministry of Justice 85.0 88.8 88.3 91.0 93.4 Ministry of Internal Administration 96.7 93.1 101.7 106.6 93.8 Ministry of Defense 93.7 100.6 98.0 95.9 97.7 Ministry of Agriculture and Fisheries 83.8 78.0 79.3 85.6 45.3 Ministry of Education and Higher Learning 98.2 93.4 93.7 102.9 101.0 Ministry of Health 89.2 100.7 82.3 100.8 95.1 Ministry of Infrastructure and Transports 87.2 80.5 66.1 76.7 79.6 Ministry of Economy, Growth and Competitiveness 28.7 45.9 78.8 85.2 86.9 Ministry of Culture 92.5 77.4 82.5 54.7 89.2 Ministry of State Reform and Public Administration 85.9 54.6 71.0 74.4 79.4 Ministry of Labor and Solidarity 145.9 85.7 82.0 84.6 82.4 PIP 68.5 59.0 63.6 74.4 81.1 Source: Ministry of Finance and Public Administration. Note: 1 Preliminary accounts. 19 Table 0.13: Organic Classification of Executed Expenditures, 2002–07 (% of GDP) 2002 2003 2004 2005 20061 20072 Presidency of the Republic 0.12 0.12 0.12 0.13 0.10 0.10 National Assembly 0.55 0.59 0.57 0.59 0.56 0.56 Head of Government 0.44 0.51 0.42 0.40 0.36 0.45 Supreme Court 0.03 0.03 0.03 0.03 0.03 0.04 General Attorney 0.02 0.01 0.02 0.02 0.03 0.03 Accounts Court 0.05 0.04 0.05 0.04 0.05 0.06 Ministry of Foreign Affairs 0.97 0.86 0.79 0.79 0.69 0.72 Ministry of Finance and Public 11.47 8.66 8.95 9.09 9.85 9.56 Administration Ministry of Justice 0.56 0.57 0.56 0.57 0.51 0.63 Ministry of Internal Administration 1.00 1.02 1.08 1.21 1.16 1.47 Ministry of Defense 0.68 0.71 0.70 0.69 0.60 0.61 Ministry of Agriculture and Fisheries 0.52 0.62 0.55 0.52 0.22 0.48 Ministry of Education and Higher Learning 4.89 5.62 5.51 5.56 5.34 5.37 Ministry of Health 1.79 1.78 1.67 1.90 1.63 1.75 Ministry of Infrastructure and Transports 0.23 0.35 0.32 0.33 0.32 0.30 Ministry of Economy, Growth and 0.09 0.10 0.14 0.13 0.12 0.18 Competitiveness Ministry of Culture 0.09 0.09 0.18 0.12 0.11 0.13 Ministry of State Reform and Public 0.06 0.06 0.05 0.04 0.03 0.03 Administration Ministry of Labor and Solidarity 0.22 0.23 0.22 0.22 0.15 0.18 Total Recurrent Expenditure 23.78 21.97 21.93 22.38 21.85 22.64 PIP 13.36 10.27 10.77 13.58 13.01 15.54 Total Expenditures 37.14 32.24 32.70 35.96 34.86 38.18 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts; 2 Approved budget. Table 0.14: Organic classification of executed expenditures, 2002–07 (CVE million) 2002 2003 2004 2005 20061 20072 Presidency of the Republic 85 92 101 113 102 108 National Assembly 400 471 465 528 565 587 Head of Government 322 404 347 354 364 477 Supreme Court 19 27 23 22 26 44 General Attorney 13 11 15 21 25 27 Accounts Court 33 30 38 40 46 58 Ministry of Foreign Affairs 706 686 648 699 698 760 Ministry of Finance and Public Administration 8,346 6,885 7,353 8,069 10,006 10,050 Ministry of Justice 406 456 462 503 522 665 Ministry of Internal Administration 726 808 885 1,071 1,182 1,541 Ministry of Defense 496 566 575 614 611 640 Ministry of Agriculture and Fisheries 381 495 454 465 224 508 Ministry of Education and Higher Learning 3,559 4,472 4,524 4,935 5,426 5,649 Ministry of Health 1,299 1,414 1,373 1,684 1,655 1,837 Ministry of Infrastructure and Transports 170 282 260 290 322 320 Ministry of Economy, Growth and 69 76 118 118 118 185 Competitiveness Ministry of Culture 66 70 146 103 112 137 Ministry of State Reform and Public 42 47 38 32 31 31 Administration Ministry of Labor and Solidarity 161 183 184 196 152 191 Total Recurrent Expenditure 17,300 17,474 18,007 19,857 22,188 23,814 PIP 9,719 8,167 8,845 12,053 13,211 16,340 Total Expenditures 27,019 25,641 26,853 31,910 35,400 40,154 Source: Ministry of Finance and Public Administration. Notes: 1 Preliminary accounts; 2 Approved budget. 20 RELATIONSHIP AMONG GPRSP-1, MTEF, AND THE BUDGET To assess the alignment of the annual budget with the GPRSP-1,17 this review examines the relationship between the 5 stated goals of the GPRSP-1 and the 2005 and 2006 PIP (programming and execution). In this exercise, only investment data are analyzed for two reasons: the link between GPRSP-1 and the budget is mostly established through the PIP, and the way the budget is prepared did not allow relating changes in current expenditures to the GPRSP- 1. The government did not have a comparable (one-to-one) programming framework for both the GPRSP-1 and PIP action plans, to analyze PIP expenditure in a more detailed manner in light of GPRSP-1 targets. Data for 2005 and 2006 concerning the projects and programs in the PIP implementation table were grouped under 5 categories associated with the 5 GPRSP-1 pillars. Total investments for each of the 5 pillars of the GPRSP-1 document are included in table 2.13. Overall, the financial programming associated with priority actions of the GPRSP-1 was very ambitious. For example, in a scenario of average annual execution of PIP of CVE 8.9 billion during 2002–04 (8.8 billion CVE in 2004), the GPRSP-1 program suggested the need for an investment budget of 14.2 billion CVE for 2005 (an increase of 61 percent with respect to 2004). This amount also is above the amount in the MTEF (CVE 13.5 billion).18 Furthermore, the budget allocation went beyond both MTEF and PRSP. The 2005 budget allocated CVE 16.5 billion to PIP (16 percent above the GPRSP-1 and 21 percent above the MTEF). As a result, the execution rate of PIP was only 73 percent, with the obvious implications of not completing, delaying, or cancelling planned projects. Annual budget allocations (and executions) per pillar were not consistent with the GPRSP- 1 financial programming, indicating that the GPRSP-1 did not guide budget preparation. The GPRSP-1 financial programming suggested that infrastructure was a clear priority by allocating more than 60 percent of the resources to Pillar IV (Infrastructure).19 In both 2005 and 2006, infrastructure had the largest share (approximately 40 percent), however, not as high as intended under the GSPRS-1. On the other hand, GPRSP-1 financial programming indicated that no more than 1 percent should be allocated to Pillar I. Budget allocations and executions show otherwise. Pillar I received approximately 15 percent in 2005 and 2006. Moreover, allocations for Pillar III were approximately twice the allocations indicated in the GPRSP-1 in 2005, and three times larger in 2006. 17 The GPRSP-1 was the main guiding document for Cape Verde’s economic and social development for the period 2004-07. Its five pillars coincided with both the Great Plan Options (Grandes Opções do Plano, or GOP, 2002–15) and with the National Development Plan (Plano Nacional de desenvolvimento, or PND, 2002–05). The GPRSP-1 was meant to consolidate the structural planning documents (GOP and PND) and to link them to the annual budgets through PIP. 18 Investment Budget MTEF: 2005, CVE 13.5 billion; 2006, 15 CVE billion; and 2007, CVE 15.5 billion. 19 As previously indicated, the pillars are (i) to promote good governance to reinforce efficacy and guarantee equality; (ii) to promote competitiveness aimed at sustaining economic growth and create employment; (iii) to develop human capital; (iv) to improve basic infrastructure to promote land use planning practices, regional development, and environmental protection; and (v) to improve efficacy and sustainability of social protection systems. 21 Table 0.15: GPRSP-1 and the Public Investment Program, 2005–06 (CVE million) Pillar 2005 2006 PRSP % Aloc % Exec. % PRSP % Aloc. % Exec. % I 150 1.1 2,480 15.0 1,658 13.8 145 1.1 2,620 16.1 1,369 10.4 II 1,536 10.8 2,330 14.1 1,288 10.7 1,554 10.9 1,136 7.0 1,078 8.2 III 1,996 14.1 4,598 27.8 3,952 32.8 1,419 10.0 5,706 35.0 3,611 27.3 IV 9,573 67.4 6,358 38.5 4,640 38.5 8,742 61.6 6,060 37.2 6,497 49.2 V 943 6.6 744 4.5 516 4.3 963 6.8 769 4.7 656 5.0 Total 14,198 100.0 16,511 100.0 12,053 100.0 12,823 90.4 16,290 100.0 13,211 100.0 Source: Ministry of Finance and Public Administration and IMF. MTEF (2005-2007) and GPRSP-1 were not aligned. The authorities prepared the first global MTEF in 2004, which covered 2005–07, the same period as the GPRSP-1. However, these two were prepared on different tracks with very little coordination. While GPRSP-1 was prepared using pillars and objectives, the MTEF was prepared according to a classification that was a mix between administrative functions and objectives (for recurrent and investment budgets). In the case of the priority sectors, education and health (two of the few sectors in which comparison is possible), it should be expected that GPRSP-1 benchmarks should match MTEF projections. However, deviations are huge. MTEF projections for education and health deviate 40 percent and 100 percent, respectively, from the GPRSP-1 goals. It can only be assumed that that MTEF did not guide the GPRSP-1 preparation, and thus did not support the government’s fiscal strategy, which intended that allocations to sectors and ministries reflect political priorities and that projects and programs are adequately and reliably funded through the budget. Table 0.16: GPRSP-1 Benchmarks for Priority Sectors and MTEF Projections, 2004–07 (% of total expenditures) 2004 2005 2005 2007 GPRSP-1 benchmarks 20 22 22.5 23 Education MTEF projections 32.1 32.4 32.8 GPRSP-1 benchmarks 6.3 6.5 7 7 Health MTEF projections 17.3 14.7 14.8 Source: Ministry of Finance and Public Administration. It is difficult to correlate sectoral allocations between MTEF and annual budgets. The MTEF projects recurrent and investment budgets by a classification that matches neither the administrative classification of the budget nor the programs of PIP. While it is difficult to compare sectoral allocations, it is possible to compare total allocations for recurrent and investment budgets. Budget allocations consistently surpassed MTEF projections (table 2.15). This divergence is particularly concerning since, under a fixed exchange rate regime, fiscal policy remains the primary instrument of macroeconomic management. In sum, results suggest that the MTEF did not define the framework for public expenditure planning, for 2005, 2006, and 2007. 22 Table 0.17: MTEF and Annual Budget Allocations, 2005–07 (% of GDP) 2005 2006 2007 MTEF 20.2 20.1 19.4 Recurrent budget Budget allocation 23.1 22.1 22.9 MTEF 14.3 14.6 13.8 Investment budget Budget allocation 21.1 17.5 19.2 Source: Ministry of Finance and Public Administration. There is not yet a legal framework for MTEF preparation, which may explain why the process is incipient. The current planning law dates from 1985 (Law 52/II). The draft of the budget planning law (approved by the Council of Ministers in April 2006) introduces the several planning instruments: the Economic and Social Development Plan (PDES), which will replaces the PND; the global MTEF; and the sectoral MTEFs.20 After its first round of discussions, Government withdrew the law from the Parliament. Debate about critical concepts, such as decentralization and regionalization, brought about the decision to revise the draft law, which is expected to start being redrafted soon. In 2004 four line ministries started preparing sectoral MTEFs with international technical assistance (Health; Education; Agriculture, Fisheries and Environment; and Social Protection). Only Education finalized the MTEF with success. The difficulties in preparing the sectoral MTEFs resulted largely from lack of capacity, shortage of human resources at the sectoral level, and lack of coordination between the Budget General Directorate (DGO) and the General Directorate for Planning (DGP). Furthermore, the preparation of the sectoral MTEFs overlapped the preparation of the global MTEF, and the interaction between the two processes was inefficient. The global MTEF set ceilings that were respected as truly the upper limits. As a result, the global MTEF was revised to accommodate the revised ceilings. The authorities are currently finalizing the revision of the MTEF (2008-2010) which they plan to revise on a rolling basis. Furthermore, several sectors are preparing their sectoral MTEFs. To ensure the effectiveness of both the global and the sectoral MTEFs it is critical to move forward with the legislation that supports their implementation (Budget Framework and Budget Planning Law) and to build capacity at the sectoral level for MTEF preparation. Furthermore, adequate integration between MTEF and the budgetary process should be ensured. With regard to the GPRSP-1-2, we recommend that its priorities to be reflected into the annual budgets in order to achieve the proposed GPRSP-1-2 goals. Recommendations: The authorities are currently finalizing the revision of the MTEF (2008-2010)21 which they plan to revise on a rolling basis. Furthermore, several sectors are preparing their sectoral MTEFs. To fully benefit from these exercises, it is recommended: 20 This law will also reflect the 1992 Constitutional revision. 21 The first draft was completed in August 2007. 23 • Move forward with the process leading to the adoption of the Budget Framework Law, which establish MTEF and program budgeting. • Prepare MTEF on a rolling basis by reviewing it annually and insure adequate integration between the MTEF and the budget (box 2.1). • Build capacity at the sectoral level for MTEF preparation. • Correct the pragmatic dichotomy that was found between GPRSP-1 and PIP for better conversion of GPRSP-1 plans into action. • Better integrate and coordinate between sector planning and budgeting teams to avoid dispersion of resources; and better prepare the intrasectoral scenarios that are linked to the economic and financial realities or real political priorities. To these ends, further integration between planning and budgeting should be facilitated by introducing single- sector coordination units in all line Ministries (Direcção Geral de Planeamento, Orçamento, e Gestão, or DGPOG). Box 0.1: Medium-Term Expenditure Framework A traditional MTEF aims at improving the decision-making process through indicative distribution of resources in a multiyear framework to better express decisions resulting from implementation of public policies. It also aims to structure decision-making to encourage efficiency and focus on expenses related to government objectives. As such, the MTEF can be defined as an instrument that organizes allocation decisions based on government priorities and its opportunity costs. A simplified MTEF should involve three basic processes: 1. Estimation of available public resources for a determined multiyear period, taking into account a fiscal and macroeconomic scenario. 2. Estimation of current costs associated with already implemented public policies, by programs and sectors, in a multiyear framework. 3. Interaction between revenues and expenditure to align public policies to resources available, always in a multiyear framework and rolling basis, thus improving resource allocation and funding mechanisms. Steps: 1. Establish macroeconomic estimates. 2. Align public policies to government objectives, considering the fiscal restrictions. 3. Establish policies, resources, and expenditure by sector. 4. Reconcile resources and expenditure. 5. Reconcile policies and expenditure. In general, steps 1 and 2 aim at improving macroeconomic estimates and to establish fiscal policies in a multiyear framework and rolling basis. Step 3 is related to the sector development policies. Steps 4 and 5 are linked to the reconciliation of existing or new expenditure and available resources (or both). 24 Public Finance Management ORGANIZATIONAL STRUCTURE MFAP is responsible for planning, budgeting, financial management, accounting, and internal auditing. MFAP is in the process of institutional reform. In 2001 Law 30 introduced the separation of Treasury and Public Accounting functions, previously performed in the General Directorate of Treasury (DGT), and established the General Directorate of Budget (DGO), responsible for the recurrent budget. The following entities at MFAP play a role in the different stages of the public financial management cycle: • General Directorate of Planning (DGP) prepares and monitors the execution of the Public Investment Program (PIP). It also is responsible for the preparation of planning instruments (for example, MTEF). • General Directorate of Budget (DGO) prepares and monitors the execution of recurrent budget. It also commits and liquidates capital expenditures. • General Directorate of the Treasury (DGT) programs and manages financial resources and manages debt (internal and external). It also is in charge of the financial management of the state budget (payments). • General Directorate of State Assets (DGPE) commits and settles the procurement of goods and services and other recurrent expenditures in accordance with contracts (electricity, water, phone, fax, insurance). • General Directorate of Public Accounting (DGCP) commits and settles outlays, except capital expenditure and these expenditures linked to the operations of public institutions. It also is in charge of implementing the National Plan for Public Accounting. • General Directorate of Tax (DGCI) collects income tax revenues. • General Directorate of Customs (DGA) collects customs revenues. • General Inspectorate of Finance (IGF) undertakes internal control and auditing of the financial activities of the executive. The current organizational model of public finance management is highly centralized under MFAP. The first steps of decentralization were taken in 2007. Decentralization started in 2007 with the transfer to line ministries of the process of commitment and liquidation of expenditure, which used to be the exclusive responsibility of MFAP. Hereafter, other budget execution processes will be gradually transferred to the line ministries. Once decentralization is completed, the general directorates of MFAP should be in charge of establishing the guidelines for line ministries. General directorates should only undertake budget execution except when centralization provides operational benefits with regard to costs or control (usually nondiscretionary expenditure such as salaries, debt, and transfers to municipalities). Moving forward with the decentralization of budget execution will require capacity building in line ministries. 25 The MFAP General Directorates that are more in need of reforming their attributions are DGCP and DGPE, because implementing budget execution should not be the responsibility of central accounting structures. As part of decentralization DGCP and DGPE should be freed from budget execution activities. DGCP should concentrate on implementing the National Plan of Public Accounting and on structuring the patrimonial accounting. DGPE should be fully responsible for identifying state assets and their evaluation criteria to register them correctly. To successfully implement this reform, it is critical, first, to evaluate the personnel structure of DGCP and DGPE. Second, it is critical to effectively integrate the accounting and state assets modules in the Integrated System of Budgetary and Financial Management (Sistema Integrado de Gestão Orçamental e Financeira, or SIGOF). 1.1 Decentralization should be accompanied by a strategic plan of organizational attributions that reflects the intended decentralization. The following organizational systems could be institutionalized: • Planning and Budgeting, • Financial Management, • Accounting, • Internal Control, • Assets, • Human Resources. Each of these organizational systems should be supported by an administrative structure and technological systems (such as modules in SIGOF). Administratively, the General Directorates of MFAP should have coordination responsibility, supported by the line ministries or decentralized structures (or both). Annex 2 describes the responsibilities and structure of each organizational system. BUDGET PROCESS The current legal framework that guides the preparation and execution of the central government budget is comprised of three main instruments: (a) The Budget Framework Law (Law 78/V/98) defines the general guidelines, principles, procedures, deadlines, and instruments related to the entire process of budget preparation and execution (box 3.3); (b) the annual budgetary laws (Finance Laws) provide guidance on year-based rules, especially on recruitment of human resources, policies, endowments (captive), and transfers to local governments22; and (c) the decrees of budget execution, published annually, provide guidance on rules and procedures related to budget management. These include rules for hiring staff and remuneration, percentage of salary increase, and norms for the external acquisition of goods and services. In addition, MFAP prepares and distributes to the line ministries the Guidelines for State Budget Preparation (Directivas para a Elaboração do Orçamento do Estado). The guidelines include a manual for budget preparation, which is reviewed annually, and establish priorities for budget 22 The annual budgetary laws also regulate tax policy issues. This confluence goes against the principle of exclusivity, one of the oldest budgetary principles. According to data provided by MFAP, this practice will end once the new Budget Framework Law is approved. 26 preparation operational procedures. They also indicate the timeframe for sector budget preparation and submission to the DGO. 1.2 The budget elaboration cycle is carried out according to a reliable schedule. The main phases can be summarized as follows: • First phase: (ends March 31): Starts with the definition of its macroeconomic parameters (predicted GDP growth, inflation).23 Next, MFAP assembles estimates of revenues (DGCI and DGA) and expenditure (DGO). • Second phase: (ends July 15): Once fiscal goals and the level of expected revenues have been defined, DGO and DGP prepare the first analysis of expenditure ceilings. After their discussion and validation by DGO and DGP, the Minister of Finance submits the ceilings to the Council of Ministers. Once they are approved by the Council, MFAP elaborates the call circular and sends it to the line ministries, specifying ceilings for recurrent and capital expenditure. Then each ministry distributes its overall ceiling to its subordinate units and provides the details on economic classification.24 • Third phase: (ends September 15): Sector proposals usually bid for values higher than the initial ceilings, resulting in a negotiation. MFAP, through the DGO, consolidates all the proposals and revises them according to policy objectives. In the event of an unsuccessful negotiation, DGO will make the adjustments as needed. Once the negotiation is finished, the MFPA completes the budget and submits it to the Council of Ministers for approval. • Fourth phase: (ends October 15): The government presents the draft budget bill to Parliament. • Fifth phase: Parliament votes on the draft budget bill within 60 days of its presentation. 1.3 The Budget Framework Law mentions the existence of a recurrent budget and multiyear Public Investment Programs (Programa Plurianual de Investimentos Públicos, or PIP). However, PIP has been prepared on an annual basis and included in the budgetary law with a common designation of “investment budget.� Although PIP should derive from the GPRSP-1 and MTEF, PIP clearly diverges from these two instruments (chapter 2). 1.4 The existing recurrent budget constitutes what is known internationally as an “entry budget.� Expenses are classified according to their nature and represent activity inputs, while no attention is paid to results. It thus is difficult to evaluate sectoral objectives and goals. This situation will change after approval of the new budgetary law, which also will establish the use of the program budgeting technique for recurrent expenditures. The 2009 budget should incorporate a pilot exercise of program budgeting of recurrent expenditures. As included in the draft project proposal of the Budget Framework, all budgets should be prepared under the program budgeting principle. However, the transition between an entry budget and a program budget is not simple and requires a number of conditions: among others, a monitoring and evaluation (M&E) system in place, changes to the IT system, definition of implementation 23 For the 2008 budget, the parameters and the macroeconomic scenario were prepared by the recently established Macroeconomic Unit at the MFAP. 24 The main expenditure items always are salaries and remuneration. For instance, salaries and remuneration corresponded to 81% of the recurrent budget of the Ministry of Health in 2007. 27 methodologies, and capacity building. The transition could be facilitated if the government decided to undertake a pilot exercise in 2 ministries with the 2009 budget. To this end, the M&E system needs to be strengthened. The authorities are aware of limitations of the existing system25 so have recently channeled human and financial resources to M&E. 1.5 The volume of investments planned for a certain year is directly linked to the availability of funds at the national level for each line ministry. The line ministries, once they receive their respective ceilings, propose inscription or maintenance of projects in the state budget, with respective financial values. The inclusion of new projects or the expansion of existing ones is conditioned by the availability of national counterpart funds26 for the project itself and for the eventual implications for the recurrent budget. The implementation of some investment projects have large implications for the recurrent budget (for example, construction of a hospital). Accordingly, the corresponding recurrent expenditures need to be taken into account. Box 0.1: Summary of the Draft Budget Framework Law 25 Government acknowledged in the first APR (July 2006) that “GPRSP-1 M&E (implementation of structures, data gathering, flow of information among sectors, the INE, and the STAD) has shown little progress.� 26 Contribution of the national budget to finance an externally funded project. This contribution varies according to the donor. 28 The main changes proposed by the Draft Budget Framework Law are: • Budgetary principles and rules. Almost all internationally accepted budgetary principles are included and described in the proposal, which includes new principles such as participatory budget and the request for program budgeting for all types of expenditure. • Management instruments. The proposal introduces the Financial Comptroller, who is to be responsible, under budget execution decentralization, for the legality, regularity, and execution of the different phases of revenue collection and expenditure. • Budgetary execution regimes. The proposal specifies three different types of budgetary execution: (i) integrated services, under the juridical-financial regime of administrative autonomy, with the Financial Comptroller, the Authorizing Officer, and the Treasury payment system responsible for budget execution; (2) autonomous services and funds, which are the responsibility of its managers, with expenses covered by own revenues (non-earmarked) and procurement of goods and services through public tender; and (3) social security, which falls under the responsibility of the Instituto Nacional de Previdencia Social (INPS) (in the case of the subsystem of contributions) and of a management entity (in the case of the subsystem of noncontributions). Revenues and expenditures are treated jointly (both subsystems), in articulation with DGT. No management unit can undertake funding operations without prior authorization by the government, and any financial balances can be used only with government authorization. • Public accounts reporting regime. The proposal specifies in detail not only which reports must be submitted to which institutions (accounts from the Social Security System, accounts from the National Assembly, accounts from Justice, and so forth––all are to be integrated subsequently into the general State accounts), but also which format, content, scope, and deadlines are to be considered. As with the current law, the proposal provides discipline to the entire process of State budget preparation. It includes in detail the deadlines, schedule, and stages involved in budget preparation by the government, as well as the deadlines and level of responsibility according to which the Law should be discussed and submitted to votes in the National Assembly. The procedures and internal deadlines for budget preparation by the government are described extensively, which is not so common in other countries. 1.6 To improve the efficiency of the investments and their successful implementation, the MFPA should monitor physically and financially the more relevant projects funded with external resources. Such monitoring would help the MFPA to make the sectoral allocations more efficiently and would prevent the underestimation of recurrent expenditures requirements that emerge from the investment projects. Furthermore, it could prevent misallocation of resources. For example, facing insufficient resources, a sector may allocate the funds to less priority projects or proceed with projects that will require additional central government resources. 1.7 Steps were taken to unify budget preparation under the current legal framework.27 With the current law, recurrent and capital budgets may be prepared almost autonomously by different directorates at the MFPA. However, such duality compromises the global vision required when preparing the budget. MFAP is aware of that limitation and thus is pursuing internal coordination, so that ceilings are defined under the rationale of a unified budget. As part of this process, line ministries are unifying their planning and budgeting units, creating a new unit called the General Directorate of Planning, Budgeting, and Management (Direcção Geral de 27 Art. 11, items “b� and “e� of Lei de Bases do Orçamento do Estado, Law 78/V/98. 29 Planeamento, Orçamento, e Gestão, or DGPOG),28 responsible for all activities related to budget preparation and execution. 1.8 The functional classification of public expenditure does not accord entirely with international standards. The functional classifier is not in line with the Classification of Government Functions (COFOG), initially proposed by the United Nations and adopted in the IMF’s Government Finance and Statistics Manual of 2001. This discrepancy hinders the direct international comparison of expenditures. It would be beneficial if Cape Verde would use the primary structure included in COFOG of 10 functions and 69 subfunctions and to make adjustments to the third level, below the subfunction.29 1.9 Some revenues and expenditures are not included in the budget. These omissions contravene the budgetary principles of universality and non-earmarking of revenues. Some autonomous institutes do not include in the budget their own revenues and the expenditures financed by them. Furthermore, oil-related subsidies have not been consistently recorded in the budget. Finally, part of the outlays related to salaries and remunerations of revenue collection institutions are directly deducted, causing an underestimation of the overall envelope of revenues and staff/remunerations expenditures. Recommendations • Update the functional classifier of expenditure according to COFOG. • Implement the DGPOGs in the remaining line ministries. • Provide a module in SIGOF to assist with budget preparation. • Unify the coordination of budget preparation in the DGO. REVENUE MANAGEMENT 1.10 Revenue collection is the responsibility of the General Directorate of Tax (DGCI) and General Directorate of Customs (DGA).30 Detailed revenue maps are sent daily to DGT (by type of tax, fee, and others) of fiscal and customs revenues collected the previous business day, indicating deposits to the Treasury account. For DGA, the Import Declaration Form is filled out electronically, including the code identifying duties or customs revenues. Revenues are filed directly in the customs offices (fiscal houses)31 for daily consolidation and deposit into the Treasury account. Deposits are made in two temporary bank accounts of the Treasury, one for VAT-related deposits and another for various types of customs revenues. Twenty-five percent of the revenues collected remains in a specific DGA account and is not included in the state budget. 28 Only the Ministries of Agriculture and Labor have implemented the new DGPOG so far. Other ministries, such as Health and Education, are dependent on changes made to their statutes to be able to implement the organic change. 29 Note that this is an old recommendation, already included in the basic document and recommendations of the Country Financial and Accountability Assessment (CFAA). 30 DGA includes three divisions (Praia, Mindelo, and Espargos) and four delegations (Aeroporto da Praia, Porto Novo, Assomada, and S. Filipe). 31 There are 10 fiscal houses distributed in the 3 divisions. 30 1.11 The fiscal revenue collection is administered by DGCI through a specific payment form on which the taxpayer provides data of the amount to be paid, including the economic classification of the revenue (the form includes bar codes). Payments are made in the commercial banks through check, cash, or direct debit.32 Despite modern payment methods, the taxpayer can still pay without providing the classification code, thus duplicating the work of DGCI, which must identify the revenue and prepare and send the detailed maps to the Treasury. Moreover, preparation of these maps still relies on the hard-copy receipts of the payment invoices. 1.12 Revenue collection comprises both advanced and old payment forms, including manual processes for identifying revenue and reporting revenues collected. Upgrading such models should be analyzed jointly with developing the accounting, Treasury management, and revenue management modules to be implemented through SIGOF. The fundamental premise is that all government revenues, understood to encompass all income into the state budget, must be collected through the banks consistent with the principles of account unification and standardization. 1.13 It is important to promote standardization of revenue collection, starting with the creation of a Single Payment Form, which will include all revenues in the state budget. The process will be as follows: • Once the revenue payment form is filled out, the citizen will proceed to pay in an accredited commercial bank. • The accredited commercial bank will deposit the amount in the Treasury’s temporary bank account and send the document to the clearinghouse within 48 hours. • Once the document is cleared, the Treasury bank account will be debited or credited. • The incorporation of resources in the Single Treasury Account will be reconciled daily against liabilities and gross receipts to be categorized. • According to a deadline to be defined by DGT, all data included on the Single Payment Form will be sent by the commercial bank to the Operational Nucleus for Information Systems (NOSI), which will classify it according to SIGOF for accounting and fiscal purposes. • In accordance with the revenue collection code, the resources can be included in SIGOF as Treasury availabilities or as availabilities for the institutions responsible for the revenue collection. Figure 0.1: Government revenue collection system 32 The electronic payments are made in the DGCI offices located in Sal, São Vicente, Praia, and Santa Catarina. 31 Government Revenue Collection System (1) SIGOF Single Payment Revenue Accounting Form (2) Entity Tax Code Commercial Bank (Own Revenue) Filed with Revenue Collection DGT Data (Treasury Revenue) (5) Classified (7) Revenue (3) N.O.S.I Result of Bank Periodical Filing Clearance for Revenue Classification BCV National Treasury (6) Central Bank Single DGT Gross Revenues Account To be classified (4) SIGOF Credit – Single Account Source: Albuquerque and others, Gestão das Finanças Públicas (2006). Recommendations: 1.14 Government revenues and expenses should be collected and paid through commercial banks in accordance with the principle of account unification. The collection, accounting, and consignment of revenues should be automatic, avoiding manual operations as much as possible. The following actions also are recommended: • Establish a mandatory economic classification in a single payment form for all government revenues. Collection should be undertaken by accredited commercial banks (comply with the rules in the provision manual to provide revenue collection services). • Update the DGA payment procedures with more automated and reliable methods. • Create a manual that includes all revenue collection rules to be applied by the accredited commercial banks so that all collection data required (tax identification number, or NIF, value, date, and revenue code) are entered concomitantly with payment by citizen. • Require mandatory direct transfer to NOSI of files containing all the revenue collection data in a format to be defined by the government, so that it is processed in SIGOF, making it available to the other government agencies (DGCP, DGCI, DGT, and DGA). • Establish revenue classification mechanisms that enable individual identification of resources available. • Create operational procedures that ensure correct revenue accounting in all its stages, even after the resources have been paid into the account, when rectifications, compensations, and restitutions can be made. 32 • Eliminate sending revenue collection documents to the clearinghouse. After implementation of the Single Treasury Account, within the context of SIGOF, the commercial bank should transfer the revenues directly to the Single Treasury Account and the data file to NOSI, according to the deadlines included in the service provision manual. PUBLIC EXPENDITURE EXECUTION 1.15 Today, the budgetary and financial registration is performed on SIGOF, which contains the economic classification of expenditure, NIF of the beneficiaries, and banking data for payments and retentions. The budget in SIGOF already includes subtraction of 10 percent of the initial allocation on all line items (headings).33 There are six phases in the expenditure stream: (i) release of a requisition by the concerned ministry; (ii) validation of the requisition by the DGCP - this operation, which consists of checking the availability of the appropriations, is called cabimento; (iii) return of a copy of the cabimento to the technical ministry for order of work or supply; (iv) realization of the amount due by the DGCP (liquidação) - this consists of calculating the amount due and certifying the effectiveness of the expenditure; (v) release of an order to pay by the DGCP cobrança; and (vi) payment is made by DGT. Decentralization of budget execution started in 2007 with the transfer of the cabimento and liquidation phases to the line ministries. Once decentralization is complete, line ministries will be in charge of all phases of execution, with the exception of payment, which will continue to be the responsibility of DGT.34 1.16 The government has systematically built up arrears in the past, due in part to the ability of making individual cabimentação. The public administration does not adopt cabimentação as an estimated value of expenditure for the whole fiscal year. Cabimentação is always individual for each expenditure incurred at a given moment. In the instance of an electricity bill, the expenditure (cabimentação) applies to each monthly bill, not to the estimated value of that expenditure for the entire fiscal year. This method is not recommended because it generates the risk of a public entity assuming higher commitments than the allocation for the fiscal year. If civil servants do not become aware that the allocation was insufficient until the end of the fiscal year, Treasury will be under pressure to make payments that it had not accounted for. This is the reason that in the past the central government built up arrears so consistently. In the 2008 budget, a rule was adopted to not make commitments after November to prevent pressures toward the end of the fiscal year (Cape Verde’s fiscal year is the same as the calendar year). 33 That percentage is revised throughout the fiscal year in accordance with revenue collection inflow. Additional funds are usually released for execution. 34 Before the decentralization of the cabimentação, this operation was conducted in three general directorates of MFAP. (i) DGCP: Expenditures such as salaries and remunerations, external transfers (interest rate payments, membership fees to international organizations, embassies), and domestic transfers (domestic debt, transfers to funds and autonomous services, subsidies); (ii) DGPE: Expenses such as goods and services and other contractual recurrent expenditures (electricity, water, telephone, fax, insurances), that is, these related to everyday transactions of public institutions; and (3) DGO: Investment and other recurrent expenditures. 33 TREASURY MANAGEMENT 1.17 The basic rules of the Treasury payment system are included in Law 10 of 1996. In this law replaced the rules dating back to colonial times (Regulamento da Fazenda Pública) (1901). Law 10 established the rules and procedures to be adopted in financial planning, treasury management, accounting of state revenues and expenditures, budgetary control and management, and payment methods. The DGT is the administrative unit that manages the Treasury. DGT is responsible for financial planning and payment execution, and centralizing all the operations of payment, as well as all revenue collected and its respective accounting.35 DGT also is responsible for preparing the annual Treasury plan, which contains cash flow in the main categories of recurrent and capital expenditure. DGT determines the gross funding requirements and plans the issuing of Treasury bills and bonds, as well as available funds. 