Report No. 29036-AO Angola Public Expenditure Management and Financial Accountability February 16, 2005 Africa Region PREM1 Document of the World Bank ... 111 ACKNOWLEDGEMENTS The Public Expenditure Management and Financial Accountability Review (PEMFAR) for Angola was prepared by the W o r l d Bank in close collaboration with the Govemment o f the Republic o f Angola. The Govemment team was led by former Deputy Finance Minister Manuel Nunes Jr., under the guidance o f H.E. JosC Pedro de Morais, the Minister o f Finance o f Angola. The core Government team consisted o f Mr. Manuel N e t 0 Costa, Director o f Studies and International Relations at the Ministry o f Finance, and Mmes. Madalena Ramalho and Rosa Baptista, both advisers to the Minister o f Finance. Within the Ministry o f Finance, valuable inputs were received from Messrs. Job Graqa, Deputy Finance Minister; Leone1 Silva, National Director for Treasury; Alcides Safeca, National Director for Budget; Hermenegildo Gaspar, National Director for Taxes, and their respective staffs, as well as from Ms. Joana Cordeiro, National Director for Accounting, and her staff. Within the Ministry o f Planning, important contributions were made by the former Deputy Planning Minister Eduardo Severim de Morais, and by the Director o f Studies and Planning, Mr. Pedro Luis da Fonseca. The report also benefited from inputs from staff o f the National Bank o f Angola, including Mmes. Marinela Amaral, Administrator, and Clotilde Mariano, Adviser to the Governor, as well as Mr. Sergio SerrZo, Accounting Director. Finally, k e y inputs were received from Mr. Francisco de Lemos JosC Maria, the Director o f Finance o f Sonangol, the national o i l company, and h i s staff. On the World Bank side, the report was prepared by a team l e d by Jorge Araujo, and composed o f Francisco GalrZo Cameiro, who coordinated the final stages o f preparation o f the document, Iraj Talai (Country Financial Accountability Assessment Coordinator), Charles McPherson, Helena GrandZo Ramos, RenC Costa, Uche Mbanefo, Maria Teresa Benito-Spinetto, and Asmara-Lua Achcar. Peter Macnab and D a v i d Reading (DFID-financed consultants) prepared background materials for Chapter 7 and Annex 2 o f the report, and Slaheddine Ben-Halima and Christine D e Mariz (core members o f the Bank’s team for the Country Procurement Assessment Report for Angola) prepared Annex 1 o f the PEMFAR. The report benefited from generous financial support and substantive comments from the United Kingdom’s Department for International Development (DFID), as well as from the European Commision-financed Public Expenditure and Financial Accountability (PEFA) program. The DFID team was led f i r s t by Peter Dearden and subsequently by Harry Hagan, with participation from Habib Rab and Jonathan Hargreaves. Support from the PEFA team, especially Nicola Smithers and Michael Boniakowski, i s acknowledged with thanks as well. The peer reviewers o f this report were Jose Leandro (European Commission), Richard Allen, JoZo do Carmo Oliveira, Parminder Brar and Gert Van Der Linde. Valuable comments were also received from a World Bank’s Quality Assurance Group (QAG) panel on the Bank’s Analytical and Advisory Activities in Angola, l e d by Florent Agueh, and consisted o f B e n Varon, Michael Gillette, and Roberto Zagha. Helpful advice was provided at different stages o f preparation o f this report from colleagues at the Poverty Reduction and Economic Management Unit 1 at the Bank’s Africa Region, including Hinh Dinh, Abebe Adugna, Bernard Myers, and Peter M o l l . iv Close collaboration with the Angola teams at the International Monetary Fund’s Africa Department and Fiscal Affairs Department (FAD) permeated the entire period o f preparation o f the report. The P E M F A R team benefited from frequent interactions with Gonzalo Pastor (IMF Mission Chief for Angola), and his team, especially Magnus Alvesson, and Alfred0 Torrez. The former IMF Resident Representative in Angola, Carlos Leite, also participated in some o f the meetings o f the P E M F A R mission in April 2003. The P E M F A R team also gained valuable inputs from joining a FAD team that visited Angola in August 2003, led by HClio Tollini and consisted o f Wander L u i z and Carlos Duarte. Emmanuel Akpa, Darius Mans, Philippe L e Houerou, Laurence Clarke, and Johannes Zutt provided guidance and support to the team throughout the PEMFAR exercise. Ligia Murphy and Debra Cubitt provided assistance in document formatting, editing, and processing. The support o f the Angola Country Team at the Bank, particularly o f the Luanda Country Office, i s gratefully acknowledged. -1- Table o f Contents EXECUTIVE SUMMARY ............................................................................................... i Policy Choices and Fiscal Outcomes ....................................................................... i A Dual Public Expenditure System ........................................................................ .. 11 H o w Can Normality be Restored? A Two-Pronged Reform Strategy................... 111 ... A Short-Term Action Plan to Strengthen the Formal System ............................... iv Ring-Fencing and Phasing-Oat “Non-conventional” Spending Mechanisms ....... vi INTRODUCTION ............................................................................................................. 1 1 . AN OVERVIEW OF ANGOLA’S FISCAL OUTCOMES................................ 5 A. Angola’s Long-Term Fiscal Choices .......................................................... 5 High Expenditure-to-GDP Ratios ............................................................... 5 Low Shares o f Capital Expenditures .......................................................... 6 Extreme Dependence on Oil ....................................................................... 7 B. Linking Choices to Fiscal Outcomes .......................................................... 8 Fiscal Deficits ............................................................................................. 8 Domestic Arrears ........................................................................................ 9 Large Extra-Budgetary and Quasi-Fiscal Expenditures ............................ 9 Overcoming the Plight o f “Unexplained Discrepancies ’’ ........................ 11 C. Main Sources o f Deficit Financing ........................................................... 11 A History of High Inflation ....................................................................... 12 A Dollarized Economy .............................................................................. 13 The Road Ahead ........................................................................................ 14 2 . AN OVERVIEW OF ANGOLA’S PUBLIC FINANCIAL MANAGEMENT SYSTEM .............................................................................................................. 15 A. Organization and Components o f the Conventional Public Financial Management System ................................................................................. 15 Overview ................................................................................................... 15 B. A Non-Conventional Spending System Disconnects Budgeting and Policy ................................................................................................................... 18 C. Main Weaknesses of the Conventional Budgetary Process ...................... 19 D. The Non-Conventional Spending System................................................. 20 E. A Two-Pronged Reform Strategy ............................................................. 21 3 . BUDGET PREPARATION ............................................................................... 23 A. Overview of The Budget Preparation Process in Angola ......................... 23 The Budget Preparation Cycle.................................................................. 23 B. The Credibility o f the OGE....................................................................... 25 Realism o f Macroeconomic Assumptions and Projections ....................... 25 Reliability and VeriJiability of Cost Information., .................................... 26 -L- Recommendations for Reform ................................................................... 26 C. The Coverage o f the OGE......................................................................... 27 The Revised 2003 OGE............................................................................. 29 Recommendations for Further Reform ..................................................... 29 D. Budgeting o f Capital Expenditures ........................................................... 30 Ongoing Reforms and Actions Contemplated in the PMFP ..................... 31 Recommendations for Further Reform ..................................................... 31 E. The Status o f Budget Units and Budget Classification Issues .................. 32 The Multiplicity of Budget Units ............................................................... 32 Budget Classijkation Issues ..................................................................... 33 4 . BUDGET EXECUTION .................................................................................... 34 A The Institutional Framework for Budget Execution ................................. 34 The Stages of the Budget Execution Process ............................................ 35 B. Compliance with the Budget Law............................................................. 35 C. Cash Planning and Management ............................................................... 36 Expenditure Processing Arrangements ..................................................... 38 Consolidation of the Treasury Single Account (Cur) .............................. 39 D. Revenues management.............................................................................. 39 E. Public Expenditure Management .............................................................. 