1.18 Even though Decree-Law 29/1998 determines the centralization of available funds in the government account, in practice such centralization applies only to Treasury resources. The “sovereign structures� and autonomous institutes still maintain balances in commercial bank accounts. Furthermore, most grants also are still kept in specific bank accounts outside of government control.36 At the moment, the big challenge for the government is to close these commercial bank accounts and transfer their balances and future financial transactions to the g Central Bank account. As it stands, DGT can have a cash-flow shortage while other government entities have available funds in commercial accounts. 1.19 The DGT has all that is necessary for the effective control of resources (legislation and IT system) but the accounting and treasury management modules need to be incorporated into SIGOF. Because these modules have not yet been incorporated, specific accounts for earmarked revenues (these with specific objectives) are opened in the Central Bank. There also is a special account in the Central Bank that accepts foreign currency in which project funds are deposited, making it difficult for DGT to follow up on all entries and exits related to each project. This problem will be solved with the creation of the accounting module in SIGOF. As a temporary measure, subaccounts of the special account could be opened, enabling transactions of each project to be individualized.37 35 Budgetary outlays are separated into eventual and permanent transfers. Permanent transfers, which do not require previous authorization from services, are budgetary and financial execution of the state commitments included in the budget, whose regularity, amount, and payment date are known and planned beforehand. Some examples include permanent monthly transfers to autonomous funds and services, local municipalities, public enterprises, private institutions, embassies and consulates, subsidies, contractual expenditures (water, electricity, telephone) public debt interest payments, internal and external financial liabilities, and membership fees for international organizations. Eventual transfers, which do require prior authorization, are budgetary and financial execution of state commitments included in the budget, whose regularity, amount, and payment date are not known and planned beforehand. Decree 26 of 1996 establishes that no expenditure commitment in the eventual transfer classification can be made without prior authorization of MFAP. 36 The budgetary registration of an expenditure financed through grants normally happens only after its effective execution. 37 The investment programs included in the budget should also be included in the demonstration called Funds Origin and Application Map (Mapa de Origem e Aplicação de Fundos). Its main objective is to evaluate the financial balance between the predicted revenues and expenditures according to the funding sources. 34 1.20 The basis for centralizing cash resources is being established, and the government should do it continuously. A first step is to centralize the resources of autonomous entities and sovereign structures in a Single Treasury Account (mother account), subdivided into specific subaccounts. This step would enable the government to have the overall position of the availabilities. Meanwhile, the effective control of the financial resources of the different administrative structures will be achieved only by implementing the Single Treasury Account in SIGOF. Doing so would enable DGT to individually control the financial balances of all of the entities, thus creating the conditions to eliminate all of their commercial bank accounts or the subaccounts kept in the Central Bank. 1.21 A common concern associated with the Single Treasury Account is the fear that “combining� funds will reduce transparency and the individual control of the returns of each of the deposits. In fact, as it often is said, “money bears no stamp,� meaning that resources available in a bank account are part of a single balance sheet. The fear that with a single account movements of money will not monitorable is unfounded. It is possible to run a parallel control that follows up on the origin of each deposit, as well as any transaction carried out in the Single Treasury Account. SIGOF can achieve this through accounting records. Therefore, after the introduction of the accounting module in SIGOF, the origin of the resources, their transactions, the breakdown of the daily balances by funding source, and the units responsible for these transactions can be accessible not only by the central government but also by the autonomous institutes and sovereign bodies. 1.22 In this model, transactions made through SIGOF will enable identification of the beneficiary of each deposit and the party responsible for each payment. When a specific government entity needs to verify its own institutional balance, it will have access to it through SIGOF and not through the Central Bank account holding the entire government balance. This process should occur concomitantly with the upgrading of the existent payment system mechanisms. The Central Bank should be capable of monitoring the account holding the commercial banks’ reserves and the Single Treasury Account through an IT system, which is not the case at the moment. In addition, access to the Central Bank system should be made available to DGT, so that it can monitor transactions in the single Central Bank account. 1.23 A fundamental principle of Treasury management is that revenues and expenditures be transacted through the banking system, strictly following the principle of cash unity. The opening of bank accounts for government resources should be restricted, unless under exceptional circumstances, and require MFAP approval. To this end, the legal documents must prohibit public institutions’ transactions of resources in commercial bank accounts outside the Single Treasury Account. However, given the long process to implement the Single Treasury Account in SIGOF, some exceptions should be permitted. These comprise the accounts of the government structures located in areas that have no technical resources to use SIGOF, the accounts in foreign currency (in commercial banks outside the country), and the accounts aimed at responding to special circumstances, to be regulated by the Treasury. 35 Recommendations: 1.24 The government should implement the model of the Single Treasury Account. For this to happen, it will be necessary to: • Prioritize inclusion of the accounting module in SIGOF. • Map and transfer the balances of all government entities, including institutions and sovereign structures, enabling the control of resource availability without loss of autonomy for the institutions, which may implement their expenditures, in all stages, through SIGOF. • Publish legislation prohibiting transactions of resources outside the Single Treasury Account, except in cases in which the specific operational characteristics do not allow them to be conducted through the Single Treasury Account system. In such cases, the resources may be deposited, according to MFAP criteria in banks designated by the government. • Provide the Central Bank with an IT system that will enable DGT to monitor transactions in the Single Treasury Account in the Central Bank, mainly for financial reconciliation. • Create mechanisms for electronic transfer of large amounts, allowing transfers of resources between the DGT and the banks. PAYMENTS 1.25 Management of payments is centralized at DGT38 because the principle of a Single Treasury Account is not yet in place in SIGOF. As a result, and given the scarce resources, DGT ends up being involved in the operational activities of expenditure execution. Once the principle of a Single Treasury Account is set up, decentralization of the payment can be implemented, leaving DGT with the sole responsibility of financial planning: matching credits and debits. DGT would be responsible for establishing the withdrawal limits for line ministries, which would issue payment orders against the Single Treasury Account. This procedure can be done in SIGOF, which will contain all the data needed to credit the beneficiary’s account once the financial amount is made available by the DGT for the period. Thus, the payment order will work as an electronic check. 1.26 SIGOF should be prepared to adopt several payment order modalities, generated according to the need of each manager and the payment modality selected. Depending on the level of integration of the financial system, withdrawals from the Single Treasury Account can be made at different times according to the various payment modalities that currently exist in the Treasury. Payment orders, once issued and authorized by the Disbursement Officer and Financial Comptroller, must be grouped by the end of the day into different files consistent with 38 The DGT is legally equivalent to a banking institution, including participating as a member of the banking clearinghouse. Payments are made through checks issued exclusively by DGT; they must be nominal and crossed, and withdrawn against the Treasury account held in the Central Bank. The Treasury check must be presented for payment to a banking institution within 30 days after it is issued to be credited within two business days. In addition, the DGT makes nominal transfers, indicating the date and value for the credits to be made effective by the banking institution. The transfer request must be made two business days in advance. 36 the banking institutions responsible for the payments. Once the banking institution has the files and after the transfer of resources by the Treasury, the bank will credit the beneficiary within the agreed period. In specific situations, payment by the Treasury can be made directly to the beneficiary through integration with the Central Bank system. Payment orders not authorized by the end of the day by the Disbursement Officer and Financial Comptroller will be canceled before the files are consolidated. 1.27 SIGOF will perform transactions and control the single account. Because the Treasury does not have an agency, for payment orders to reach the final beneficiaries, it must continue to use the accredited banking institutions. In this model, after being debited from the Single Treasury Account, the financial resources corresponding to the payment orders issued each day will be credited to the beneficiary banking institutions early the following day. As a general rule, government payments should follow four steps (figure 3.2): • Registration of creditor and its banking address. • Issue of payment order in “D,� after the regular liquidação. • Transfer of information from the treasury to the financial agent (3a); financial resources (3b) corresponding to the payment orders issued the day after the payment orders have been issued (D+1). 1.28 Reception of credit by the beneficiary in the period established. To prevent public resources from being incorrectly credited, the validation of the beneficiary’s bank account data should occur in real time, concomitant with its registration. If the validation cannot be implemented in real time, a normative rule by the Central Bank should make it mandatory for the banking institutions to verify (NIF) whether the beneficiaries of the credits are the owners of the bank accounts. If they are not, the bank will return the funds to the Single Treasury Account according to the conditions in the normative rule, providing the number of the payment order that is to be canceled. Only after the creditor and its banking address have been registered can the manager issue the payment order. Furthermore, the issuance of the payment order in SIGOF does not guarantee that the creditor will receive the funds, because it is mandatory to have the signatures and authorizations of two people from the line ministries (Disbursement Officer and Financial Comptroller). 1.29 The ideal model would be one in which SIGOF has been programmed to certify electronically that the authorization has been issued by the authorized agent. If the electronic authorization mechanism is not adopted, it is recommended that to wind up the payment process, the Payment Orders List from the unit be printed and the signatures from the responsible persons be collected and delivered to the appropriate branch for supervision (in this case, a branch for each banking institution). To not delay issuing payments, it is advisable to appoint alternates. 37 Figure 0.2: Disbursements - Payment orders Creditor register (1) Creditor/Beneficiary SIGOF (4) DGT Single Account $ OP File Financial Agent OP Entity Entity (3a) Authorized (2) Payment $ (3b) BCV DGT Total Amounts Transfer Single Account Debt Source: Albuquerque and others, Gestão das Finanças Públicas (2006). 1.30 The system should include the option of Intra-SIGOF payments whenever one government unit must make a payment to another government unit that also is part of SIGOF. The government unit that requests the service will undertake the cabimentação and, after receiving the invoice from the service provider, make its liquidação. The commitment will be paid through an Intra-SIGOF payment order; only an accountancy registry is made, with no transaction in the Single Treasury Account of the Central Bank, only the account recording in both government units. 1.31 Given the proposed models for the transfer of revenue collection and expenditure execution, Treasury participation in the clearinghouse should be re-evaluated. According to international best practice, creation of clearinghouses in a payment system is intended to reduce the liquidity risk of participants. Such risk is unlikely to affect the Treasury because Treasury programming will ensure a sufficient positive balance in the account. In addition, the liquidation and collection processes do not take long (approximately two days). This length of time should not give rise to a large balance problem for the Treasury. As the government usually operates with large amounts, the direct transfer to the Treasury account at the time of revenue collection and direct transfer of payments reduce not only the cost to the Treasury but also systemic risk. 38 Recommendations: • Replace checks with electronic payment orders in SIGOF, with generation of files for each beneficiary banking institution. • After elimination of checks and direct transfer of the collection, the Treasury can stop participating in the clearinghouse. DEBT MANAGEMENT 1.32 Debt management has always been the responsibility of MFAP. DGT centralizes management of its own issued total debt, both domestic and external. However, the Central Bank also is authorized to issue public securities for monetary policy purposes. In practice, the Central Bank supports the DGT in operations related to debt administration, establishing a productive relationship between the government’s fiscal and monetary policies. Furthermore, the legislation enables the government to eventually finance any insufficient cash flow by borrowing up to 5 percent of recurrent revenues from the previous fiscal year, so long as it is settled before end of the fiscal year. 1.33 DGT has had specific software for registering public debt since 1996 but started using it effectively only after 2004. Recently, all data from the database and registries were lost and had to be re-entered, which demonstrates the need for creating backup security mechanisms. Data on external debt was re-entered and its validation was concluded in early 2008. Currently, the DGT is inserting the data related to domestic debt, after which it will be able to evaluate debt sustainability. Currently, neither the domestic nor the external guaranties are recorded in the system. The software already permits estimating debt service payments and generating quarterly data, and, given its integration with SIGOF, contributes to the budget preparation process. At present, only the DGT has access to the software, and the debt has not been registered with regard to accounting. 1.34 Treasury bonds debt is almost entirely pre-fixed, with maturity varying from 18 to 30 months. The government operates with a liquidity reserve to administer debt-related obligations. That reserve, kept in the Central Bank, increases Treasury liquidity without generating returns. Securities with a maturity that exceeds one year are denominated National Treasury Notes (Obrigações do Tesouro Nacional, or OTNs) and these with a less-than-one-year maturity, National Treasury Bills (Bilhetes do Tesouro Nacional, or BTNs). There are no securities indexed to exchange or inflation rates, and all debt is under Central Bank custody. In addition, the securities also are offered directly to citizens through the banking institutions. When the Central Bank carries out debt auctions, there is a direct dialogue with the Treasury to agree on the interest rate to be used. The operational expenditure of the Central Bank is not included in the state budget. 1.35 Recently, the government recognized “hidden debt� between some public administration structures and some public institutions and enterprises, as well as debt with the private sector. Part of this debt had been included in the budget, but had not been settled. Another part was related to contracts for the provision of goods or services, for which there was 39 no budgetary authorization.39 To prevent the need of assumption of other agencies’ debts, during the preparation of 2008 Budget, DGO worked jointly with autonomous institutes and line ministries to obtain a more realistic budget.40 However, it is clear that there is no restriction to the possibility of the Treasury incurring debt via state enterprises. It is common practice to recognize debt based on “protocols,� which represent debt agreements between the government and its creditors, such as the Postal Services and the Instituto de Fomento à Habitação. Usually, this is floating debt: short-term negotiated debt with preferential rates and conditions. Payment is discharged as part of the budgetary cycle of line ministries. However, because the SOEs are not part of that budget, payments to them are not intrabudgetary operations. In fact, public enterprises do not normally receive resources from the budget. There are some isolated cases in which a specific public enterprise was capitalized through transfer of resources from the state budget. Recommendations • Publish an annex in the Budget Law that specifies the guarantees provided (including estimation of the amount). • Clarify the relationship between Treasury and Central Bank with respect to remuneration of reserves and services provided. • Centralize the issuance of titles at the Treasury, with the Central Bank participating in the secondary market for monetary policy purposes. ACCOUNTING 1.36 The reform of the accounting system started with Law 29 of 2001 and was enhanced with the Law 96 of 2006. The latter implements a chart of accounts (Plano Nacional de Contabilidade Pública, or PNCP). It determined that the PNCP should adopt the double-entry accounting method from 2007 on. Delays resulted in its not being passed until 2006 through Law 10. Law 10 indicates that only in 2008 can the method starts being effectively used, along with the other accounting changes established by Law 29/01. The system being used is still composed of independent and disintegrated elements based on the simple-entry method, without publishing periodic and consolidated statements containing the state’s equity status. The new legislation establishes that public accounting should enable the (a) periodic preparation of statements of accounts and other tables in all dimensions (for individual entities and consolidated): assets, operational, financial, budgetary, and outside-equity; (b) annual organization of the consolidated general balance; and (c) integration of other economic and financial data of the public sector administration in national accounts. It will also provide data of transactions related to: • Availabilities. • Relations with third parties, or more specifically between the state in a restricted sense and all the services and entities that, on behalf of the state, collect revenues, make expenditures, and administer or keep state goods. • Inventories and assets. 39 An example used is the debt of the Health Ministry with the State enterprise supplier of medicines related to an essential outlay, which had not been given sufficient resources because of budgetary limits. 40 The 2008 Budget includes larger provisions for the payment of utilities (Electra) and drugs (EMPROFAC). 40 • Net Wealth and public debts. • Asset elements related to public social welfare, facilitating follow-up of its financial balance and net wealth position. • Costs and losses, benefits and profits. • Operations of budgetary execution. • Outside-equity position related to contingent and future responsibilities. • Annual operational, financial, and extraordinary results. 1.37 Implementation of the PNCP will be a fundamental stepping stone in the reform of state administration and organization of public accounts. The PNCP will require a cultural change to promote innovative practices. Most of the public-accounting-related legislation in place until 2001 dated as far back as 1901. The reforms must be followed by changes in the state organizational structure and its competencies, and the integration of the accounting module in SIGOF, which is essential. Irrespective of delays in implementation, such reforms generally do not materialize over a short period. Furthermore, the proposal for the PNCP structure should follow International Public Sector Accounting Norms (IPSAN). Taking into consideration that other countries in general are trying to implement IPSAN, it is important that implementation guidelines for the accounting system in Cape Verde follow it. In practice, IPSAN aims at harmonizing criteria for recognition of revenues and expenditures and assets and liabilities. 1.38 To register revenues and expenditures on an accrual basis,41 acts and facts to be registered through accountancy must have a direct correlation in the accounting system of SIGOF. Correlation will enable the recognition of revenues and expenditure under the accruals optic and the registration of budgetary events. The registration of a specific “document� characterizing in a different manner the various acts and facts would achieve that. The document would be defined as the representation of the acts and facts implemented by the administrator, directly or indirectly, that have budgetary, financial, equity, economic, or management effects. By using this model, characterization of the various documents would record all events that occurs in the public administration and would become a strong element of SIGOF. Their reflection in the budgetary accounts and their respective classifications would be a consequence of the document registration itself, when correlated with the budget. As such, SIGOF would enable, for example, the registration of contracts, fiscal notes, receipts, and other documents, accounted for automatically by the system. 1.39 To register revenues and expenditures on an accrual basis, 42 acts and facts to be registered through accountancy must have a direct correlation in the accounting system SIGOF. That will allow the recognition of revenues and expenditure under the accruals optic and the registration of events of a budgetary nature. The registration of a specific “document� characterizing in a different manner the various acts and facts would achieve that. The document would be defined as the representation of the acts and facts implemented by the administrator, directly or indirectly, that have budgetary, financial, equity, economic, or management effects. By using this model, characterization of the various documents would represent everything that occurs in the Public Administration and would become the strong element of SIGOF. Their reflection in the budgetary accounts and their respective classifications would be a consequence 41 Currently, there is a mix of accrual and cash basis. 42 Currently, there is a mix of accrual and cash basis. 41 of the document registration itself, when correlated with the budget. As such, SIGOF would allow, for example, the registration of contracts, fiscal notes, receipts, and other documents, accounted for automatically by the system itself. 1.40 To facilitate accounting registration of the acts and facts and eliminate the need for instructing the accounts to be debited and credited, SIGOF will make use of a standardized table of codes. Each code in the “table of accounting events� summarizes the accounting transactions of a specific administrative act or fact. Therefore, a code of accounting events will be associated with each event in a document. The DGCP will be the entity responsible for updating and maintaining the table, as well as for creating the respective accounting scripts associated with each accounting event. For the system to correctly enter the fact that is being registered in the document, to the user must indicate the situation. For instance, a fiscal note may correspond to a service expenditure, consumption material, permanent material, or other situation, each having a different accounting registration. The user must indicate in each document the situation involved. 1.41 In Cape Verde, it is still the responsibility of DGCP to prepare the accounts and proceed to the cabimentação and liquidação of large part of public expenditure. As such, the DGCP is involved in tasks that are not usually the responsibility of an accounting unit. The mandate of DGCP should be to prepare the accounting data necessary for the different reports, providing the responsible agents with data regarding the budgetary execution of revenues and expenditure, as well as the state net wealth position, to facilitate the decision-making processes. Figure 0.3: System Fact and Accounting Registration Representation of the fact in the system Accounting registration Situation 1 Accounting Document Event A For Situation 2 example, Accounting Fiscal Note, Event B Receipt Situation 3 Accounting Event C Source: Albuquerque and others, Gestão das Finanças Públicas (2006). Recommendations • Develop specific software and undertake the inventory of the assets of the State. • Implement the new chart of accounts, including the patrimonial flows. • Register in SIGOF documents representing acts and facts that have budgetary, financial, equity, economic or management implications. INTERNAL CONTROL 1.42 The legal framework regulating the mandate and activities of the IGF was revamped in 2005. The General Finance Inspectorate (IGF), an entity subordinated to MFAP, 42 began its activities in 1987 with Law 130. In 2005, with the approval of a new organic diploma, the juridical framework necessary for undertaking inspections and control of acts and of economic and financial management of the entities was strengthened. Even though its autonomy is limited by its subordination to MFAP, the IGF was able to start the financial control of laws, decrees, and Cabinet resolutions. Based on the auditing manual, IGF also started observing compliance with normative instructions and determining to what extent the interests of finance are protected. 1.43 The IGF operational budget decreased from 2000 to 2004, but has improved with the introduction of the new organic “diploma� (act) in 2005. One administrative difficulty that IGF faces relates to the per diem table that establishes the amounts to be reimbursed to inspectors, which has not been revised since 1991. However, the IGF inspectors are compensated through a 13 percent risk bonus, which they receive in exchange for exclusive dedication, and through the higher remuneration of MFAP versus the rest of the public administration. 1.44 Financial limitations and the limited number of technical staff prevent the IGF from making more than about 40 inspections a year, jeopardizing the principle of permanent internal auditing. The current staff of the IGF comprises 30 inspectors; however, 13 are on work leave in other entities of the Public Administration. Effectively, there are only 17 inspectors, organized in 7 working teams. Each team manages to complete on average five inspections a year. A typical inspection usually takes two months, spanning the preliminary preparation phase, an external visit, preparation of the project report, provision of the right to take exemption, and drafting the final report. In 2005 and 2006 the IGF undertook the 35 inspections that had been planned. For 2007, 42 inspections were planned, including all the 22 municipalities. In total, in 2007, 54 inspections were conducted; however, 22 inspections were related to legal and procedural conformity of accounting documents only. Furthermore, the new Budgetary Framework Law (not yet approved) proposes that the IGF should prepare audits to the quarterly accounts submitted to the National Assembly. 1.45 Line ministries are seldom inspected because of the large scope of the IGF mandate. The 2005 organic diploma establishes that the IGF must conduct annual inspections in all municipalities, and every two years in the embassies and consulates abroad. In practice, because of travel costs, there is a concentration of activities in the city of Praia. Moreover, the IGF does not have historical indicators of the incidence of irregularities by sector that could help establish priorities for the line inspections. IGF argues that its working plan reflects its perceptions about which areas of the government are more relevant in terms of expenditure and occurrence of irregularities. The IGF annual report demonstrates that follow-up visits are being made to entities previously audited to evaluate whether IGF’s previous recommendations are being implemented. 1.46 The 2005 organic diploma introduces better articulation between the IGF and the Court of Accounts (Tribunal de Contas, or TdC). However, additional improvements are required. Since 2005, the annual activities plans of IGF and TdC are discussed jointly at the beginning of the fiscal year to prevent redundancies and to cover the greatest possible number of entities. However, the inspectors from IGF and TdC have not yet been trained to use SIGOF and therefore do not as yet take advantage of the benefits of the system in relation to integration and availability of timely data regarding budgetary execution. On the other hand, the diploma does not establish that the IGF reports that present administrative or criminal irregularities be 43 systematically sent to TdC or Public Prosecutor’s Office (Procuradoria-Geral da República, or PGR). 1.47 Delays in juridical processes undermine the usefulness of IGF audits. The PGR does not investigate in a timely manner the facts raised by the audits, thus jeopardizing the effectiveness of the system and generating a perception of impunity. It has happened that, given the lack of verification by PGR of the allegations contained in the writ, the accused manage to reverse the situation, instead accusing the IGF of frivolous behavior, which affected the morale of the inspectors. 1.48 The modus operandi of IGF is not based on a results approach. Rather, issues of legality and punctuality of documents submitted are the core of the inspections. In 2007 IGF undertook, for the first time and in an incipient manner, a results audit, as opposed to the formal audit, when trying to quantify the results of programs for combating AIDS and poverty. It is known that a good Internal Control System must go beyond the administrative dimension and include as well the budgetary, economic, financial, equity, normative, and management dimensions along with evaluation of programs and projects. In addition, the internal audit process should inform the administration concerning operations of entities inspected and recommend ways to improve administrative controls. 1.49 The IGF is part of the MFPA. A more centralized system would enable a greater dissemination of knowledge regarding internal audits. The General Inspectorate should be subordinated directly to the Prime Minister’s office so that it would give the Prime Minister direct access to the data generated by inspections and audits, promote a higher degree of independence for IGF when it undertakes inspections in the MFAP, and provide IGF with greater authority when it works in the line ministries. Recommendations • The final inspection reports that detect some administrative irregularity or crime should be systematically sent to the TdC or PGR, respectively. • Adopt a risk management strategy, including development of a historical trend series pinpointing Public Administration sectors that have irregularities more frequently. • Provide adequate resources to IGF so that can competently fulfill its mandate. EXTERNAL CONTROL 1.50 A draft of a new organic law that will empower the General Audit Office43 (TdC) has been prepared since 2004. It awaits final discussion and voting in the Parliament. TdC has prepared and submitted to the government two proposals for changes in its organic law, one in 2001 and another in 2004. In May 2007, after three years, the draft law was finally approved by the Council of Ministers and submitted to Parliament. The law is still waiting to be discussed and voted in the “especialidade� in the Parliament – the law was approved in the 43 Independent external entity with regard to the Executive Authority in accordance with its Creation Law dating from 1993. 44 “generalidade� in early 2008. Furthermore, for the norms to become effective after its approval, it will still be necessary to publish three complementary diplomas (acts): regulations relating to supporting services, safe boxes, and fees. 1.51 The possibility of approval of a new organic law has given rise to considerable expectation in TdC regarding the expansion of its attributions and reform of its administrative structure. The main changes included in the proposal relate to (a) introduction of flexibility in the ex-ante control; (b) adoption of economic, productivity, and convenience criteria in the control; (c) introduction of successive inspections for the whole state public sector; (d) categorization of financial irregularities; (e) restructuring of auditing reports; and (f) presentation of an assessment of the General State Accounts. Appointment of the president of TdC will remain the responsibility of the President of the Republic, based on proposals presented by the Council of Ministers. The president of TdC, as well as other judges, will have guaranteed security of tenure, except in special circumstances, with predetermined mandates of 5 years, renewable for another 5. The new organic law establishes the increase in the number of judges from the current 3 to 5. They are to be selected among the auditors of TdC, IGF inspectors, or judges of the Judicial Authority. 1.52 The new project of organic law establishes that the government will have 12 months after the end of the fiscal year to submit the State Accounts to the General Assembly. MFAP is responsible for preparing and sending the annual state accounts to the Prime Minister, who submits them to the National Assembly within the deadline specified above. The National Assembly will then send them to TdC within five business days. TdC will have three months to provide an assessment regarding legality and financial compliance, as well as recommendations to the National Assembly or government when necessary. 1.53 Even though the new legislation has not as yet been enacted, TdC is updating preparation and auditing of the General State Accounts (with the support of external technical assistance). TdC has made good progress in catching up with the backlog of state audits. In December 2007, TdC submitted 2001–05 accounts to the Parliament, thus becoming current on audit of accounts. However, this impressive achievement was possible only thanks to TA financed by the EU and the World Bank. The remaining challenge is the audit of the municipalities, for which TdC will need technical and financial assistance. While the clearance of the backlog requires extra resources, it is important that, in the future, TdC’s fulfillment of its mandate will not depend on the availability of donor financing but rather should be a priority in a budget that seeks out good governance. It should be noted that the legal framework for TdC is yet to be approved by Parliament. 1.54 TdC charges fees for all the processes assessed by the institution. TdC has a team of 17 technicians, with 4 in the area of previous control,44 5 who audit accounts, and 8 who inspect municipalities, embassies and consulates, autonomous institutions, and sovereign structures. The value of these fees varies in accordance with the type of process. The fee can go up to 2 percent 44 The ex-ante control, undertaken through a stamp imposition, covers staff recruitment, attribution of reforms and pensions, returns, contracts for the provision of services and goods, acquisitions above CVE 7.5 million, as well as public works for an estimated cost above CVE 4 million. Contracts financed by external resources, the domestic counterpart of which is below that value, are not objects of previous inspection by TdC. In 2006, more than 300 processes were refused and sent back, with a request for additional documents necessary for analysis by TdC. 45 of the value of the contracts, with a ceiling of CVE 100,000. These fees are deposited in a specific bank account, known as the TdC “safe box� and are not channeled through the state budget. Use of these resources throughout the fiscal year must be approved by the TdC plenary. Beginning in 2008, the fees are to be incorporated in the state budget, earmarked to TdC. 1.55 TdC’s cooperation with IGF improved over the past few years, with joint coordination meetings at the onset of each fiscal year. However, TdC still does not receive the IGF inspection reports, which present administrative irregularities, because IGF cannot send them directly to TdC. That is the responsibility of the Ministry of Finance, IGF’s supervising structure, to do so whenever it is considered relevant. 1.56 TdC representatives have reported that they have never felt under pressure to modify or fail to publish verdicts of guilt. There were already verdicts by authorities at the municipal level, but the last sentence against a minister was passed in the 1990s. This lack of sentences probably results from the fact that ministers are not required to report on their activities, even though they can be accused and condemned if denounced before TdC. TdC authorities consider that, given the municipalities’ autonomy and limited capacity of control by the IGF, the most important step to be taken in the institution would be to strengthen the analysis of the municipalities, given their autonomy and limited capacity of control by the IGF. 1.57 Formal independence of TdC from the Executive Authority is threatened by the lack of budgetary resources, which cover only staff remunerations and office rent. All travel costs have to be financed by external funds or through TdC fees. For TdC representatives, this reflects a government option that does not reflect the priority of TdC in the state budget, thereby limiting its administrative authority. 1.58 The new organic law extends TdC’s scope of action beyond the strict control of formal legality by including the criteria of economy, productivity, and convenience. TdC will start auditing results, verifying whether the beneficiaries of government programs are benefiting from them. For example, they will confirm with the students enrolled in public schools whether they are receiving the meals to which they are entitled (in terms of both quantity and quality) under the school nutrition program. 1.59 TdC receives international support in several ways. Besides grants from the World Bank for institutional capacity building and development, TdC receives support from audit courts in other countries, primarily Portugal. Brazil and France also make available places in training courses for auditors and technicians and promote visits by the Cape Verdean judges to their institutions. TdC also participates in the annual meetings of the International Organization of Supreme Audit Institutions (INTOSAI), which works to ensure access to internationally accepted auditing standards.45 1.60 A critical factor limiting the effectiveness of TdC is the recurrent absence of a representative from the Public Prosecutor’s Office (PGR) in TdC plenary sessions held on Thursdays. Despite continual requests by TdC, not only is there no PGR representative attending the meetings, but also PGR retains some processes for several years before issuing an 45 Cape Verde for many years has been on a list of defaulters to the institution, because the MFAP centralizes payments of fees to international organizations but does not make timely payments. 46 opinion. Delays in the investigations are the exclusive responsibility of the PGR, and in many cases this implies prescription of penalty, preventing application of the respective penalties. This impunity promotes frustration in the judges and technical staff of TdC. Recommendations • Decrease from one year to six months the time for the executive power to submit national accounts to the General Assembly. • Ensure that the judges, auditors, and technical staff of TdC have online access to SIGOF for consultation purposes. 47 Fiscal decentralization in Cape Verde status and perspectives46 STATUS OF DECENTRALIZATION Decentralization was launched in 1989. During the post-independence period, Cape Verde kept the structure inherited from the colonial era. Fifteen Councils (conselhos), were the centers of public activity outside the capital. Each council could include several parishes, themselves divided into various “areas� (several villages or hamlets). In each island, a delegate appointed and accountable to the central government represented the central government. In 1989, Law 47/III/89 created “decentralized territorial collectivities,� with a decree-law (Decret-Law 52- A/90) that stated the principles of administrative, financial, and patrimonial autonomy of the municipalities. With the move to multiparty democracy in 1991, the political decentralization option was strengthened47 and the first local elections held. The elected municipalities’ territories match the 15 former conselhos, to which 2 additional municipalities were added in 1991 and 5 more in 2005. According to their 1995 Statute, municipalities are defined as “territorial collective entities with representative bodies emanating from their respective population, and which pursue the interests of these populations.� The municipalities have administrative, financial, patrimonial, normative, and organizational autonomy. They can establish municipal taxes and fees, manage and part with their physical assets, issue licenses, manage public works and destroy public buildings, and attract investment. Furthermore, political decentralization in Cape Verde is embedded in the Constitution, which under the title of “Local Power� states the main principles of decentralization (box 4.1). The issues of solidarity among municipalities, regional imbalance, and risks of inequalities between islands and municipalities have remained persistent concerns in the country. Indeed, municipalities and their economic potential are vastly diverse (table 4.1). However, these inequalities are not accompanied by strong tensions among regions. Regional identities in the country are relatively strong (with the main distinction between the Barlavento and Sotavento islands, and some islands’ identities) but Cape Verdeans perceive themselves foremost as Cape Verdeans. Furthermore, the regional identities are not associated with any ethnic or religious divides.48 46 This chapter combines general considerations on the rules and their applications regarding municipal finances, and examples from data and information collected during the field visit in 5 municipalities in November 2007: 2 “small� municipalities, Mosteiros (Fogo Island) and Santo Domingos (Santiago Islands); 1 medium-sized municipality, Santa Cruz (Santiago Island); and the two largest municipalities in the country, Praia (Santiago) and São Vicente (São Vicente). Since only data on central government transfers are readily available for all municipalities, it was not possible to give a consolidated picture of municipal finances as a whole. A rather comprehensive but now outdated study on municipal finances from 1996 to 2002 is the only document that attempted to synthesize and collect municipal budgets so far). However, the five municipalities visited represent a wide variety of contexts. They also contain more than half of the total population. 47 Decret-Law no 122/91. 48 “Afrobarometer Report� ( 2005). 48 Box 0.1: “Local Power� in the Cape Verdean Constitution Central administration: • Recognizes the existence of elected representative local governments, specifically municipalities (Inframunicipal or supramunicipal entities may also be created). • Ensures solidarity among local governments. • Grants technical, material, and human support to local governments. • Controls (tutela49), but the control is restricted to legality (that is, the control only checks that legal and regulatory rules are complied with. It does not check the opportunity of the decision. Municipalities: • Have their own finance and assets, and fiscal powers. • Have an elected assembly and a collegial executive accountable to the assembly. • May delegate powers to community organizations. RESPONSIBILITIES OF THE MUNICIPALITIES Following the first local elections, various mandates were specifically transferred to the municipalities through protocols or decrees. These were public works (through various protocols), Retail Trade (Decree Law 31/92), Collective Road Transports (Decree Law 68/94), Social Promotion (Decree Law 12/94) and Water Supply. The 1995 Statute of Municipalities describes in detail municipalities’ responsibilities.50 These include local planning, basic sanitation, rural development, primary health care, preschool and primary education, urbanism, road transportation, social promotion, culture, sports, tourism, environment, local trade, civil protection, public works, professional training, local security, and municipal investments. Furthermore, the law specifically states that attributions not legally pertaining to the central government belong to the municipalities, which is a very bold statement in favor of decentralization. Transitional rules for implementation specify that attributions will be delegated progressively. Municipalities have become active in their areas of responsibilities, but organization of the transfer has been sketchy at best. Weaknesses of the protocols signed between the central government and the municipalities for transferring some of these responsibilities are well documented. Most of the devolution protocols left gray areas. For instance, the decentralization of social promotion did not clarify the responsibility for managing social equipments. In addition, the process was somewhat improvised: lack of planning and evaluation of needed resources and staff, lack of organization of coordination of activities and phasing of the transfers). 49 Tutela is the control means of the government (or its delegates) over its administration. 50 Law No. 134/IV/95, July 3, 1995. 49 Table 0.1: Diversity of Municipalities Poverty rate Vaccina- ISLAND and Total pop. Pop.<=17 Infant head family Female Male tion rate Municipality (2007) (2007) mortality (2001-2) <= 1 yr. BOA VISTA––Boa Vista 5.621 1.973 12 24 8 30.6 98 BRAVA- Brava 6.333 2.783 22 29 18 22.5 70.8 FOGO ISLAND 38.192 31 48 22 29.5 Fogo––Santa Catarina 4.874 2.417 Fogo––Mosteiros 9.853 4.638 16.8 53.6 Fogo––S. Filipe 23.465 11.071 28.3 MAIO––Maio 7.868 3.626 20 26 15 18.1 60.6 SAO NICOLAU ISLAND 13.192 23 34 15 17.6 79.2 S. Nicolau––Tarrafal 5.007 2.205 S. Nicolau––Rª Brava 8.185 3.31 SAO VICENTE––S. Vicente 77.535 30.088 21 33 13 22.2 73.4 SAL- Sal 18.966 7.161 10 24 5 20.1 92 SANTIAGO ISLAND 279.449 29 42 18 Santiago––São Lorenço dos Orgões 8.903 4.48 Santiago––RªG. Santiago 9.27 4.977 Santiago––São Salvador do Mundo 10.507 5.529 Santiago––S. Domingos 14.118 7.084 2.2 37.3 Santiago––São Miguel 17.339 8.937 7.2 32.2 Santiago––Tarrafal 22.078 11.098 6.3 57 Santiago––Santa Cruz 28.814 15.287 7.5 58.3 Santiago––Santa Catarina 46.576 23.457 12.3 53.9 Santiago––Praia 121.844 52.279 35.5 69.6 SANTO ANTAO ISLAND 49.163 34 51 26 Santo Antão––Paúl 8.782 3.935 0 90.3 Santo Antão––Porto Novo 18.393 8.291 8.6 70.3 Santo Antão––Rª Grande 21.988 9.425 23.8 63.3 TOTAL 496.319 224.051 27 40 18 20.2 63.1 Sources: Population per municipality and youth population DGAT, Poverty rate PRSP, infant mortality and vaccination health statistics 2005. NEW DRAFT LAW ON DECENTRALIZATION Municipalities today are an evolution of post-independence conselhos. After independence, Cape Verde had 15 conselhos. At the time of the first local elections, conselhos became municipalities but 2 more were created to take into account their large population (division of the conselho of Praia in 2 municipalities and creation of a distinct municipality in the northern part of the Fogo Island). Only 2 municipalities in the country have more than 50,000 people. This is obviously linked to the relatively low population (less than half a million people), but also to the geographic specificity of the country: there is at least one municipality per island, whatever its size. 50 Over the past decade, the issue of whether Cape Verde should create more levels of local governments or create more municipalities came to the forefront of political debate. Local factors and international examples fed the debate on regionalization. Local factors included the desire to open up new opportunities for elected positions and possibly local employment as well as existing regional cultural identities. The debate on the creation of more municipalities was driven by the population of some of the subentities of large municipalities, which felt somewhat “left out�––be it from the political debates or the concrete municipal implementation. The creation of a new municipality was perceived as a solution to this problem. In fact, according to Afrobarometer 2005, 67 percent of Cape Verdeans were in favor of creating more municipalities, with some differences among islands. Only 48 percent of respondents from Sao Vicente agreed (Sao Vicente has only 1 municipality, and its population is relatively concentrated). In contrast, the proportion reached 75 percent for respondents from the interior of Santiago––outside Praia–– and the Island of Fogo. Five new municipalities were created in 2005. However, the debates on regionalization or creation of new municipalities seem to have come to an end in 2007. Following broad-based discussions with political parties, municipalities, civil society, and decentralization experts, the government confirmed that the creation of a regional local government level will not take place. Instead, efforts will be made to strengthen administrative regions as well as intermunicipal cooperation. This is clarified in the framework Law on Decentralization, which has been approved in the Council of Ministers and submitted to Parliament. The new draft framework law on decentralization puts an end to a looming debate on whether there should be more than one level of local government, and attempts to give clear guidance on the creation of new municipalities and the organization of future transfers. It includes the following innovations and clarifications: • Regarding the municipalities: (a) affirmation of the principle that there is only one level of local governments the municipality; (b) creation of new municipalities is conditional on a feasibility study; (c) reaffirmation of the broad area of responsibilities of the municipalities, but with more specificity about these that are managed “in coordination with central authorities�; (iv) rules for concluding public-private partnerships; and (v) dispositions to ensure dissemination of information to the public (via internet) and participation of the public in policy design and evaluation, as well as the possibility of “class actions� or “popular initiatives.� • Regarding the transfer or delegation of responsibilities: The draft law attempts to avoid the pitfalls noted above by proposing a roadmap and a pilot phase for all future transfers. It details the rules for delegating, temporarily or definitely, new responsibilities to the municipalities. The central government has a duty to transfer responsibilities to municipalities, because they are closer to the population, for social and economic development and collective needs of the population. These transfers can be made for all municipalities, or depending on their socio-economic and capacity status, decentralized asymmetrically. All transfers of responsibilities are concomitant with transfers of resources. Transfer of new responsibilities is done for an experimental phase of 3 years that can be prorogated by 2 more years. Transfer is done according to a negotiated Convention that indicates financial resources transferred and performance indicators, and includes an annex related to a capacity building program. Control of the transfer 51 processes and results are the responsibility of both the central government and the municipalities, and are subject to audits by private companies. The government will also take stock of past transfers and propose remedies if needed. Local governance seems to be well entrenched and relatively strong. Local elections took place in Cape Verde in 1991, 1996, 2000 and 2004, and are planned for 2008. They are the subject of intense debates. Elections results reflect the bipartisanship of the political system, although some municipalities have elected “independent� candidates. According to the 2005 Afrobarometer survey, local governance is rated positively by the population. Of all elected political actors, the Presidentes das Camaras (the mayors) receive the highest satisfaction rate. On average, 56 percent of respondents approved the way their mayor had performed in the past 12 months. Only 24 percent disapproved their mayor’s performance. Furthermore, the commitment of central government to decentralization is taken for granted by all actors, and indeed, does not seem to be an issue as often is the case in countries newly decentralized. RESOURCES OF THE MUNICIPALITIES This subsection draws on data collected in a sample of 5 municipalities, even though comprehensive data are available for only 4 of them––Praia being the notable exception. Indeed, although Praia, the largest city and the capital of the country, is included in the sample, it did not transmit much data on its budget. This is a source of concern, since among all municipalities, it is the one that would be expected to have the highest capacity to produce timely and comprehensive financial statements, and it also is the place in which one-quarter of the population lives. Furthermore, the self-reporting of this subsample of municipalities does not match data collected at the center. For instance, all municipalities have received “Contracts- Programs� allocation in 2005 per the transfers data obtained at the Ministry of Finance, but only one of them reported such transfers. This could be a matter of correctly labeling these funds but, nonetheless, is another source of concern since this kind of discrepancy in reporting should not happen. Local taxes do not represent a significant proportion of municipal revenues, except for the two largest cities. Only Praia and Sao Vicente, the largest municipalities in the country, have a share of local taxes in their budgets above 10 percent. Annex 2 illustrates the main sources of resources for the five municipalities studied. Data are from 2005 except for Praia, 2004. Table 0.2: Source of Revenues (% of total, as reported by municipalities) Mosteiros Praia S. Cruz S. Domingo S. Vicente Property tax 3.21 10.71 2.67 1.28 15.89 Other revenues 2.47 10.68 12.57 7.33 7.87 Contracts-Programs 23.43 0.00 0.00 0.00 0.00 Loans 4.40 0.00 0.00 0.00 10.43 Source: All data from 2005 budget except Praia (2004). 52 Central government transfers represent more than 50 percent of municipalities’ revenues, especially for the smaller municipalities.51 (figure 4.1). This means that a large proportion of local governments are heavily dependent on the central government for their resources, which limits their autonomy. The situation also is highly contrasted regarding the source of the central government transfers. In the authors’ sample, only one municipality benefited in 2005 from “contracts-programs,� as opposed to the formula-based transfers. The “contracts-programs� are ad hoc and based on an agreement between one or more municipality and the central government to achieve specific objectives. Figure 0.1: Resource Composition, 2005 (2004 for Praia) (%) Mosteiros S.Domingo S.Cruz S.Vicente Praia 2004 0 20 40 60 80 100 Own ressources State Transfers Loans Beyond these two main categories (own revenues and central government transfers), land sales appear as a significant revenue source for 3 of the municipalities surveyed––including the two largest (figure 4.2). This dependence on land sales is a source of concern (see below), since land is not a sustainable source of revenue. 51 There are exceptions, however. For example in the case of São Vicente central government transfers represent 15 percent of the resources. 53 Figure 0.2: Transfers or Land Sales (CVE), 2005 (2004 for Praia) 700000000 600000000 500000000 400000000 State Transfers Sale of land 300000000 Total 200000000 Total State Transfers 100000000 Sale of land 0 Praia S.Vicente S.Cruz S.Domingo Mosteiros The 2005 local finance law52 increased the source of internal revenues for municipalities from approximately 33 types of taxes and duties to approximately 45. Internal revenues include a broad range of local taxes, collected directly by the municipalities. It is important to note a distinction in Cape Verdean law between impostos and taxas. Impostos are created by the National Assembly and cannot be created by municipalities. Taxas can be created by both the National and municipal assemblies. Table 0.3: Principal Internal Sources of Revenues Direct taxes − Fees for getting various statements, such as − Single tax on property proof of residence − Municipal tax on vehicles − Publicity − Indirect taxes − Sale of services − Commercial licensing − Location of cars and machines − Road transport license − Collective transportation services − Markets and fairs fees − Electricity − Fees for checking accuracy of − Vehicle inspections? measurement tools − Property location maps − Secretary services − Leisure and cultural services − Various taxes and other − Other recurrent resources penalties − Budget balance − Public works − Capital resources − Sanitation and hygiene − Sale of land − Kindergarten quotas − Sale of machine and equipment − School transportation − Central government revenues collected by − Fines municipalities Source: Law 79/VI/2005. Following an agreement concluded in 1998, municipalities in Cape Verde collect their own taxes––contrary to many countries in which local taxes are still collected centrally, deconcentrated by the administration, and then transferred back to municipalities. The decision 52 Law N° 79/VI/2005. 54 to delegate tax collection to municipalities addressed to some extent the municipalities’ request, and has the advantage of avoiding delays and potential conflicts between the center and local governments regarding the amounts effectively perceived and reversed. Furthermore, delegation can also act as a strong incentive for municipal services to strengthen their tax collection capacity. In fact, even though municipalities could delegate tax collection to the central government for a fee of up to 5 percent of the collected taxes,53 no municipality uses this option. The delegation of tax transfers also had some downsides, and has, at least initially, created difficulties, especially for the smallest municipalities. The main difficulties arose from the lack of preparation for this delegation and the fact that the voluntary transfer of tax collection personnel to municipalities was not successful (most of the staff chose to go back to the central ministry instead of becoming municipal employees). The latter left municipalities without skilled staff to exercise their new mandate. Another important issue was linked to a structural constraint. Most land registries, at least at the time of the transfer of tax collection, were outdated. Consequently, there was no clear rule for evaluating the value of the existing patrimony, thus making the patrimony tax, an important potential source of revenue, difficult to collect. To date, the state of the land registry or the taxpayers registries vary greatly from one municipality to the next, thus increasing the discrepancies between their abilities to raise their own revenues. Municipalities’ efforts to strengthen their fiscal units and computerization can substantially raise their own revenues. The example of the municipality of Sao Vicente shows that, at least for some municipalities (Sao Vicente has strong economic activity), this is indeed so. The municipality tripled its tax income since 2001. Land sales as a source of municipal financing Land sales represent an important source of revenues for several municipalities.54 International experience shows that municipal land sales provide one option for financing urban infrastructure investment. In countries in which land is owned by the public sector, land is by far the most valuable asset on the municipal balance sheet. Selling land or long-term leasing rights to land use while investing the proceeds in infrastructure facilities can be viewed as a type of portfolio asset adjustment. Various countries have turned to land sales and leasing to finance infrastructure. From a local perspective, land sales present advantages. They typically are free from the intergovernmental restrictions that require higher-level approval for increases in local tax rates or user fees and that restrict local government borrowing. However, financing municipal infrastructure investment through land sales creates special risks that are not recognized in most intergovernmental fiscal frameworks. Three main risks emerge: (a) the use of proceeds to finance operating budgets; (b) risk exposure is exaggerated by the highly volatile nature of urban land markets; and (c) the greatest financial sector risk stems from municipal borrowing based on inflated land values offered as collateral to banks. 53 Local Finance Law No. 79/VI/2005. 54 This section is based on G.E. Peterson, Land Leasing and Land Sale as an Infrastructure-Financing Option, World Bank Policy Research Working Paper 4043 (November 2006). 55 To avoid excessive risk-taking by local governments, sound intergovernmental fiscal management will require tighter regulation of municipalities’ financial leveraging of land assets. The authors’ fieldwork in Cape Verde did not show evidence that land is used as collateral for loans. However, it seems that the practice is to sell, not to lease, land. Since there are no regulations on the use of the proceeds of land sales, municipalities use at least some of them to finance operating budgets. Another issue is the sale of land that has no clear legal status. The central government considers that it is then the property of the country, whereas the municipalities consider it their property. This ambiguity led to some sharing agreements whereby a percentage of the sale goes to the center and a percentage goes to the municipalities. Nevertheless, some cases led to serious disputes. Revenue projections Optimistic revenue projections are the rule and can create problems in the execution phase. In the sample, most municipalities were overly optimistic in their planning phase regarding the amounts of resources, they actually would get (table 4.4), thus creating problems in the execution phase. Table 0.4: Resources Available as a % of Resources Expected at the Planning Stage (%) 2004 2005 2006 Mosteiros 67.60 68.91 94.63 Praia 52.09 N/A N/A Santa Cruz 40.18 43.68 48.69 São Domingos 56.19 49.48 53.92 São Vicente 62.95 69.15 115.72 The period 2004–06 provides evidence, with the exceptions being ad hoc rather than attributed to wise planning. In 2004 and 2005, most of the five municipalities of the sample showed a poor rate of tax collection compared to their previsions. For 2006, two municipalities stand out: Mosteiros’s resources were very close to the resources expected at the planning stage, and Sao Vicente’s resources were even higher. However, for both, the explanations are more ad hoc than a consequence of wise or conservative planning. Mosteiros benefited from (a) a unplanned transfer of an ecological tax by the government; (b) a higher balance of its budget from 2005; and (c) a more aggressive policy in selling land. Sao Vicente benefited from a strong increase in the Municipal Finance Fund (MFF) as well as proceeds from land sales. 56 Most municipalities explain the discrepancy between their resources and their planning stage expectations using the following reasons: • Lack of skilled personnel resulting in ambitious but unfounded targets and unrealistic projects. • Absence of an efficient tax collection system. • Uncertainty about the amounts of central government transfers from one year to the next. Central government transfers Two types of central government transfers coexist in Cape Verde. The first one, a formula- based transfer without ties, was created by the 1998 Local Finance law and modified by the 2005 Local Finance law. The second one, called “contracts-programs,� is a contract between the central government and one or more municipalities to provide resources for a specific goal, possibly over several years. The contracts-programs suffer from various weaknesses and need to be reformed. The Municipal Finance Fund (MFF) is a functioning formula-based transfer. The bold decision to create a formula-based transfer guaranteed to the municipalities a minimum of predictable, non-earmarked resources. Historically, especially for the municipalities that are either small or have low resource potential, this formula-based transfer has been the main source of revenues. The first formula, the Financial Equilibrium Fund (FEF), was in place until 2005.55 The FEF amount could not exceed 7 percent of total tax income (indirect and direct taxes, with the exception of the tourism tax, the ecological tax, and the income generated by airport and port services) generated in the preceding year. Criteria for fund allocation were dual: (a) criteria linked to the population and the geographic area of the municipality––the larger these factors, the higher the allocation; and (b) criteria linked to tax collection capacity in the municipalities, that is, the allocation was inversely correlated to the amount of taxes collected. The fewer taxes collected, the higher the allocation. The main weaknesses of the system became obvious within a few years. Municipalities criticized the ceiling of only 7 percent of total tax incomes. They also criticized the fact that the FEF allocations, transferred each month in equal amounts, were sometimes transferred late. However, more fundamentally, the formula was blamed for two major flaws: (1) poverty levels were not taken into account in the formula, which thus had no criterion to improve equity among municipalities; and (2) the inverse relationship between tax collection amounts and level of allocation became a disincentive for municipalities to improve their tax collection system, especially if their expected revenues were low. Consequently, the 2005 Local Finance law proposed a new formula (MFF), which succeeded in addressing past criticisms. The MFF amounts to 10 percent of all direct and indirect taxes collected the year before the last at the time of budget preparation. This amount is then split between two “funds�: the Common Municipal Fund or CMF (75 percent of the MFF), 55 Regulated by a Decree from December 13, 1996. 57 and the Municipal Solidarity Fund, or MSF (25 percent of the MFF). The CMF is distributed between municipalities according to the following formula: 20 percent distributed equally, 50 percent according to the resident population of the municipality, 15 percent according to the youth population, and 15 percent according to the geographic area of the municipality. The MSF is calculated so that the poorest municipalities get a bigger share of the transfers. Only the municipalities that have a level of municipal tax collection per capita lower than the national average and a proportion of their population below the poverty line higher than the national average are eligible for that part of the MFF. The allocation takes into account both factors to determine respective amounts due to each eligible municipality. The amounts of the formula-based transfers have drastically increased with the 2005 Local Finances Law. This occurred partly through a more generous calculation (higher share and broader scope of government’s internal revenues going to the municipalities), and partly because fiscal revenues in the country also increased. For all but one municipality, the amounts have increased, oftentimes substantially (annex 5, first series of figures with amounts of the formula- based transfers since 1993). In terms of distribution among municipalities, the picture varies based on the basis. The new Local Finances Law (2005) led to a higher share of resources being transferred to larger municipalities. However, per capita, the situation is more complex, and, generally, the smaller municipalities fare relatively well. The formula-based transfers are generally timely and their amounts seem to be a little bit more predictable than were those of the FEF. The MFF amounts are transferred monthly (1/12th each month), and only 1 municipality complained of delays in the transfers. Since the bulk of the transfers no longer depend on the local tax collection capacity, amounts seem to be more stable––although they will vary depending on the change in the country’s overall tax collection from one year to the next. This also happens because some of the indicators used for the formula are less often computed. However, the share of the formula-based transfer designed as a “solidarity� mechanism does take into account tax collection and thus could continue to create a disincentive for some municipalities to improve their tax revenues. The new MFF seems to be satisfactory for its principal beneficiaries, but over time adjustments may be needed. The government’s commitment in transferring non-earmarked resources to the municipalities was clear in the change of formula, which significantly raised the amounts transferred. However, the formula is relatively complex, which may make its yearly adjustments difficult. Hence, MFF amounts seem to have stayed identical in 2006 and 2007. Furthermore, it might be useful in the next iteration of the formula to build in a mechanism to encourage own revenue collection. As with most formulas, experience will show whether this one is the most appropriate for the country. “Contracts-programs� “Contracts-programs� are a much more traditional and less transparent instrument than the FFM. Contracts-programs were created as a way to decentralize the Multiannual Program of Public Investments. The former are signed between one or more municipalities and the central 58 government, around specified objectives. The contracts-programs are then transferred in 3 tranches: 30 percent, 30 percent, and 40 percent of the total. The contracts-programs are criticized because of the lack of clear and well defined criteria for their use . The Local Finances Law states that the government regulates conditions of use and access for contracts-programs. However, the government has not regulated them across the board, but only ad hoc. The fact that they are the product of negotiations between parties gives no predictability to this source of transfer, which can be an important part of some municipalities’ budgets. The draft law on decentralization details the types of “conventions� that will be signed for future transfers, but it is not clear whether these criteria will apply to all new contracts- programs. Some clarity will be needed to first move away from the perception that contracts- programs are politically used, and second give municipalities more predictability for planning. Beyond these needed clarifications, municipalities should be encouraged to develop public- private partnerships to maintain, upgrade, and extend their infrastructure. Past evidence gives some credit to the criticism toward partiality. Table 4.5 shows the respective shares of the formula-based transfer and the contracts-programs allocated to each municipality in 2005 and 2006. Since contracts-programs are not allocated on the basis of a formula, it is not surprising that these shares differ. However, it is interesting to note that some municipalities seems to benefit more from these transfers over these two years, which can give some credit to the criticism of their partiality. This assessment is nuanced by the fact that only two years are being considered in these tables. Table 0.5: Comparison Between Share of Formula-based Transfers and Share of Contracts-programs Amounts Received by Each Municipality, 2005–06 (%) 2006 (Jan.-Oct.) Share FEF Share CP Difference CP/ FEF Boa Vista 8.66 9.02 0.60 Brava 11.33 17.86 1.96 Maio 6.71 10.78 0.63 Mosteiros 13.02 5.36 7.88 Paúl 5.57 3.43 0.38 Porto Novo 3.08 10.96 -2.56 Praia 5.52 3.43 -7.66 Ribeira Grande 3.54 3.81 -1.89 S. Vicente 3.05 3.43 -4.61 Sal 6.36 3.80 0.27 Santa Catarina 9.92 5.31 6.54 Santa Cruz 6.08 4.19 0.37 São Domingos 2.83 3.43 -1.66 São Filipe 5.08 3.43 4.07 São Miguel 3.51 3.43 -2.09 São Nicolau 2.96 4.92 -0.08 Tarrafal 2.79 3.42 -2.15 2005 Share FEF Share CP Difference CP/ Formula 59 Boa Vista 4.33 5.66 -0.51 Brava 6.46 8.94 1.33 Maio 8.43 12.02 0.06 Mosteiros 7.37 4.10 3.32 Paúl 4.07 3.56 -0.56 Porto Novo 4.08 4.15 -1.84 Praia 10.07 14.47 4.40 Ribeira Grande 4.25 3.69 0.00 S. Vicente 6.72 3.94 -3.56 Sal 4.38 4.38 -0.25 Santa Catarina 6.35 4.51 -0.38 Santa Cruz 8.25 4.70 3.60 São Domingos 4.33 3.77 -3.27 São Filipe 7.65 7.27 2.48 São Miguel 4.32 7.64 -2.78 São Nicolau 5.41 3.93 -1.48 Tarrafal 3.53 3.28 -0.56 Other sources of revenues Borrowing The 2005 Local Finances law opened the possibility for municipalities to borrow, either to resolve short-term difficulties or to invest. The recourse to credit as a way to momentarily solve a problem of liquidity seems to be generalized. In this case, the law restricts the amounts of short-term borrowing to less than 10 percent of the revenues collected in the previous economic year. Borrowing for investment requires a formal government approval. Moreover, the amortizations and the interest of the loans cannot exceed 15 percent of municipal recurrent revenues or 25 percent of the value of the investments made for the municipality during the previous year. Finally, the 2005 Law allows municipalities to borrow for the medium- to long- term outside the country, provided the terms of borrowing meet certain conditions (interest rates and services offered by the external creditor must be superior to the internal market). Such borrowing seems to be used more and more by municipalities. Examples of expenditures financed by long-term borrowing in the municipalities visited are the construction of the municipality main office and a market (Santa Cruz); construction of the municipal delegation and municipality main office and acquisition of three vehicles (Sao Domingos); upgrading a sport complex (Sao Vicente); and purchasing a truck and a compressor (Mosteiros). Opening the possibilities for local governments to borrow creates a fiscal risk that the central government must minimize through proper oversight. It is important to ensure that the restrictions and conditions set by the regulations are indeed followed by the municipalities, which means that control institutions must be sufficiently endowed to perform these checks. “Decentralized cooperation� Proceeds from decentralized cooperation––twinning agreements with local governments from foreign countries––may represent an important part of municipalities’ budget. Cape 60 Verde has very strong ties to many foreign communities, and decentralized cooperation is one opportunity for the Cape Verdean municipalities to receive additional resources (in some cases, in-kind through equipment and technical assistance). Sister-city and cooperation agreements with municipalities or regions in other countries feed into the development budget and may represent an important part of these budgets. From a fiscal decentralization viewpoint, these external resources come with some caveats. (1) These external resources are not always properly accounted for in the budget, thus diminishing budget comprehensiveness. (2) There is an issue of communication with the central administration. Even though data are sketchy, the resources from decentralized cooperation seem to finance primarily capital expenditures. This is not a problem per se. However, newly built infrastructure may have an impact on the central government’s expected role once the infrastructure is built––providing teachers for a school for instance. The problem comes from the absence of a framework that could ensure a smooth coordination between the municipalities and the central and deconcentrated administration to ensure that such impacts are planned for. (3) These external resources can create dependency and economic vulnerabilities. A law is being drafted to clarify the modalities and exchanges of information that must accompany decentralized cooperation. In any case, all donor funds should be fully disclosed and recorded in municipal budgets to ensure budget comprehensiveness. Share of centrally collected taxes Another source of revenues for the municipalities comes from a few taxes, centrally collected, that are then reversed to them. However, these transfers suffer from delays. The most important of these shared taxes is the 50 percent share of the rent for the use of airport zones by the public enterprise, National Company of Airport Security. This tax has potential but so far does not seem to be duly channeled to the municipalities. The Ecological Tax also is transferred to the municipalities but not with enough regularity. To enable timely transfer of centrally collected taxes, clear legal agreements, technological capabilities, and judicious oversight must be in place. Another option would be to evaluate the possibility of the municipalities themselves collecting these taxes. Accountability of municipalities is constrained by two main factors. (1) The legal framework makes municipalities the principal actor for basic service delivery. However, few transfers took place, and it is thus unclear what to expect from a given municipality. Municipalities are legally entitled to engage in all mandates legally theirs, but for these which were not organized and planned, the situation remains vague. (2) Financial management suffers from serious weaknesses. Thus, it is difficult for the central government and citizens alike to get a clear picture of municipal finances. 61 MAIN EXPENDITURE ITEMS OF THE MUNICIPALITIES56 An unclear scope of mandates lowers accountability. The main weakness of decentralization implementation in Cape Verde is precisely that there is no clear framework that delineates the roles and responsibilities of the local and the central governments. This omission makes it difficult to define adequate levels of expenditure and revenue assignment. The municipalities are legally endowed with most basic service delivery responsibilities. However, the structure of the state and the flows of information and resources were not necessarily adapted to this new reality. Unclear reporting aggravates the problem. Another weakness regarding accountability is that, despite a set of rules that mandate detailed information on municipal finances, most municipalities do not present their budgets in a form that enables scrutiny. The analysis of the budget breakdown for four municipalities exemplifies this lack of transparency (table 4.6). It shows that municipalities are involved primarily in construction–– roads, electrification, and social housing) but that their activities span various areas. However, table 4.6 also shows the limit of the existing budget classification––or rather the limits of the municipality’s capacity to implement it (see section on financial management). The budget combines line items and “programs� items, which precludes the possibility of getting a comprehensive view of total expenditure at the sectoral level. For instance, it is not possible to identify staff costs per sector, or to identify which sector benefited from the uses of “machines and equipment.� Table 0.6: Principal Budget Items (total costs in 2005) S. Vicente S. Cruz Mosteiros S. Domingos SALARY AND STAFF COSTS 37 54 29 58 CONSTRUCTION including: 29 6 33 9 Roads 8 0 27 2 Rural electrification 1 1 3 0 Sport infrastructure 13 0 3 3 Social housing 4 2 0 0 Diverse construction and rehabilitation 3 3 0 0 Others 4 SOCIAL AFFAIRS (sport, culture, recuperacao de habitacao, festivals.) 8 6 4 9 MACHINES and EQUIPMENT 4 6 4 4 PETROL 3 3 6 6 SUBSIDIES FOR WATER SUPPLY 2 3 0 0 Subtotal of main uses of the budget 83 78 76 86 Other uses of the budget 17 22 24 14 Total 100 100 100 100 In the sample, recurrent expenditures constitute the bulk of municipal expenditures, leaving few resources for investments (figure 4.3). In fact, personnel costs represent 29 percent 56 Of the municipalities in this sample. 62 to 58 percent of the overall budget. There are no mandatory expenditures in the country (except for the amortization of previous loans), nor any provision for a compulsory ratio between recurrent and capital expenditures. Figure 0.3: Share of Capital and Recurrent Expenditures in 2005 Budget Distribution between capital and recurrent expenditures 100 90 80 70 57 66 60 86 84 50 40 30 20 43 34 10 16 14 0 S.Vicente S. Cruz Mosteiros S.Domingo Capital Capital Recurrent expenditures Figure 0.4: Share ofexpendituresand Recurrent Expenditures (staff and non staff) in 2005 budget (%) Distribution between capital and recurrent expenditures (% total, 2005) 1.2 1 0.363 0.2964 0.8 0.5418 0.5796 0.6 0.2736 0.297 0.4 0.3182 0.2604 0.2 0.43 0.34 0.14 0.16 0 S.Vicente S. Cruz Mosteiros S.Domingo Capital expenditures Recurrent expenditures- non staff Recurrent expenditures- staff BUDGET PREPARATION AND EXECUTION Since the adoption of the 2005 Local Finances law, municipal budget preparation and execution should follow the same standardized classifications as that of the central government (economic, organic and functional classifications). However, implementation is lacking. Past municipal budget forms were outdated and lacked standardization, which made budget monitoring difficult. However, few municipalities seem to be able to complete and make 63 available the full range of financial tables and annexes requested by the Local Finances Law. In fact, as shown above, the budget data collected in the field also show that the presentation of the budget itself does not follow, as mandated, an economic, organic, and functional classification. This consistency with the central budget should be a priority and will require capacity building. It could also be useful to select a subset of the most essential tables identified by the law and ensure that at least these are produced annually by all municipalities. 