40 F. Payment Arrears and Price Subsidy Arrears ............................................. 42 G. Execution of the Public Investment Program ........................................... 44 H. Debt Management Issues .......................................................................... 45 5 . BUDGET ACCOUNTING AND REPORTING .............................................. 47 A. Angola’s Public Accounting System ........................................................ 47 Non-Financial Assets and Liabilities ........................................................ 48 B. Main Issues Identified in Angola’s Public Accounting System ............... 49 C. Angola’s Integrated Financial Management Information System ............ 50 D. The Recent Push for Reforms ................................................................... 53 E. Main Areas for Further Reform ................................................................ 53 6 . AUDIT AND CONTROL SYSTEMS ............................................................... 55 A. Internal Control, Financial Accountability, and Fiduciary Risk............... 55 Expenditure Controls ................................................................................ 56 B. Internal Audit: Overview .......................................................................... 58 Legal and Institutional Framework .......................................................... 58 Auditing Standards .................................................................................... 58 Reporting and Follow-up Mechanisms ..................................................... 59 C. Reforming the Internal Control and Audit Functions ............................... 59 Ongoing Reforms and Actions Contemplated in the PMFP ..................... 60 Recommendations for Further Reform ..................................................... 60 D. The External Audit Function: Overview ................................................... 61 Legal and Institutional Framework .......................................................... 61 Auditing Standards and Coverage ............................................................ 62 Organization and Staf$ng ......................................................................... 62 Quality and Timeliness o f Reporting ........................................................ 63 -5- Government Follow-up to the Tribunal’s Recommendations ................... 63 E. M a i n Areas for Reform ............................................................................. 63 7 . THE FIDUCIARY FRAMEWORK FOR OIL REVENUE MANAGEMENT IN ANGOLA ........................................................................................................ 66 A. Background ............................................................................................... 66 B. Petroleum Taxation ................................................................................... 67 C. Treasury-Like Operations by Sonangol .................................................... 69 Overview ................................................................................................... 69 The Petroleum Account Mechanism ......................................................... 71 An Informal Compensation Mechanism .................................................... 72 D. Petroleum Sector Governance ................................................................... 74 E. Summary and Recommendations ............................................................. 75 8 . LINKING BUDGETING AND POLICY .......................................................... 80 A. The Need for a Medium-Term Approach to Budgeting ........................... 80 B. Existing Obstacles to Linking Budgeting and Policy ............................... 81 The Absence o f a Multi-Year Public Investment Program ....................... 81 The Lack o f a Medium-Term Policy Foundation for the OGE .................82 C. The Seeds for a Successful Medium-Term Approach to Budgeting and Policy ........................................................................................................ 83 The Revenue Side: The Oil Sector Financial Model ................................. 83 The Expenditure Side: The Post-Conflict Rehabilitation and Reconstruction Program ............................................................... 84 The Poverty Reduction Strategy ................................................................ 85 Recommendations for Further Reform ..................................................... 86 9 . THE PUBLIC FINANCE MODERNIZATION PROGRAM ........................ 89 A. The Public Finance Modernization Program ............................................ 89 B. Progress in Program Implementation........................................................ 90 C. M a i n Challenges Ahead ............................................................................ 91 D. Priority Areas for Reform ......................................................................... 91 E. Performance Indicators ............................................................................. 97 ANNEXES ....................................................................................................................... 98 ANNEX I.......................................................................................................................... 99 ANNEX I 1...................................................................................................................... 100 ANNEX I11..................................................................................................................... 114 -4- Charts Chart E 1: The Articulation Between the Conventional and the “Non-Conventional” Spending ... Systems .................................................................................................................................. 111 Chart E 2: A Schematic V i e w o f the Transition ........................................................................... vii Chart 1.1: Total Government expenditure as a percent o f GDP for selected countries .................. 5 Chart 1.2: Non-oil fiscal balance and o i l prices. 1996-2004 .......................................................... 7 Chart 1.3: Evolution o f the overall fiscal balance (% o f GDP) ...................................................... 8 Chart 1.4: Composition o f payment arrears (% of total change in arrears) .................................... 9 Chart 1.5: BNA quasi-fiscal expenditure (flows) ......................................................................... 10 Chart 1.6: Distribution o f quasi-fiscal expenditures by Sonangol during 2002 ............................ 10 Chart 1.7: Financing Angola’s overall deficit (cash basis) (“A o f GDP) ....................................... 12 Chart 1.8: A Snapshot o f Inflation and Dollarization .................................................................. 13 Chart 3.1: Overall Fiscal Balance on a Commitment Basic (% o f GDP): OGE vs . Actuals .......25 Chart 4.1 : Price Subsidy Arrears, Company by Company (kz) . Fiscal Year 2002 ....................... 43 Chart 4.2: Execution o f the 2002 PIE’: The “Financial” Dimension (Kz million) ........................ 44 Chart 7.1 : The Informal Compensation Mechanism at W o r k in 2002 (Kz million) .................... 73 Chart 7.2: Schematic V i e w o f the Transition................................................................................ 78 Chart 8.1 : Composition o f the PCRRP, 2003-2005 ...................................................................... 84 Chart 8.2: Poverty Reduction Strategy (ECP): Priority Areas in Terms o f Multi-Year Allocations, 2003 - 2006 ........................................................................................................................... 86 Tables Table 1.1: Angola Government expenditure by function. .............................................................. 6 Table 8.1 : Financial Model: M a i n Exogenous and Endogenous Variables .................................. 83 Table 8.2: The Share o f the PCRRP in the total Public Investment Program............................... 85 Table 9.1 : Priority Actions: Strengthening the Formal System .................................................... 92 Table 9.2: Priority Actions: Ring-Fencing and Phasing-Out the “Non-conventional” System ....95 Table 9.3: Recommended Performance Indicators ....................................................................... 97 Figures Figure 2.1 : The Articulation Between the Conventional and the “Non-conventional” Spending Systems ................................................................................................................................. 18 Figure 4.1: Cross Arrears: A Chain Reaction ............................................................................... 43 Figure 7.1: Financial Flows under Decree 30/95 .......................................................................... 72 Abbreviations and Acronyms BAI Banco African0 de Investimentos BCI Banco de ComBrcio e Industria BPC Banco de Poupanqa e CrBdito BNA National B a n k o f Angola CABGOC Cabinda Gulf Company CGE General State Accounts CFAA Country Financial Accountability Assessment CPAR Country Procurement Assessment Report CUT Treasury Single Account DMFAS Debt Management and Financial Analysis System DNA National Customs Directorate DNC National Accounting Directorate DNI National Tax Directorate DNO National Budget Directorate DNPO National Directorate for the State’s Non-Financial Assets DNT National Treasury Directorate ECP EstratBgia de Combate a Pobreza EDEL Electricity Distribution Enterprise EMTA Economic Management Technical Assistance ENE National Electricity Enterprise FAD Fiscal Affairs Department (IMF) GDP Gross Domestic Product GEREI Office o f Studies and International Relations (MINFIN) IMF Intemational Monetary Fund INF National Inspectorate o f Finance LNG Liquefied Natural Gas MAPESS Ministry o f Public Administration, Employment, and Social Security ME0 Budget Elaboration Manual MINARS Ministry o f Social Assistance and Reinsertion MINFIN Ministry o f Finance MINPLAN Ministry o f Planning MTEF Medium-Term Expenditure Framework OGE National State Budget PCE Chart o f Accounts PCRRP Post-Conflict Rehabilitation and Reconstruction Program PES Economic and Social Program PIP Public Investment Program PMFP Public Finance Modernization Program PSA Production Sharing Agreement SA1 Supreme Audit Institution SIGFE Integrated Financial Management System o f the State SIGIP Integrated Public Investment Management System UNCTAD United Nations Conference o n Trade and Development UNDP United Nations Development Program -1- EXECUTIVE SUMMARY 1. One o f the most salient features o f Angola’s public expenditure management and financial accountability framework i s the coexistence o f two parallel, but articulated, expenditure execution systems: the “conventional” system, coordinated by the National Treasury Directorate, and a “non-conventional” one centered around the national o i l company Sonangol. A standard assessment o f a country’s fiscal framework would usually concentrate o n the “conventional” system; the PEMFAR goes one step further by assessing the workings o f the “non-conventional” system as well as i t s articulation with the conventional one. The key finding i s that the justification o f the use o f the non- conventional system is gone, and its maintenance i s imposing heavy costs on the economy. The PEMFAR proposes accordingly the adoption o f a two-pronged reform strategy aiming at: (i) strengthening the formal public financial management structures and tools; and (ii) phasing-out and eventually eliminating the “non-conventional” mechanisms. POLICY CHOICES AND FISCAL OUTCOMES .. 11. The size o f the public sector in Angola i s considered high by international standards, but partly due to a 27-year long civil war government spending has had very limited impact on development outcomes. Compared to the African continent as a whole, the expenditure-to-GDP ratio in Angola (nearly 50% throughout the 1990s) i s twice as large as the region average (25.6% GDP). The composition o f government expenditures, however, remains concentrated o n activities with l i t t l e direct contribution to development. In this context, a higher share o f expenditures have been devoted to defense and public order in comparison to what has been spent on education, health, and other social sectors. This situation has started to change and social spending increased from 12.7% as a share o f total public expenditures in 2003 to 20.1% in 2004, while the share for defense and public order declined from 13.8% to 12.5% o f total expenditures over the same period. Despite this shift in the composition o f public spending by function, health and education expenditures as a percentage o f GDP in Angola (less than 2% for health and less than 5% for education) are (with the exception o f Equatorial Guinea) amongst the lowest in the African continent (with averages o f 6% and 10% for both categories, respectively).’ Such composition o f public spending i s far below the country’s needs in terms infrastructure reconstruction and provision o f essential services to the population and seems to reflect political choices s t i l l concemed with the prevalence o f a wartime budget. ’ Data from the W o r l d Bank’s 2003 African Development Indicators; the African average figures are exclusive o f South Africa and Nigeria. ... 111. Despite recent improvements in fiscal discipline and transparency, institutional weaknesses remain as an important threat to the country’s fiscal stance. A recent assessment o f Angola’s government expenditures recordings has collected evidence o f remarkable progress in the elimination o f the so-called “unexplained discrepancies” from the budget. Up until 2002, quasi-fiscal extrabudgetary expenditures were not initially budgeted but ended up being recorded ex post in the government accounts creating gaps in the published accounts that were translated into a discrepancy o f roughly U S $ l billion between stated government funds and actual revenue. Unexplained discrepancies in the fiscal accounts were, among others, substantial funds received as signature bonuses for o i l contracts and o i l royalties, and non- transparent external debt transactions. The problem reached substantial proportions in 1997 and 1999 (amounting respectively to 22.47% and 18.38 % o f GDP), and by 2002 n 2003 and unexplained discrepancies s t i l l averaged 3% o f GDP, or US$ 347 million. I 2004, these discrepancies were completely eliminated from the published accounts, but sizable quasi-fiscal operations executed by Sonangol and BNA, which create uncertainty about the country’s fiscal position, have yet to be eliminated. iv. Identified deficiencies in Angola’s public financial management impair good macroeconomic management. High govemment-to-GDP ratios and large quasi-fiscal expenditures, for example, are associated with inadequate cash and debt management practices. A weak fiscal stance, o n the other hand, feeds high inflation which by i t s turn complicates budget formulation and impairs an efficient execution o f expenditures. The perpetuation o f public financial management practices that are n o n transparent and that are characterized by insufficiently clear channels o f accountability contribute to hinder fiscal discipline and obstruct macroeconomic stabilization efforts. Ultimately, these inadequate practices reflect political choices that impose heavy costs o n the economy. In this context, a notorious anomaly arising in the case o f Angola i s the existence o f a non- conventional budget execution system that creates uncertainty as regards the country’s actual fiscal stance. PUBLICEXPENDITURE A DUAL SYSTEM v. A distinctive feature o f Angola’s public financial management system i s the coexistence o f two parallel spending processes. The “conventional” system i s coordinated by the National Treasury Directorate, while the “non-conventional” one, which aims to facilitate transactions such as the servicing o f key external debt operations, i s centered around the national o i l company Sonangol. The “non-conventional” mechanisms partly emerged as a result o f the war economy, where a large portion o f extra-budgetary transactions were believed to be related to the procurement o f military material. While expediency considerations may have played a role in a context in which national security was at stake, there i s absolutely n o reason for the maintenance o f such mechanism in the current environment o f peace. vi. The co-existence o f parallel spending systems weakens transparency and accountability, and impairs planning. The non-conventional system o f public finances communicates with the formal system, but this happens through rather informal compensation mechanisms. In practice, Sonangol have held at the source part (or the totality) o f the profit o i l and taxes due i t i s supposed to deliver to the Treasury, in order to compensate i t s e l f for transactions (e.g., subsidies o n petroleum products, oil-backed debt service, other quasi-fiscal spending). A largely informal and unsystematic “reconciliation” process between the Ministry o f Finance (MINFIN) and Sonangol ensues, and the next payment-compensation-reconciliation cycle starts. vii. Independently of how organized these non-conventional mechanisms can become, they violate the existing financial legislation and must be discontinued. W h i l e during 2002 the compensation process seemed to be rather haphazard, in 2003 i t became somewhat more organized and predictable. This in i t s e l f presents the additional danger o f creating an artificial “functionality” that could lead to the perpetuation o f a mechanism that violates basic legislation and ultimately weakens the Ministry o f Finance as the chief fiscal institution in Angola. In addition, the existence o f the parallel expenditure execution systems weakens the budgetary process and creates uncertainty as regards the actual fiscal stance o f the Government. Chart E.1. provides a stylized picture o f this complicated relationship. . Chart E 1: The Articulation Between the Conventional and the “Non- Conventional” Spending Systems “Conventional” Spending Compensation Mechanisms: “Non-Conventional” Spending System Retention of taxes, profit oil Mechanisms I I Led by DNT. Led by Sonangol. Transactions take place outside Transactions take place in the context of the SIGFE. of the SIGFE. They include: > Oil-backed debt service > Other quasi-fiscal expenditures I I I H O W CAN NORMALITY BE RESTORED? A TWO-PRONGED REFORM STRATEGY viii. There have been improvements recently in the formal public finance management system, but more needs to be done to restore normality. Noteworthy i s the progress made since 2003 o n the Government’s I F M I S (Integrated Financial Management Information System or SIGFE in Portuguese - Sistema Integrado de Gestiio das Finanqas do Estado). The SIGFE i s the main formal instrument to register the Government expenditures and revenues and generate monitoring reports. An updated version o f it i s being gradually rolled out to budget units in Luanda and the provinces for standardization and integration o f all aspects o f the revenue and expenditure management and reporting. Once fully implemented and control features activated, this system will be one o f the strengths o f public financial management in Angola. ix. The reforms need to be deepened in spite o f some expected resistance. In 2002, the Government took the initiative to implement a wide-ranging Public Finance Modernization Program (PMFP), with three components: (i)a Public Finance Management component, whose aim i s to strengthen the public financial system, i.e., budgetary, financial, non-financial assets, accounting, internal control systems, public - iv - pricing, and insurance markets; (ii) an Information Technology component; and (iii) a Training and Capacity Building component. M a i n areas o f progress thus far include the strengthening o f the commitment and verification stages o f the budget cycle. The main challenges involved in meeting the PMFP goals will be political and institutional rather than merely technical. The main issue i s how to engage organizations which are not part o f the PMFP implementation but o f whose activities l i e at the core o f the PMFP goals. This is particularly the case o f Sonangol and, to a much lesser extent, the Tribunal o f Accounts. X. A two-pronged reform strategy should be adopted to address the issue of the dual system of expenditure execution. In view o f the co-existence o f a conventional and a “non-conventional” spending system in Angola, the PEMFAR proposes a two-pronged reform strategy: 0 T o strengthen the conventional system, by expanding and enhancing the credibility and effectiveness o f the formal budget preparation, execution, accounting, reporting and control systems. 