64 Table 0.7: Budget Tables and Information Annex to be Prepared by Municipalities According to 2005 Local Finances Law Budget tables Information annexes I––Recurrent and capital revenues of the Municipality, specified according to an economic and organic Priorities and means for the fiscal policy and the classification. municipal policies for expenditures and public debts. II––Investment and operational expenditures of the Human resources management policy, in particular Municipality, specified according to an economic and training and recruitment of staff. organic classification. Evolution, in last the three years, of the status of the III––Investment and operational expenditures of the municipal public debt and its structure and composition, Municipality, specified according to a functional indicating its variation and its impact for the budget. classification. Information on treasury operations and banking IV––Revenues of the municipal autonomous services, accounts of the Municipality, with explanation on according to an organic and economic classification. account balances. V––Expenditures of the municipal autonomous Maps of evolution of the execution of revenues and services, according to a economic and organic expenditures of the Municipality in last the three years, classification. in accordance with the structure foreseen in n.º 1 of the article 36º of the 2005 law, and analysis compared to VI––Expenditures of the municipal autonomous planning stage. services, specified according to a functional classification. Earmarked revenues, with the indication of the respective shares in investment and recurrent costs VII––Consolidated budget of recurrent and capital Information on staff, forecasts for additional staff revenues and of the operational costs of the expenditures, and indication of provisional budget set Municipality and the municipal autonomous services, aside for this purpose, in accordance with the b) n.º 1 according to an economic classification. of the article 32º of the 2005 law. VIII––Consolidated budget of recurrent and capital Financial status of all municipal autonomous services. revenues and of the operational costs of the Municipality and the municipal autonomous services, The following reports must also be sent: according to an organic classification. Ways of financing a possible budget deficit and the IX––Consolidated budget of recurrent and capital amortizations. revenues and of the operational costs of the Municipality and the municipal autonomous services, Justification of the forecasts for tax revenues following a functional classification. distinguishing the status of the main taxes. X––Programs of Municipal Public Investments, structured by program, subprograms and projects. XI––Summary of the fiscal operations of the Municipality specifying the balances and the nature of its financing. Budget preparation Budget is prepared mostly incrementally on an annual basis, with steps toward program budgeting subject to constraints. The central government is very progressively moving toward program budgeting, but this shift is likely to be even slower at the local level. First, uncertainties 65 on the levels of revenues available to municipalities from one year to the next would be a major impediment to multi-annual projections. Second, municipality staffing would need to be substantially adjusted to ensure that the necessary skills are present for this more complex budget process. The municipal assembly approves the budget. There is no other approval process, since the Tutelle is restricted to a legality control. Furthermore, the autonomy of municipalities in using their budget is very high. They have are no compulsory expenditures (except for loan amortizations), and no other form of constraints. For instance, there are no compulsory ratios between capital and recurrent expenditures. In the five municipalities visited, aggregate expenditure outturn compared to original approved budget is low. Municipalities fare badly regarding aggregate expenditure outturn compared to original approved budget (table 4.8). The fact that this situation has not shown much or any improvement over the years (except for the exception of Sao Vicente for 2006) is a cause of concern. Obviously, municipalities need to strengthen their budget preparation process by analyzing past years’ discrepancies and drawing lessons from them. The poor execution rate can have several explanations: • A poor planning process, with overambitious plans for the coming years not based on a rigorous evaluation of past years’ budgets. • A poor estimation of upcoming resources, due either to the uncertainty of some central government transfers or to the difficulty of estimating future tax revenues. Most municipalities prepare their budgets by estimating either the same value or a small increase for the forthcoming central government transfers, and adjusting later according to the actual allocations. • Poor implementation capacity. Table 0.8: Variation Between Approved Budget and Executed Budget 2004 2005 2006 Mosteiros 29 49 55 Praia 75 N/A N/A Santa Cruz 56 57 62 São Domingos *45 51 46 Note: *For Sao Domingos, 2004 data are from 2003. The composition of expenditure outturn is consistent in percentage to the originally approved budget. The discrepancy between the voted and executed budget is not matched by a discrepancy in the composition of executed expenditures. If they are not able to implement all planned activities, municipalities seem to follow the priorities chosen in the voted budget and the shares of the budget allocated per sector/function are the same in the voted and executed budget. Municipal budgets are not comprehensive. As indicated in section II, one weakness related to budget comprehensiveness is linked to decentralized cooperation, which is not always in cash or on budget. This is a serious issue that also is difficult to tackle, since municipalities should not be prevented from having an active policy of attracting such cooperation. However, without curbing their initiatives or their autonomy, the central government must be able to know the flows of 66 resources coming into their budgets. In addition, their constituents must be given the means to follow the full amounts of resources available to their municipality. The new draft law on decentralized cooperation attempts to address these issues. Budget execution Municipalities must prepare quarterly budgets showing their budget execution, as well as the “management account� (the annual executed budget). Various forms and annexes must be attached to these. The 2005 Local Finances Law specifies the timeframe for the “management accounts� presentation and approval process (table 4.9) Table 0.9: Milestones in Budget Preparation and Control Timing Activity Before July 31st Municipality’s executive appreciates the proposal of municipal budget elaborated by the mayor for the following economic year. Before August 25th Mayor presents the proposed municipal budget to the Municipal Assembly. Before September Municipal Assembly approves the budget. 20th Before 1st of October President of the Municipal Assembly sends the approved municipal budget to the Minister of Finance. Before December President of the Municipal Assembly publishes the budget. 31st Before March 1st Mayor submits management accounts of the previous year to the Municipal Assembly. Before end-March Municipal executive approves the management accounts. In April Municipal Assembly approves the management accounts. Before end-June Management accounts have to be submitted the TcC. The procedures of municipal procurement are not clear. Municipalities indicated that they follow the central government’s rules, which were changed by a 2007 law. It is possible and even probable that at least some of the municipalities have not yet mastered the provisions of this new law. Regarding the transparency of procurement contracts, the field visit was rather alarming: no municipality could give any statistics on the procedures it used per type or volume of contracts. The most common procurement method used was said to be “three quotes.� This is the basic process that, depending on the amounts involved, may not provide all guarantees that the procurement led to the best outcome for the municipality. None of the municipalities visited has an internal audit unit, which would be in charge of ensuring that rules and processes are regularly complied with. Furthermore, none has an internal audit manual that could recapitulate specific rules for a professional management of financial, human, and assets resources. The daily management of transactions is done by the “Municipal Secretary,� assisted by an accountant. One aspect of municipal finances that is prone to suspicion is the use of the “safe.� Municipalities have bank accounts in commercial banks in which they deposit their internal and external revenues. However, for revenues collected day to day, and depending on their volume, municipalities are entitled to deposit them in a safe for up to 24 hours, or to keep them to cover 67 small expenditures or until they reach a certain ceiling. Then they should be transferred to the bank account. In light of the poor level of internal and external control, it is possible that rules regulating the use of these resources are not always followed. FISCAL BALANCE AND CREDIBILITY OF THE BUDGET All municipalities visited show a level of revenues slightly above the level of expenditures, which seems to show that they all respect fiscal discipline. However, the reality is more complex. Except for Sao Vicente, where the introduction of IT enabled rigorous monitoring of municipal resources and expenditures, there is no clear system for monitoring payment delays. Most municipalities thus have debts and under-reported contingent liabilities, which are either accounted for in previous budget years or pending because a negotiation process in ongoing. A further cause of concern is that the field visit did not obtain relevant data from the sources on this issue (see section below on intergovernmental debts and arrears). However, this lack of success may also be linked to the brevity of the field visit. Figure 0.5: Comparison, Resources, and Expenditures: Selected Municipalities, 2004-06 Comparison, Ressources and Expenditures 2004-06, selected municipalities 800 700 Praia 600 S.Vicente 500 400 300 S. Cruz S.Domingo 200 Mosteiros 100 0 4 5 6 4 5 6 4 5 6 4 5 6 '0 '0 '0 '0 '0 '0 '0 '0 '0 '0 '0 '0 a go go s s s e e e uz uz uz ir o ir o ir o ai nt nt nt in in Cr Cr Cr Pr ce ce ce te te te om om S. S. S. os os os Vi Vi Vi D D S. S. S. M M M S. S. Ressources Expenditures Findings of the 2004 report on contingent liabilities A 2004 report on contingent liabilities attempted to take stock of intergovernmental debts in the country and illustrates suboptimal communication with central government and poor financial management: intergovernmental debts and municipal arrears. Annex 7 includes the synthesis tables supplied in the report, but it should be noted that the numbers were to a large extent estimates due to the data limitations. In this report, the central government was a net debitor toward the municipalities as a whole, but 5 municipalities (of 17 at the time) were net debitors toward the central government. 68 Table 4.10 reviews the 2004 report’s main conclusions regarding the sources of the debts, and examines whether these potential sources of debts are still relevant today. Table 0.10: Sources of Debt Status in 2004 Comments as of today Intergovernmental debts from the central government toward the municipalities Non- or delayed reversion of taxes Ecological Data are available from 2000–04. The 2005 and 2006 syntheses of tax Only in 2001 were any amounts transfers made to municipalities transferred to municipalities for transmitted by the Ministry of this tax, which should have Finance show that various integrally benefited them. No data amounts were transferred for were available for the previous previous years for each of these 3 years but a similar problem may taxes, seeming to indicate a catch- have occurred. up or a partial catch-up on Tourism tax Same source. arrears. Fire tax No transfer recorded. Since the tourism tax is now extinct, and the fire tax is part of a single tax on results, these two taxes are no longer sources of new arrears. Reduced amounts of formula-based transfer (FEF) transferred to some municipalities FEF To compensate the debt of some FEF is now transferred integrally municipalities toward the center, and timely. the State decided to transfer However, the issue of whether the reduced FEF amounts to 9 FEF could be used as an municipalities as a instrument to constrain sanction/compensation. The municipalities in paying back legality of this measure was always their arrears is under discussion, contested. and the government has indicated that it wants to propose a legal amendment to allow this. Other debts Delays in The report gave mitigated There was no mention of the issue transferring information on this point. Delays during the field visits, but it could contracts- were recorded by municipalities, still be an issue. programs but the central government argued allocations that the delays resulted from the municipalities’ not transmitting the necessary documents to justify the transfers. Intergovernmental debts from the municipalities toward the central government Debts Income tax on municipal staff First item remains an issue–– toward collected directly by municipalities PRSC mission interview, DGCI that should have been reversed to December 2006. the center, but some municipalities had not. Debts on stamp rights. Debts Customs duties (principally from No current information on this toward Praia) not paid to DGA. However, issue. DGA report noted that the information was sketchy. Debts Concern INGRH and INERF but No current information on this toward wide discrepancies between data issue. institutes from creditor and debtors. 69 Besides the intergovernmental debts, the report also noted that municipalities accumulated significant arrears toward public and para-public enterprises. The two enterprises toward which most debts were incurred were ELECTRA and CV Telecom (see detailed table in annex 7). Table 0.11: Municipal Debts to Public and Para-public Enterprises Status in 2004 Comments as of today Debts from municipalities toward public enterprises Toward Debts estimated at 135 Debts toward ELECTRA seem to ELECTRA M CVE. have been reduced from 2006 to 2007, except for public lighting, for which arrears increased. However structural problem still remains; see section below. Toward Debts estimated at 135 No information. CV M CVE. Telecom The issue of intergovernmental debts or debts of municipalities toward public enterprises is far from being resolved, even if some measures are being taken to solve the problem. Following the 2004 report, some immediate measures were taken for the areas in which there was immediate agreement on the report’s conclusions. To completely solve the problem, two main agreements must be reached: an agreement on the exact amounts of past arrears (or a negotiation of these) and an agreement on the issues that, if not solved, will lead to the build-up of new arrears. For the agreement on past arrears, a multisectoral commission was to be created in 2007. The commission would have representatives from the Ministry of Finance (IGF, DGO, Treasury, DGAL) and the National Association of Cape Verdean Municipalities (ANMCV). Its role would be to visit each municipality, proceed to an audit, and agree on a viable solution to solve the problem. For the agreement on the issues that, if not solved, will lead to the build-up of new arrears, the situation is more complex, and nowhere more sensitive than for the debts toward ELECTRA57. The arguments from the municipalities to suspend or delay their payments to ELECTRA are two-fold. First, they would like to be duly compensated for the use by the public company of buildings that belong to the municipalities. The concession contracts do not seem to be clear on this issue, but municipalities argue that this is the common practice. Second, municipalities argue that they also build up arrears because they use electricity to supply water to their farmers. However, but the price of water is centrally regulated. This central control puts municipalities in a difficult situation, whereby their water production cost is above the authorized retail price. 57 As of September 30, 2008 the municipalities’ debt towards Electra corresponded to CVE 460 million (approximately US$ 5 million). 70 VERTICAL AND BOTTOM-UP CONTROL OF MUNICIPALITIES’ FINANCES Compared to most countries in the subregion, the central government’s tutelle (control) of the Cape Verdean municipalities is relatively light. According to the Chapter VI of the Municipality Statute (1995), the central government controls administrative, patrimonial, and financial management of the municipalities insofar as they comply with the law of the land. The only municipal acts that need government approval to be valid are these whose goal is to launch a municipal tax and adicionais (additionals). Beyond the legality control (which can be exercised by inspection), the municipalities have a duty to send the following information to the government: • Municipal “management� accounts. • Minutes of the municipal meetings. • Municipal budget. • Activity plan. • Activity report. • Twinning agreements. • Report on the state of the municipal administration. The central government does not prescribe any type of expenditures or share thereof with which municipalities need comply. This is a welcome development for local autonomy but restricts the initial control from the central government on local entities that may not yet have the capacity to single-handedly manage their finances––some countries for instance indicate a ratio of recurrent to capital expenditures or mandate some specific expenditure in strategic sectors. Both institutions in charge of ex-post control of municipal finances, the IGF (General Finance Inspectorate, under Ministry of Finance) and the TdC (Court of Accounts, or external audit institution), find it difficult to fulfill their mandates. One major constraint of both institutions is the insufficient number of qualified staff and their insufficient operating budgets. This is aggravated by the territorial discontinuity of the country, which increases considerably the inspection costs. According to the law, the IGF inspects municipalities. Irregularities are recorded in a report transmitted first to the municipality, so that it can supply additional information or contest parts of the report. Once the municipality has been given a chance to comment, the final report is transmitted to the Minister of Finance, who may send it to the TdC for further action. In preparation for the 2008 local elections, the IGF fielded inspections in all municipalities in 2007 so that newly elected officials would find a clean slate. Reports for these inspections are not yet available. Interviews with the IGF provided some information on various cases of abuses encountered. These included abuses in land sales in “tourist zones,� gap between contracts-programs objectives and actual expenditures, and payments not engaged or planned in the budget. The TdC certifies the legality and conformity of municipalities’ public accounts and transmits accounts to the Parliament. However, its limited contribution to and impact on the control of municipal finances is a real problem. The last “management accounts� analyzed 71 by the TdC date from 2000, and analysis was done for only 1 municipality (Sal) of 17. This low rate is mainly due to the financial and staffing constraints of the TdC. It should be noted that municipalities do not all transmit their accounts to the TdC in time, Hence, by the end of October 2007, the TdC had received “management accounts� for 2006 from only 10 municipalities, and 4 municipalities still had not transmitted their accounts for 2005. Finally, the impact of the control is rather limited since audits are made available after the end of the electoral mandate for which budgets were being audited. One specific concern of the TdC is the poor monitoring of human resources in the municipalities. As seen in table 9 (municipal management capacity), many municipalities were not able to transmit the requested information for their staff during the human resources census. In the sample, it can be up to 24 percent of the staff for whom no detailed information was made available. The major concerns on this front are: absent or outdated personnel files, absence of security of the process, absence of control of new entries and departures, and absence of TdC approval of some of the human resources processes for which its approval is compulsory. ACCOUNTABILITY TO CITIZENS The performance of local governments generally is rated positively by the population (Afrobarometer 2005) but with interesting variations. If road maintenance and sanitation receive a good appreciation (respectively 57 percent and 60 percent of respondents approve the ways in which their local governments deal with these issues), tax collection and use of local resources fare much lower. The findings call for proactive measures by the local assemblies and executives to inform and involve their populations in budget preparation and monitoring. In fact, for tax collection as well as use of local resources, the main finding is that citizens do not feel informed enough to answer the question, Forty-two percent of respondents indicate that they do not have enough information to assess their local governments’ performance in tax collection, and 52 percent do not have an opinion on the use of local resources. Forty-one percent of the respondents who did answer approve the performance of their municipalities in tax collection. In contrast, the rating falls to only 26 percent for these who respond concerning the performance in use of local resources. The field visit confirmed that, despite an open policy, the culture of accountability is still not entrenched, on neither the side of the municipalities’ executive or assemblies nor on the citizens’ side. Municipal authorities indicated several times that rules exist permitting citizens’ participation and that municipal assemblies’ sessions are open to the public, thus allowing citizens to raise any issues they would like at the beginning of the session. However, this passive openness falls very short of proactive outreach to encourage citizen participation. Similarly, the fact that municipal budgets are published in the Official Gazette is positive, but will not be enough to support strong accountability. The municipality of Praia’s opening of a public room in which citizens can consult the budget is a small step in the right direction. Finally, it is not always clear whether remedial measures exist for citizens, for instance, in cases of abuses by tax collection officers. 72 MUNICIPALITIES’ CAPACITY One main constraint on the municipalities is the skills mix of their staff. Table 4.13 shows that a majority of the municipal staff have no secondary training.58 The municipalities suffer from the fact that, in general, careers in the central administration and living conditions are more attractive in one of the two largest urban centers. Municipalities also have suffered from the distinction between the “common� civil service statute and the “specific� ones. Municipal staff careers and remuneration follow the “common� rule, while many central administrations, in particular, the financial and fiscal ones, are ruled by “specific� schemes, with higher remunerations. Finally, in some cases, municipalities inherited an excess of personnel who lacked the needed profiles to fulfill the new mandates. The solution is likely to lie in a combination of measures. The current civil service reform attempts to reduce the number of “specific� statutes, which should reduce discrepancies within the civil service. A capacity building program for some municipal staff is an option, but it needs to be well-targeted. Not every staff can fully benefit from specific training, and the main issue will remain the retention rate of the staff once trained. Finally, more creative solutions can work, as the example of Sao Vicente shows with the use of retired staff for the fiscal unit. The Municipal Information System (SIM) developed by NOSI (including the financial management system also used at the center, SIGOF) is a modern tool to strengthen municipal management capacity and communication between the central level and the municipalities, and among municipalities. Financing has been secured for installing it in 12 municipalities. Implementation will continue in 2008, with an extension of the system to the 7 remaining municipalities (Santo Antão, Sao Nicolau, Boa Vista, and Maio). 58 It is important to note that the Human Resources Database, from which this table is extracted, has a low rate of returns of information from the municipalities. Thus, the level of accuracy of the data is an issue. 73 Table 0.12: Number of Staff per Education Level Praia % S. Cruz % S. Domingos % Mosteiros % Illiterate 48 26 3 9 Preschool level 10 4 Literacy course 47 19 4 4 Subtotal 105 10.67 49 11.69 7 5.74 13 10.92 Primary school 442 231 40 61 Subtotal 442 44.92 231 55.13 40 32.79 61 51.26 High School––1º Cycle 57 25 8 10 High School––2º Cycle 61 37 29 22 High School––3º Cycle 42 11 11 8 Subtotal 164 16.67 76 18.14 50 40.98 41 34.45 Gradute studies––Bacharelato 2 1 3 1 Graduate studies––Licenciatura 31 1 1 1 Graduate studies––Pós-Graduação 1 Subtotal 34 3.46 2 0.48 4 3.28 2 1.68 (No information in central database) 239 24 61 15 21 17 2 2 Total 984 419 122 119 Source: 2005 Human Resources Management Database. 74 Box 0.2: Municipality of São Vicente - Impact of Introducing IT In just a few years (the reform started in 2001), dramatic progress has been achieved in financial management by introducing IT solutions. This IT solution, developed with national expertise under the leadership of NOSI (dependent on the Prime Minister’s office), has been installed in three municipalities and is being rolled out in all municipalities of Santiago, Brava, and Fogo islands. • Electronic filing: Prior to the reform, Sao Vicente used 22 different books to store information on the city budget and other transactions. People were brought in from outside––in some cases, students on vacation–– to help with input of data. Now, 100 percent of accounting and treasury data are on computer. • Budget: Almost all municipal budget operations are performed online. • Tax: Through the greater control and supervision offered by a computerized system, the municipality has been able to triple its tax income since 2001. The increase is due to greater efficiency rather than higher statutory rates. Due to an integration of various records, a citizen in arrears on municipal taxes, for example, might not be allowed to obtain a permit for a car. It should also be noted that the municipality hired retired skilled personnel to assist it in establishing an efficient tax collection department. CONCLUSION AND RECOMMENDATIONS In Cape Verde, similarly to many other countries in which decentralization is relatively recent and in which most municipalities are relatively small, central government transfers represent a significant source of revenues for municipalities. Municipalities thus are very dependent on the central government for their finances, and it will be important to support the municipalities in diversifying their sources of incomes. Therefore, it is key to build their capacity to collect and manage local taxes, since this task is a mandate of the municipalities. Also, it is critical to assist municipalities in updating/establishing their land and taxpayers registries, and provide them with tools/methodologies to evaluate property value. Municipalities’ own-revenues come mostly from local taxes and proceeds from land sales. Most municipalities have not been able to strengthen their tax unit (staff and tools) and thus cannot exploit the whole tax potential in their locality. Local taxes have boosted revenues in some municipalities that have a broader tax basis and have strengthened their tax collection department through adding staff and IT. In smaller municipalities, taxes remain difficult to manage and do not contribute much to the total budget. In all municipalities visited, local taxes represent no more than 10 percent of revenues. As a result, in many municipalities, land sales represent a significant share of an annual budget. This dependency on land sales creates a sustainability issue when the municipalities use them to finance operating expenditures. Therefore, it is recommended to regulate the use of land sale proceeds and restrict or forbid use of land sales to finance operational expenditures. Also, borrowing is increasing as a measure to face cash flows constraints. Increased borrowing, specially if used to finance operational costs, may pose fiscal risks and therefore, it is crucial to ensure that the use of borrowing comply with regulations. Central government transfers are composed of two instruments: formula-based transfer and "contracts-programs.� The formula-based transfer was reformed in 2005 with an increase in the percentage and sources of revenues at the basis of the formula. This instrument guarantees neutrality and predictability and seems to satisfy most actors. The “contracts-programs� are a more ad-hoc way to promote investments in selected municipalities. Consequently, it important 75 to implement mechanism of control in the management of contractos-programa. The draft law on decentralization is addressing this issue for future transfers. Responsibilities in the purview of municipalities are very broad in Cape Verde, but there is not a clear delineation of roles and responsibilities between central government and municipalities. As a result, it becomes difficult to define adequate levels of expenditure and revenue assignment, and accountability also suffers. Municipalities can engage in all the areas identified as “local.� However, they have been constrained in doing so because of scarce resources and poor coordination/collaboration with central administrations. Transfers have occurred in a rather ad-hoc manner. Therefore, it is recommended to improve communication of the amounts to be transferred for a next fiscal year, even if based on estimates, and better define the responsibilities of the local administration. Municipal accountability is weakened by relatively poor financial reporting. There is lack of knowledge of municipalities’ finances because many municipalities do not prepare the mandatory tables (and others do not prepare them on a regular basis). To improve the knowledge of local finances it is critical to collect, analyze and disseminate annual consolidated data. Overall, financial management at the local level needs to be strengthened. Currently, there are several main weaknesses: (a) no evidence of accurate revenues forecast and medium-term expenditure framework; (b) wide discrepancy between aggregate expenditure outturn and original approved budget; (c) absence of sector plans and alignment with national policies; (d) no account of extra-budgetary funds in the budget (for example, decentralized cooperation); (e) very weak internal controls; (f) weak procurement procedures; (g) weak ex-post controls due to weak capacity and limited resources of the control institutions; and (h) accumulation of arrears towards state companies, especially Electra. To strengthen financial management and overcome these weakness, it is recommended to: • Request update information regarding grants from the donors and record it in the budget. • Clarify concession contract with Electra and protocols with other Institutions that provide goods and services to the administration. • Ensure that IGF continues to regularly inspect all municipalities and introduce internal audits at the municipalities. • TdC should concentrate on certifying municipal accounts for the past two years, and external support should be contracted to review all past accounts. • Improve planning instruments, including the preparation of MTEF. • Train local staff on the application of the Procurement Code. 76 infrastructure Public expenditure review BACK GROUND AND MOTIVATION Cape Verde’s geography has contributed to the relatively high cost of infrastructure services and access limitations in the archipelago. A low––approximately half a million––and dispersed population in 9 islands results in an extended infrastructure network. Cape Verde has 3 international airports, 4 aerodromes, and 9 ports. The decentralized and very fragmented provision of utility services can hardly benefit from economies of scale and poses an enormous challenge when deciding on the best technology and optimal scale of operation. The lack of efficiency gains that result from the diseconomies of scale drives up infrastructure services and local products’ costs (prices). Figure 0.1: Population Density Ribeir a Grande Il Paúl Port o h as d o Ba 0 2,0 00 1 7°N Novo r l a v en to km São Vicent e Sal São N ico lau O cean o A tlâ n tic o Boa t Vis a 1 6°N Densidade Pop. (Ha b /Km2 ) > 300 ve n to 1 50 - 30 0 So t a do h as 5 0 - 1 50 < 50 Il São Tarrafal Miguel Sant a Maio Sant a Cat ar ina Cruz 1 5°N t M os eiros ão S Domingos São Praia No rte Filipe Br ava 0 20 km 2 5°W 2 4°W 2 3°W Source: Plano Estratégico de Transportes. The country has neither (known) oil resources nor the capacity to refine oil, and the availability of water is limited. Underground water resources are limited as the islands are in the semiarid Sahel region. Water resources exist or are not scarce only in the islands of Santo Antão and Fogo. In Sub-Saharan Africa, only Djibouti has lower water resources per capita than Cape Verde. Increasingly, Cape Verde has had to rely on desalination plants for water (accounting for approximately 85 percent of production) and only to a very limited extent on extraction from underground sources (over the last 40 years, rainfalls have sharply decreased by approximately 54 percent). 77 Cape Verde is driven by a strong commitment to offer access to basic services to the whole population. Infrastructure delivery has come a long way in terms of overcoming some of the challenges imposed by geography and nature. Access rates compare favorably with similar countries (Table 0.1Table 0.1).59 It is noteworthy that Cape Verde outperforms the comparator countries on road density but clearly underperforms on sanitation facilities.60 The former reflects the policy decision of targeting high accessibility, even in low-density areas. Table 0.1: Benchmark of Access Indicators61 Sub- Energy Sector Cape Dominican South Indicator Verde Republic Guyana Indonesia Maldives Mauritus Philippines Senegal Africa Saharan Africa 1/ Households Reporting Access to 66.0 66.8 nd 80.4 nd 50.0 78.0 32.2 66.0 27.2 Electricity (% of households) and Sanitation Water Supply Improved water source 80.0 95.0 83.0 77.0 83.0 100.0 85.0 76.0 88.0 64.1 (% of population with access) Improved sanitation facilities 43.0 78.0 70.0 55.0 59.0 94.0 72.0 57.0 65.0 36.5 (% of population with access) Roads Road density, (% of arable land) 3.5 1.1 1.7 1.6 2.2 2.0 3.5 0.6 2.5 0.8 (road-km/sq-km) Source: World Bank Africa Development Indicators; WB Infrastructure Research Database; Ctry Sectoral Studies; AFTPI Processed. 1/ Average of latest obs. 2000-2004 The challenge ahead is to keep up with the growing needs accentuated by the fast-growing tourism industry and to ease constraints imposed by expensive infrastructure. Water is scarce, and electricity provision is erratic in most of the islands. Improper waste disposal, resulting in ravines being used as dump sites; transport access; and irregular ferry services are some of the constraints (Table 0.1Table 0.1). Furthermore, infrastructure services such as water and electricity are very costly, thus hindering the competitiveness of the economy. The government intends to expand and improve infrastructure substantially in the near future. According to PIP the amount of investments in infrastructure (electricity, roads, aeroports and ports) planned for 2009 are expected to reach US$ 100 million. To pursue these investments without prejudicing the country’s fiscal sustainability, the authorities will have to increase the availability of budgetary room by also mobilizing other resources. To make informed decisions on new investments and to avoid expansion’s compounding efficiency losses, it is important to learn how much has been spent on the several components of infrastructure, whether the composition of spending is appropriate, and, most importantly, where fiscal savings can take place due to a more efficient management. This chapter addresses these issues, as, so far, no analysis has been conducted on them due to the lack of an efficient M&E system. 59 Because Cape Verde is more developed than most of other countries in Sub-Saharan Africa (SSA), the comparator countries, when possible, include some of the more developed countries in the region (Senegal and South Africa), other middle-income islands economies (Dominican Republic and Guyana), and lower middle-income in East Asia (Indonesia and Philippines). 60 Due to the geography of the country there are many remote areas. 61 The set of countries includes a mix of island and middle-income economies. As possible, given the limitation in the available information, the authors use the same set of countries in the rest of the report. 78 Table 0.2: Key Infrastructure Assessment for Cape Verde, by Island Santiago Served by an international airport, although the airport is relatively small and unable to handle large volumes of passengers. Fresh water is limited, and electricity provision is erratic. Proper waste disposal is not in place, resulting in ravines being used as dump sites. Small port and airport––not able to handle large planes or vessels. Airport has no landing lights, making it impossible to land at night. Water is available, but electricity provision can be erratic. Fogo Reasonable road network is in place. Proper waste disposal is not in place, resulting in ravines being used as dump sites. New international airport opened in July 2007. Lack of water and reliable electricity are problems. Boavista Water and electricity are problematic, with most resorts investing in their own desalination plants and electricity generators. Waste disposal is not as bad as in some of the other islands. The port (at Sal Palmeira) handles both cargo and passenger ships. The international airport (at Sal) is able to handle large airplanes, and there are plans to have it upgraded. The airport at San Pedro was upgraded to an international airport (it is not operational yet). The port at Mindelo is the largest and deepest port in Cape Verde, and there are plans to develop a marina. Vicente São Roads on the island are in a reasonable condition. Water and electricity are problematic, and the lack of proper waste disposal leaves a bad impression with visitors. Although this is the second largest island, Santo Antão does not have a functional airport, and the only access is by ferry from São Vicente twice a day. The port is very small and cannot Antão Santo accommodate large vessels. Roads on the island are in a reasonable condition, although there are some very small unpaved roads that are difficult to negotiate. Water is available because of the annual rainfall, and electricity and telecommunications are generally available. Maio’s small airport is served only three times a week by a TACV flight from Praia. The airport does not have landing lights; therefore, planes are not able to land at night. The port is served by an Maio irregular ferry service to Praia. Many roads are accessible only by 4x4 vehicles. The small airport does not have landing lights, preventing planes from landing at night. The port at Tarrafal is served by ferry services to Mindelo (São Vicente) and Palmeira (Sal). There is a Nicolau São reasonable road network on the island, but water and electricity are problems. The biggest infrastructural problem is the lack of transport access. There is no airport. A very irregular ferry service from Fogo allows passenger access if and when available. Brava Source: IFC Cape Verde Sector Growth and Investment Program. According to the Transports Strategy Plan 2008, the private sector should play an increasing role in financing, constructing, and maintaining infrastructure.62 A reform program to liberalize the economy and to encourage private participation in investments begun at 62 The Millenium Challenge Compact includes the rehabilitation of several roads in Santiago Island and construction of bridges in Santo Antão Island. 79 the end of the 1990s. The reform program seeks to: (a) remove entry barriers to private investments, including reforming taxes, alleviating administrative barriers, improving supply chains, and legal reform; and (b) implement a large divestiture program of state-owned- enterprises (SOEs), emphasizing public utilities, to improve the availability and quality of basic services. The reforms have led to changes in the ownership structure of the infrastructure sectors in Cape Verde. Yet, after several years, the government retains a key role in the provision of infrastructure. The state manages and retains the majority of the capital of the electricity and water company (51 percent), owns airports and ports (also manages ports), and owns the Cape Verde Airline (TACV)63. The only sector in private sector hands is telecommunications. This chapter reviews public expenditures in infrastructure during 2001–06. Sectors covered include electricity, water and sanitation, roads, ports, airports, and air transportation. The sources of information are the national accounting system, SOEs’ financial statements, and sectoral reports prepared by the government and World Bank.64 These reports closely follow a standardized approach developed by the Bank’s Africa Region (AFR) for infrastructure public spending analysis.65 They also benefit from a set of performance indicator benchmarks and, perhaps more importantly, from a database of “Africa Infrastructure Country Diagnostic� in African countries,66 which quantitatively underpins the chapter. BRIEF OVERVIEW OF INFRASTRUCTURE DELIVERY Energy, Electricity and Water, and Sanitation Cape Verde’s petroleum market is dominated by a duopoly. There are two companies: Enacol (major shareholders are Petrogal and Sonangol; the State has a participation of 2.1 percent) and Shell (a wholly owned subsidiary of Shell International). The two companies import, store, and distribute petroleum products. Concerns with the transparency of the pricing of oil products have been raised: by Booz, Allen and Hamilton (2002) and PER (FY06).67 Furthermore, price formulation by the regulator allowed for differing mark-ups among companies, thus discouraging competition and efficiency. Other flaws in this market include high costs for import, storage, and distribution.68 To circumvent, or at least minimize these, the 63 TACV is a public company with private management. 