0 T o “ring-fence” and phase-out the “non-conventional” budget execution mechanisms, while establishing a carefully designed “path towards normalization’’ whereby Sonangol would relinquish i t s Treasury-like activities over a reasonahle period o f time. xi. The proposed reform strategy recognizes weaknesses in the conventional system, which must be dealt with even as broader reforms permit a gradual return to normalization.The first prong o f the strategy acknowledges that the formal budgetary mechanisms also possess weaknesses which need to be removed. The presence o f such weaknesses facilitates the emergence and creates incentives for the maintenance o f “non- conventional” practices. The second prong recognizes the high price o f expediency - which becomes both an excuse and a conduit for corrupt practices - and proposes a transitional period in which the “non-conventional” mechanisms would be brought increasingly under the control o f the Ministry o f Finance and “agencies o f restraint” such as the Tribunal o f Accounts. xii. While the implementation o f neither of the two prongs would be easy, reforms associated with the second prong are likely to be more politically difficult. For one, most o f the reforms under the first prong are within the sphere o f influence o f the Ministry o f Finance, whose Public Finance Modernization Program (PMFP) may become the centerpiece o f the change process. In addition, the phasing-out and eventual elimination o f the “non-conventional” system is likely to face obstruction from powerful vested interests which are currently insulated from the influence o f the genuine public financial management institutions. A SHORT-TERM ACTION PLAN TO STRENGTHEN THE FORMAL SYSTEM ... xiii. Strengthening the conventional public spending system i s a key element of the reform effort. Despite recent reform efforts, the formal public financial management continues to be impaired by severe weaknesses that must be overcome. Those debilities -v- are pervasive, affecting all stages o f the budget cycle. Their mere presence, moreover, gives rise to incentives to bypass the formal mechanisms, thereby strengthening the incentive to continue with current “non-conventional” practices. Therefore, in the short- term, the formal system should be strengthened by means o f a series o f implementable measures. xiv. To immediately improve the budget preparation stage it i s necessary to improve coordination and forecasting skills in the Government. Short-term actions in this realm involve: the rigorous preparation o f realistic macroeconomic projections and their timely submission to the National Budget Directorate (DNO) so that the budget elaboration process initiates with a proper macroeconomic framework in place; the adoption o f the O i l Diagnostic’s Financial Model to generate more accurate revenue projections; and better coordination between MINFIN and MAPESS to improve the accuracy o f wage bill estimates. xv. The main weaknesses identified with respect to budget execution to a large extent refer to the non-adherence to the Organic Budget Law. The OGE i s unrealistic and therefore not easy to comply with in the first place; there i s a cash rationing process and a cumbersome month-to-month cash management that makes planning impractical for the budget units. As a result, some o f the latter find i t legitimate to commit above and beyond the authorized commitment ceilings and quarterly plans, and sometimes even above the approved annual budget. Shortage o f cash results in a selective payment o f invoices and lack o f transparency as well as creation o f arrears. In this respect, short-term measures include: ensuring the consolidation o f the Treasury’s Single Account (CUT); introducing a reasonable degree o f flexibility in the cash management process; training key staff on the modus operandi o f the SIGFE; and enforcing the sequencing o f the budget execution process established in the Organic Budget L a w (commitment, verification and payment). xvi. I n what respects accounting and reporting the main weaknesses are mostly associated with the incompleteness of the process o f implementing the new Accounting Law and the Chart of Accounts. Contrary to the Organic Budget Law, which stipulates that all budgetary operations have to be recorded according to the double entry system, public accounting i s s t i l l cash-based and single entry. The present accounting information system does not allow the closing o f accounts and the preparation of the General State Accounts (Contu Gem1 do Estado, or CGE) as stipulated in the Organic Budget Law. To correct this situation the Government should provide training on accounting and reporting practices o n a regular basis for relevant staff in the budget units; establish accounting and financial management procedures and routines for non- financial assets; and complete adoption o f an accounting manual as well as the new Chart o f Accounts. xvii. Strengthening audit and control systems i s crucial f o r both prongs of the reform strategy proposed in the PEMFAR. A major flaw in Angola’s public financial management system i s that audit and control systems, although to a large extent regulated by law, do not operate in practice. The main weaknesses in this area that need to be tackled include: strengthening the National Inspectorate o f Finance (INF), which i s in -vi - charge o f the internal control and audit functions but i s too weak to exert any real impact in terms o f enforcing compliance with the Organic Budget Law; and passing and enacting the required regulations for the internal control function in Angola, which have not yet been passed, despite the fact that the Organic Budget L a w was approved in 1997. RING-FENCING AND PHASING-OUT ‘“ON-CONVENTIONAL” SPENDING MECHANISMS xviii. Sonangol’s activities on behalf o f the Treasury are at the core o f the %on- conventional” mechanisms o f budget execution. These include quasi - fiscal activities and the servicing o f o i l backed loans on behalf o f the Government. All those transactions are performed outside o f the formal budgetary framework, the SIGFE. An ever-present risk therefore exists that their total executed amount will exceed what was allocated to them in the OGE, which further weakens aggregate fiscal discipline. The fact that they are not subject to the automatic controls o f the SIGFE also implies that lines o f accountability are blurred. Further, their existence gives rise to complex and largely informal compensation mechanisms between Sonangol and MINFIN. xix. T h e PEMFAR proposes a sequenced transition process, or “path towards normalization”, whereby %on-conventional” execution mechanisms would b e eliminated. This “path towards normalization” must involve the establishment o f clear reporting mechanisms from Sonangol to both the Ministry o f Finance and the Central Bank o n all transactions it carries out o n behalf o f the Treasury; strengthened internal control mechanisms (both within the Ministry o f Finance and Sonangol) to ensure that such transactions are conducted in accordance with legitimate fiscal policy objectives and acceptable public expenditure management practices; and improved accounting procedures, whereby the National Accounting Directorate at MINFIN obtains from Sonangol all the information it needs to prepare the yearly General State Accounts as well as Quarterly Budget Execution Reports. xx. “Normalization” refers in this context to a situation in which the “non- conventional” activities would have been eliminated over an agreed time period. This would represent the arrival point o f a transition process involving the k e y institutions - MINFIN, the BNA and the Ministry o f Petroleum (MI”ET). Chart E.2 summarizes the main milestones o f the transition process. Current Situation Coexistence of a formal budget execution system with “expedient” execution mechanisms. Compensation mechanisms between Sonangol and MINFIN. Weak oversight increasing fiduciary risk. BNA not fully exerting its role as foreign exchange authority. Ongoing reforms at Sonangol, MINFIN, and BNA. xxi. U n d e r these circumstances, the main challenge for the Government’s PMFP lies precisely on the second prong o f the proposed strategy. N o t only some o f the key institutional actors currently fall outside the scope o f the PMFP, but also the implementation o f the needed reforms may face s t i f f political resistance from influential quarters in Angola. Therefore, strong political support at the highest levels o f government i s required for the PMFP to succeed in this task. xxii. Difficulties are greater with the second prong as political economy considerations are bound to play a m a j o r role. The process o f dismantling the “non- conventional” processes over a period o f time will involve dealing with a powerful vested interests and traditional sources o f rent-seeking. In addition, it will also require the establishment o f an appropriate institutional hierarchy, one in which the Ministry o f Finance would restore i t s undisputed role as the prime public financial management institution in Angola. , - vii - C h a r t E 2: A Schematic View o f the Transition The Transition Ring-fencing of expedient mechanisms. Elimination of fuel price subsidies. Assessment and of ongoing quasi-fiscal activities by Sonangol, followed by transfer to appropriate institutions. Implementation of a debt management strategy, reducing recourse to oil- backed loans. Capacity-building at MINFIN, MINPET, BNA. The Arrival Point Full compliance with Organic Budget Law. Full compliance with Organic Law of the Central Bank. Transfer of concessionaire role to MINPET. -1- INTRODUCTION THENATURE AND SCOPE OF THE PEMFAR The basic starting point for a sound public expenditure assessment strategy i s the general proposition that the institutional framework and the management practices prevailing in a given country have a direct impact on budgetary outcomes at three levels: (i) aggregate fiscal discipline; (ii)the degree to which resource allocation and use reflect strategic priorities; and (iii) the efficient and effective use o f the resources thus allocated (World Bank, 1998). The PEMFAR focuses primarily o n issues related to the country’s institutional framework; as a result, it emphasizes issues more directly connected with level (i), although level (ii) and (iii) issues - related to strategic prioritization and operational efficiency - are dealt with in a limited way, that is, only to the extent to which they are affected by budgetary choices2 ’. More specifically, the PEMFAR concentrates on identifying the institutional weaknesses present in Angola’s public finance framework that give rise to severe fiscal problems - such as the consistent pattern o f extremely large unexplained discrepancies in the fiscal accounts. Therefore, the main objective o f the P E M F A R i s to provide a comprehensive and integrated assessment o f Angola’s fiduciary framework. In this process, i t also helps to fill a major knowledge gap in the critical areas o f public expenditure management and financial accountability in Angola4. The PEMFAR’s policy recommendations are aimed at getting the basics right5.Although i t is clear that institutional reforms need to be customized to the country’s needs, some fundamental principles should guide the formulation o f any proposed reform program: budgetary comprehensiveness and discipline; transparency and accountability; an adequate balance between restraint and flexibility; and a medium-term approach linking policy, planning, and budgeting. * F o r example, when operational performance is jeopardized by inadequate or untimely provision o f funding. F o r a recent overview o f level (ii) and (iii) issues as they affect the health and education sectors in Angola, see UNDP/IOM/UNICEF/WHO (2002). Angola’s one and only Public Expenditure Review was published in 1993. See The W o r l d Bank (1993). A Country Financial Accountability Assessment (CFAA) was never produced in Angola; a Country Profile of Financial Accountability (CFPA) was produced by the Bank in 1997. See Section I V for a fuller bibliographical discussion. “In other words, reformers should focus o n the basics o n w h i c h reform i s built, not o n particular techniques”. W o r l d Bank (1998), p. 8. -2- N ANGOLA THEPEMFAR APPROACH I The overall approach o f the PEMFAR has been to analyze the country’s legal, institutional, and regulatory framework, comparing them with the actual, prevailing practices. Accordingly, the diagnostic work as well as possible solutions have focused both on possible gaps in the legal-institutional framework and o n issues o f actual implementation of legislation, regulations, and procedures. The main goals o f this approach have been to maximize the country’s ownership o f the reform program proposed herein, while at the same time building upon existing efforts by the Angolan Government. Chart 1 below summarizes that approach. -3- Chart I.1: The PEMFAR approach in Angola -4 Legalllnstitutional Framework De Facto Implementation Budget Organic Law (9/97) Compliance with Law 9/97 and Decree 4/96 other instruments. New Accounting Law "Checks & balances" of the Others system in the various stages of the budget cycle. i Diagnostic Adequacy of legal/institutional framework vis-a-vis relevant international comparators and domestic needs. 1 I Adequacy Diagnostic pliance implementation performance vis-a-vis and prevailing IegaVinstitutional framework. \ 1 PEMFAR Recommendations Improvements in legal/institutionalframework Improvements in compliance and implementation performance A w Public Finance Modernization Program (PMFP) Comparison between PEMFAR recommendations and actions already contemplated in the PMFP. Recommendations for a revision of the PMFP. Implementation Support Economic Management Technical Assistance Credit. Other instruments. -4- ORGANIZATION OF THE PEMFAR The report i s organized around five major themes: Overview o f Angola’s Fiscal Outcomes; Assessment o f Angola’s Public Expenditure Management Framework; Assessment o f Angola’s Financial Accountability Framework; Update o f the Country Procurement Assessment Report (in annex 2); and, Linking Policy, Planning and Budgeting in a Credible and Transparent Resource management Framework. The fiduciary and accountability aspects o f the PEMFAR cover the following areas o f the central government: Institutional and legal fi-amework; Budget development; Budget Execution and Monitoring; Public Accounting and Financial Reporting; Intemal Control System and Records; and External Fiscal Reporting and Legislative Oversight. The P E M F A R contains nine chapters, organized in three parts: Context, Practices, and Policy Reform. Three annexes, focusing o n macroeconomic and fiscal information; public procurement; and o i l revenue management issues are also included. -5- 1. AN OVERVIEW OF ANGOLA'S FISCAL OUTCOMES Angola's long-term fiscal choices have been based on large public expenditures as a share of GDP. Despite the end o f the conflict in 2002, the composition ofpublic spending continues to be concentrated on activities with little contribution to development. I n fact, the expenditure-to-GDP ratio remains one o f the largest in Africa while the share o f capital spending is one o f the lowest. Identified deficiencies in the public financial management system have contributed to weaken the country's fiscal stance. As a result, the country 's fiscal choices have materialized in the form o f persistent fiscal dejkits, the accumulation o f domestic arrears, the existence of sizable quasi-fiscal expenditures, and high inflation. A concise analysis o f such choices and outcomes is what this chapter sets out to do. A. ANGOLA'S FISCAL LONG-TERM CHOICES High Expenditure-to-GDPRatios 1.1 Partly due to a 27-year long c i v i l war, the size o f the public sector in the Angolan economy i s considered high by international standards. The level o f public spending as a share o f GDP in Angola i s one o f the highest in Africa averaging as much as 55% o f the GDP between 1996 and 2003 (see Chart 1.1). The highest share was registered in 1999 (82%) reflecting the scaling-up o f the government's war efforts against the armed opposition movement UNITA. With the end o f the conflict in 2002, this ratio has been declining steadily, reaching 37.5% o f GDP in 2004, which is s t i l l high by international standards (see World Bank, 1993). Chart 1.1: Total Government expenditure as a percent of GDP for selected countries -6- 1.2 The challenge o f reconstruction will demand a shift in the composition o f public spending by function toward activities with a more direct contribution to development. Due to the war situation, more than h a l f o f the Government’s total expenditures have been historically devoted to general administration and defense. I n this context, a higher share o f expenditures have been devoted to defense and public order in comparison to what has been spent o n education, health, and other social sectors (see Table 1.1). This situation has started to change and social spending increased from 12.7% as a share o f total public expenditures in 2003 to 20.1% in 2004, while the share for defense and public order declined from 13.8% to 12.5% o f total expenditures over the same period. Despite this shift in the composition o f public spending by function, health and education expenditures as a percentage o f GDP in Angola (less than 2% for health and less than 5% for education) are (with the exception o f Equatorial Guinea) amongst the lowest in the African continent (with averages o f 6% and 10% for both categories, respectively). Table 1.1: Angola Government expenditure by function, 1996 - 2004, in percent of total’ 1996 1997 1998 1999 2000 2001 2002 2003 2004 General public services 18 15.5 17.8 3.7 7.3 27.8 33.2 47.9 30.0 Defense, Public Order and Safety 52.6 12.7 33.7 31.3 14.8 15.5 15 13.8 12.5 Social Spending (including health & education) 12.2 11 11.4 4.8 15.6 19.6 15.6 12.7 20.1 Economic Affairs and Services 2.9 3.3 3.6 1.6 3.4 7.4 5.2 8.9 5.7 Interest payments (committed) 25.9 10.7 19.9 11.1 9.3 0.2 2.1 16.8 31.7 Residual (unclassiRed/extra-budgetary) -1 1.6 46.8 13.6 47.5 49.7 29.5 28.9 - Low Shares of Capital Expenditures 1.3 Capital expenditures in Angola are also amongst the lowest in Sub-Saharan Africa. They have dropped by more than 50% over the course o f the 1990s, from as high as 35% to an average o f 12% o f total government expenditures, which i s less than h a l f o f the average registered for SSA countries (25.3%). Excluding the Democratic Republic o f Congo and Sudan, Angola has the lowest level o f capital expenditures over total government expenditures in the whole o f SSA.