64 The information relies on the data provided by the government. In some cases, the team was able to crosscheck figures and facts from different reports. With respect to SOEs, cost and revenue figures rely on the companies’ financial statements. 65 C. Briceño-Garmendia and V. Foster, “More Fiscal Resources for Infrastructure? Evidence from East Africa� (World Bank 2006). 66 Cape Verde is 1 the 24 pilot countries in the ongoing “Africa Infrastructure Country Diagnostic� knowledge work program, developed by the Bank in collaboration with the Infrastructure Consortium. 67 The oil companies report import prices on a cost-plus-freight basis, with freight computed as a constant premium. The companies should make explicit the fob and the freight costs (PER FY06). 68 Cape Verde was an important hub for refueling for both maritime and air transport. Today, most of that traffic has been reallocated to other stops (mainly Dakar). Furthermore, the storage capacity of fuel seems to be placed far from the current consumption centers. 80 government is setting up a logistics company that will import, store, and distribute; and the regulator is working on a revised (more transparent) formula for price-setting.69 The water and electricity company (ELECTRA) is the incumbent operator for electricity. It supplies electricity to the whole country and water to the islands of Sao Vicente, Sal, Boavista, and the city of Praia. Municipalities provide water to the remaining islands and cities of Santiago Island. In 1999 ELECTRA became a public-private partnership through the sale of 51 percent of the share value of the production assets of electricity and water. The government maintained 34 percent of the share value, and the municipalities retained 15 percent. In July 2006, as a result of a series of failed negotiations between the private stakeholders and the government, the State regained the position of major shareholder (51 percent) in May 2008. Private stakeholders retained 34 percent; the municipalities retained 15 percent. During 2001–06, production and connections grew rapidly, while installed capacity decreased. Between 2002 and 2006, installed capacity decreased approximately 4.4 percent. As a result, blackouts were more frequent in duration and quantity. During this period, energy losses remained unchanged. Energy production is largely dependent on diesel plants––very costly for a country dependent on oil. The authorities are planning to replace the gas oil generation by fuel oil generation plants (roughly, the price of gas oil––diesel––is double the cost of fuel oil, FO380) and to invest in renewable energy to improve technical efficiency and reduce oil dependency.70 Table 0.3: ELECTRA - Power Statistics Clients Production Losses Black-outs Installed Capacity (MW) Island GWh % Hours / year 2002 2006 2002 2006 2001 2006 2001 2006 2002 2006 Total Diesel Eolic Thermal Total Santo Antão nd nd 7.16 7.82 12.0% 19.9% 14.72 2.97 3.67 3.72 3.72 S. Vicente nd nd 45.95 48.37 15.4% 16.9% 16.17 14.18 20.64 18.35 0.90 0.77 20.02 S. Nicolau nd nd 3.08 3.29 13.0% 13.0% nd 8.60 1.39 2.52 2.52 Sal nd nd 20.59 21.85 9.9% 6.2% 2.83 10.88 13.02 9.23 0.30 9.53 Boavista nd nd 2.38 2.73 8.6% 6.5% 10.17 16.17 1.04 1.63 1.63 Maio nd nd 1.27 1.36 16.3% 9.2% 24.82 42.55 0.74 0.71 0.71 Santiago nd nd 78.18 88.82 32.3% 29.4% 189.33 752.43 29.79 27.77 0.90 28.67 Fogo nd nd 4.06 5.00 23.2% 24.4% 5.00 11.50 1.87 2.64 2.64 Brava nd nd 1.67 1.76 18.2% 15.7% nd 52.25 1.17 0.66 0.66 Total 60,274 82,880 164.33 181.00 22.7% 21.5% 263.03 911.53 73.31 67.22 0.90 1.97 70.09 Source: Electra - Relatorio e Contas Tariffs for electricity and water71 in Cape Verde are among the highest for African countries (figure 1.2). According to the Investment Climate Survey (2006), firms in Cape Verde perceive the poor performance of the power sector as the greatest constraint on their operations and growth.72 Over 60 percent of firms say that the power sector is a major or very severe obstacle to operate and grow. Furthermore, the number of power outages and losses due to power outages in Cape Verde are high compared with its peers. 69 This PER does not expand the analysis on the petroleum sector, as it is beyond its scope. 70 The new business plan for 2008–11 indicates that the company’s strategies will be to reduce the use of costly fuels and to interconnect some of the subnetworks in Santiago. 71 According to several specialists, including one of the technical advisors of the regulator agency, there are some concerns about the efficiency of technology used for water desalinization. 72 See Cape Verde Investment Climate Assessment (2007), 9. 81 Figure 0.2: Effective Electricity and Water Tariff Electricity tariff at 100kWh (USc/kWh) Effective water tariff at 10m3 (USc/m3) 35 300 30 250 25 200 20 150 15 100 10 50 5 0 0 C d' i a re Be o C Eth ri a M gal in de w n Ke ho N wi Se ya Le da R s R ni p oi Fa e a t r k er ot io D ne n an so Iv al ig Bu e V a e ap C Source: Banerje and others, Africa Infrastructure Country Diagnostic (2008); Briceño-G and Shkratan, Africa Infrastructure Country Diagnostic (2008). Figure 0.3: Losses Due to Power Outages - A comparison D o m inic an R e public G uya na P hilippines S e ne ga l S e ne ga l M aurit ius Indo ne sia P hilippine s M a urit ius So ut h A f ric a G uya na Indo ne s ia S o ut h A f rica M a ldiv e s C a pe V e rde C ape V e rde 0 5 10 15 0 10 20 30 40 50 Median losses due to pow er outages Median number of pow er outages Average losses due to pow er outages Average number of pow er outages Source: Cape Verde ICA (2007). Even though production costs vary significantly across islands, prices are the same across the country. The decentralized structure of electricity production leads to differences in the production costs across islands (caused by differences in the inputs used and the scale of generators). However, prices are the same owing to a cross-subsidization among islands on the grounds of the principles of equity and social justice. Demand also varies significantly across islands. Moreover, due to the overall small size of the system, demand is very skewed. One large 82 individual firm (hotel) can drive up substantially the demand needs in a specific island. As a result, if one large firm (hotel) builds its own generator, the impact on the economies of scale, and thus efficiency gains, is very high. Prices of electricity, water, and petroleum products are established by the regulator, Agencia de Regulação Economica (ARE). Water production and access (connections) increased in the recent years. In total, from 2001 to 2006, connections increased approximately 50 percent. The increase in connections has contributed to raise the percentage of population with access to improved water sources to 80 percent (World Bank Indicators 2006). However, the increase of production and connections was accompanied by an increase in network losses.73 As with electricity, prices are the same across all islands, regardless of the provider, ELECTRA, or municipalities. Table 0.4: Water Supply in Cape Verde Capacity Production Losses Connections Island M3/day x000 M3 2002 2006 2002 2006 2002 2006 2002 2006 S. Vicente 3,600 6,320 1,145 1,379 23.0% 32.8% nd nd Electra Praia 11,532 6,913 1,908 2,124 33.5% 31.1% nd nd Sal Palmeira 1,477 2,880 519 744 19.5% 21.0% nd nd Sal Rei Boavista 320 230 67 63 1.9% 35.0% nd nd Total 16,929 16,343 3,639 4,310 27.6% 30.0% 19,585 29,038 Municipalities Fogo and Brava nd nd 848 nd nd nd 3,534 nd Maio nd nd 73 nd nd nd 900 nd San Antão nd nd 189 nd nd nd 3,029 nd Total nd nd 1,037 nd nd nd 6,563 nd Source: Electra - Relatorio e Contas and municipality studies Overall sanitation services increased only from 41 percent in 2000 to 43 percent in 2004 (World Bank Indicators). Cape Verde’s percentage of population with improved sanitation facilities is higher than Sub-Saharan Africa’s (36.5 percent), but lower than those of Senegal (57 percent) and all the other comparator countries (Table 0.1Table 0.1). The difference between the rural and urban is very acute. Approximately three times more population in the urban areas has access to improved sanitation facilities. These indicators are expected to have improved in the recent past due to the recent investments in treatment plants and solid waste treatments. 73 The central administration has virtually no information on water supply from the municipalities; hence, it is difficult to do a countrywide assessment of the sector. 83 Figure 0.4: Sanitation Services 0.7 0.6 0.5 Urban (2004) Rural (2004) 0.4 Total (2000) Total (2004) 0.3 0.2 0.1 0 Access Source: World Bank Indicators. Transport Roads The road network in Cape Verde is state owned and managed. The Roads Institute - Instituto da Estrada (IE) has jurisdictional responsibility over the national roads network. Furthermore, IE is in charge of the maintenance, which is funded by a Road Maintenance fund. Roads maintenance is discussed in a section below. Table 0.5: Roads Network (2005) Kms by Quality Kms by Type Type / place Total Good Fair Bad 1 2 3 4 5 6 National 250 231 521 83 5 296 354 124 140 1,002 Santo Antão 33 17 146 5 104 12 75 196 S. Vicente 20 13 9 9 12 20 1 42 S. Nicolau 38 16 41 69 14 12 95 Sal 26 15 36 5 41 Boavista 4 2 24 20 3 5 2 30 Maio 21 12 15 48 48 Santiago 94 105 165 38 204 16 63 43 364 Fogo 14 48 107 132 29 8 169 Brava 3 14 7 10 17 Other types 63 118 424 7 40 97 178 283 605 Total 313 349 945 83 12 336 451 302 423 1,607 Source: Instituto da Estrada (Undated) Note: 1. Betao Betuminso, 2. Bicouche, 3. Calçada de Paralelos, 4. Calçada Portuguesa, 5. Empedramento, 6. Terrra Batida 84 Compared to other similar countries, Cape Verde has a dense road network. Road density (measured as percentage of arable land)74 is 3.5 in Cape Verde, higher than the comparator countries (for instance, Maldives, 2.2; Mauritius, 2.0; Indonesia 1.6). The dense road network is justified by principles of equity and social justice to offer access to health, education, and other basic services to all the population. The other objective is to interconnect the domestic market. However, approximately 60 percent of the roads are in bad condition.75 The quality of the network varies considerably across the country. For instance, while the network in Sal seems to be in good or fair condition, the entire network in Brava is in either fair or bad condition. The latter suggests that the Cape Verde’s policy of extending the network to low-density areas, in which economic returns are low, may have raised maintenance costs to an unsustainable level, or that maintenance was not been planned and accounted for (this will be discussed further below). Air transport To overcome the isolation imposed by geography on the country and its inhabitants, air transportation has always played a critical role in providing connectivity among the country’s islands and from the archipelago to the rest of the world.76 In the face of the recent tourism boom, air transportation is playing a growing role in the economy. As a result, the government has systematically channeled significant financial and human resources to the aeronautical sector.77 Air transportation, domestic and international, is offered by Cabo Verde Airlines (TACV)– –with a fleet of 5 aircrafts. Even though the government decided to privatize the company several years ago, TACV continues to be state owned. The process has been lengthy and has stalled several times for various reasons. The conclusion of the privatization process is expected in 2008. According to the Transports Strategy Report, while international flights are break-even flights, interisland flights are nonprofitable. TACV’s passengers and seats per km have increased steadily since 2000. However, as the seats per km increased much faster than passenger per km, the cabin factor decreased (passenger/seat ratio) sharply in 2002. It increased in 2005. From 2002 to 2004, cabin factor averaged 62 below what is usually assumed for large carriers: 70 percent.78 This apparent excess of capacity suggests that the company may be using oversized aircraft for a given demand, or that some routes do not have enough demand.79 74 According to the World Bank Indicators, in 2003 the total land in Cape Verde was equal to 4,033 km2, of which 11.4% is considered arable land (460.3km2). Thus, road density is the ratio of road length over arable land (km2): 1607km/460.3km2=3.5. 75 According a Transport strategy report (2008), the corresponding figure is 70%. 76 TACV’s nonstated duties are to (a) provide regularly scheduled inter-island air services; (b) fly to the cities hosting the larger diaspora communities, and (c) connect Cape Verde with the European cities that provide the bulk of tourists. 77 Cape Verde’s Air Transportation Sector Phase II Report, 3. 78 See Stephen Holloway, “Straight and Level: Practical Airline Economics� (Ashgate Publishing: 1997). 79 We do not have enough information to assess the cause of the apparent excess of capacity. The choice of an oversized aircraft may be justified by leasing, fuel, and maintenance costs. 85 Figure 0.5: TACV - Performance Indicators, 1997–2005 1400000 0.74 0.72 1200000 0.7 1000000 - x1000 CVE- 0.68 800000 0.66 600000 0.64 0.62 400000 0.6 200000 0.58 0 0.56 1997 1998 1999 2000 2001 2002 2003 2004 2005 Seat-Km Passangers-Km Cabin Factor Source: TACV. The privately owned airline, Cabo Verde Express (CVE) provides inter-island charter services for tourists.80 CVE transports tourists who wish to complement their sun and sea vacations with a visit to the archipelago’s other islands. Halycon Air will be the new local carrier that will provide inter-island air transportation (pending its certification by the Civil Aviation Agency). Halycon Air will be a direct competitor to TACV. Airports and Ports Cape Verde has 3 international airports that have operated since 2007 (Sal, Praia and São Vicente) and 4 aerodromes.81 The international airports have been recently renovated including buildings, facilities, and runways. A fourth airport (Sao Pedro Airport in Sao Vicente Island) is being remodeled to also become an international airport. The remaining three (Fogo, Maio, and Sao Nicolau) cater only to domestic flights. Sal and Praia are the only 24-hour operational airports that also satisfy Category I of the US Federal Air Administration and the International Civil Aeronautics Organization (ICAO). This safety certification for both air traffic control and airport operations requires costly investments in facilities and training. Airports are managed by a state-owned enterprise, Empresa Nacional de Aeroportos e Segurança Aérea (ASA). The company has corporative governance and has followed standard business procedures since 2001. Activities include the support of civil aviation, air traffic management, services of departure, arrival and over-flight of aircraft, management of terminals and freight couriers, and the Flight Information Region that represents the main source of revenues for the company. 80 Its fleet is 3 L410 with capacity to 19 passengers each. 81 The International Airport of Sal is a hub for international and domestic flights. The airport of Praia (Santiago Island) offers flights to Europe since 2006. The airport of Rabil (Boavista Island) was remodeled with private investments and in 2007 started operating with international flights. 86 Traffic is concentrated in international airports. The Airport of Sal, which handles almost half of the passengers, has served as regional hub. South African Airlines was one of the first airlines offering scheduled services from Sal to various destinations. The cancellation of the flight from Johannesburg to North America via Sal to route instead via Accra and Dakar resulted in a significant loss of intra-transit passengers for Sal (-114,413 passengers, or a decline of 12 percent). However, owing to the booming tourism industry, new scheduled and charter airlines have been entering the market.82 Figure 0.6: ASA - Distribution of Aircrafts and Passengers, 2006 Aircrafts Aircraft Passengers Passengers AD-MAIO AD-MAIO AD-S. NICOLAU AD-S. NICOLAU AD-FOGO AD-FOGO AIAC-SAL AD-BOAVISTA AD-BOAVISTA ASP-S.VICENTE AIAC-SAL ASP-S.VICENTE ADP-PRAIA ADP-PRAIA Source: ASA. Table 0.6: Passengers (arrivals + departures + in transit) AEROPORTOS 2001 2002 2003 2004 2005 2006 AIAC-SAL 708,711 781,539 829,716 1,007,561 877,007 772,600 ADP-PRAIA 264,823 248,864 247,538 253,303 269,625 354,100 ASP-S.VICENTE 132,693 124,952 120,504 125,648 125,497 135,080 AD-BOAVISTA 47,351 41,420 44,144 39,072 38,467 72,756 AD-FOGO 44,904 39,601 39,408 38,485 40,561 41,108 AD-S. NICOLAU 24,997 25,064 23,885 20,063 18,664 20,556 AD-MAIO 16,652 12,543 10,909 11,079 10,692 10,401 TOTAL 1,240,131 1,273,983 1,316,104 1,495,211 1,380,513 1,406,601 Source: ASA Table 0.7: Passengers in-transit Variation Passengers 2002 2003 2004 2005 2006 2006/2005 All airports 190,236 206,471 215,421 212,689 126,596 (86,093) Source: AAC, Relatorio de Actividades Ano 2006. Cape Verde has 9 ports, of which 3 ports receive international traffic (Praia, Porto Grande, and Porto da Palmeira). Maritime transportation is of great importance in Cape Verde due to the archipelago characteristics of the country. All islands have ports that allow maritime access. The ports of Fogo, Brava, and Santo Antão are very small, which limits the types of vessels traveling to these islands. Of the three international ports, only Porto Grande and Praia83 have 82 Cape Verde’s Air Transport Sector Phase II Report, Partnership to Mobilize Investment. 83 Remodeling (expansion) of the Port of Praia will start in 2008 (and go until 2010). The cost is estimated in Euros: 75 million Euros (Plano Estrategico dos Transportes 2008–20). 87 capacity for receiving large ships and the movement and storage of containers. Overall traffic has increased significantly in past years (table 5.8). The authorities intend to use these ports as regional hubs to transport passengers and cargo.84 Table 0.8: ENAPOR - Traffic Statistics Year 2004 2005 2006 Load (tons) 1,308,126 1,533,132 1,712,281 Long-distance 852,976 991,537 1,115,739 General 112,743 106,322 136,691 Container 213,622 257,157 331,451 other 84,964 106,444 137,284 Bulk 122,756 166,374 166,899 Fuel 300,866 343,833 325,432 Transit 18,025 11,407 17,982 Coastal traffic 455,150 541,595 596,542 Loading 217,062 253,999 270,299 Unloading 212,651 256,145 283,445 Bulk (liquid) 25,437 31,451 42,798 Passengers 550,460 640,502 676,646 In 258,797 302,119 315,803 Out 263,416 300,675 311,219 Transit 28,247 37,708 49,624 Source: Enapor Relatorio e Contas All ports are state owned and managed by the state port authority, ENAPOR, under the service port concept. As with TACV, the government long ago decided to privatize ENAPOR management, but here as well the process has met many challenges. In July 2006, the TA contract signed with Booz Allen Hamilton for the privatization ended without producing the desired results. A new privatization advisor has been recruited. The Millennium Challenge Compact financing of the rehabilitation of the port of Praia is tied up with the privatization of port management. Privatization is expected for 2008. Table 5.8 shows that traffic has increased significantly in the last years. The main inter-island ferry service has been privatized, including a provision for subsidy to ensure minimum service on some non-profitable (low-demand) routes. These connections are important to connect the domestic market and also from a social point of view. Schedule reliability of the ships that combine passengers and cargo has been greatly criticized by local inhabitants and tourists. 84 An appropriate market analysis is being conducted as part of PEP Africa Phase 2 assessment. A 2004 report prepared by Don Breazele and Associates, Inc. for the US Trade and Development Agency on behalf of ENAPOR indicates that “while the development of Cape Verde as a marine and air transport hub has been a strategic focus for the country for some time, the actual potential appears quite limited� [quoting Cape Verde Maritime Transport Draft Report, 4]. 88 INFRASTRUCTURE SPENDING Adequacy Measured based on execution, infrastructure spending in Cape Verde is extremely high: US$401 per capita in 2006.85 This figure is large even when compared to some other countries with larger per capita income such as Chile and Turkey (Table 0.9). ELECTRA expenditures per se correspond to 5.9 percent of GDP and approximately US$138 per capita.86 To get a cross- country comparison, TACV expenditures have been excluded from Table 0.9. TACV’s spending represented 9.8 percent of GDP in 2006, raising the total infrastructure spending to 26.9 percent of GDP in 2006. This high level of spending does not necessarily indicate over-investment, but rather inadequacy of past infrastructure investments. For instance, in the road sector, past expenditures on both investment and maintenance have been minimal. Most investments now are for rehabilitation of the network to bring it up to a maintainable standard. Table 0.9: International Benchmarking of Infrastructure Spending Levels, 1998–2004 In fra s tru c tu re In c o m e P e r S p e n d in g 1 / 2 / C a p ita la te s t ye a r (U S $ - C o u n try a v a ila b le (% G D P ) (C u rre n t U S $ ) p e rs o n ) Uganda 2004 1 9 .0 7 .2 % 250 T a n za n ia 2004 3 0 .9 9 .2 % 297 Rw anda 2005 1 6 .8 7 .1 % 237 C a p e V e rd e 2006 4 0 1 .8 1 7 .0 % 2 ,3 5 9 O n -b u d g et 2006 9 7 .2 4 .1 % E L E C TR A 2006 1 3 8 .6 5 .9 % AS A, E N AP O R an d R M F 2006 1 6 6 .0 7 .0 % K e n ya 2003 4 7 .8 1 0 .5 % 430 In d o n e s ia 2004 9 3 .2 7 .9 % 1 ,1 6 7 T u rk e y 2004 2 4 0 .1 5 .7 % 4 ,2 1 0 C h ile 1998 3 1 7 .3 6 .4 % 4 ,9 2 3 S o u rce : F isc a l C o st B a se lin e , W B P P I D a ta b a s e , A F TP I P ro ce s se d . N ote: 1/ Infrastruc tu re in clude s ele ctricity, w ater an d san ita tio n, a irpo rts , po rts , an d roa ds ), and te le co m m un ic ations . T e leco m s are no t inc lude d in C ap e V erde . 2 / O verall s pen ding in clud es o pe ra tion s an d m ainten an ce, inve stm en ts fo r bo th p ub lic a nd p riva te sec to r; un le ss othe rw is e in dica ted. The public sector in Cape Verde so far has played a critical role as provider, policy-maker, and planner of the infrastructure business. However, the government’s capacity for achieving improvements in infrastructure delivery through increased fiscal spending is limited by its overall objective of fiscal discipline. Moreover, within the boundaries of fiscal discipline, an increase in the envelope for infrastructure would imply reduction in other sectors, perhaps of equal priority to the overall goals of growth and poverty reduction. Cape Verde’s spending on infrastructure is already high (table 1.9). Hence, continued improvement in infrastructure 85 This figure includes spending on electricity, water, sanitation, and roads; and excludes TACV. Spending is defined as the sum of the recurrent and capital expenditures (on- and off-budget, special funds). SOEs’ spending includes salaries, inputs, outsourcing, taxes, depreciation, debt service, and subsidies if any. 86 Public spending in 2006 may be even higher due to the recapitalization of ELECTRA. The Government share of the recapitalization reached almost 3.5% of GDP. 89 delivery should not be achieved solely through an increase in total spending but also through improvements in efficiency. International experience shows that authorities should work on three fronts to circumvent these constraints. The fronts are (a) coordinating with development partners so that infrastructure receives high priority in public expenditures; (b) improving efficiency of the infrastructure expenditures so that better service provision can be achieved with the same given level of resources, and (c) exploring and implementing additional options to involve the private sector in management, ownership, and financing infrastructure operations in the state- owned firms. The share of the private sector in infrastructure expenses in Cape Verde is relatively small. The private sector operates the fixed and mobile telecommunications through Cape Verde Telecom and holds approximately 30 percent of ELECTRA’s capital. In addition, there are a few performance-based maintenance contracts in roads.87 There are no precise data on private spending in infrastructure. The World Bank PPI database shows that private participation in ELECTRA represented 8 percent of GDP, or US$48 million, in 2001 (private participation in ELECTRA was reduced in 2006) and 4.1 percent in Cape Verde Telecom in 1995. The government so far has entrusted a major role to the private sector only in the telecommunications sector, aiming to mobilize private sector capital and management skills. This is not surprising as it is a worldwide trend. Privatization processes in telecommunications have been successful globally, mainly because the industry has proved to be easy to regulate and has grown rapidly. Budget process There are two main channels for public expenditures: on-budget and off-budget. On-budget expenditures are included in the central government budget and monitored as the budget is executed. Off-budget are the four SOEs that deliver infrastructure in Cape Verde: ELECTRA, TACV, ASA, and ENAPOR.88 They generate and spend a large share of the financial resources within their respective sectors. Figure 0.7Figure 0.7 is a road map to the allocation of public resources among various infrastructure sectors, institutional process, and various administrative entities. The main special fund relevant to infrastructure is the Road Maintenance Fund, which essentially is a mechanism to pass through road user levies to road maintenance. The fund became operational in 2006 and is implemented by a mechanism that transfers the proceeds from the point of collection directly to the fund. It has a minimum revenue guaranteed by the Treasury (CVE 300 million).89,90 The accounts of the new fund are public, and harmonization and accountability with public accounting are ensured. The mandate of the Road Agency is to maintain the national road network. The government decided that, pursuant to Cape Verde 87 A performance-based maintenance contract establishes that part of the operator’s revenue is conditional on some observable variable. For instance, for roads, it is the quality of the maintenance executed. Incentives for an efficient operation are transmitted via a conditional transfer. 88 Note that even though ELECTRA is a PPP with public management, this report considers ELECTRA a state- owned enterprise. 89 In 2006 customs transferred taxes to the Maintenance Road Fund. In 2007 and 2008 the Road Maintenance Fund was subsidized by the Treasury (transfers were however below the CVE 300 million). 90 The Council of Ministers approved a tax of CVE 7 per liter that entered into effect in January 2009. 90 legislation, public works should always be procured, coordinated, and supervised by the MITM through DGI. Thus, the Road Agency has no formal competence to procure, coordinate, or supervise road investments. Despite this fact, in 2007 the road agency was managing the construction of new roads.91 Finally, infrastructure expenditures made by local governments, mainly in the water sector, have been negligible, as funding has been largely provided from the central government budget. Figure 0.7: Institutional Responsibilities for Infrastructure Delivery92 Infrastructure Service Policy Umbrella Regulatory Functions Delivery Network expansion of electricity Ministry of Economic, Growth Economic Regulation Agency, water supply and waste water and Competitiveness General Directorate of Public Investment: 55% collection. Assets are transferred to Industry and Energy Total Spending: 15% Public O&M: 1% ELECTRA or municipalities (water) (Technical Regulation) National Roads: construction of new Ministry of Infrastructure and Instituto de Estradas On-Budget Shares on: roads. Transport Maintenance of Urban and Rural Municipalities Instituto de Estradas Roads Port and airport expansions and assets Ministry of Infrastructure and Instituto Maritimo e Portuario are transferred to ENAPOR and ASA Transport Maintenance of National Roads Road Fund Instituto de Estradas Airport Operations: Airport and Ministry of Infrastructure and Civil Aviation Agency Aviation Security (Aeroportos e Transport Segurança Area - ASA) Public Investment: 45% Air Transport - TACV Ministry of Infrastructure and Civil Aviation Agency Total Spending: 85% Transport Public O&M: 99% Electricity generation, distribution: Ministry of Economic, Growth General Directorate of Off-Budget Shares on: ELECTRA and Competitiveness Industry and Energy Water supply: ELECTRA, Ministry of Agriculture and Economic Regulation Agency Municipalities Environment Cellular services, fixed telephony, Ministry of Infrastructure and National Communication internet services: Cabo Verde Transport Agency Telecom Seaports and maritime transport: Ministry of Infrastructure and ENAPOR Transport Sanitation (Sewerage Services): Ministry of Infrastructure and Economic Regulation Agency ELECTRA Transport, Municipalities Source: Fiscal Cost Baseline Sectoral and Economic Allocation Important resources have been allocated to infrastructure when compared to the social sectors. During 2001–06, allocation to infrastructure corresponded to on average 16 percent of the budget During this same period, allocation to education and health combined averaged 24 91 This fact can hide a lack of lack of technical staff to oversee this contract at the DGI. On the other hand, it was mentioned during the field mission that the road agency needs to be reinforced in terms of technical staff. 92 On- and off-budget figures are measured on an execution basis (2006). 91 percent. Allocation to infrastructure in Cape Verde, as percent of total budget, also is high when compared to Kenya, Rwanda, Tanzania, or Uganda. 92 Table 0.10: Budget Allocation of Public Expenditure Between Infrastructure and Social Spending - A comparison Years Kenya Rwanda Tanzania Uganda Cape Verde Infrastructure Social sectors Infrastructure Social sectors Infrastructure Social sectors Infrastructure Social sectors InfrastructureSocial sectors 2001 11.7 40.6 11.7 35.3 16.4 30.7 15.4 26.1 2002 11.8 38.8 9.8 31.7 12.8 33 17.9 23.9 2003 14.9 38.6 10.5 27.5 12 31 14.8 22.8 2004 13.6 39.2 7.7 20.4 13.1 34.8 14.7 25.2 2005 17.1 37.5 11.7 25 11.9 25.9 2006 13.8 21.6 Source: Africa Infrastructure Country Diagnostic (2007). Note: Social sectors include education and health. During the analysis period, only one-third of public spending was channeled to asset creation. The remaining was spent on operation and maintenance (O&M) (salaries, inputs, and routine maintenance costs) (Table 0.11Table 0.11). Asset creation has been financed in almost equal shares by SOEs and the central government budget, because the financial positions of most SOEs do not allow them to carry out major infrastructure investments. On the other hand, almost all O&M expenditure is carried out by the SOEs with resources from tariffs and user fees. Finally, before the effective creation of the road agency in 2006, road maintenance was financed with on-budget expenditure. Table 0.11: On- and off-budget Infrastructure Expenditures Nominal GDP Shares (%) 2001 2002 2003 2004 2005 2006 INFRASTRUCTURE SPENDING 25.93% 32.88% 25.57% 26.82% 26.52% 26.85% Total Gross Investment 7.27% 10.10% 4.50% 6.60% 5.83% 6.96% Total O&M 18.66% 22.78% 21.07% 20.23% 20.70% 19.89% On-Budget 3.72% 3.01% 2.78% 3.78% 2.91% 4.12% Gross Investment 3.36% 2.78% 2.34% 3.51% 2.58% 3.85% O&M 0.36% 0.23% 0.45% 0.28% 0.33% 0.27% Parastatals/SOE 22.21% 29.87% 22.79% 23.04% 23.62% 22.61% Gross Investment 3.91% 7.32% 2.16% 3.09% 3.25% 3.11% O&M 18.30% 22.55% 20.63% 19.95% 20.36% 19.50% Road Fund - IE 0.13% O&M 0.13% Source: Fiscal Cost Baseline, WB PPI Database Air Transport (TACV) spending on infrastructure represents approximately 40 percent of the total of infrastructure spending (11 percent of GDP). From 2001–05, on average 97 percent of this spending was allocated to O&M. TACV spending on gross investment increased in 2006; yet the major share of the spending continued to be allocated to O&M (95 percent). This situation was reversed in 2007 with the introduction of private management. Furthermore, on average 20 percent of the total spending on infrastructure was allocated to electricity.93 Spending on electricity has decreased over time (as percentage of GDP), but it represented on average 5.4 93 It is important to note that ELECTRA does not produce a separated cost accounting sheet for each product. Therefore, the separation of electricity and water expenses is based on some assumptions. For instance, the electricity consumed by desalinization was priced using the share of the electricity consumed by desalinization multiplied by the oil cost used in generation. Another alternative is to price the energy by using the tariff that final users of electricity paid for it (opportunity cost). However, there is a circularity problem as the cost of electricity also depends on how the cost allocation of electricity and water is made. Finally, the allocation of common cost is another important issue. It is assumed that common cost in electricity and water was equivalent to the share of the revenues generated by each activity. An alternative can be to use the book value of the assets. 93 percent of GDP during the analysis period. Spending on roads has been comparatively low (on average 0.8 percent of GDP from 2001–05) but it increased in 2006 to 2.48 percent of GDP, because of the IDA Credit (Cape Verde Road Sector Support Project). Roads is the sector in which O&M is the lowest, and in which the difference between asset gross investment and O&M is larger, suggesting that maintenance is very low for the level of gross investment. Table 0.12: Public Expenditures by Infrastructure Sector Nominal GDP Shares (%) 2001 2002 2003 2004 2005 2006 Infrastructure Public Spending 25.57% 32.59% 25.06% 26.20% 26.28% 26.33% Total Gross Investment 6.94% 9.84% 4.02% 6.01% 5.61% 6.57% Total O&M 18.63% 22.75% 21.04% 20.19% 20.67% 19.76% Electricity 6.05% 8.92% 4.42% 4.44% 3.82% 4.92% Gross Investment 2.76% 5.17% 0.43% 0.31% 0.16% 0.47% O&M 3.29% 3.74% 3.99% 4.12% 3.66% 4.45% Water Supply and Sanitation 3.09% 2.67% 2.20% 3.77% 2.08% 2.14% Gross Investment 2.03% 1.42% 0.71% 2.44% 0.87% 0.85% O&M 1.06% 1.25% 1.49% 1.33% 1.21% 1.29% Roads 0.86% 0.62% 1.23% 0.75% 0.68% 2.48% Gross Investment 0.65% 0.55% 0.97% 0.63% 0.51% 2.33% O&M 0.20% 0.07% 0.26% 0.11% 0.18% 0.15% Airports 3.54% 4.77% 4.21% 4.86% 5.83% 4.71% Gross Investment 0.51% 1.64% 1.53% 2.14% 2.87% 1.89% O&M 3.03% 3.13% 2.67% 2.72% 2.96% 2.82% Seaports 1.75% 1.78% 1.95% 2.13% 2.04% 2.24% Gross Investment 0.00% 0.00% 0.17% 0.37% 0.24% 0.49% O&M 1.75% 1.78% 1.78% 1.76% 1.80% 1.76% Air Transport 9.50% 13.03% 10.99% 10.23% 11.21% 9.82% Gross Investment 0.23% 0.27% 0.18% 0.11% 0.38% 0.54% O&M 9.28% 12.75% 10.81% 10.11% 10.83% 9.28% Other infrastructures 0.79% 0.80% 0.06% 0.03% 0.62% 0.02% Gross Investment 0.76% 0.78% 0.03% 0.00% 0.60% 0.00% O&M 0.03% 0.02% 0.03% 0.03% 0.02% 0.02% Source: Fiscal Cost Baseline Donor resources play an important role, particularly in the roads and water sectors. According to OECD DAC statistics (table 5.13), Cape Verde received substantially more donor funding for infrastructure than its peers, measured as a percentage of GDP. From 2001 to 2004, ODA for infrastructure averaged 2.39 percent of GDP, while the corresponding figure for Sub- Saharan Africa (SSA) was 0.73 percent.94 Donors concentrate support in the road sector (57 percent) and in water supply and water resource management (36 percent). Donors finance 24 and 70 percent of the total spending on roads and water management, respectively. Even though the available data is not sufficiently detailed, it is known that donor funding is primarily allocated to finance investments. 94 According to the DGP, donor funding for infrastructure during that period was lower: 1.3% of GDP. 94 Table 0.13: Comparison of Official Development Assistance for infrastructure (% of GDP) Annual Sub-Saharan Cape Verde Senegal South Africa Average Africa 1995-1999 2.73% 2.58% 0.03% 0.77% TOTAL 2000-2004 2.39% 1.76% 0.04% 0.73% 1995-1999 0.76% 0.65% 0.00% 0.15% Energy 2000-2004 1.54% 0.25% 0.01% 0.15% Water and 1995-1999 0.52% 1.22% 0.02% 0.21% Sanitation 2000-2004 0.74% 0.95% 0.02% 0.20% 1995-1999 0.06% 0.05% 0.01% 0.04% Telecom 2000-2004 0.05% 0.01% 0.00% 0.02% 1995-1999 1.38% 0.65% 0.01% 0.37% Transport 2000-2004 1.54% 0.54% 0.01% 0.33% Source: OECD DAC Statistics. AFTPI Processed. Table 0.14: Sources of Public Spending for Infrastructure (A) Total Public (B) Public Resources 1/ Channelled Share of Nominal GDP Donor O$M Spending Investment Funded 2/ (including wages) (A+B) Electricity 4.45% 0.42% 0.05% 4.92% Energy w/o electricity 0.01% 0.16% 0.18% Water Supply 1.26% 0.58% 0.13% 1.96% Water resource management 0.01% 0.09% 0.24% 0.35% Sanitation 0.03% 0.14% 0.00% 0.18% Roads 0.15% 1.74% 0.58% 2.48% Airports 2.82% 1.89% 4.71% Seaports 1.76% 0.48% 0.01% 2.24% Air Transport 9.28% 0.54% 9.82% Other infrastructures 0.02% 0.02% TOTAL PUBLIC 19.78% 6.05% 1.02% 26.85% INFRASTRUCTURE Source: Fiscal Cost Baseline Notes: 1/ Includes budget, parastatals, and special funds, 2/ This is a very rough proxy for external financing in infrastructure based on figures provided by the Ministry of Finance. INVESTMENT SYSTEM95 Three aspects of Cape Verde’s Public Investment System (PIS) are of utmost concern for infrastructure spending: (a) project identification, appraisal and selection; (b) project implementation and completion; and (c) maintenance provision for existing assets. An evaluation of these three aspects of the PIS can shed light on how to improve the process of appraising and selecting projects to guarantee not only that projects with the highest priority and economic returns are financed in the first place but also that funds for their completion and subsequent maintenance are annually set apart and protected. The implicit consequence is that projects without appropriate economic returns––among other things because they are not aligned with development and economic strategies—are excluded from the MTEF and annual budgets, thus conserving funds to finance priority projects. 95 The structure of this section based on Rajaram and others (2006). 95 Project identification, appraisal, and selection Cape Verde has taken important - although still preliminary - steps toward improving its public investment system. The GPRSP-1 has been successfully incepted. It explicitly recognizes infrastructure as a key sector and provides strategic guidance on how infrastructure enters the overall development priorities.96 In particular, GPRSP-1 stated three guiding principles for transport sector development: (a) increased cost recovery via user fees; (b) improved accessibility of poor areas; and (c) a more active pursuit of public/private partnerships (PPP) to expand services. However, even though it recognized the risk embedded in contingent liabilities for PPP and existing SOEs, GPRSP-1 remained vague regarding the channels of private involvement. The Strategic Plan for Transports 2008–20, the latest planning document for infrastructure, indicates that the transports strategy is based on a market philosophy that espouses: (i) free entry in the market and competition; (ii) cost recovery prices; and (iii) private sector participation in the financing, construction, maintenance, and management of infrastructure. The GPRSP-2 reinforces the role of infrastructure as a lever for both growth and poverty reduction in Cape Verde. However, the GPRSP-2 as the GPRSP-1 does not indicate the channels through which the private sector will be involved. In infrastructure, as in the other sectors, the several planning instruments are not consistent. The MTEF (2005-2007) has played a very limited role in guiding annual budget preparation. Analysis of the GPRSP-1, MTEF, and annual budgets reinforces the view that these instruments are not aligned. For instance, in 2006, the GPRSP-1 indicated CVE 233 million for energy sector development, whereas the budget allocation and execution was approximately three times greater than what the GPRSP-1 and MTEF had envisaged (allocation was CVE 779 million and execution was CVE 716 million). In contrast, for the transport sector, budget allocation was between one-third and one-half (2005 and 2006, respectively) of the amount planned in the GPRSP-1 and MTEF.97 Table 0.15: Comparison GPRSP-1, MTEF, Annual Budget (investment budgets - CVE million) 2005 2006 2007 GPRSP-11 323 233 125 MTEF2 310 270 280 Budget allocation3 813 779 1033 Budget execution3 473 716 na Ratio execution/GPRSP-1 (%) 147 308 na Ratio allocation/MTEF (%) 262 289 369 Notes: 1 Energy sector development. 