* In a country where there i s an urgent need to build up infrastructure and provide essential services to the population, financing public expenditure that “crowds in” private investment should be a way to support non- o i l growth and create a larger revenue base in the future. Source: IMF, W o r l d Economic Outlook databank. Source: IMF staff and Angolan Ministry o f Finance estimates. * Data from the W o r l d Bank African Development Indicators, 2003, p. 2003. -7- Extreme Dependence on Oil 1.4 Angola’s main source o f fiscal revenue i s through the taxation o f the o i l sector, including the state-owned o i l company Sonangol. As a result, fiscal revenues have been excessively vulnerable to international crude o i l price volatility and have not always been able to keep pace with expenditures. In the absence o f a consistent mechanism to stabilize revenues, the Government i s at the mercy o f fluctuations in international o i l prices - a problem made worse at times o f o i l market slumps. 1.5 On the other hand, during the war period the non-oil economy remained very dependent on o i l price developments. As a result o f the dependence o n oil, pro-cyclical spending related to o i l prices have led to “boom-bust” cycles with detrimental effects on the non-oil sector. The 1998 drop in o i l prices to $12 per barrel, for example, led to an immediate improvement o f the non-oil fiscal balance which was followed by a major deterioration the next year when the price o f o i l had returned to the level observed in 1997, before the price had fallen (see Chart 1.2). The non-oil economy i s apparently growing since 2002 with an estimated annual growth rate o f 8.8% in 2004, which i s again reflecting the impacts o f higher o i l production and rising international o i l prices. Chart 1.2: Non-oil fiscal balance and oil prices, 1996-20049 r 40 I 1’ 0% I -50% Q n 30 Crude oil price Q) 0 - .- (3 0 - I - ($US/bl.) E 20 -100% li 0 C -Non-oil fiscal 15 10 -150% balancehon-oil GDP -200% 1.6 Diamonds are another natural-resource source o f tax revenues in Angola. The country i s the world’s fourth largest diamond producer, with close to 12% o f the world’s export value in 2003. In 2004, Angola exported 6.63 million carats o f diamonds, worth U S $ 763 million. Fiscal contributions from the sector have improved dramatically in recent years, from US$44.6 million in 2002 to U S $ 112 m i l l i o n in 2003. Preliminary figures for 2004, however, point to a sharp reduction in Government revenues, to US$ 69.5 million. The potential for further growth in diamond tax revenues has yet to be fully exploited by the Government as informal production and smuggling remain significant in this sector while transparency remains very limited. Source: IMF staff and Angolan Ministry o f Finance estimates -8- B. LINKING CHOICES TO FISCAL OUTCOMES Fiscal Deficits 1.7 Fiscal deficits are pervasive in Angola even in periods o f rising o i l prices, despite some recent improvements. The increase in o i l prices observed in 1999, for example, did little to revert a long trend o f fiscal deficits and accumulation o f arrears. In 2000, as o i l prices continued to rise, the budget deficit on a commitment basis was s t i l l as high as 8.6% o f GDP, which translated into a large overall surplus o n a cash basis due to the continued accumulation o f domestic and extemal arrears. Despite a new drop in o i l prices in 2001, which worsened the budget o n a cash basis, the budget deficit o n a commitment basis was further reduced thanks to a reduction in domestic and extemal arrears. Between 2001 and 2002, the fiscal deficit (on commitment basis) more than doubled reaching 9% o f GDP, while the non-oil fiscal deficit remained very high, at an average o f about 38 % o f GDP in 2001-02. Preliminary estimates indicate a substantial decline in the fiscal deficit between 2003 and 2004, mainly reflecting exogenous factors, an apparent improvement in fiscal discipline, and policies to reduce fuel subsidies" (see Chart 1.3). C h a r t 1.3: Evolution o f the overall fiscal balance (YOo f GDP)" 30% I I 20% 10% o Overall balance 0% (coimit.basis) Overall balance -10% (cash basis) 0 Changes in -20% 1 arrears( net) -30% -40% I--J- U lo The sustainability o f this progress, however, m a y b e l i m i t e d by some associated developments: external debt levels have continued t o rise following the accumulation o f more, expensive oil-backed loans; international reserves remain low; the potential cost competitiveness o f A n g o l a n goods has been reduced; and, with inadequate expenditure control mechanisms, continuing payment arrears and rising debt levels, the fiscal position remains vulnerable to changes in o i l prices. 11 Source: IMF staff and Angolan Ministry o f Finance estimates. -9- Domestic Arrears 1.8 The accumulation o f arrears i s another marked feature o f Angola's public finances management system. Since the 1990s, the composition o f the government arrears has alternated between withholding domestic payments and withholding external interest payments'* (see Chart 1.4). Such practice i s also used by Sonangol with regard t o i t s tax obligations to the Government. This results in a vicious circle in which Sonangol's tax arrears to the Government mirror the Government's delays in reimbursing Sonangol for the i m p l i c i t subsidy to petroleum products sold domestically. This pattern o f arrear accumulation i s accentuated in periods o f o i l price declines and vice versa. Chart 1.4: Composition o f payment arrears (% of total change in arrears) 'i 300% 1 , , 1 Domestic payments I -200% -300% L a r g e Extra-Budgetary and Quasi-Fiscal Expenditures 1.9 Until very recently, fiscal indiscipline combined with limited comprehensiveness o f the budget had resulted in large extra-budgetary and significant quasi-fiscal expenditures. In 2002, for example, the Angolan fiscal accounts revealed the existence o f sizable quasi-fiscal expenditures carried out by the Central Bank (BNA), as w e l l as large fiscal expenditures incurred outside the framework o f the State budget. Quasi-fiscal e x p e n d i t ~ r e s 'were ~ estimated to b e in the order o f 6 % o f GDP, w h i l e extra-budgetary fiscal expenditures reached 14.7 % o f GDP. W h i l e the extra-budgetary expenditures have n o w been eliminated there are s t i l l concerns regarding the existence o f sizable quasi- fiscal operations performed by both BNA and Sonangol. 1.10 BNA's quasi-fiscal expenditures have generally fallen under t w o broad categories: (i)the servicing o f o l d on-balance sheet government obligations to residents (public sector salary supplements and liquidation o f the state-owned bank CAP-related expenditures), and (ii)the servicing o f off-balance sheet items related t o foreign debt and 12 The numbers for 1999 have been revised by IMF staff in order to account for Sonangol's tax arrears, estimated to having reached 13.2 percent o f GDP in 1999. 13 Resulting mainly from a number o f financial operations undertaken by BNA o n behalf o f the Government as well as by Sonangol. - 10- other foreign-currency denominated liabilities o f the central government. The negative figure in Chart 1.5 below shows that all quasi-fiscal expenditures were transferred as credit to the government by end-2002. Chart 1.5: BNA quasi-fiscal expenditure (flows) I I n millions o f Kwanzas As a percentage o f GDP I 3000 1 Millions of Kwanzas =Public sector salary supplements -2 I C A P -related e x p e n d i t u r e s 1.11 Sonangol also performs treasury-like operations on behalf o f the Government in a significant amount. In 2002, for example, the distribution o f quasi-fiscal expenditures by Sonangol was concentrated o n subsidies to petroleum products which amounted to 46% o f total quasi-fiscal expenditures (see Chart 1.6). In addition to this, in order to compensate i t s e l f for the expenditures incurred, Sonangol unilaterally withholds tax payments to make up for the quasi-fiscal expenditures. This practice creates uncertainty over the revenue situation and the Treasury’s cash position and i s against the basic principles o f sound financial management. Chart 1.6: Distribution o f quasi-fiscal expenditures by Sonangol during 2002 % of total Subsidies 45 I Delivery of 40 Petroleum Products 35 0 State Enterprises 30 25 0 Govemment 20 Ministries 15 IUNITA relocation 10 camps 5 iid Other 0 2002 - 11 - Overcoming the Plight o f “Unexplained Discrepancies” 1.12 A recent assessment o f Angola’s government expenditures recordings has collected evidence o f remarkable progress in the elimination o f the so-called “unexplained discrepancies” l4 from the budget. Up until 2002, quasi-fiscal extrabudgetary expenditures were not initially budgeted but ended up being recorded ex post in the government accounts creating gaps in the published accounts that were translated into a discrepancy o f roughly U S $ l billion between stated government funds and actual revenue. Unexplained discrepancies in the fiscal accounts were, among others, substantial funds received as signature bonuses for o i l contracts and o i l royalties, and non-transparent external debt transactions. The problem reached substantial proportions in 1997 and 1999 (amounting respectively to 22.47% and 18.38 % o f GDP), and by 2002 unexplained discrepancies s t i l l ,averaged 3% o f GDP, or US$ 347 million. In 2003 and 2004, these discrepancies were completely eliminated from the published accounts. c. SOURCES OF DEFICIT MAIN FINANCING 1.13 Angola’s external commitments remain a serious source o f concern. Angola’s total external debt i s currently estimated at US$ 8.9 billion, or roughly 70% o f GDP (including late interest). Out o f this total, approximately US$ 3.4 billion represent debt owed to Paris Club bilaterals, US$ 2.1 billion to non-Paris Club bilaterals (including oil- backed debt to Brazil), US$ 2.8 billion to commercial banks and US$ 890 million as suppliers credit, which are mostly securitized, or oil-guaranteed, debt. 1.14 Some “stylized” facts about the deficit financing modalities practiced in Angola are worth noticing (see Chart 1.7): 0 First, grants and debt relief have historically played a marginal role in financing Angola’s deficit on a cash basis. 