2 Combustibles and energy. 3 Energy sector development. 96 Cape Verde Joint IDA-IMF Staff Advisory Note on the Poverty Reduction Strategy Paper. Report No. 31019, December 7, 2004. 97 All figures refer to investment. 96 Table 0.16: Transports: Comparison GPRSP-1, MTEF, Annual Budget (investment budgets - CVE million) 2005 2006 2007 1 GPRSP-1 6,955 5,481 417 MTEF2 6,730 6,650 6,870 Budget allocation3 1,945 2,116 2,186 Budget execution3 2,121 2,656 Na Ratio execution/GPRSP-1 (%) 31% 49% Na Ratio allocation/MTEF (%) 29% 32% 32% Notes: 1 Transport infrastructure development. 2 Transports and communications. 3 Transport infrastructure development. The Cape Verde investment system has yet to provide operational links between the GPRSP-1 and a medium-term expenditure plan for infrastructure. After the preparation of the global MTEF for 2005–07, four sectoral MTEFs followed (agriculture, education, health, and social protection). There is no indication that a sectoral MTEF on infrastructure will be prepared on the basis of the latest global MTEF even though public investment in infrastructure is considered a priority. Furthermore, the global MTEF does not include sector-specific chapters, but only indicates global ceilings per ministry. Outcome targets also are missing for infrastructure sectors. This unfortunately implies that Cape Verde lacks institutional processes to identify key infrastructure bottlenecks and sector-wide investment approaches that can act as platforms for systematic participation of relevant stakeholders (donors, public, and private sector representatives). The definition and follow-up of outcomes for infrastructure, as well as for some other sectors, are deficient, largely because of the poor M&E system. The institutional arrangements for monitoring the GPRSP-1 were not fully in place at its inception, and, while improved, they remain weak. The government acknowledged in the first APR (July 2006) that “GPRSP-1 M&E (implementation of structures, data gathering, flow of information among sectors, the INE and the STAD) has shown little progress.�98 Reflecting the slow progress on M&E implementation, M&E use was generally limited. For example, the delay in the implementation of the 2006 Unified Survey of Well-Being Baseline Indicators (QUIBB) (owing to a lack of funding) meant that these data, which should have been a key input to GPRSP-1 implementation and to the Transport strategy, were not available. Box 5.1 shows Chile’s model– –a best practice approach to project identification, appraisal, and selection. On a more operational level, screening and economic evaluation of infrastructure projects are particularly deficient. The Ministry of Finance and the Ministry of Infrastructure do not have a unit or staff whose mandate is to conduct sound and verifiable cost-benefit analyses and complement the work of the Treasury in ensuring that proposals for financing remain consistent with overall debt sustainability considerations. The infrastructure sectors lack sector-specific policies to determine priorities; selection criteria; and economic, social, and financial analyses as 98 INE is the National Statistics Institute; STAD is the Technical Secretariat for Development Support. 97 part of investment project preparation. Furthermore, ex-post there is no impact evaluation to track deviation with respect to targets (also missing) and needs. Box 0.1: Chile’s Model of Investment Planning Chile has found a successful model based on a bottom-up approach of investment planning in each of the line ministries combined with strong oversight by the Budget Department of the Ministry of Finance. Chile has a long tradition of public expenditure evaluation, which is part of a broader government-wide performance measurement and evaluation system. Approximately 60 percent of government expenditure is subject to some form of evaluation. All investment projects are subject to cost-benefit analyses. Other key tools include cost-effectiveness analysis, shadow price estimation, training in project design and evaluation, and an online-integrated project database. Project selection is done at the ministry or agency level, and only projects that meet a minimum social rate of return are eligible for public funds. Chile also is advancing ex-post performance monitoring and evaluation. Source: IMF (2005). Project implementation and completion However, even if the projects with the highest economic returns are planned and selected, their implementation can be jeopardized by Cape Verde’s limited capacity to execute investment. In fact, despite recent efforts to improve the execution of investment budget allocation, at least one-third of annual investment allocations are not executed (table 5.16). Execution ratios are lowest in the energy, seaports, and water sectors and fluctuate worrisomely for airports, roads, and sanitation. Specific steps for improving project completion include further decentralization of financial management and budget execution (initiated in 2007) through SIGOF. However, it may take a couple of years before Cape Verde can capitalize the benefits of this measure. Better capital execution requires building stronger institutions, developing managerial capacity to deal with simultaneous and complex projects, and improving procurement practices with lenders and donors––all of which require time, sustained effort, and political commitment. Table 0.17: On-budget Investment (Execution) by Sector Ratio (%) Realized/Budgeted 2001 2002 2003 2004 2005 2006 Energy 31% 46% 56% 54% 31% 55% Water 41% 24% 32% 51% 39% 86% Water resource management 23% 28% 18% 50% 15% 41% Sanitation 72% 44% 13% 130% 78% 43% Roads 49% 26% 55% 46% 49% 98% Airports 1088% 33% 5681% 37% 1064% 39% Seaports 1% 54% 10% 13% 1% 10% Telecommunication and Post 2% 79% 5% 12% 2% 35% Other infrastructures 88% 79% 120% 121% 89% 144% TOTAL ON-BUDGET 58% 39% 44% 63% 59% 73% INFRASTRUCTURE Source: Fiscal Cost Baseline Furthermore, a major constraint on the public management, planning, and execution in infrastructure is lack of capacity. In certain areas, there is not enough staff with the necessary technical expertise. For instance, the supply of civil engineers in the country is small, and growth in their number has been limited (until last year) by the absence of a local engineering school. The increase in private sector demand for engineers resulting from the recent construction boom has made it very difficult for the Ministry of Infrastructure to recruit and retain qualified staff. 98 Cape Verde’s lack of capacity in infrastructure compromises adequate evaluation of projects, timeliness of implementation, and overall efficiency of the investments. Maintenance Provision for Existing Assets Provisioning for subsequent maintenance of created assets is the most efficient way of fully utilizing the economic returns and benefits of infrastructure investment projects. The process of appraising, selecting, and completing public investment projects needs to be linked in the appropriate way, that is, not only to the budget cycle but also to mid-term planning. The mid- term planning must include provisions for maintaining created assets. If Cape Verde cannot afford or guarantee maintenance of a new investment, it might be better off by not pursuing it in the first place but rather focusing on maintaining and rehabilitating the otherwise eroding existing infrastructure without spreading its resources too thinly. This issue is particularly important and challenging for road network projects appraised, implemented, and maintained by the central government. Due to lack of maintenance, some sections of the road network have deteriorated gravely. In other cases, lack of investment has left the basic network incomplete. These conditions are explained in part by both the lack of a systematic program for maintenance and the lack of financing. The latter leads to frequent postponements in the face of competing priorities. In 2006 a second-generation Road Maintenance Fund (RMF) was incepted abolishing the existing and non-functioning road fund (only part of its funds were reaching the RMF). However, the former’s full-fledged functioning has been characterized by difficulties. In 2006 almost all revenues from the road maintenance funds were used to finance 6 performance-based contracts covering maintenance needs for 35 percent of the road network. New roads are being constructed, and authorities expect to continue expanding the network in coming years. Thus, the challenge of provisioning for a satisfactory maintenance level will increase significantly, raising additional concerns about the ability of the Cape Verde’s institutions to provision for road maintenance. EFFICIENCY OF SPENDING Financial Viability of the State-Owned Enterprises When rationalizing Cape Verde’s scarce fiscal resources, identification and quantification of implicit and/or ad-hoc financial transfers to State-owned enterprises are necessary steps. While SOEs sometimes cover their own financial inefficiencies by postponing maintenance or reducing quality of service, more frequently than not, they do it by inducing financing transfers, for instance, by defaulting or accumulating debts. From the fiscal viewpoint, well-functioning SOEs would create at worst limited––at best no––fiscal costs. Given that the central government is a main (when not the only) stakeholder of SOEs, it bears all risks of inefficient financial behavior. Cumulative debts and triggered contingent liabilities would lead to periodic capitalization and debt swaps with important fiscal implications. In a macroeconomic context, this implies that fiscal resources are very inefficiently allocated, essentially because they are allocated without any economic and/or strategic rationale. These lumpy and ad-hoc fiscal 99 interventions deviate funds from priority infrastructure (and non-infrastructure) projects, instead throwing scarce resources to zero- or negative-return uses. Table 0.18: Financial Indicators for SOEs99 Indicator 2001 2002 2003 2004 2005 2006 ELECTRA Equity (%GDP) 1.4% -0.4% 0.2% -0.4% -0.4% 2.8% Liabilities (%GDP) 11.4% 16.8% 16.1% 16.0% 14.4% 8.7% Net Income (%GDP) -1.4% -1.7% -1.2% -0.9% -0.6% -1.4% Liabilities / Net Income - - - - - (6.1) TACV Equity (%GDP) 0.2% -1.8% -1.2% 0.2% -0.1% -0.1% Liabilities (%GDP) 7.0% 8.2% 7.3% 4.9% 5.2% 4.6% Net Income (%GDP) -0.3% -2.0% 0.5% 0.0% -0.3% -0.5% Liabilities / Net Income - - 15.81 - - - ASA Equity (%GDP) 8.5% 8.5% 10.2% 10.6% 11.6% Liabilities (%GDP) 3.4% 4.2% 8.4% 6.8% 8.2% Net Income (%GDP) 0.7% 0.2% -0.1% 0.0% 0.0% Liabilities / Net Income - 4.7 16.7 (147.0) 537.6 541.8 ENAPOR Equity (%GDP) 3.4% 3.4% 3.0% 3.0% 2.9% 2.6% Liabilities (%GDP) 1.9% 1.8% 1.9% 1.8% 1.5% 1.7% Net Income (%GDP) 0.1% 0.2% 0.0% 0.0% 0.1% 0.1% Liabilities / Net Income 13.61 8.20 46.93 47.97 13.96 12.33 Aggregate Equity (%GDP) TOTAL financial losses 9.7% 10.6% 13.0% 12.9% 16.8% of Liabilities (%GDP) 30.2% 29.5% 31.1% 27.9% 23.2% infrastructure SOEs amounted Net Income Net Income Liabilities / (%GDP) - -2.7% - -0.4% -70.5 -0.9% - -0.8% - -1.8% (13.1) to an annual average of 1.5 Source: Companies' Relatorios e Contas percent of the GDP for 2001-06 (table 5.18). The almost systemic nature of these losses raises flags concerning the financial health of the infrastructure SOEs. It also gives a good sense of the urgency of taking institutional and policy corrective Figure 0.8: World Crude Prices and Household measures––such as tariff redesign or Electricity Prices in Cape Verde adjustments—as opposed to ad-hoc emergency interventions. It is the 7,000 30.0 government of Cape Verde––in its 6,000 capacity of sole or major 25.0 stakeholder—that ultimately 5,000 20.0 assumes the total (or a substantial CV / KWh CV / Barrel 4,000 part) of the financial losses. The 15.0 combined net income deficit of 3,000 ELECTRA and TACV in 2002 10.0 peaked at 3.7 percent of the GDP 2,000 cost recovery. 1,000 Oil Electricity 5.0 ELECTRA is not able to recover 0 - basic operational costs (table 20 -4 20 -1 20 -2 20 -3 20 -4 20 -1 20 -2 20 -3 20 -4 20 -1 20 -2 20 -3 20 -4 20 -1 20 -2 20 -3 20 -4 -1 02 03 03 03 03 04 04 04 04 05 05 05 05 06 06 06 06 07 20 5.20). Several reasons contribute to ELECTRA’s poor financial performance: depend-ence on a technology that uses expensive fuels, poor bill collection,100 technical losses, and inadequate price-setting (this last is the responsibility of the regulator). 99 Equity has been estimated by subtracting the short- and long-term liabilities from the asset book value. 100 While it has improved, ELECTRA is still far from being a financially viable company.101 Improvements include the closing of gas-oil-input-based plants; greater bill collection recovery (even though there are still standing issues, such as the nonpayment of public illumination, arrears from municipalities, and electricity thieves), and some savings in service costs. Inadequate tariff-setting has long been a critical issue. The lack of adjustments in response to oil price increases (70 percent of electricity cost is fuel cost) resulted in financial deterioration and the resulting depletion of the company's capital (figure 5.9). The company began to accumulate a tariff deficit due to increasing fuel costs and other charges, partially because of the institutional failure of the Multisectoral Regulatory Agency (ARM) to adjust tariffs when fuel price increases were above the benchmark (agreed in January 2000). The Regulator (ARM) was never established; therefore, no adjustment mechanism was put in place (the Economic Regulatory Agency, or ARE, has been functioning since August 2004). This lack of adjustment gave latter origin to the “tariff deficit�: US$13.2 million over the next 5 years, in installments of US$2.63 a year. The financial situation of ELECTRA has had and will continue having fiscal implications, if measures to turn the company around are not undertaken. Its capital erosion in 2003 and 2004 led to a recapitalization (CVE 1.4 billion), for which the central government was called to participate as a shareholder. Another important fiscal cost was the direct subsidy to ELECTRA (covering the difference between the gas-oil purchase price and the CV 37.1 liter).102 In April 2006, the authorities eliminated the oil subsidies. While doing so put an end to the open-ended commitment of the direct subsidy, it exacerbated the under-pricing structure. The tariff increase for water and electricity tariffs in June 2006 and later in February 2007 were insufficient to compensate the increase in the price of fuel. As a result, a new tariff deficit has accumulated. In September 2007, ELECTRA’s treasury difficulties prompted a request for an urgent payment of the tariff deficit for May 2006–February 2007 in the amount of CVE 550 million (US$6.3 million). ARE implemented an automatic adjustment mechanism for electricity and water in February 2007, whose adjustment is triggered by adjustments in petroleum prices. None of these mechanisms is being applied as stated by law. TACV also is unable to recover basic operational costs. In the case of TACV, a highly competitive market caps its ability to increase tariffs, bringing to light important cost and market strategy inefficiencies.103 Should TACV wish to remain in the market as a financially viable and competitive enterprise, it should aggressively tackle its high production costs, look at best practices in the industry, improve managerial practices, and enhance revenues by, for instance, a web-based platform for ticking. Some improvements have been achieved with new management. They include operational restructuring that has led to increased efficiency and savings in fuel and 100 Bertolini and others (2006) report that the billing collection rates for public agencies and public companies are 65% and 57%, respectively. As of April 2006, the total for overdue electricity and water consumption bills was CVE 1.9 billion, or US$21 million. 101 Fernando Lecaros, “Evaluation of the Financial Situation and Perspectives for ELECTRA, 1: “Although there have been major improvements in controlling service costs, some of which are still underway, the operating savings have not been sufficient to turn the company around and making it profitable.� 102 In 2005 this subsidy was CVE382 million. 103 According to the Sterling Merchant Bank (TACV manager) Assessment Report, March 2007, “TACV has factually been operating in bankruptcy in all years since 2002 except 2004.� 101 crew costs, new commercial agreements, more rigorous control of overhead costs, and the sale of noncore assets and some retrenchment.104 Table 0.19: Cost coverage105 Annual Average, 2001–06 (% of operating firm’s revenues) Cost-Recovery Company Operating Overall (a) Overall (b) ELECTRA -1.5% -27.6% -33.0% Electricty -4.7% -31.6% -37.7% Water Supply and Sanitation 7.8% -15.9% -19.7% TACV -3.1% -4.4% -7.6% ASA 39.2% 15.7% 15.7% ENAPOR 30.6% 9.5% 8.7% Source: Fiscal Cost Baseline and Relatiorio e Contas Note: Operating cost: raw inputs, labor, taxes and subsidies ; Overall (a): operating cost plus depreciation, financial cost and extraordinary results; Overall (b): overall cost without subsidies, government's grants and extraordinary results ASA and ENAPOR achieved full cost recovery on average (2001–06), but have been unable to generate revenues to fund all their capital needs. This is an encouraging result, in particular when neither company received any operative subsidy from the Treasury. However, it seems that a share of capital expenditures was financed through on-budget resources. Similarly, for ports, the parastal ENAPOR is well positioned on cost recovery as it generates revenues that are larger that its cash operating costs. However, the company is not able to finance with own resources expansions and the modernization of the principal ports (the expansion of the ports in Praia and Palmeira will be financed by the central government––largely with foreign aid). In both sectors, the low scale of operation and sunk costs of these industries reinforce the inability of both companies to support large capital expenditures. Thus, they often require support from national or/and subnational governments. Table 0.20: Revenues and (operative) costs allocation, 2004–06106 2004 2005 2006 Airport Revenue / Revenue / Revenue / Cost Revenues Cost Revenues Cost Revenues cost cost cost Sal 47.5% 33.8% 95.9% 43.4% 27.1% 74.2% 38.5% 25.3% 78.5% Praia 10.5% 4.1% 52.8% 13.6% 5.0% 43.9% 15.9% 8.6% 64.8% S. Vicente 6.3% 1.0% 22.0% 4.8% 1.2% 28.8% 7.7% 1.4% 22.0% Boavista 4.2% 0.2% 5.5% 3.5% 0.0% 0.0% 6.1% 0.7% 14.1% Fogo 1.0% 0.2% 22.6% 0.9% 0.2% 23.3% 0.7% 0.4% 76.1% S. Nicolau 2.2% 0.2% 10.4% 1.8% 0.2% 10.8% 0.6% 0.3% 53.0% Maio 0.7% 0.2% 33.9% 0.6% 0.2% 35.0% 0.6% 0.1% 23.3% Total airports 72.3% 39.6% 73.8% 68.5% 33.8% 58.6% 70.2% 36.9% 62.8% Air traffic control 27.7% 60.4% 294.6% 31.5% 66.2% 249.6% 29.8% 63.1% 253.2% Total 100.0% 100.0% 116.3% 100.0% 100.0% 104.1% 100.0% 100.0% 103.4% Source: ASA - Relatorio e Contas Ground airport activities do not generate enough revenues (user fees) to guarantee an operational break-even (table 5.20). The most important source of revenues comes from air 104 Growth and Competitiveness Project Aide-Mémoire, May 2007. 105 Capital contributions financed by on-budget expenditures have been excluded from the analysis. 106 ASA reports do not provide revenues and costs. 102 traffic control (ATC).107 Using the cost allocation prepared by the company, cost recovery ratios by airport show that not all airports achieved cost recovery. Air traffic operations and safety oversight subsidize airport operations to ensure full cost recovery. Nevertheless, despite the fact that revenues from air traffic control have grown since 2004, cost recovery figures deteriorated during this period. the company does not provide detailed information in its financial statements on costs and revenue allocation per port,. Hidden Cost of ELECTRA108 Financial assessment of utilities falls short of capturing economic losses linked to suboptimal operations. Although embedded in cost structures, when quantified, the operational inefficiencies not only provide a lower bound of the opportunity cost of flawed administration and functioning but also give a sense of the fiscal impact of poor performance. Such operational inefficiencies are not captured in accounting reports despite the fact that in practice they represent hidden costs (HCs). In other words, they are unintended transfers from the public sector to the private sector of the economy in the form of under-pricing (transfer via price), collection inefficiencies (transfer via pending/uncollected bills), and unaccounted losses (transfer via pilferage and waste).109 These HCs are estimated by comparing actual indicators of a functioning SOE against ideal norms of cost recovery, distribution losses, and collection ratios (Ebinger 2006).110 Annex 1 provides the details and data used in the estimation. During 2001–06, Cape Verde’s HCs amounted to annual averages of approximately 1 percent of GDP in the electricity sector and 0.3 percent of GDP. Cape Verde’s HCs are high in comparison to those of East African countries. Only Tanzania in the electricity sector, with prices recently affected by a drought-triggered power crisis, creates higher HCs. Overall, as in other international experiences, electricity HCs are significantly higher than those created by the water sector. 107 The country is strategically located in the route from Europe to South America so all aircraft crossing the Cape Verdean skies pay a fee for safety and navigation oversight. 108 We conduct this analysis based on a methodology developed by J. Ebinger (2006) to calculate hidden costs for water, electricity and gas sectors. 109 There other types of transfers or quasi-fiscal costs, for instance, engrossed payrolls and subsidized input prices. See Briceño-Garmendia and Smits 2008. 110 Assumptions include standardized technical losses of 10% and 20% in electricity and water, respectively. A collection rate for the ideal company is assumed at 100%. 103 Figure 0.9: Hidden cost in Electricity and Water Distribution (annual averages for 2001–06 in GDP shares) 2.0% 1.0% Electricity Water 1.0% 0.5% 0.0% 0.0% Cape Burundi Kenya Rw anda Uganda Tanzania Cape Burundi Kenya Rw anda Uganda Tanzania Verde Verde Unaccounted Losses 2/ Under-Pricing 3/ Collection Inefficiencies 4/ Unaccounted Losses 2/ Under-Pricing 3/ Collection Inefficiencies 4/ Source: Based on data from ELECTRA (2006) using Hidden Cost methodology. For international comparators, see Briceno-Garmendia and Foster, 2007; Benitez, 2007. Unaccounted losses and collection inefficiencies have been consistent sources of HCs in both the electricity and the water distribution sectors. ELECTRA’s power operations are characterized by substantial energy losses (technical and commercial––approximately 22 percent of the energy produced, but figures differ across islands. These energy losses are in general below African standards but much higher than in developed countries. Electricity revenues also are affected by collection inefficiencies ranging between 88.0 percent and 96.5 percent of total electricity billed. In recent years, efforts were made to reduce unaccounted-for water and collection inefficiencies. Nevertheless, unaccounted-for water ratios remain at 20.0 percent and collection inefficiencies at approximately 91.0 percent. These, with increased production, still create annual average HCs equivalent to 0.2 percent of GDP. Under-pricing presents a huge variability, accounting for more than 60 percent of total HCs. The HC can be interpreted as a lower boundary of the fiscal gains that ELECTRA can accrue. That is, the hidden costs in the provision of electricity and water represent fiscal savings that the company could achieve by reducing technical losses, improving collection, and eliminating under-pricing. Naturally, there is no free lunch. These fiscal gains need to be backed up by new investments in the network and by a better management and collection policy. The resultant gross fiscal gains could achieve 1.3 percent of GDP per year. This is the lower limit. These gains could be higher if cost-saving technologies in electricity and water were implemented. 104 Table 0.21: Cape Verde’s Hidden Cost in Electricity and Water Distribution, 2001–06 (GDP shares) Electricity 1/ 2001 2002 2003 2004 2005 2006 % of Nominal GDP Unaccounted Losses 2/ 0.40% 0.35% 0.27% 0.28% 0.22% 0.45% Under-Pricing 3/ 0.84% 0.82% 0.24% 0.00% 0.00% 0.49% Collection Inefficiencies 4/ 0.18% 0.21% 0.34% 0.10% 0.25% 0.36% ELECTRICITY HIDDEN COST 1.42% 1.38% 0.84% 0.39% 0.47% 1.30% Water Supply 1/ 2001 2002 2003 2004 2005 2006 % of Nominal GDP Unaccounted Losses 2/ 0.10% 0.09% 0.12% 0.11% 0.11% 0.12% Under-Pricing 3/ 0.18% 0.28% 0.11% 0.00% 0.00% 0.09% Collection Inefficiencies 4/ 0.06% 0.06% 0.12% 0.03% 0.08% 0.10% WATER HIDDEN COST 0.34% 0.43% 0.35% 0.14% 0.19% 0.32% Source: Based on data from ELECTRA (2006) using Hidden Cost methodology. For international comparators, see Briceno-Garmendia and Foster 2007; Benitez 2007. Notes: 1 Indicators based on methodology described in Measuring Financial Performance in Infrastructure by Ebinger 2006. 2 Unaccounted Losses = (End-User Consumption)*(Average Cost Recovery Price)*(Total Loss Rate-Normative Loss Rate)/(1-Normative Loss Rate). 3 Under-pricing = End-User Consumption*(Average Cost Recovery Price––Average Actual Tariff). 4 Collection Inefficiencies = End-user Consumption*Average Actual Tariff*(1–– Collection Rate). EMERGING MESSAGES AND RECOMMENDATIONS The government has recognized the crucial role of infrastructure and is committed to improving sector performance and investments in forthcoming years. Therefore, as more scarce resources will be deployed, to create fiscal savings, better management of existing assets should be ensured. This is not an easy task, as many of the fiscal gains depend on costly investments. Therefore, authorities in Cape Verde will need to play a key role in improving the sector performance and the quality of the infrastructure services. This review provides some direction to improve service provision at a competitive and efficient rate. The planning of new infrastructure investments should be framed in the context of a MTEF for infrastructure. Over the past few years, various measures have been taken to modernize the budget process, adapt it to international norms, and make it more efficient. A MTEF that includes infrastructure will be an important tool to galvanize discussions around policy priority options and to encourage consultation with stakeholders during the planning and budgeting phases. To ensure the effectiveness of this instrument it is important to articulate it with the strategic plan for infrastructure (Master Plan) and with the budget. Also, it should be prepared on a rolling basis. Empirical evidence in other countries indicates that medium-term projections under the MTEF process are good guidance for annual budgets if fully integrated into budget formulation. Furthermore, in order to increase the planning of new infrastructure it is critical to improve the definition of outcomes for infrastructure. The master plan should include 105 indicators of output as well as outcome at different levels (project, company, region and country) that are measurable and possible to monitor over time. New projects, above a certain amount, should be subject to a cost-benefit analysis and to a careful analysis of economic viability. The Ministry of Finance should have a unit with the mandate to conduct cost-benefit analyses and ensure consistency with debt sustainability. The establishment of this unit, or reinforcement of an existing unit (by adding staff and broadening its mandate), could prevent the misuse of scarce resources. For example, a resources unit could be mobilized that is responsible for collecting the infrastructure investment needs from the sectors and for securing financing. Furthermore, this unit, which should work closely with the Treasury with regard to financing, could prevent the launch of projects that cannot be concluded, whether due to lack of resources or social or environmental implications. Feasibility evaluations are particularly important as Cape Verde already devotes a substantial amount of disputed public resources to infrastructure, at levels higher than those of countries with higher per capita income. Financial transfers amount to an annual average of 1.5 percent of GDP, without considering tax exemptions or special tax treatments that could not be tracked due to lack of data. The adequacy of the overall spending does not seem to be an issue. Concerns arise about the efficiency of the spending, given the low (or lack) efficiency gains due to the low economies of scale and large sunk costs among other factors. The room for increasing prices or tariffs is limited. For instance, ELECTRA and TACV face binding market conditions. In the case of ELECTRA, it is bound in the short run to diesel-based generation and is not able to cover operational costs, much less capital expenditures. As for TACV, operating in a very competitive market caps prices; thus, its viability depends on reducing costs. Furthermore, the expansion of infrastructure poses additional challenges to the already difficult financing of the maintenance of existing assets. Currently, only one-third of the road network is being maintained. With regard to projects in execution it is important to understand the reasons behind the delays with their implementation. It is critical to review and evaluate projects in execution in order to understand whether the low execution rate is due to procurement, overly optimist budgeting or low absorption capacity. Often times, too optimistic budgeting can lead to the impossibility of concluding projects (there are several cases of massive infrastructure investments in Africa that were never concluded due to unrealistic budgeting). To prevent that, it is important to rationalize the portfolio, coordinate budget cycles, and develop a system that evaluates realistically the financing conditions including direct contributions. Lack of maintenance of the existing assets can lead to the rapid deterioration of the assets. In Cape Verde approximately 60 percent of the roads are in bad condition due to lack of maintenance with some parts of the road network severely deteriorated. To prevent further deterioration, as rehabilitation costs are much larger than maintenance costs, it is crucial to ensure the financial autonomy of the Road Maintenance Fund. To that end, it is crucial to ensure that the fuel levy is collected and transferred to the Road Maintenance Fund. In addition, technical capacity should be enhanced at the Roads Institute. The inefficiencies of some state companies can jeopardize the freeing-up of resources for new investments. The aggregated financial losses of the state companies reached an annual 106 average of 1.6 percent of GDP during 2001-2006. The almost systemic nature of those loses raise concerns about the financial sustainability of those companies and their implication to the State budget. Therefore, it is recommended to develop a monitoring and evaluation system of the performance of those companies on a regular basis (including the publication of annual reports). Moreover, partnering with the private sector should be pursued as a way to reduce fiscal risks and free public resources. The participation of the private sector could bridge the gap between available and required funding. Partnership with the private sector also would allow sharing the risk and would free scarce public resources for sectors in which the private sector has no interest, such as social projects. 107 References Afro barometer. 2005. D. Reis, F. Rodrigues, and J. Semedo. “Atitudes Em Relação À Qualidade Da Democracia Em Cabo Verde.� Working Paper 53. Albuquerque, C., and others. 2006. “Gestão das Finanças Públicas.� Coleção Gestão Pública. ANMCV. 2003. Analise Financeira dos Municipios de Cabo Verde de 1996 a 2002. Banco Central de Cabo Verde. Relatório do Conselho de Administração. Various years. Banerjee, S., V. Foster, Y. Ying, H. Skilling, and Q. Wodon. 2008. “Achieving Cost Recovery, Equity and Efficiency in Water Tariffs: Evidence from African Utilities.� Africa Infrastructure Country Diagnostic. Briceno-Garmendia, C., and V. Foster. 2007. “More Fiscal Resources for Infrastructure? Evidence from East-Africa.� Africa Infrastructure Country Diagnostic. Briceno-Garmendia, C., and M. Shkaratan. 2008. “Cost Recovery and Efficiency in Electricity Tariffs: Evidence from African Utilities.� Africa Infrastructure Country Diagnostic. Briceno-Garmendia, C. and Smits, K., 2008 Fiscal Costs of Infrastructure in Sub-Saharan Africa, Africa Infrastructure Country Diagnostic. Ebinger, J. 2006. “Measuring Financial Performance in Infrastructure.� World Bank. Thornton, G. 2007. Tourism Draft Report. IFC (International Finance Corporation). 2007. “MCC Cape Verde Sector Growth and Investment Program: Phase II Report. Air Transport Sector.� IMF (International Monetary Fund). 2003a. “Article IV Consultation, First Review under the Poverty Reduction and Growth Facility, and Request for Waiver of Performance Criteria.� _____. 2003b. “Second Review of the Poverty Reduction and Growth Facility Arrangement.� _____. 2004. “Under the Three-year Arrangement under the Poverty Reduction and Growth Facility, and Request for Waiver of Performance Criteria. _____. 2005. “Staff Report to the 2005 Article IV Consultation, Sixth Review under the Poverty Reduction and Growth Facility, and Request for Waiver of Performance Criteria.� _____. 2008. “Third Review under the Policy Support Instrument-Staff Report.� Macamo, J. L. 2007. “Diagnóstico da Descentralizacao Fiscal em Cabo Verde.� Field Report. Ministry of Finance and Public Administration. 2005. “Etude de la Dette Publique Interne, Etat/Municipalités/Instituts Autonomes/Entreprises Publiques et Parapubliques.� Rapport Provisoire. _____. State Accounts. 2002 to 2006. Ministry of Infrastructure and Transports. 2008. “Plano Estratégico de Transportes, 2008–2010.� PEP Africa. 2007. “Cape Verde Maritime Transportation. First Revision of Draft Report. Pereira and others. 2005. “Economia e Finanças Públicas.� Escolar Editora. 108 Potter, B., and J. Diamond. 1999. “Guidelines for Public Expenditure Management.� International Monetary Fund. Rajaram, A., T. Minh Le, and N. Biletska. 2006. “Public Investment Efficiency: Guidance Note on an Indicator-Based Approach.� World Bank. Mimeo. Holloway, S. 1997. “Straight and Level: Practical Airline Economics.� Burlington, VT: Ashgate Publishing. Sterling Merchant Finance, Ltd. 2007. “Privatization of TACV: Potential International Partners and Networks.� Unidade de Acompanhamento Económico. Several reports. World Bank. 2002. “Financial Modeling of Regulatory Policy.� _____. 2003. CFAA Report. _____. 2004. “Cape Verde: The Role of the State.� _____ and IMF. 2004. “Cape Verde Joint IDA-IMF Staff Advisory Note on the Poverty Reduction Strategy Paper.� Report 31019. _____. 2005. Road Sector Project. Report No. 32170. _____. 2006a. “Cape Verde: The Challenge of Increasing Fiscal Space to Meet Future Pressures.� _____. 2006b. “Cape Verde: Infrastructure Reform, Regulation and Competitiveness.� PPIAF. _____. 2007a. Growth and Competitiveness Project, Aide-Mémoire. May. _____. 2007b. “Cape Verde Investment Climate Assessment.� _____. 2007c. “Evaluation of the Situation and Perspectives for ELECTRA.� Draft Working Paper. 109 Annex 1: Joint Bank-Fund Debt Sustainability Analysis [w1] Prepared by the staffs of the International Development Association and the International Monetary Fund Approved by Sudhir Shetty and Carlos Braga (World Bank) and David Nellor and Philip Gerson (IMF) November 7, 2008 The risk of debt distress in Cape Verde remains low. Nevertheless, medium-term fiscal policy will reverse the public debt decline of recent years. The total public debt-to-GDP ratio is projected to rise until 2012 and then decline thereafter—a path opposite that projected in the 2007 DSA. The temporary rise in external debt will be only partially offset by continued decline in domestic debt. Despite the rise, debt ratios remain manageable in all scenarios. Foreign direct investment (FDI) will finance most of the external current account deficit, which will narrow as Cape Verde transforms itself into a services exporter. The main risks to the debt outlook are currency exposure and contingent liabilities. The risk of debt distress remains low under the baseline as well as alternative scenarios that take those risks into consideration. A. BACKGROUND111 1. This DSA reviews the evolution of Cape Verde’s public debt since the 2007 DSA112 and analyzes the projected debt path for the period 2008–28. Using the Fund-World Bank debt sustainability framework (DSF), it projects the baseline economic scenario and performs stress tests to assess whether debt distress will stay low. The thresholds for public external debt distress are those for countries like Cape Verde that have sound policies and institutions (Table 1).113 The baseline scenario was updated based on discussions with the authorities during the fifth review of the Policy Support Instrument (PSI) (September–October 2008). The discussions centered on the 2009 budget and the medium-term fiscal framework the authorities submitted to Parliament in October 2008 along with the 2009 budget. 2. Since the last DSA Cape Verde has continued to reduce public debt as a percentage of GDP and to change its composition (Table 2). Total public debt (domestic plus external) was reduced by 10 percentage points in 2007. Net domestic debt was pushed down to the original PSI benchmark of 20 percent of GDP two years ahead of schedule; it is likely to reach 14 percent of GDP by year-end, thanks to expenditure restraint as well as buoyant revenues. The proportion of domestic debt in total debt was also reduced, reflecting efforts to reach out to 111 This analysis includes only central government debt and guarantees; it excludes municipalities and state-owned enterprises. 112 IMF Country Report 08/37 and World Bank Report 44350-CV. 113 Cape Verde’s score on the World Bank’s Country Policy and Institutional Assessment (CPIA) was upgraded in 2007 from 4.1 to 4.2. Its average score for 2005–07 was 4.13, above the 3.75 floor for strong performers. 110 development partners for concessional financing, making it possible to replace domestic with mostly concessional external borrowing. All external funds borrowed in 2007 were concessional. Cape Verde’s main external creditors are IDA and the African Development Fund (Table 2). While the credit crunch in Europe is making it hard to roll over the nonbank private external debt, this totaled only 8 percent of GDP as of the end of 2007 and is mainly long-term. Table 1. Cape Verde: Central Government External Debt Ratios Baseline Scenario 1 Thresholds 2008 2018 2028 NPV of debt as a percentage of: GDP 50 25 26 21 Exports 200 56 42 30 Revenues 2 300 99 104 83 Debt service as percentage of: Exports 25 5 5 4 2 Revenues 35 8 11 11 Source: Ministry of Finance; and staff estimates. 1 Based on Cape Verde's 2005-07 classification as a strong performer. 2 Excluding grants. 3. The depreciation of the dollar in 2007 and 2008 was favorable to Cape Verde, but it revealed open currency positions (Tables 2 and 5). The nominal external debt-to-GDP ratio declined by 5 percentage points despite the fact that the dollar value of the country’s nominal external debt grew by US$ 58 million in 2007 (4 percent of GDP). This is because the nominal GDP measured in dollars grew by 20 percent boosted by the appreciation of the escudo relative to the dollar. The open currency exposure to the dollar results from the fact that the external liabilities of the Treasury are denominated mainly in US$ and SDR (which contains dollars), and the net foreign assets of the central bank are mostly in euros. This raises questions about whether the authorities should swap part of their foreign reserves in euros for dollars to cover the outstanding open positions or should prefer that future loans be denominated in euros. The authorities have made commitments in the PSI and PRSC series to improve debt management, and the Fund and the Bank together will provide technical assistance (TA) on debt management in addition to the TA Cape Verde receives from Portugal. 111 Table 2. Cape Verde: Central Government Debt, 2004–08 2004 2005 2006 2007 2008 Actual Proj. (Percent of nominal debt) Total 100 100 100 100 100 Nominal external debt 66 64 65 68 74 Multilaterals 51 50 51 57 64 Official bilaterals 13 12 12 10 8 Commercial 2 2 1 1 1 Domestic debt1 34 36 35 32 26 (Units indicated) Total US$ million 842 802 919 959 936 Percent of GDP 89 81 75 65 52 Nominal external debt US$ million 553 513 598 656 694 Percent of GDP 60 51 50 45 38 Domestic debt1 US$ million 288 289 322 303 243 Percent of GDP 29 30 25 20 14 Memorandum item: GDP (US$ million) 918 1,006 1,203 1,443 1,826 Source: Cape Verdean authorities, staff estimates and projections. 1 Net of deposits and obligations with the Trust Fund. B. MEDIUM-TERM BASELINE SCENARIO 4. The long-term macroeconomic scenario is projected to revolve over the next 20 years around two axes: economic transformation toward a service-based economy, and accumulation of international reserves and government deposits at the BCV. 5. The growth forecast is designed to test the robustness of the conclusions of this DSA. Because of the financial and commodity price shocks in 2008, short-term growth was revised downward and inflation upward compared to the previous DSA (Table 3). For the outer years, the previous assumptions are maintained: real GDP is expected to grow by 5 percent in the long term (5 to 20 years), which is a prudent 2 percentage points below the historical average (1 standard deviation). Growth will be driven by the transformation into a service-exporting economy, financed mostly by FDI. Moreover, the projections do not consider the growth- promoting effect of public investment in infrastructure. 