0 Second, the government has alternated between net external borrowing and domestic financing in the second h a l f o f the 1990s, probably reflecting the country’s ability to borrow in international capital markets as well as the t e r m structure o f amortization repayments o f pre-existing loans. 0 Third, as new o i l field discoveries were made in deep and ultra-deep sea waters during the late 1990s, Angola benefited from one-time signature bonuses associated with the concession o f ultra-deep water Blocks 31, 32 and 33 amounting to nearly US$ 1 billion in 1999. 0 Fourth, the accumulation o f domestic and external arrears, and the receipt o f o i l signature bonuses allowed the government to accumulate large net claims against the banking system in 1999 and especially in 2000. l4 The discrepancy behind “unidentified” expenditures is calculated as the difference between recorded inflows (revenues plus below-the-line financing) and recorded outflows (executed expenditures as recorded by the formal budgetary system). -12- C h a r t 1.7: Financing Angola’s overall deficit (cash basis) (% o f GDP)” 20 7 Oil signature bonuses 0 Eternal borrow ing(net -20 ‘ A History o f H i g h Inflation 1.15 Angola has struggled for years with an unstable macroeconomy, but there have been recent signs o f progress. A succession o f failed or only partially successful stabilization plans have marked Angola’s economic policy history since i t s first major attempts to stabilize in 1987. On the inflationary front, for example, there was a pattern o f successive inflationary peaks during the period up to mid-1996 that was temporarily broken by the adoption o f a stabilization plan, only to be followed by an even higher peak after some months (see Chart 1.8). There was a “structural break” in the inflation rate time series around the introduction o f the “Nova Vida” Plan in 1996. Before that, inflation was not only four digits, but also very volatile, due to the continuous adoption o f new (and unsuccessful) plans. With the “Nova Vida” Plan, and the introduction o f a fixed exchange rate (to the dollar), inflation was halted for a while, but crept up again in 1997 as i t became clear that fiscal adjustment was again being postponed. 1.16 Inflation has declined substantially more recently, but it is s t i l l considered high. The decline in the rate o f inflation was made possible with a major change in macroeconomic policy implementation, popularly known as the “hard kwanza” policy. Under this new policy, the Government has promoted the liberalization o f the foreign exchange market, an active absorption o f domestic liquidity by central bank intervention (notably in foreign currency), development o f government securities, and improvements in fiscal control. As a result, the annual rate o f inflation f e l l from around 100% in 2003 to some 30% by end-2004. Over the same period, the exchange rate depreciated by less than 10% in relation to the U S dollar, implying a real appreciation in effective terms o f roughly 20%. l5Source: IMF staff and Angolan Ministry o f Finance estimates. - 13 - Chart 1.8: A Snapshot o f Inflation and Dollarization l6 iOO% 350% 300% A Quarterly inflation 250% 200% 150% 100% 50% 0% 1.17 Fiscal deficits and their monetization are widely known among the main causes o f inflation. The insufficient control o f public spending, including notably large extra budgetary expenditures and the sizeable operational deficit o f BNA, have induced large increases in base money. Additionally, in the past, favored interest groups, including Sonangol, have used arbitrage and other tactics to benefit fkom high inflation, for example, by delaying payments in domestic currency for o i l and other sales received in hard currency. Until 2002, this combination o f affairs had actually created positive incentives for high inflation. In order to guarantee the sustainability o f recent progress on the inflationary front, the Government should strengthen currently ongoing efforts to improve control on public spending and monetary growth. A Dollarized Economy 1.18 A recent IMF s t ~ d y noted '~ that broader monetary aggregates including dollar denominated assets seem to display a closer relationship with inflation than reserve money since the issuance o f kwanzas has declined over time. The high level o f dollarization o f the Angolan economy has been the result o f high and persistent inflation and U S dollars have become a main medium o f exchange. Such pattern aggravates the fiscal deficit in such a way that i t causes a revenue loss to the government in terms o f seigniorage. The IMF estimates that 80% o f demand deposits are dollar-denominated and that this phenomenon was exacerbated during 1999 and 2000 as inflation increased. Most o f the monetary growth in 1999 and 2000 came from increased net foreign assets that had the cumulative effect o f expanding the stock o f n e t credit to the government. l6Source: IMF Staff, Angolan authorities. l7Angola, Selected Issues and Statistical Appendix, African Department, July 11, 2003. - 14- The Road Ahead 1.19 I t i s possible to relate medium-term poor fiscal outcomes with weak public * expenditure' management (PEM) practices, ranging from deficient accounting to the lack o f a clear, prioritized, spending program. High government-to-GDP ratios and large quasi-fiscal expenditures, for example, are associated with inadequate cash and debt management practices. A weak fiscal stance, on the other hand, feeds high inflation which by i t s turn complicates budget formulation and impairs an efficient execution o f expenditures. The perpetuation o f public financial management practices that are non transparent and that are characterized by insufficiently clear channels o f accountability contribute to hinder fiscal discipline and obstruct macroeconomic stabilization efforts. Ultimately, these inadequate practices reflect political choices that have imposed heavy costs o n the economy. Therefore, the conclusion seems inescapable that aggregate fiscal discipline and macroeconomic stability cannot be established without thoroughly addressing the remaining pervasive deficiencies in public financial management in Angola. l8 The analysis will focus henceforward o n expenditures rather than revenues. The Bank i s starting work o n a Country Economic Memorandum for Angola, in w h i c h issues o f o i l dependency and vulnerability to o i l price changes w i l l figure prominently. - 15 - 2. AN OVERVIEW OF ANGOLA’S PUBLIC FINANCIAL MANAGEMENT SYSTEM Perhaps the most salient feature o f Angola’s public expenditure management and financial accountability framework is the coexistence o f at least two parallel, but articulated, public spending systems: the “conventional” system, coordinated by the National Treasury Directorate, and an “non-conventional ” one, which aims to facilitate transactions such as the sewicing o f key external debt operations, centered around the national o i l company Sonangol. A standard assessment o f a country’s fiscal framework would usually concentrate on the “conventional” system; the PEMFAR goes one step further by attempting to understand the workings o f the %on-conventional ’’ system as well as its articulation with the conventional one. The PEMFAR proposes accordingly the adoption o f a two-pronged strategy: (i) strengthening the formal public Jinancial management structures and tools; and (ii) phasing-out and eventually eliminating the %on-conventional mechanisms. ” A. ORGANIZATION AND COMPONENTS OF THE CONVENTIONAL PUBLIC FINANCIAL MANAGEMENT SYSTEM Overview 2.1 The basic premise for public financial management in Angola i s expressed in the Constitution, which stipulates that the State guides the development o f the economy and ensures the rational and efficient use o f all productive capacity and national resources. I t further pursues that the National Assembly i s responsible for review and approval o f the National Plan and o f the General State budget, as well as o f reports o f their execution. 2.2 O n the basis o f the authority vested in it, the National Assembly presides over the entire budget cycle - from approval o f the State budget to the examination o f the final State account^'^. O n the executive side, the Ministry o f Finance (MINFJN) plays the main role in public finance management, from budget preparation, budget execution and accounting o f all public transactions in Angola to the administration o f the State non- financial and financial assets. Each spending unit has the responsibility for the execution o f i t s budget under oversight and control o f the MINFJN and to provide monthly reports o n the execution. 2.3 MINFIN i s organized in six operational National Directorates (Budget, Accounting, Treasury, Taxes, Customs and Non-Financial Assets), six technical support offices (International Studies and Economic Relations - GEREI; Legal; Prices and Competition; Computer Technologies; National Inspectorate o f Finance - INF; and the l 9 See Box 2.1 for details. - 16- General Secretary). MINFIN has Provincial Delegations in the 18 provinces o f the country. Internal control i s under the responsibility o f the National Inspectorate o f Finance. 2.4 The other two ministries involved in public finance management are the Ministry o f Planning (MINPLAN), and the Ministry o f Public Administration, Employment and Social Security (MAPESS). The Central Bank o f Angola (BNA) hosts the single Treasury Account, while most in-country treasury operations are carried out by the Government’s fiscal agent, the Savings and Credit Bank (BPC). BPC has extensive territorial coverage (46 branches connected on line, in 14 o f the 18 provinces). In the remaining four provinces, budgetary payments are s t i l l being done by another public bank, Banco Comercio e Industria (BCI). Extemal control o f state financial accounts i s the responsibility o f the Tribunal o f Accounts, which was created in 1996 (Law 5/96). 