112 Table 3. Cape Verde: Macroeconomic Baseline Assumptions, 2008–28 Average 5 Years Ahead 6-20 Years Ahead (1998-2007) 2007 DSA Current DSA 2007 DSA Current DSA Real GDP growth rate (percent) 7.2 7.3 6.6 5.2 5.1 Inflation rate (percent) 3.1 2.3 3.4 2.0 2.0 Exports of goods and services (percent of GDP) 32 47 48 58 65 Imports of goods and services (percent of GDP) 65 78 71 82 78 Financing needs (percent of GDP)1 … -1.0 -2.5 -2.7 -0.4 Grant element of new external borrowing … 28 16 10 9 Public revenue and grants (percent of GDP) 30 29 28 28 27 Primary public deficit (percent of GDP) 4.9 1.6 2.7 1.3 0.8 Source: National authorities, staff estimates and projections. 1 Current account plus foreign direct investment. 6. The economic transformation is marked by an increase in imports and service exports and by a decrease in reliance on remittances and other current transfers. Cape Verde is expected to break its past dependence on aid and remittances as it continues to transform itself into a self-propelled economy. While the fuel and food shock increased the import bill in 2008, the restraint in recurrent Balance of Payments Source of Financing, 1995-2007 expenditures created 1995 2007 fiscal space that has Tourism ODA Tourism enabled the government 4% FDI 12% Remittances 19% 14% 49% ODA to expand social transfers 38% to protect the vulnerable without putting pressure on the balance of Remittances 46% FDI 18% payments. As a result, Source: National authorities foreign reserves will stay The transformation of the Cape Verdean economy above 3 months of 100 imports and continue to Balance of payments 80 grow through the forecast horizon. FDI 60 Imports of goods Exports of goods and services will drive and finance GDP percent and services the current account 40 deficit and keep debt- Current transfers generating flows close to 20 Debt-generating balance. The debt- flows generating inflows 0 Current account needed to finance the -20 current account deficit 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 are likely to be largely unaffected if FDI is below baseline projections owing to the self stabilizing dynamics of the current account relative to FDI (imports would decline in tandem with FDI thanks to its high import content). Despite a possible deceleration caused by global financial turmoil, the prospects for FDI inflows in the medium term continue to be bright. New commitments approved by the Investment Promoting Agency support the expectation that the private investments will materialize. 113 7. The baseline scenario projects that international reserves will continue to accumulate, as will government deposits at the BCV. The prudent fiscal policy implemented in the PSI is assumed to continue through the forecast horizon, with foreign reserves building up. This assumption is based on two facts: a. In October 2008 the authorities submitted to Parliament a medium-term fiscal framework for 2009–2011 that indicates a prudent fiscal policy. Although the policy reverses the recent decline in public debt, it preserves a stable debt path that allows for public investments in infrastructure and social transfers. b. The authorities announced in the Letter of Intent for the 5th PSI review that they intend to continue with a PSI for at least four more years, until 2013 (a 1-year extension of the current PSI followed by a request for a new 3-year PSI). Based on these facts, net domestic borrowing is projected to be contained in the next 20 years, allowing net domestic debt114 to land softly at about 11 percent of GDP. This fiscal restraint is needed to accomplish the authorities’ goal of increasing reserve coverage by 0.1 month of prospective imports each year, reaching 5.7 months by 2028 (equivalent to 41 percent of GDP). Financing the reserve accumulation requires that the Treasury make annual deposits of about 1.2 percent of GDP at the BCV. Using the balance sheet approach, this result assumes that the authorities’ efforts to develop the domestic securities market will allow the domestic private sector to absorb about 19 percent of GDP in Treasury securities by 2028 (Table 4). % GDP The accumulation of net foreign assets (NFA) and Treasury deposits 45 40 Balance sheet of the Bank of Cape Verde (projections) Treasury deposits Treasury Treasury 35 Treasury deposits 15 deposits Treasury deposits NFA 12 30 deposits 9 6 5 41 NFA 25 NFA NFA M0 M0 M0 M0 37 20 NFA 28 33 M0 19 20 21 22 22 23 15 10 T-bills/bonds BCV-bills 5 5 5 TCMF and 5 4 other assets Net worth 4 0 2008 2008 2013 2013 2018 2018 2023 2023 2028 2028 As sets Liabilities As sets Liabilities As sets Liabilities Assets Liabilities Assets Liabilities 8. The baseline scenario assumes a faster rise than the previous DSA in the share of nonconcessional external borrowing. While Cape Verde will continue to have access to concessional loans from IDA and others,115 this DSA assumes that Cape Verde will 114 Net of government deposits. 115 Cape Verde is a “blend country� as it is eligible to IBRD and IDA funds (under the “small island exception�). 114 increasingly take out nonconcessional loans to finance growth-enhancing public investments. It is assumed that the average grant element of all external borrowing will decline to less than 10 percent by 2028. This assumption is justified by the recent graduation of Cape Verde from the U.N.’s least-developed country category and nonconcessional loans envisaged with the European Investment Bank (EIB), the IBRD, and the OPEC Fund. This assumption is useful for probing the resilience of the debt path to less favorable borrowing terms.116 To further test resilience in stress scenarios, the grant element of marginal debt117 is negative because it is assumed that under stress conditions the country would be charged a risk premium of 100 basis points above the market rate.118 Table 4. Balance Sheet Approach: Intersectoral Positions with the Treasury, Selected Items (in percentage of GDP) Treasury Bank of Cape Verde 2008 2028 2008 2028 Assets 7 15 Assets 31 43 Deposits at the BCV 5 15 Net Foreign Assets 22 41 Deposits in banks 2 0 Claims on the Treasury 3 0 Other items (net) and TCMF 7 1 Liabilities1/ -20 -26 Liabilities -31 -43 With the BCV -3 0 Money base -19 -23 With banks -11 -6 BCV bills (sterilization) -4 -4 With non-banks -7 -19 Deposits of the Treasury -5 -15 Equity capital -4 -1 Net domestic debt -14 -11 Domestic Banking Sector Domestic non-banking sector 2008 2028 2008 2028 Assets 71 85 Assets … … Net Foreign Assets 2 0 Treasury 7 19 Reserve money and cash 12 15 Bank deposits 69 84 Treasury securities 11 6 Cash 7 8 BCV bills (sterilization) 4 4 Private credit 46 60 Other items (net) and TCMF -4 -1 Liabilities -71 -85 Liabilities … … Deposits -69 -84 Bank loans -46 -60 Deposits of the Treasury -2 0 Source: IMF and IDA staffs' projections. 1/ Excludes TCMF. C. EXTERNAL DEBT SUSTAINABILITY BASELINE SCENARIO 9. Although the recent decline in external debt will be temporarily reversed because borrowing to finance public investments will accelerate, it will remain below the 116 The Fund and the Bank will provide TA to Cape Verde on debt management to enable the authorities to ensure that future nonconcessional borrowing is consistent with debt sustainability, especially because the nonconcessional borrowing will double the external interest bill through 2028. 117 “Marginal debt� is debt taken to cover the gap created by the shock simulated in the stress scenarios. 118 CIRR (Commercial Interest Reference Rate). 115 threshold. In the previous DSA, external debt was expected to decline continuously. The reason for the difference is the new funds Cape Verde recently secured for public investments, especially from the EIB and the IBRD. The finding that this temporary rise in external borrowing will not jeopardize debt sustainability repeats the finding of the 2007 DSA that a 5- year scaling-up of nonconcessional borrowing is consistent with debt sustainability. The average grant element of the new borrowing will be especially low during 2009–2011 when the EIB loan will be disbursed. The debt service ratios will rise gently but stay below the stress thresholds. This rise in debt service indicators results from the decline in concessional financing and the assumed shortening of amortization periods. Because in this DSA the grant element of new borrowing is projected to decline faster than in the previous one, the rise in debt service ratios will be frontloaded rather than backloaded, as it was in the previous DSA. ALTERNATIVE AND STRESS SCENARIOS 10. The risk of external debt distress is low even with depreciation and an abrupt worsening of borrowing terms. The debt ratios remain far below their thresholds in all alternative and stress scenarios, including the scenario where all new borrowing is 200 basis points above the baseline rates (Figure 1 and A2 Table 6). This finding reinforces the conclusion that nonconcessional borrowing is unlikely to jeopardize debt sustainability. The extreme scenario is a currency depreciation, which highlights the need to hedge open currency positions to support the peg. This result is a corollary to the cautionary note about the country’s currency exposure (¶ 3).119 In the historical scenario, the external debt ratio rises for a longer period of time because FDI is less than in the baseline, but it also declines faster in the outer years because the historical scenario implies faster growth and a smaller external deficit. The historical scenario should be interpreted with caution because it does not take into account that in a highly open economy like Cape Verde the current account self-stabilizes to some extent to fluctuations of FDI and growth. D. TOTAL DEBT SUSTAINABILITY BASELINE SCENARIO 11. The trajectory of total public debt contrasts with the previous DSA because it reverses the decline observed in recent years. In the baseline scenario, the NPV of total public debt as a percentage of GDP is expected to rise until 2013 and decline thereafter. In the previous DSA it was expected to decline in the short term and stabilize in the outer years. The rising trajectory is expected in spite of the faster decline in domestic debt because external borrowing is expected to be larger than in the previous DSA in order to finance public investment in infrastructure. Yet, public debt is sustainable because the baseline scenario maintains the assumption of the previous DSA that the fiscal policy pursued in the PSI and PRSC series to preserve sustainability will continue through 2028. In particular, the government is expected to hold domestic debt at about 11 percent of GDP, which will require it to decelerate public investment over time to make space to pay the interest on current 119 The hump-shaped path for the historical scenario (Figure 1, red dotted line) was preserved for completeness of this DSA. However, it is not informative because the baseline assumptions on the current account and GDP growth are more pessimistic than the historical scenario. 116 nonconcessional borrowing. The expectation is supported by the medium-term fiscal framework the authorities submitted to Parliament in October 2008 and by the depth of their commitment to sound policies. Sales of coastal land to tourism developers will also enhance fiscal performance. Therefore, we find that debt is sustainable in the baseline scenario. ALTERNATIVE SCENARIOS AND STRESS TESTS 12. Although the macroframework is robust to alternative assumptions and shocks, the DSA highlights the importance of fiscal discipline. The alternative scenarios tested are120 (i) real GDP and primary balance at historical averages; (ii) primary balance unchanged from 2008; and (iii) permanently lower GDP growth.121 All debt ratios remain within sustainable levels under all the alternative scenarios. In the extreme stress test for the debt-to-GDP ratio, the annual fiscal deficit is 10 percent of GDP for 2009–2010, which is 7 percent of GDP larger than the baseline primary balance. This simulates a situation where, for example, hypothetical contingent liabilities equivalent to 7 percent of GDP materialize in two consecutive years. For the debt service ratio, the extreme stress is a 30 percent depreciation of the escudo. All debt ratios remain manageable during the forecast horizon under all stress tests. 13. While conclusive information is not yet available, contingent liabilities arising from state-owned enterprises may be a risk for the debt outlook. This risk was taken in consideration in the risk assessment of this DSA under the extreme stress test for the debt-to- GDP ratio. This test shows that all debt ratios remain manageable even if contingent liabilities amounting to 7 percent of GDP materialize in two consecutive years. There are explicit and implicit risks: the explicit guarantees122 provided to state-owned enterprises represent a fiscal risk of 4.5 percent of GDP at the end of 2008. The implicit fiscal risks are currently being assessed by the government, especially the amounts needed to recapitalize some state-owned enterprises. For example, the electricity and water supplier, Electra, had losses equivalent to 1.3 percent of GDP in 2007, which erased about half of its net worth, and it may suffer further losses in 2008 and the following years until the more efficient generators now under construction start operating. The government is taking action supported by the PSI to assess these fiscal risks by compiling an aggregate balance sheet and the net gains or losses of the largest state-owned corporations; it will report the fiscal risks to the Council of Ministers early in 2009. This should encourage prompt action to address the risks, such as allowing the private sector to participate in infusing capital into state-controlled enterprises. E. CONCLUSIONS 14. The DSA concludes that the risk of debt distress is low and highlights Cape Verde’s strengths as well as vulnerabilities. Even with extreme shocks, public debt is on a sustainable path, given continued fiscal discipline and the economic transformation caused by expansion of service exports and FDI. While the decline in public debt observed in recent years is being reversed, the DSA shows that using nonconcessional funds to expand public investment will not jeopardize debt sustainability as long as the expansion is temporary and recurrent expenditures 120 Table 8, alternative scenarios A1–3. 121 Assumes that real GDP growth is at the baseline minus one standard deviation divided by the length of the projection period. 122 Domestic debt issued by state owned enterprises. 117 remain controlled. This conclusion holds even if the expansion of public investments in infrastructure does not generate the expected growth returns, because no growth-enhancing effect of infrastructure is assumed. It is, however, critical that Cape Verde strengthen its debt management. In particular, it needs to conduct DSAs regularly to set a borrowing envelope for the next year’s budget and an MTFF consistent with debt sustainability. Two important vulnerabilities identified in the DSA also need to be addressed: the public sector’s unhedged currency exposures and the contingent liabilities for state-owned enterprises. The authorities are preparing to address these vulnerabilities as they firm up their debt management. 118 Table 5. Cape Verde: External Debt Sustainability Framework, Baseline Scenario, 2005-2028 1/ (In percent of GDP, unless otherwise indicated) Actual Historical 0 Standard Projections Average 0 Deviation 2008-2013 2014-2028 2005 2006 2007 2008 2009 2010 2011 2012 2013 Average 2018 2028 Average External debt (nominal) 1/ 68.9 60.8 60.5 49.7 48.9 46.9 46.6 44.8 42.7 46.6 35.2 33.7 33.7 o/w public and publicly guaranteed (PPG) 51.0 49.7 45.5 38.0 39.7 39.3 39.3 38.4 37.0 30.7 22.5 Change in external debt -10.7 -8.1 -0.3 -10.8 -0.8 -2.0 -0.3 -1.8 -2.1 -1.9 1.4 Identified net debt-creating flows -11.1 -15.5 -9.9 -9.5 -1.0 -2.2 -2.4 -3.5 -3.6 -3.9 0.0 Non-interest current account deficit -0.6 2.2 7.3 7.3 4.1 9.7 9.4 8.9 8.9 8.7 8.5 9.0 4.6 2.3 3.9 Deficit in balance of goods and services 27.8 25.4 28.7 24.7 26.4 23.4 23.2 23.0 22.3 15.4 9.9 Exports 36.5 41.6 43.2 31.9 44.0 45.2 47.2 48.8 49.5 50.7 47.5 61.5 71.8 64.6 Imports 64.3 67.0 71.9 65.3 68.6 71.6 70.6 72.0 72.5 73.0 71.4 76.9 81.8 78.4 Net current transfers (negative = inflow) -27.8 -24.2 -21.4 -25.4 2.0 -14.7 -17.0 -14.6 -14.7 -14.8 -14.3 -15.0 -11.6 -7.8 -10.5 o/w official -4.6 -4.1 -4.4 -4.4 -5.4 -3.1 -3.2 -3.3 -2.9 -1.7 -0.5 Other current account flows (negative = net inflow) -0.7 1.0 -0.1 -0.3 0.0 0.1 0.4 0.4 0.5 0.8 0.2 Net FDI (negative = inflow) -7.5 -9.2 -8.9 -6.6 2.9 -8.5 -8.0 -8.0 -8.5 -8.9 -9.1 -8.5 -6.8 -1.2 -5.1 Endogenous debt dynamics 2/ -3.0 -8.5 -8.2 -10.7 -2.4 -3.1 -2.8 -3.3 -3.0 # -1.7 -1.0 Contribution from nominal interest rate 4.0 2.8 1.9 2.0 1.7 1.8 1.8 1.8 1.8 # 1.6 1.4 Contribution from real GDP growth -4.7 -6.2 -3.5 -2.8 -2.8 -3.1 -3.0 -2.8 -2.7 # -1.7 -1.5 Contribution from price and exchange rate changes -2.3 -5.0 -6.6 -9.8 -1.3 -1.8 -1.5 -2.2 -2.1 # -1.7 -0.9 Residual (3-4) 3/ 0.4 7.4 9.6 -1.3 0.2 0.3 2.0 1.7 1.5 # 1.6 2.0 o/w exceptional financing 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 # 0.0 0.0 PV of external debt 4/ ... ... 42.9 36.2 36.2 35.1 35.5 34.8 33.9 30.3 32.4 In percent of exports ... ... 99.4 82.3 80.3 74.5 72.9 70.3 66.8 49.3 45.2 PV of PPG external debt ... ... 27.9 24.5 27.0 27.5 28.2 28.4 28.1 25.8 21.3 In percent of exports ... ... 65 56 60 58 58 57 55 42 30 In percent of government revenues ... ... 110 99 109 112 117 117 115 104 83 Debt service-to-exports ratio (in percent) 17.9 11.2 8.0 8.1 7.0 6.1 5.7 5.8 6.0 5.8 4.7 PPG debt service-to-exports ratio (in percent) 8.5 5.6 4.7 4.7 4.3 3.9 3.7 4.0 4.3 4.7 4.0 PPG debt service-to-revenue ratio (in percent) 13.2 9.8 7.9 8.3 7.9 7.5 7.5 8.1 8.9 11.5 11.3 Total gross financing need (billions of U.S. dollars) 0.0 0.0 0.0 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.5 Non-interest current account deficit that stabilizes debt ratio 10.1 10.3 7.6 20.5 10.2 10.8 9.2 10.4 10.6 6.4 0.9 Key macroeconomic assumptions Real GDP growth (in percent) 6.5 10.8 6.9 7.2 2.5 6.0 6.1 7.0 7.1 6.8 6.7 6.6 5.0 5.0 5.0 GDP deflator in US dollar terms (change in percent) 2.9 7.9 12.2 4.4 10.6 19.4 2.8 3.8 3.4 5.0 5.0 6.6 3.5 2.7 3.4 Effective interest rate (percent) 5/ 5.5 4.8 3.8 17.4 26.9 4.2 3.8 4.0 4.2 4.3 4.5 4.2 4.8 4.5 4.9 Growth of exports of G&S (US dollar terms, in percent) 24.1 36.3 24.5 17.3 13.2 28.8 12.1 16.0 14.4 13.9 14.6 16.6 12.2 8.5 11.2 Growth of imports of G&S (US dollar terms, in percent) 0.6 24.6 28.7 14.3 12.3 20.8 13.8 9.4 12.9 13.0 12.7 13.8 9.8 8.3 9.4 Grant element of new public sector borrowing (in percent) ... ... ... ... ... 10.2 14.4 19.0 17.7 16.4 14.4 15.4 9.2 5.7 8.2 Government revenues (excluding grants, in percent of GDP) 23.7 23.9 25.5 24.7 24.8 24.5 24.2 24.3 24.4 24.9 25.6 25.1 Aid flows (in billions of US dollars) 7/ 0.1 0.1 0.1 0.1 0.2 0.2 0.2 0.3 0.4 0.6 1.0 o/w Grants 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 0.1 o/w Concessional loans 0.0 0.0 0.0 0.0 0.1 0.1 0.2 0.3 0.3 0.5 0.9 Grant-equivalent financing (in percent of GDP) 8/9/ ... ... ... 5.1 5.9 4.6 3.2 3.5 3.2 2.3 1.1 2.0 Grant-equivalent financing (in percent of external financing) 8/9/ ... ... ... 61.0 52.5 54.8 45.1 47.1 46.9 42.9 26.3 37.9 Memorandum items: Nominal GDP (billions of US dollars) 1.01 1.20 1.44 1.83 1.99 2.21 2.45 2.74 3.07 4.76 10.59 (PVt-PVt-1)/GDPt-1 (in percent) 3.1 5.0 3.5 3.7 3.6 3.1 3.7 1.6 1.4 1.6 Source: Staff simulations. 1/ Includes both public and private sector external debt. 2/ Derived as [r - g - r(1+g)]/(1+g+r+gr) times previous period debt ratio, with r = nominal interest rate; g = real GDP growth rate, and r = growth rate of GDP deflator in U.S. dollar terms. 3/ Includes exceptional financing (i.e., changes in arrears and debt relief); changes in gross foreign assets; and valuation adjustments. For projections also includes contribution from price and exchange rate changes. 4/ Assumes that PV of private sector debt is equivalent to its face value. 5/ Current-year interest payments divided by previous period debt stock. 6/ Historical averages and standard deviations are generally derived over the past 10 years, subject to data availability. 7/ Defined as grants, concessional loans, and debt relief. 8/ Grant-equivalent financing includes grants provided directly to the government and through new borrowing (difference between the face value and the PV of new debt). 9/ Numbers not comparable with the 2007 DSA, which mistakenly excluded grants. 119 Table 6. Cape Verde: Sensitivity Analysis for Key Indicators of Public and Publicly Guaranteed External Debt, 2008-2028 (In percent) Projections 2008 2009 2010 2011 2012 2013 2018 2028 PV of debt-to GDP ratio Baseline 25 27 28 28 28 28 26 21 A. Alternative Scenarios A1. Key variables at their historical averages in 2008-2028 1/ 25 26 26 28 29 30 32 23 A2. New public sector loans on less favorable terms in 2008-2028 2 25 29 31 32 33 33 34 34 B. Bound Tests B1. Real GDP growth at historical average minus one standard deviation in 2009-2010 25 27 28 29 29 29 27 22 B2. Export value growth at historical average minus one standard deviation in 2009-2010 3/ 25 30 38 38 36 34 27 21 B3. US dollar GDP deflator at historical average minus one standard deviation in 2009-2010 25 30 33 34 34 34 31 26 B4. Net non-debt creating flows at historical average minus one standard deviation in 2009-2010 4/ 25 25 21 23 24 24 25 21 B5. Combination of B1-B4 using one-half standard deviation shocks 25 25 26 27 28 28 28 23 B6. One-time 30 percent nominal depreciation relative to the baseline in 2009 5/ 25 38 39 40 40 40 36 30 PV of debt-to-exports ratio Baseline 56 60 58 58 57 55 42 30 A. Alternative Scenarios A1. Key variables at their historical averages in 2008-2028 1/ 56 58 56 57 59 60 53 32 A2. New public sector loans on less favorable terms in 2008-2028 2 56 65 65 66 67 66 55 47 B. Bound Tests B1. Real GDP growth at historical average minus one standard deviation in 2009-2010 56 60 58 58 57 55 42 30 B2. Export value growth at historical average minus one standard deviation in 2009-2010 3/ 56 72 98 92 87 81 53 36 B3. US dollar GDP deflator at historical average minus one standard deviation in 2009-2010 56 60 58 58 57 55 42 30 B4. Net non-debt creating flows at historical average minus one standard deviation in 2009-2010 4/ 56 55 45 47 48 48 41 30 B5. Combination of B1-B4 using one-half standard deviation shocks 56 54 52 54 54 54 44 31 B6. One-time 30 percent nominal depreciation relative to the baseline in 2009 5/ 56 60 58 58 57 55 42 30 PV of debt-to-revenue ratio (excluding grants) Baseline 99 109 112 117 117 115 104 83 A. Alternative Scenarios A1. Key variables at their historical averages in 2008-2028 1/ 99 105 108 114 120 124 130 89 A2. New public sector loans on less favorable terms in 2008-2028 2 99 118 126 133 136 137 136 132 B. Bound Tests B1. Real GDP growth at historical average minus one standard deviation in 2009-2010 99 111 116 121 121 119 107 86 B2. Export value growth at historical average minus one standard deviation in 2009-2010 3/ 99 122 157 155 149 141 109 83 B3. US dollar GDP deflator at historical average minus one standard deviation in 2009-2010 99 119 136 142 142 140 126 101 B4. Net non-debt creating flows at historical average minus one standard deviation in 2009-2010 4/ 99 100 86 94 98 100 100 83 B5. Combination of B1-B4 using one-half standard deviation shocks 99 101 104 112 115 116 112 91 B6. One-time 30 percent nominal depreciation relative to the baseline in 2009 5/ 99 154 158 165 165 162 146 117 120 Table 6. Cape Verde: Sensitivity Analysis for Key Indicators of Public and Publicly Guaranteed External Debt, 2008-2028 (continued) (In percent) Projections 2008 2009 2010 2011 2012 2013 2018 2028 Debt service-to-exports ratio Baseline 5 4 4 4 4 4 5 4 A. Alternative Scenarios A1. Key variables at their historical averages in 2008-2028 1/ 5 4 4 3 4 5 6 6 A2. New public sector loans on less favorable terms in 2008-2028 2 5 4 4 4 4 4 3 3 B. Bound Tests B1. Real GDP growth at historical average minus one standard deviation in 2009-2010 5 4 4 4 4 4 5 4 B2. Export value growth at historical average minus one standard deviation in 2009-2010 3/ 5 5 6 9 8 8 7 5 B3. US dollar GDP deflator at historical average minus one standard deviation in 2009-2010 5 4 4 4 4 4 5 4 B4. Net non-debt creating flows at historical average minus one standard deviation in 2009-2010 4/ 5 4 3 2 2 3 4 4 B5. Combination of B1-B4 using one-half standard deviation shocks 5 4 3 2 3 3 4 4 B6. One-time 30 percent nominal depreciation relative to the baseline in 2009 5/ 5 4 4 4 4 4 5 4 Debt service-to-revenue ratio Baseline 8 8 8 7 8 9 11 11 A. Alternative Scenarios A1. Key variables at their historical averages in 2008-2028 1/ 8 8 7 7 8 10 16 16 A2. New public sector loans on less favorable terms in 2008-2028 2 8 8 8 8 8 8 7 7 B. Bound Tests B1. Real GDP growth at historical average minus one standard deviation in 2009-2010 8 8 8 8 8 9 12 12 B2. Export value growth at historical average minus one standard deviation in 2009-2010 3/ 8 8 10 14 14 15 15 11 B3. US dollar GDP deflator at historical average minus one standard deviation in 2009-2010 8 9 9 9 10 11 14 14 B4. Net non-debt creating flows at historical average minus one standard deviation in 2009-2010 4/ 8 8 6 3 4 6 9 11 B5. Combination of B1-B4 using one-half standard deviation shocks 8 8 6 5 6 7 11 12 B6. One-time 30 percent nominal depreciation relative to the baseline in 2009 5/ 8 11 11 11 11 13 16 16 Memorandum item: Grant element assumed on residual financing (i.e., financing required above baseline) 6/ -4 -4 -4 -4 -4 -4 -4 -4 Source: Staff projections and simulations. 1/ Variables include real GDP growth, growth of GDP deflator (in U.S. dollar terms), non-interest current account in percent of GDP, and non-debt creating flows. 2/ Assumes that the interest rate on new borrowing is by 2 percentage points higher than in the baseline., while grace and maturity periods are the same as in the baseline. 3/ Exports values are assumed to remain permanently at the lower level, but the current account as a share of GDP is assumed to return to its baseline level after the shock (implicitly assuming an offsetting adjustment in import levels.) 4/ Includes official and private transfers and FDI. 5/ Depreciation is defined as percentage decline in dollar/local currency rate, such that it never exceeds 100 percent. 6/ Negative numbers indicate interest rates higher the market rates. Applies to all stress scenarios except for A2 (less favorable financing) in which the terms on all new financing are as specified in footnote 2. 121 Figure 1. Cape Verde: Indicators of Public and Publicly Guaranteed External Debt under Alternatives Scenarios, 2007-2028 1/ a. Debt Accumulation b.PV of debt-to GDP ratio 6 20 18 50 5 Threshold: 50 16 14 40 4 12 3 10 30 8 2 20 6 4 1 2 10 0 0 2008 2013 2018 2023 2028 0 Rate of Debt Accumulation Grant element of new borrowing (% right scale) 2012 2017 2022 2027 Grants and implicit grants (% of GDP) d. PV of debt-to-revenue ratio (excluding grants) 125 c.PV of debt-to-exports ratio 175 Threshold: 200 Threshold: 300 150 100 125 75 100 75 50 50 25 25 0 0 2012 2017 2022 2027 2012 2017 2022 2027 e.Debt service-to-exports ratio f.Debt service-to-revenue ratio 15 25 Threshold: 25 Threshold: 35 20 10 15 10 5 5 0 0 2012 2017 2022 2027 2012 2017 2022 2027 Baseline Historical scenario Most extreme shock: 1/ 2007 DSA Source: Staff projections and simulations. 1/ The most extreme stress test is the test that yields the highest ratio in 2018. In figure b. it corresponds to a depreciation shock; in c. to a borrowing cost shock; in d. to a depreciation shock; in e. to a export shock; and in picture f. to a depreciation shock. 122 Table 7. Cape Verde: Public Sector Debt Sustainability Framework, Baseline Scenario, 2005-2028 (In percent of GDP, unless otherwise indicated) Actual Estimate Projections Standard 2008-13 2014-28 Average 2005 2006 2007 deviation 2008 2009 2010 2011 2012 2013 Average 2018 2028 Average Public sector debt 1/ 84.0 65.4 62.5 52.2 51.5 51.3 51.4 50.4 48.8 50.9 41.9 33.1 39.0 o/w foreign-currency denominated 53.7 47.2 42.7 38.7 39.6 39.2 39.2 38.4 37.0 30.7 22.5 Change in public sector debt 0.4 -18.6 -2.9 -10.3 -0.7 -0.2 0.2 -1.0 -1.6 -1.5 0.0 Identified debt-creating flows 6.6 -13.4 -10.8 -8.9 -0.6 -1.0 -0.6 -0.8 -1.3 -1.2 0.2 Primary deficit 4.2 3.1 -0.9 4.9 5.6 -0.4 4.1 3.2 2.8 3.6 2.8 2.7 0.8 0.9 0.8 Revenue and grants 30.0 29.4 30.3 29.7 2.2 29.5 29.9 28.2 26.5 27.0 27.0 28.0 26.9 26.6 26.8 of which: grants 6.3 5.5 4.8 4.7 5.0 3.7 2.4 2.7 2.6 2.0 0.9 Primary (noninterest) expenditure 34.2 32.5 29.4 29.0 33.9 31.4 29.3 30.7 29.9 27.7 27.5 Automatic debt dynamics 2.7 -16.1 -8.7 -7.0 -3.6 -3.5 -3.4 -3.9 -3.8 -1.9 -0.7 Contribution from interest rate/growth differential -5.4 -9.6 -4.6 -3.9 -2.6 -2.9 -3.0 -3.0 -2.8 -1.4 -0.6 of which: contribution from average real interest rate -0.2 -1.4 -0.3 -0.4 0.4 0.4 0.3 0.3 0.4 0.6 1.0 of which: contribution from real GDP growth -5.1 -8.2 -4.2 -3.5 -3.0 -3.4 -3.4 -3.3 -3.1 -2.0 -1.6 Contribution from real exchange rate depreciation 8.1 -6.5 -4.1 -3.0 -1.0 -0.6 -0.4 -0.9 -1.0 ... ... Other identified debt-creating flows -0.3 -0.4 -1.3 -1.5 -1.1 -0.6 0.1 -0.4 -0.3 -0.1 0.0 Privatization and land sales (negative) -0.3 -0.4 -1.3 -1.5 -1.1 -0.6 0.1 -0.4 -0.3 -0.1 0.0 Recognition of implicit or contingent liabilities 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Debt relief (HIPC and other) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Other (specify, e.g. bank recapitalization) 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 Residual, including asset changes -6.2 -5.2 7.9 -1.4 -0.1 0.7 0.7 -0.3 -0.3 -0.3 -0.2 Other Sustainability Indicators PV of public sector debt 30.3 45.4 46.0 38.5 38.8 39.5 40.4 40.4 40.0 39.6 37.0 31.9 35.1 o/w foreign-currency denominated 0.0 27.2 26.2 25.0 27.0 27.5 28.2 28.4 28.1 25.8 21.3 o/w external ... 27.2 26.2 25.0 27.0 27.5 28.2 28.4 28.1 25.8 21.3 PV of contingent liabilities (not included in public sector debt) 1.6 5.6 5.1 4.5 4.1 3.7 3.3 3.0 2.7 1.7 0.8 Gross financing need 2/ 9.7 10.5 9.4 4.5 4.2 8.3 7.0 6.4 7.4 5.6 5.6 PV of public sector debt-to-revenue and grants ratio (in percent) 101 155 152 131 130 140 152 150 148 137 120 PV of public sector debt-to-revenue ratio (in percent) 128 190 181 156 157 161 167 166 164 148 124 o/w external 3/ … 114 103 101 109 112 117 117 115 104 83 Debt service-to-revenue and grants ratio (in percent) 4/ 7.3 6.2 5.4 5.6 5.5 5.6 5.5 5.7 5.9 6.0 6.3 Debt service-to-revenue ratio (in percent) 4/ 9.2 7.6 6.5 6.7 6.6 6.4 6.1 6.3 6.5 6.5 6.6 Primary deficit that stabilizes the debt-to-GDP ratio 3.8 21.7 2.0 9.9 4.7 3.4 2.6 4.7 4.4 2.3 0.9 Key macroeconomic and fiscal assumptions Nominal GDP (local currency) 89.2 105.6 116.1 132.1 146.2 161.6 178.1 198.9 221.7 343.6 764.2 Real GDP growth (in percent) 6.5 10.8 6.9 7.2 2.5 6.0 6.1 7.0 7.1 6.8 6.7 6.6 5.0 5.0 5.0 Average nominal interest rate on forex debt (in percent) 1.1 1.1 1.1 1.3 0.3 1.4 1.5 2.0 2.3 2.5 2.8 2.1 3.2 4.0 3.4 Average real interest rate on forex debt (in percent) 2.9 -2.5 -6.3 3.1 10.5 -10.8 5.8 3.8 2.4 0.7 0.8 0.5 2.6 5.9 3.7 Real exchange rate depreciation (in percent, + indicates depreciation) 0.2 -5.2 -16.1 -3.4 21.2 -23.1 ... ... ... ... ... ... ... ... ... Inflation rate (GDP deflator, in percent) 2.9 7.9 12.2 4.4 10.6 19.4 2.8 3.8 3.4 5.0 5.0 6.6 3.5 2.7 3.4 Growth of real primary spending (deflated by GDP deflator, in percent) 0.1 0.1 0.0 0.0 0.2 0.0 0.2 0.0 0.0 0.1 0.0 0.1 0.0 0.1 0.0 Grant element of new external borrowing (in percent) ... ... ... … … 10.2 14.4 19.0 17.7 16.4 14.4 15.4 9.2 5.7 ... Sources: Country authorities; and staff estimates and projections. 1/ Central government. Debt figures are net of deposits at central bank. 2/ Gross financing need is defined as the primary deficit plus debt service plus the stock of short-term debt at the end of the last period. 3/ Revenues excluding grants. 4/ Debt service is defined as the sum of interest and amortization of medium and long-term debt. 5/ Historical averages and standard deviations are generally derived over the past 10 years, subject to data availability. 123 Table 8. Cape Verde: Sensitivity Analysis for Key Indicators of Public Debt 2008-2028 Projections 2008 2009 2010 2011 2012 2013 2018 2028 PV of Debt-to-GDP Ratio Baseline 38 39 39 40 40 40 37 32 A. Alternative scenarios A1. Real GDP growth and primary balance are at historical averages 38 39 42 45 46 47 57 77 A2. Primary balance is unchanged from 2008 38 34 31 29 26 22 14 4 A3. Permanently lower GDP growth 1/ 38 39 40 42 43 43 46 63 B. Bound tests B1. Real GDP growth is at historical average minus one standard deviations in 2009-2010 38 40 42 44 45 45 46 48 B2. Primary balance is at historical average minus one standard deviations in 2009-2010 38 46 53 54 53 52 47 40 B3. Combination of B1-B2 using one half standard deviation shocks 38 43 48 49 49 49 47 45 B4. One-time 30 percent real depreciation in 2009 38 49 49 49 49 48 46 45 B5. 10 percent of GDP increase in other debt-creating flows in 2009 38 39 50 50 49 47 42 34 PV of Debt-to-Revenue Ratio 2/ Baseline 131 130 140 152 150 148 137 120 A. Alternative scenarios A1. Real GDP growth and primary balance are at historical averages 128 132 148 168 169 173 212 293 A2. Primary balance is unchanged from 2008 128 115 111 110 95 83 51 13 A3. Permanently lower GDP growth 1/ 128 132 143 158 157 159 170 236 B. Bound tests B1. Real GDP growth is at historical average minus one standard deviations in 2009-2010 128 133 149 165 165 166 170 179 B2. Primary balance is at historical average minus one standard deviations in 2009-2010 128 153 190 203 196 192 176 152 B3. Combination of B1-B2 using one half standard deviation shocks 128 143 171 186 182 180 175 173 B4. One-time 30 percent real depreciation in 2009 129 164 173 185 180 177 170 169 B5. 10 percent of GDP increase in other debt-creating flows in 2009 128 130 176 187 180 175 155 128 Debt Service-to-Revenue Ratio 2/ Baseline 16 14 13 14 14 15 18 20 A. Alternative scenarios A1. Real GDP growth and primary balance are at historical averages 16 14 14 16 18 19 32 59 A2. Primary balance is unchanged from 2008 16 14 8 7 6 3 0 0 A3. Permanently lower GDP growth 1/ 16 14 14 14 15 16 23 44 B. Bound tests B1. Real GDP growth is at historical average minus one standard deviations in 2009-2010 16 14 14 15 16 17 24 33 B2. Primary balance is at historical average minus one standard deviations in 2009-2010 16 14 21 26 21 23 28 28 B3. Combination of B1-B2 using one half standard deviation shocks 16 14 18 21 19 21 26 32 B4. One-time 30 percent real depreciation in 2009 16 16 17 18 19 20 28 40 B5. 10 percent of GDP increase in other debt-creating flows in 2009 16 14 13 14 14 15 18 20 Sources: Country authorities; and staff estimates and projections. 1/ Assumes that real GDP growth is at baseline minus one standard deviation divided by the length of the projection period. 2/ Revenues are defined inclusive of grants. 124 Figure 2. Cape Verde: Indicators of Public Debt Under Alternative Scenarios, 2007-2028 1/ Baseline Primary balance at 2008 level Most extreme shock: primary balance 2007 DSA 60 PV of Debt-to-GDP Ratio 50 40 30 20 10 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 250 PV of Debt-to-Revenue Ratio 2/ 200 150 100 50 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 45 Debt Service-to-Revenue Ratio 2/ 40 35 30 25 20 15 10 5 0 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 Sources: Country authorities; and staff estimates and projections. 1/ The most extreme stress test is the test that yields the highest ratio in 2018. 2/ Revenues are defined inclusive of grants. 125 Annex 2: The Organizational System Organizational System (Planning and Budgeting, Financial Management …) Administrative Structure IT Support Planning and Budgeting System The Planning and Budgeting System comprises a group of structures with specific functions that will act in an integrated way to make viable the development of actions and decision-making processes related to the preparation, monitoring, and evaluation of plans, programs, and budgets. The Planning System will be characterized by the group of actions and decisions conducive to the preparation, approval, administration, and evaluation of the Plan. Analogously, the Budgeting System will be understood as the group of actions and decisions conducive to the preparation, approval, administration, and evaluation of the budget. It is suggested that the Planning and Budgeting System cover, among other things, the following lines of responsibility, beyond the preparation of socioeconomic studies and surveys: • Preparation of the national strategic plan. • Formulation of national, sector, and regional plans aimed at promoting economic and social development. • Preparation of budget guidelines and annual budgets. • Administration of the planning and budgeting process. • Promotion of links with municipalities, so as to encourage the harmonization of norms and related tasks. With decentralization, DGO and DGP, as well as the sector structures responsible for planning and budgeting functions, will be integrated into the Planning and Budgeting System. Other entities can also be integrated into the organizational system of planning and budgeting as long as their functions are related to planning and budgeting. • DGP, as the central structure in the Planning System, will have the following responsibilities: • Coordination, orientation, and supervision of the Plan and special development projects preparation. • Identification, analysis, and evaluation of government strategic capital investments and their articulation with private investments, as well as provision of institutional and management support in their implementation. • Coordination and orientation of evaluation activities of the Plan and special development projects. 126 • DGO, as the central structure in the Budgeting System, will have the following responsibilities: • Coordination, consolidation, and supervision of the budget guidelines and budget proposal preparation in articulation with DGP. • Establishment of the necessary norms required to prepare and implement budgets. • Proposal of measures to improve the Planning and Budgeting System. • Monitoring, without damage to the responsibilities of other structures, the physical and financial budget execution. • Preparation of studies and surveys aimed at developing and improving the budget process. • Technical orientation, coordination, and supervision of the budget sector units. • Establishment of the functional, institutional, and programmatic classifications of public expenditure, in articulation with DGP. • Planning and coordination of the activities related to data technologies for budgeting. The Planning and Budgeting System sector structures will play the role of establishing the links between central structures and decentralized budgetary units. They will have, among others, the following responsibilities: • Planning, coordinating, and supervising execution of activities related to planning, budgeting, financial management, and accounting systems in that ministry. • Promoting articulation with the central structure, as well as data and guidance regarding the ministerial structures, insofar as the established norms and regulations are concerned. • Coordinating, preparing, and consolidating the ministry’s plans and programs, as well as forwarding them to a higher decision-making body. • Monitoring and promoting projects and evaluation of activities. The Financial Management System The Financial Management System will have DGT as its central structure and the State Financial Balance as its main function. In this context, it will include the following activities: financial planning, control of duties and assets, and control of guarantees and obligations that are the responsibility of Treasury, as well as technical and normative orientation insofar as budget and financial execution is concerned. DGT, as the central structure, and the line structures will be part of the Financial Management System. The line structures will receive normative orientation and technical supervision from the central structure of the System, regardless of subordination to the body under which the administrative structure is integrated.123 The structures in charge of activities of the Financial Management System will have the following responsibilities: 123 It is important that there exist technical subordination to DGT of its representatives in fiscal councils or equivalent structures in institutions directly or indirectly controlled by the government. 127 • Ensuring the financial balance of the Treasury. • Managing the real estate and financial assets of the Treasury. • Preparing the financial plans of the Treasury, administering the Single Treasury Account, and contributing to the preparation of the public expenditure funding policy. • Managing the Treasury’s public and external debt. • Controlling debt resulting from credit operations under the direct or indirect responsibility of the Treasury. • Managing Treasury credit operations. • Maintaining control of commitments that represent, directly or indirectly, state payments to international organizations. • Preparing and disseminating norms of financial planning and financial and budget execution, as well as promoting the monitoring, systematization, and standardization of public expenditure execution. • Promoting integration among government sectors regarding issues of financial planning and administration. The Public Accounting System DGPC will play the role of central structure in the Public Accounting System, which will have as its main objective the registration of the acts and facts related to the state budget, finance, and equity administration. The system will provide data regarding the following: • Operations undertaken by governmental structures or entities and their consequences affecting the equity position of the state. • Resources of the current budget and its changes, estimated and effective revenues collected, expenditure submitted to commitment analysis, and liquidated and paid expenditure against these resources, as well as the availability of financial resources. • The position of all these who, regardless of circumstances, collect revenues, have expenses, and administer or keep goods and chattels belonging to state institutions or that are currently entrusted to it. • The asset position of public institutions and its variation. • The costs of Public Administration programs and units. The Accounting System will comprise the activities of registration, treatment, and control of the state budget, financial, and asset administration operations. The main objective will be to prepare accounting statements. To implement asset accounting in the public sector, these operations that are a result of debits and credits of a financial nature and are not integrated into the budget execution will also be a target for registration, individualization, and accounting control. As far as the administrative structure of the Public Accounting System is concerned, it will be composed of DGCP (as the central structure) and line structures. The line structures are the internal management units of the ministries and will be the object of normative orientation and technical supervision by the central structure in the System and of subordination to the body 128 under which the administrative structure is integrated. It will be the responsibility of the units to integrate the Public Accounting System to do the following: • Maintain and improve the National Plan of Public Accounts. • Establish norms and procedures to promote adequate accounting registration of acts and facts related to the budgetary, financial, and asset management of Public Administration structures and entities. • Based on illegal or irregular acts or facts detected, undertake the relevant registration and adopt the necessary measures to encourage responsibility of the agent by reporting this fact to the authority to whom the agent is subordinated and to the structure responsible for the Internal Control System. • Establish, maintain, and improve the data systems that render possible the accounting of acts and facts related to the state budgetary, financial, and asset management, as well as prepare management data for ministerial supervision and decision making. • Prepare quarterly preliminary accounts and General State Accounts. • Promote integration with other government structures insofar as accounting issues are concerned. The Internal Control System The Executive Internal Control System will be implemented by an entity, which could be called State General Inspectorate (IGE), having as its main function the evaluation of the public administrator’s management. IGE will be directly subordinated to the Prime Minister’s office to provide it with the necessary authority in carrying out its responsibilities. The System will use as its main instruments auditing and inspection and will have the following objectives: • Evaluating compliance with the objectives established in the plans, insofar as the execution of government programs and budgets is concerned. • Undertaking the concomitant control of budgetary execution and supervising the juridical and financial correctness of budgetary and financial management, legality, regularity, economy, effectiveness, and good financial management. • Verifying legality and evaluating results concerning the effectiveness and efficiency of the Public Administration’s budgetary, financial, and asset management, as well as the use of public resources by private entities. • Controlling credit operations, guarantees, and state duties and assets. • Supporting external control as part of its institutional mission. The Internal Control System will evaluate executive actions through supervision of accounting, financial, budgetary, and operational acts and assets and will support external control as part of its institutional mission. The Executive Internal Control System will have as its administrative structure IGE and the line structures, which will be under the normative orientation and technical supervision of IGE, without detriment to their subordination to the body under which the administrative structure is to be integrated. They will: 129 • Evaluate the accomplishment of the goals established in plans. • Supervise and evaluate the execution of government programs, including decentralized actions paid with resources from public budgets, as far as the level of achievement of the goals and objectives and the management quality are concerned. • Evaluate the execution of budgets. • Control state credit operations, guarantees, duties, and assets. • Provide data about the physical and financial position/situation of projects and activities included in the budget. • Audit management of public resources under the responsibility of public and private entities. • Discover illegal or irregular acts or facts, undertaken by public or private agents, while using public resources and, when appropriate, inform the responsible accounting unit. • Undertake audits in the accounting, financial, budgetary, and human resources systems, as well as other administrative and operational systems. The Assets System DGPE will be the central structure of the Assets System, which will have as its main objective the administration of state assets. The System will cover, among other elements, the incorporation, maintenance, inspection, and alienation of property, and DGPE will have the following main responsibilities: • Administer state holdings and see to their maintenance. • Adopt the necessary measures to promote regulation of state assets. • Promote control, inspection, and maintenance of such State property as is used to promote public service. • Establish norms for utilization and rationalization of state property used to promote public service. • Incorporate holdings into state assets. • Promote, directly or through third parties, evaluation of state holdings. • Promote alienation of state property that is not used to promote public service, in accordance with current legislation. • See to the rental and leasing of state holdings. • Authorize the occupation of state property as established by law, promoting the respective registration. Establish the guidelines that authorize use of State property. • Undertake the steps needed to promote the acquisition of property beneficial to the state. • Adopt the necessary administrative procedures to discriminate, reclaim, and reintegrate state property. • Oversee the use of common-use public goods by adopting the necessary inspection procedures. • Promote the donation or free cessation of state property when in the public interest. 130 • Undertake the demarcation and identification of state property; formulate a policy relative to registration of state property, elaborating the plan and incorporating generic values. • Formulate the policy of administrative and patrimonial collection by executing, in accordance with the law, the necessary actions to optimize its collection. • Maintain under its control and responsibility the legal documents, titles, and other processes linked to state property under the domain of or owned by the state. The Human Resources System The main objective of the Human Resources System will be to propose the formulation of policies and guidelines for the administration of human resources. In the Cape Verde administrative structure, it would be important to have a specific entity in charge of: • Exercising normative responsibility insofar as the public human resources are concerned. • Proposing formulation of policies and guidelines for administration of human resources, including such elements as social security, benefits, labor relations, career trajectory, remuneration, size of labor force, and implementation of public competition. • Planning, supervising, and providing orientation regarding activities related to human resources in the state. • Proposing and implementing actions with providers, insofar as the administration of human resources is concerned. • Implementing human resources auditing activities, as well as data analysis, monitoring, and supervision aiming at identifying irregularities in the implementation of the legislation on human resources management. 131 Annex 3: The Risks of a Payment System To better evaluate the need of the Treasury to participate in the clearinghouse, it is necessary to understand the internationally accepted standards regarding payment systems. A payment system can be considered as the group of rules, procedures, instruments, and operational systems that act in an integrated way to transfer resources from the payer to the recipient, with the objective of closing a specific obligation existent among the parties. Such systems act by linking the payment orders issued by the nonbanking agents, the banking institutions, and the Central Bank.124 Note should be taken that, with the exception of payments made in kind, the majority of the other financial transactions—done through checks, credit cards, electronic transfers, or credit documents—are transformed into a few interbank transfers of high value against the account holding the banking reserves that each bank keeps at the Central Bank. Every payment system must be constructed to minimize risks. The biggest source of risk in the payment process is the difference between what is established in the contract and settlement of the operation. That difference generates the possibility of the other party becoming insolvent before the settlement occurs. This means that the debtor party becomes a defaulter before repayment of commitments. This is what constitutes the so-called credit risk. When this happens with a financial institution, beyond the credit risk there also is the image risk toward its clients and the market. It can happen that the owing entity will pay with a small delay, which forces the crediting institution to finance its temporary imbalance (liquidity risk). Such fact can cause instability in the bank’s Treasury, forcing it to look for funding in the market. The risk of liquidity together with credit risk can result in a third risk: the systemic risk. This occurs when an unstable situation leads to a domino effect, causing the institution that is supposed to receive a payment to refrain from paying its own commitments and so on in the same chain. For banking institutions that could have an effect on several or all of the financial institutions linked to the payments system, which means that even these banks that are not directly linked to the initial problem could suffer the consequences of a chain reaction. Another source of risk is created when the design of the legal basis sustaining the payments system does not guarantee the conclusion of payments as established in the contracts, according to the principle of irrevocability and non-conditionality of transactions. This is the so-called legal risk. In reality, international literature indicates other risks for the participants in the payments systems. However, in general, risks can be classified into three different categories: operational, legal, and financial, as shown in Table 1. 124 In this document banks are the financial institutions which hold banking reserves in the Central Bank. The banking reserves are deposits in kind which the banks keep at the Central Bank to fulfill the mandatory reserves as well as make interbanking payments. 132 Table 1: Main Risks of Payments Systems Type Definition Risk of loss resulting from failure or inadequate procedures, people, internal systems, or external events. Operational Risk of loss resulting from a legal basis (laws or regulations) not properly conceived, which means not consistent with the functioning of the settlement system, mainly in what contracts, rights, and other guarantees are concerned. Legal Credit risk: risk of one party not liquidating an obligation in its full value, be it upon expiration or any subsequent stage. It includes both the risk of loss because of unrealized revenues, unliquidated contracts with the defaulter participant (risk of reposition cost), as well as the risk of loss of bills delivered or payments made to the defaulter before the completion of default (principal risk). Financial Liquidity risk: risk of a party not paying an obligation in its full value at the established deadline and only at a undetermined future date. Selection of the payments system must have the objective of creating a structure that resists the noncompliance of one or more participants. It is crucial not to transform the monetary authority (in this case, the Central Bank) into the hostage of a systemic risk. To that end, it is necessary to monitor, administer, and minimize the risks. Therefore, as a general rule, the setup of a good payments system aims at minimizing the risks of non-settlement of the contracted transactions. The first element to take into consideration is the period from contracting to settlement, which should be the shortest possible. In summary, reducing the possibility of losses while the payment hasn’t become final, before the effective transfer of resources between the banks involved, and on the other side, the transfer of goods, service, or negotiated asset is the key rule of a good payment system. Keeping in mind that the environment in which the transactions in the financial system occur is a potential source of disturbances, these guarantees are essential for the stability of financial markets. The main risk control mechanisms currently adopted in a system of settlements originate in the experiences of countries that constitute the G-10 and the recommendations of international organizations that operate in that area (the World Bank, IMF, and BIS). It begins with the creation of a system for interbank transfers of large amounts (Large-Value Transfer Systems, or LVTS), taking a robust system to a higher degree of protection, so that it is responsible for these payments that are associated with a high potential for systemic risk. In these systems, choice of the settlement form is conditioned by the choice of reducing the difference between settlement and minimal cost, represented by the lower demand for reserves. The option will lie between systems that settle imbalances through the multilateral net value (Deferred Net Settlement, or DNS) and systems that settle for the gross value in real time (Real-Time Gross Settlement, or RTGS). Management of the systemic risk associated with financial transfers between institutions can be facilitated by introduction of transfer systems for large amounts that segregate transactions 133 involving large amounts. There are several types of such systems; they differ in the way that the values to be settled are computed, as well as the time at which this settlement occurs, as indicated in table 2. Table 2: Types of Transfer Systems for Large Amounts Characteristics of By gross value By net value settlement Specific schedule Settlement through gross value in specific Differed settlement by the net value (differed). schedule. (LDL). Continuous (in real time). Settlement through gross value in real time n.a. (LBTR). n.a. = Not applicable The System of Deferred Settlement through the Net Multilateral Value (Chamber) reduces the opportunity cost of maintaining idle rabbets by the participants, decreasing the liquidity risk. Payment instructions are settled at the end of the period with the transfer of the net multilateral value of these instructions. The main advantage of the LDL systems is the so-called low demand for banking reserves because the multilateral compensation of the values among participants reduces the volume of resources to be transferred. If the differed settlement through the net value decreases the liquidity risk, the difference in settlement times creates an implicit credit concession from the receiving to the paying institution, which exposes the participants to credit risk. One way to solve that problem is to implement mechanisms that facilitate the conclusion of payment even before the transfer of resources to the Banking Reserves account. Some examples of such mechanisms include the bilateral and multilateral limits for risk exposure, the establishment of guarantees, and rules for spreading the losses in situations of noncompliance of one or more participants. The guarantees play the role of covering the exposure of a participant to credit risk. Usually the minimum guarantee is calculated as a percentage over the total of the highest bilateral limit open to the participant. The rules for spreading losses are concerned with the way the value owned by the participant declared insolvent by the chamber is divided. In the Settlement System by Gross Value in Real Time (RTGS), the liquidity risk is high because for the settlement of operations, the banking reserves accounts are immediately affected. Because the payments concern large amounts and are settled in gross, this type of settlement requires greater banking reserves. The liquidity risk can still occur when, at the time established as per contract for the settlement, the entity responsible for the payment does not release the funds, either in the form of idle rabbets or in the form of credit granted by the Central Bank. However, it is possible to eliminate the settlement mismatch, together with the credit risk, by ensuring the conclusion of payments throughout the day, thus reducing the potential for systemic risk. It is fundamental to create mechanisms to respond to the need for higher liquidity to prevent checking problems in which the lack of resources to comply with a payment prevents a substantial number of other instructions from being discharged, increasing the possibility of liquidity risk. 134 Annex 4: Evolution of Formula Based Transfer per Municipality Size See first series of figures with amounts Small municipalites, formula-based transfers evolution ('000 Escudos) 120.000 Santo Antão - Paúl 100.000 Santo Antão - Porto Novo S. Nicolau - Rª Brava 80.000 Boa Vista - Boa Vista 60.000 Maio - Maio Santiago - S. Domingos 40.000 Fogo - Mosteiros Fogo - S. Filipe 20.000 Brava - Brava 0.000 93 95 97 99 01 03 05 07 19 19 19 19 20 20 20 20 Medium m unicipalities, form ula-based transfers evolution ('000 Escudos) 200.000 Sal - Sal 150.000 S. Antão - Rª Grande Santiago - Tarraf al 100.000 Fogo - S. Filipe 50.000 Santiago - Santa Cruz Santiago - Santa Catarina 0.000 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 Largest m unicipalities, form ula-based transfers evolution ('000 Escudos) 250.000 200.000 150.000 Santiago - Praia 100.000 S. Vicente - S. Vicente 50.000 0.000 93 95 97 99 01 03 05 07 19 19 19 19 20 20 20 20 Another consequence of the change in the formula seems to be that largest municipalities receive a higher share of the total amounts transferred. (see second series of figures). 135 Annex 5: Revenues for Five Municipalities The following Figure shows, in volume, the composition of revenues for five municipalities, broken down into own resources, central government transfers and loans. Ressources com position in volum e Mosteiros S.Domingo S.Cruz S.Vicente Praia 0 100000 200000 300000 400000 500000 600000 700000 000 000 000 000 000 000 000 Ow n ressources State Transfers Loans Total 136 Annex 6: Intergovernmental Arrears––Some data Extract from the 2004 report on contingent liabilities Debts from the central government toward the municipalities Debts from the municipalities toward the central government Net balance Municipalities Faveur DGT Faveur Etat Communes TE 2000-4 IT 2000-4 IE 2001-4 FEF 2003-4 Others Total DGCI DGA others Institutes Total Praia 35,400,684 9,647,897 62,658,999 16,500,000 8,510,195 132,717,775 119,615,825 146,744,323 0 3,518,171 269,878,319 137,160,544 São Domingos 30,270,604 0 0 4,900,000 45,000,000 80,170,604 24,846,460 0 0 11,894,895 36,741,355 43,429,249 Santa Cruz 17,750,113 0 0 5,600,000 108,907 23,459,020 38,792,065 0 0 52,906,752 91,698,817 68,239,797 São Miguel 23,087,946 0 0 0 6,307,394 29,395,340 0 0 0 2,036,922 2,036,922 27,358,418 Tarrafal 12,611,548 223,046 0 960,000 162,583 13,957,177 2,798,840 0 0 20,590,399 23,389,239 9,432,063 Santa Catarina 22,402,849 4,760 0 1,200,000 22,283,600 45,891,209 9,721,517 0 0 15,174,300 24,895,817 20,995,392 Mosteiros 11,979,946 0 0 0 120,780 12,100,726 0 0 0 889,678 889,678 11,211,048 São Filipe 14,274,844 150,459 0 0 13,900,520 28,325,823 0 0 0 260,000 260,000 28,065,823 Brava 10,919,931 0 0 0 116,700 11,036,631 0 0 0 0 0 11,036,631 Maio 16,597,258 0 0 0 43,660 16,640,918 0 0 0 4,345,849 4,345,849 12,295,069 Boa Vista 15,657,558 2,607,801 0 14,895,959 0 33,161,318 0 0 40,500,000 96,343 40,596,343 7,435,026 Sal 10,627,136 53,954,200 0 0 0 64,581,336 0 6,454,827 0 611,604 7,066,431 57,514,905 São Nicolau 15,727,607 0 0 0 3,820 15,731,427 0 0 0 560,416 560,416 15,171,011 São Vicente 25,339,805 5,501,654 14,181,997 0 2,500,000 47,523,456 2,138,704 0 0 6,128 2,144,832 45,378,624 Porto Novo 19,164,697 65,760 0 3,640,000 3,021,190 25,891,647 12,328,873 105,044 0 9,576,921 22,010,838 3,880,809 Ribeira Grande 14,015,222 140,405 0 3,750,000 1,691,211 19,596,838 19,959,030 0 0 7,386,680 27,345,710 7,748,873 Paùl 12,935,625 0 0 1,650,000 2,950,000 17,535,625 4,050,470 0 0 6,515,040 10,565,510 6,970,115 Non allocated 17,945,700 0 0 0 17,945,700 17,945,700 Total 308,763,373 90,241,680 76,840,996 53,095,959 106,720,560 635,662,568 234,251,784 153,304,194 40,500,000 136,370,098 564,426,076 301,252,793 230,016,301 137 Municipalities INPS Public and Para-public Companies TOTAL ELECTRA ENAPOR Imprensa N. CV Telecom Garantia CABNAVE ENACOL Subtotal Praia 38,247,210 712,741 212,865 19,415,713 1,678,414 60,266,943 60,266,943 São Domingos 60,996 2,741,503 1,150 294,832 55,883 3,093,368 3,154,364 Santa Cruz 2,329,195 2,408,567 494,162 119,856 10,829,605 15,978 13,868,168 16,197,363 São Miguel 1,666,182 38,546 325,761 978 2,031,467 2,031,467 Tarrafal 4,364,293 57,289 167,683 3,366,022 178,837 8,134,124 8,134,124 Santa Catarina 1,404,860 5,172,515 59,979 61,015 1,591,803 346,857 18,900 7,251,069 8,655,929 Mosteiros 2,212,731 43,153 17,398 533,194 546,381 3,352,857 3,352,857 São Filipe 1,412,699 154,177 81,712 327,727 214,781 2,191,096 2,191,096 Brava 79,258 2,778,842 8,579 100,841 965,632 366,593 4,220,487 4,299,745 Maio 3,030,746 412,415 77,534 114,205 73,868 3,708,768 3,708,768 Boa Vista 1,546,060 15,400 924,147 2,485,607 2,485,607 Sal 10,590,595 89,342 98,660 1,945,298 12,723,895 12,723,895 São Nicolau 1,718,808 5,200 248,167 564,511 2,536,686 2,536,686 São Vicente 26,043,819 50,436 65,125 777,230 160,985 10,000 27,107,595 27,107,595 Porto Novo 12,210,245 911,907 40,825 431,051 6,729 13,600,757 13,600,757 Ribeira Grande 10,060,316 292,688 402,267 1,020,010 52,000 11,827,281 11,827,281 Paùl 8,767,066 55,973 242,008 9,065,047 9,065,047 Total 3,874,309 134,972,197 3,358,241 1,088,410 42,734,662 5,230,805 62,000 18,900 187,465,215 191,339,524 138 Debt toward ELECTRA in 2006-07 Items in italics show a catch-up on the arrears. Municipality/ Municipal Service 31-12-2006 30-09-2007 Difference 2007–06 Subtotal Sub total Public Subtotal Sub total Public Difference Public lighting lighting lighting 2007–06 C.M. BOAVISTA 6,233,995 5,020,072 604,151 6,706,039 -3,943,877 1,685,967 C.M. BRAVA 8,191,931 7,097,190 971,592 9,022,437 -5,295,092 1,925,247 C.M. CALHETA DE S.MIGUEL 3,164,780 2,179,991 749,474 2,891,605 711,614 -1,703,692 C.M. MAIO 7,961,377 2,259,842 7,574,599 3,294,150 647,530 1,034,308 C.M. MOSTEIROS 5,765,012 4,669,334 2,291,833 6,059,871 -2,082,642 1,390,537 C.M. PAÚL 16,823,398 6,837,579 10,650,037 9,348,012 -3,662,928 2,510,433 C.M. PORTO NOVO 12,766,831 9,166,177 3,156,972 11,537,087 -7,238,949 2,370,910 C.M. RIBEIRA GRANDE 25,130,560 19,353,057 5,301,257 26,042,547 -13,139,813 6,689,490 C.M. RIBEIRA GRANDE 3,441,398 n/a 4,256,857 n/a 0 (SANTIAGO) 815,459 C.M. S.DOMINGOS 6,599,503 2,160,673 6,466,004 2,772,603 478,431 611,930 C.M. S.FILIPE 8,332,765 7,505,403 28,376 9,389,816 -6,419,976 1,884,413 C.M. S.LOURENÇO DOS 360,716 360,716 0 565,111 204,395 ORGÃOS -156,321 C.M. S.VICENTE 81,414,960 72,858,729 385,615 93,929,095 -59,958,979 21,070,366 C.M. SANTA CRUZ 7,088,975 5,228,823 1,667,900 6,463,843 -4,186,055 1,235,020 C.M. Sta. CATARINA 12,905,062 8,870,327 3,322,654 11,750,008 -6,702,727 2,879,681 C.M. Sta. CATARINA (FOGO) 32,341 2,107 30,660 8,908 5,120 6,801 C.M. TARRAFAL (SANTIAGO)) 10,684,854 7,557,746 3,675,669 9,876,560 2,318,814 -4,690,371 C.M. TARRAFAL (S.NICOLAU) 1,625,248 1,248,143 504,142 2,035,559 787,416 -333,690 C.M. PRAIA 70,349,981 27,606,106 41,251,461 42,296,896 -14,407,730 14,690,790 C.M. S.NICOLAU (Rª BRAVA) 6,748,618 6,087,560 134,413 7,729,822 -4,971,943 1,642,262 C.M. SAL 30,038,400 20,538,695 1,668,791 26,069,053 -22,839,251 5,530,358 Sub––total Municipalities 325,660,705 216,608,270 94,692,457 287,789,022 -159,787,496 71,180,752 �GUA BRAVA 5,406,334 7,793,009 2,386,675 139 Municipality/ Municipal Service 31-12-2006 30-09-2007 Difference 2007–06 Subtotal Sub total Public Subtotal Sub total Public Difference Public lighting lighting lighting 2007–06 SEPAMP 1,554,252 1,236,592 -317,660 SERV. AUT. AG E SANEAM–– 782,002 623,164 Calheta S.Miguel -158,838 SERV. AUT. AG E SANEAM–– 45,047 155,442 S.Nicolau 110,395 SERV. AUT. AG E SANEAM–– 16,706,239 15,494,203 Maio -1,212,036 SERV. AUT. AG E SANEAM–– 6,968,250 7,162,963 Porto Novo 194,713 SERV. AUT. AG E SANEAM–– 410,543 480,733 Ribeira Grande 70,190 SERV. AUT. AG E SANEAM––S. 2,773,924 3,524,814 Domingos 750,890 SERV. AUT. AG E SANEAM–– 5,481,690 9,466,523 Sta. CATARINA 3,984,833 SERV. AUT. AG E SANEAM–– 4,213,514 5,673,873 Sta. Cruz 1,460,359 SERV. AUT. AG E SANEAM–– 352,894 2,796,895 Tarrafal 2,444,001 Sub––total Municipal Firms 44,694,689 54,408,211 9,713,522 TOTAL 370,355,394 216,608,270 149,100,668 287,789,022 -150,073,974 71,180,752 140 Annex 7: Use of PEFA Framework for Municipal Finances The following tables use the PEFA methodology to synthesize some of the information from the chapter. Table 10 on fiscal decentralization is directly extracted from the PEFA indicators, and table 11 is using an adapted version of the PEFA tables for local governments. Table 1: PEFA Ratings for Intergovernmental Fiscal Relations Indicator Score Brief Explanation PI-8. Transparency of Inter- C Governmental Fiscal Relations (a) Transparency and objectivity B Score = B: The horizontal allocation of most transfers from in the horizontal allocation central government (at least 50 of transfers) is determined by amongst sub-national transparent and rules based systems. governments. (b) Timeliness and reliable C Score = C: Reliable information to SN governments is information to SN governments issued before the start of the SN fiscal year, but too late for on their allocations. significant budget changes to be made. Score = D: Reliable estimates on transfers are issued after SN government budgets have been finalized, or earlier issued estimates are not reliable. (c) Extent of consolidation of D Score = D: Fiscal information that is consistent with central fiscal data for general government fiscal reporting is collected and consolidated for government. less than 60 (by value) of SN government expenditure OR if a higher proportion is covered, consolidation into annual reports takes place with more than 24 months delay, if at all. Table 2: PEFA methodology adapted to municipalities General situation Situation in our sample of municipalities A. Credibility of the Budget PI-1 Aggregate expenditure N/A Very high discrepancy. outturn compared to original approved budget. PI-2 Composition of N/A Even though total expenditure outturn is expenditure outturn well below planned expenditures, the compared to original composition of actual expenditures approved budget. matches very closely, in percentage, the voted budget. PI-3 Aggregate revenue outturn N/A Discrepancies vary but are generally compared to original high––in our sample, between 40 to 115 approved budget. over three years of data. PI-4 Stock and monitoring of Arrears are important especially Only one municipality in the sample has expenditure payment toward the company providing a working system to monitor possible arrears. electricity, but besides poor arrears. For four of the municipalities monitoring there are some visited, arrears toward ELECTRA in fundamental disagreements in 2006 stem from 7 to 27 of their yearly interpreting the causes of the expenditures. arrears. 141 General situation Situation in our sample of municipalities B. Comprehensiveness and Transparency PI-5 Classification of the Municipal budgets according to Due to capacity constraints, most budget. Local finances law should be municipalities are not able to comply available according to economic, with all requirements for budget organic and functional presentation. classification. PI-6 Comprehensiveness of Local Finances Law prescribes a Due to capacity constraints, most information included in very comprehensive set of municipalities are not able to comply budget documentation. documents to accompany the with all requirements for budget budget. presentation even though efforts are made in this sense. PI-7 Extent of unreported Two main areas seem to be No specific evidence collected during municipal operations. prone to poor reporting in the the mission. budget: proceeds from decentralized cooperation––that are sometimes in-kind––and possibly the use of “the safe� prior to transfer in bank accounts. PI-10 Public Access to key fiscal The IGF and TdC are in charge 2 municipalities out of the five in the information. of ex post controls but suffer sample have not sent their management from capacity constraints and accounts to the TdC for 2005, and one also late reporting from some has not sent it for 2004 municipalities–– Only one municipality out of 5 has tried Local budgets are published in to organize easier access to budget data Official Gazette. for citizens (opening of a room where budget can be freely consulted). C(a) Policy-Based Budgeting PI-11 Orderliness and Participatory budgeting has not participation in the annual been experimented in the budget process. country. PI-12 Multi-year perspective in Municipalities still very much Municipalities of the island of Santiago fiscal planning, use an incremental approach to were supported by the Austrian expenditure policy and budgeting. cooperation in elaborating Local budgeting. Development Plans that should help them in shifting to a multi-year planning approach. C (b) Predictability and Control in Budget Execution PI-13 Transparency of taxpayer Municipalities are responsible Main weaknesses are due to outdated obligations and liabilities. for collecting local taxes, and land and taxpayers registries. Anecdotic some taxes are decided by them evidence shows that tax collection Clarity and comprehensiveness officers sometimes act in a discretionary of tax liabilities are good, but manner. taxpayers access to information on tax liabilities and administrative procedures are less so, and there does not seem to be an appeal mechanism. PI-14 Effectiveness of measures Main weaknesses are due to One municipality computerized its fiscal for taxpayer registration outdated land and taxpayers service and is able to link taxpayer 142 General situation Situation in our sample of municipalities and tax assessment. registries. There does not seem arrears with non delivery of other to be systematic planning and documents––the IT system is being monitoring of tax audit rolled out across municipalities. programs. PI-15 Effectiveness in collection Tax payments are still a low One municipality boosted its revenues of tax payment. proportion of municipal significantly by hiring retired fiscal resources, due to lack of agents to establish its fiscal unit and capacity in fiscal units, poor tax computerizing its fiscal unit. registries, difficulty in estimating patrimony values and low tax basis. PI-16 Predictability in the Central government transfers Only one municipality mentioned one availability of funds for come in two forms: a formula delay in transfer of formula based commitment of based one, transferred monthly, transfers. However, contracts programs expenditures. and thus predictable but for are believed to be prone to political which municipalities would wish considerations and lack conditions and to know the yearly amount criteria for eligibility. earlier for planning purpose; an ad-hoc “contract-program� system for investment purposes, which depends on negotiations between the center and the municipality (ies) and once signed is allocated in three tranches (30, 30 and 40). PI-17 Recording and Borrowing is allowed and management of cash regulated. balances, debt and guarantees. PI-18 Effectiveness of payroll Relatively low at local level controls. according to TdC. PI-19 Competition, value for Municipalities are subject to the No data could be collected on the types money and controls in newly enacted national of processes most used for awarding procurement. Procurement Code. contracts. There does not seem to be a formal mechanism for complaint in case of abuses. PI-20 Effectiveness of internal N/A No municipality in our sample has an audit controls for non- internal audit unit, and there are no salary expenditure. procedure manual. However, all municipalities in the sample restricted their expenditures to their revenues limits without external pressure (there are no ex ante controls). PI-21 Effectiveness of internal The IGF is in charge of internal audit. audit but lacks funds and personnel to undertake missions, which are made more expensive by territorial discontinuity. However, all municipalities were inspected in 2007 in anticipation of 2008 local elections. 143 General situation Situation in our sample of municipalities C (c) Accounting, Recording and Reporting PI-22 Timeliness and regularity N/A In some municipalities, this can be done of accounts reconciliation. online and is thus not a problem. PI-23 Availability of information N/A N/A on resources received by service delivery units. PI-24 Quality and timeliness of Municipalities must produce Municipalities seem to comply. in-year budget reports. quarterly reports. PI-25 Quality and timeliness of Municipalities produce annual financial management accounts yearly. statements. They need to be submitted to local assembly by March 1st, and to the TdC by end of June. They rely on national accounting standards. Delays have been reported for some municipalities in transmitting their accounts to the TdC. C (iv) External Scrutiny and Audit PI-26 Scope, nature and follow- The TdC is responsible for the up of external audit. external audit of municipalities. Accounts need to be transmitted by the end of June following the last fiscal year. Unfortunately, the TdC is seven or more years behind in its municipal accounts review. PI-27 Legislative scrutiny of the The municipal assembly annual budget. approves the budget. PI-28 Legislative scrutiny of The law requires the mayor to N/A external audit reports. comment on the report, but does not specify whether the report once finalized needs to be sent to the Municipal Assembly. 144 Annex 8: Hidden-Cost Deficit Methodology Model -- Calculating Hidden Costs in the Energy and Water Sectors125 The model to calculate “hidden costs� in the infrastructure sector is described below. While there are more detailed and complex ways in which a model could be developed to reflect a loss in specific countries and sectors, this model has been formulated specifically to provide an insight into three key components of hidden costs: poor collections, tariffs set below cost-recovery levels and losses above normative levels. The intention in developing this model was to devise a simple-to-use methodology with which to monitor trends and to benchmark across sectors and countries without the need for intensive data collection efforts. Let H be the “hidden costs� in the (electricity, gas or water) sector, defined as: H = R*––R ……………………………. (1) Where, R* is the expected revenue in a system operating with tariffs that cover costs, where bills are paid and where losses are within normal levels expected for a system of that age and design. R is the actual revenue. Qs, is the volume (electricity, gas or water) supplied to the transmission network. Qe, is the end-user consumption (of electricity, gas or water) Lm, are the losses in transmission and distribution (of electricity, gas or water). Lm includes normative losses, Ln where Ln are these losses that are expected in a system of that design and age as well as losses due to system inefficiencies outside norms and theft. Then, Qs = Qe + Lm …………………………… (2) And, R* = (Qs––Ln) Tc ……………………….. (3) Where Tc defined as the cost-recovery tariff, is the long run cost of operation and maintenance and includes a reasonable allowance for investment and normative losses. And, R = (Qs––Lm) Te Rct ................................ (4) Where Te is defined as the weighted average end-user tariff, and Rct is the rate of collection of billed amounts. Substituting for R* and R in (1), using (3) and (4), then H = (Qs––Ln) Tc––(Qs––Lm) Te Rct H = (Qe + Lm––Ln) Tc––Qe Te Rct H = Qe (Tc––Te) + (Lm––Ln) Tc––Qe Te (Rct––1) 125 Quoted from Ebinger, J. (2006), Measuring Financial Performance in Infrastructure, World Bank Working Paper 145 H = Qe (Tc––Te) + Tc (Lm––Ln) + Qe Te (1––Rct) ………………….. (5) Total losses, Lm, can also be defined as lm Qs, or the rate of total losses multiplied by the volume (of electricity, gas or water) supplied to the system. Normative losses, Ln, can in turn be defined as ln Qs, or the rate of normative losses multiplies by the volume (of electricity, gas or water) supplied to the system. Therefore, Lm––Ln = (lm––ln) Qs ………………….. (6) Where Qs is defined in (2) as: Qs = Qe + Lm Substituting for Lm, we have, Qs = Qe + lm Qs Qs (1––lm) = Qe Qs = Qe/(1––lm) .…………………………. (7) Substituting for Qs in (6) using (7) gives: Lm––Ln = (lm––ln) Qe/(1––lm) ………………………….(7) Hidden costs defined in (5) can therefore be expressed as: H = Qe (Tc––Te) + Qe Tc (lm––ln)/(1––lm) + Qe Te (1––Rct) ………………………..(8) In performing the calculation if any of the three components of hidden costs defined above has a value of less than or equal to zero, it is set to zero. Please note that tariffs set below cost-recovery levels as a matter of policy are considered an explicit subsidy and should be deducted from the overall calculation. In this case the correct measure of hidden costs would be H = R*––R––T Where, T is the amount of capital or other transfer. 146 Electricity End User Average Cost Normative Average Actual Loss Collection Year Consumption Recovery Tariff Loss Rate Tariff Rate Rate (GWh) (ECV/KWh) (%) (ECV/KWh) 2001 95.62 20.6 23% 10% 14.5 91% 2002 117.07 19.6 20% 10% 14.5 91% 2003 132.50 18.4 18% 10% 16.9 88% 2004 147.31 16.4 19% 10% 16.7 97% 2005 161.68 14.8 17% 10% 15.9 91% 2006 164.77 21.5 22% 10% 18.5 88% Water Supply End User Average Cost Normative Consumption Loss Average Actual Collection Year Recovery Tariff Loss Rate (million of Rate Tariff (ECV/m3) Rate (ECV/m3) (%) m3/year) 2001 2.30 256.2 29% 20% 201.7 91% 2002 2.61 277.0 28% 20% 200.2 91% 2003 2.85 302.2 29% 20% 270.8 88% 2004 2.85 258.5 30% 20% 270.3 97% 2005 2.93 252.0 31% 20% 282.0 91% 2006 3.00 325.6 30% 20% 294.6 88% Source: Our estimation based on ELECTRA's financial Statements 147 Annex 9: Accounting for Infrastructure––Methodology This type of exercise is very demanding on information. Data was gathered during a diagnostic mission to the field (Praia, October 1-5 2007). Data was subsequently subject to an extensive desk review. However, data limitations still impend detailed analysis of the magnitude and scope of the public spending. Many questions cannot be answered without the assistance of the technical staff at the field. This report lacks of details about the infrastructure spending at the sub-national governments, maintenance policies/executions at the SOEs, and road maintenance. Moreover, the government does not produce information in the format that satisfies the needs of the exercise. For instance, some capital expenditure in electricity, water and water management resources are included in a single project, thus disentangling the allocations can only be done under some assumptions. Data sources: On-budget spending • Investments: the department of planning, Ministry of Finance and Public Administration, which collects annual data on executions (from different sources of financing: treasury, donors, etc. More than 750 projects exist for each fiscal year. They are grouped by sectors (electricity, energy, water & sanitation, ports, airports, roads). • Recurrent spending: mostly salaries and inputs from the Ministry of Infrastructure. The spending was allocated by sector of activity using the share of public investment per sector. Off-budget spending • Investments: each SOE produces a financial statement including the investments at each period. In some cases, these investments matched capital expenditures (interest payments, depreciation and amortization). In other cases, the team was not able to track the differences and assumed the maximum of both to define capital expenditures. In any case, to avoid duplication, investment financed with on-budget resources were not counted as off-budget investments. • Recurrent spending: based on the financial statement for each SOE Road maintenance • Based on the IE’s 2006 annual report. For 2001–05, the source of information is the voted/executed projects collected by the Department of Planning. 148