2.5 Accounting and reporting i s done by both Budget Units and the National Accounting Directorate (DNC) at MINFIN producing budgetary and management accounts. The SIGFE has become the main instrument to generate those accounts and reports and an updated version o f it i s being installed and gradually rolled out for standardization and integration o f all aspects o f the revenue and expenditure management and reporting. Once fully implemented and control features activated, this system will be one o f the strengths o f public financial management in Angola, not encountered in many comparable countries in the region. 2.6 The mechanical and automated internal controls over expenditures, as part o f features o f the SIGFE, are complemented by the internal audit function established at the MINFIN and vested in INF, which i s a specialized department o f the Minister o f finance. A s discussed in Chapter 6, IN F needs strengthening to fulfill i t s function to i t s fullest, as other controls are not robust and a compliance culture i s lacking. 2.7 As for the external and independent controls, the external audit o f state accounts i s the responsibility o f the Tribunal o f Accounts, which was created in 1996 (Law 5/96) and has been gearing-up to fulfill the functions o f an Independent Supreme Audit Institution, yet to be fulfilled in Angola. 2.8 Finally, the cycle i s completed with the scrutiny o f the reports on the budget outturn by the parliament and the jurisdictional verification o f the management accounts by the Tribunal (see B o x 2.1). - 17- Box 2.1. Legislative Oversight in Angola The Parliament is at the very beginning and at the very end o f the financial accountability cycle as far as public finances are concerned. I n Angola, the National Assembly is expected topreside over the entire budget cycle -from approval of the Orgamento Geral do Estado (OGE) to the examination o f the final State accounts (Conta Geral do Estado, or CGE).. I a) The National Assembly at the Budget Approval Stage T h e Commission for Economy and Finance (51h Commission) o f the National Assembly has the power and responsibility o f monitoring budget preparation on behalf o f the National Assembly. In the execution o f this function, i t holds discussions with all the Ministries when the budget arrives at the National Assembly, and before i t is presented for a vote. During these reviews and discussions the Commission compares the proposed budget with the outcome for the preceding year, to see whether the proposed budget in reasonable in the light o f past experience. After the review, the Commission prepares a report, which it presents to the National Assembly for discussion in Plenary Session at the time Government i s presenting the budget for parliamentary approval. The budgets for 2002 and 2003 arrived in Parliament b y October 3 1 o f the preceding year, as required by law, and were dealt with as described below. I b) Legislative Oversight o f the Budget Execution Process The Budget execution process is monitored by the SIh Commission itself. If deemed necessary, the Plenary can also examine specific matters, based on the opinion issued by the Commission. Again, i f needed, the Commission i s empowered to summon the fiscal authorities for clarification o f issues. The Organic Budget Law (Law 9/97) establishes that quarterly Budget execution reports are to be submitted to the Assembly. In practice, shorter progress reports (“Balancetes Trimestrais de Execupio) are presented to the Commission 45 days after the end o f each quarter. T h e information is presented cumulatively; therefore, the 2”d Balancete contains information on the semester, and the 4’h shows the yearly Budget execution. Delays are common in the presentation o f such Balancetes. These Balancetes are treated as internal documents; occasionally, reference to their contents i s made in the press. K e y obstacles to proper Parliamentary oversight are l o w capacity and lack o f expertise on budgetary matters - particularly to analyze priorities at the sectoral and provincial levels - on the part o f the Commission itself. In addition, the Commission’s work seems to focus more in scrutinizing the budget proposal rather than on budget execution per se. The Commission currently has only t w o “tecnicos” to carry out all its research and advisory work. It plans to recruit four more, bringing the total to six advisers with different specializations T h e Constitutional L a w also establishes that the Govemment i s authorized to contract external loans to address cash flow problems (“emprestimos de tesouraria”), as long as the loans are liquidated in the course o f the year they were contracted. F o r loans w i t h longer maturity, approval by the National Assembly i s required. In practice, however, there are two modalities o f approval: “explicit authorization”, in which loans are individually examined b y the Assembly; or “implicit authorization”, in which the total amount o f foreign loans to be contracted in a given year is recorded as part o f the line item “external financing” in the State Budget. I t seems that “explicit authorization” is not a common practice; however, from a legal viewpoint, the 51h Commission i s empowered to request full information on the loans before they are contracted, and the authorities must attend to such requests. c) Closing the Cycle: Examination of the Conta Geral do Estado In addition to the quarterly balances for the last quarter o f the year, the Commission i s supposed to receive from the Accounts Tribunal, not later than 15 months following the end o f the fiscal year, an audited financial statement, together with a summary audit report. This i s expected to happen for the first time around in early 2004. d) Follow-up of Audit Findings One of the most important traditional duties of the Parliamentary Accounts Committee (PAC) in other countries is tc follow-up the recommendations ofthe SAI, and insists that they be implemented by the government. I n Angola, where the SAI is a court of law, with powers to impose penalties for accountingfailures, this traditional role of the PAC ma) prove to be less important. - 18 - B. A NON-CONVENTIONAL BUDGETING SPENDING SYSTEM DISCONNECTS AND POLICY 2.9 The most salient feature o f Angola’s public expenditure management and financial accountability framework i s the coexistence o f at least two parallel, but articulated, expenditure execution systems. The “conventional” system, coordinated by the National Treasury Directorate, and an “non-conventional” one, which aims to facilitate transactions such as the servicing o f key external debt operations, centered around the national o i l company Sonangol, among other “Treasury-like” operations carried out by the latter. 2.10 Because o f such dual system, a significant share o f total expenditures i s executed outside the formal budgetary system. Although the numbers vary considerably in that respect, the Government’s 2002 Report o n Budgetary and Financial Execution o f the OGE2’ indicates that approximately 28.4% o f total expenditures (or 14.8% o f GDP) were executed outside o f the formal budgetary framework, the SIGFE. Given that, presumably, the entirety o f the OGE i s loaded in the SIGFE as the fiscal year starts, this should give a reasonable approximation o f the degree o f coverage o f the OGE in 2002 - a little more than 70% o f total Government expenditures captured in the 2002 Execution Report. 2.11 Sonangol i s at the core o f the “non-conventional” mechanism. What singles out Sonangol’s role in Angola’s public financial management processes i s the degree o f complexity and fbnctionality o f its articulation with the Treasury. In practice, Sonangol holds at the source part (or the totality) o f the profit o i l and taxes due it i s supposed to deliver to the Treasury, in order to compensate i t s e l f for transactions (e.g., subsidies o n petroleum products, oil-backed debt service, other quasi-fiscal spending2’). A largely informal and unsystematic “reconciliation” process between MINFIN and Sonangol ensues, and the next payment-compensation-reconciliation cycle starts. Figure 2.1. provides a stylized picture o f this complicated relationship. Figure 2.1 : The Articulation Between the Conventional and the “Non-conventional” Spending Systems “Conventional” Budget Compensation Mechanisms: “Non-conventional” Execution System Retention o f taxes, profit oil Expenditure Execution b Mechanisms ‘ LedbyDNT. Treasury-like operations Transactions take place in Led by Sonangol. the context o f the SIGFE. Transactions take place outside o f the SIGFE. They include: > Oil-backed debt service > Other quasi-fiscal 2.12 In this context, not only the links between policies and budgeting become weak, but the foundation for sound public expenditure policymaking remains fragile. As briefly 2o See DNC~MI N FN I (2002). 21 See Chapter 7 for details. - 19- discussed in Chapter 1, the OGE has been conceived to a large extent as a wartime budget, and the composition o f Government expenditures essentially reflect that (with the share o f defense and public order expenditures as high as all the social sectors combined). This i s suggestive that public expenditure policies - including fuel price and utility tariff subsidies - are established in a largely ad hoc manner, and without the underpinning o f a medium-term spending plan or strategy. Such situation represents a serious hinder to planning initiatives because an effective medium-term expenditure framework cannot work if a solid budgetary management process - particularly as regards budget execution - i s not in place22 2.13 The formal budgetary system must be strengthened to guide public expenditure policy. The co-existing spending mechanisms weaken the budgetary process and create uncertainty as regards the actual fiscal stance o f the Government. As a result, there is not much room for effectively planning ahead and the reality i s that the Government’s Public Investment Program (PIP) remains merely as a l i s t o f projects to be carried out over the course o f a given year when i t should be the starting point for a better integration between policies and their funding. In the same fashion, the OGE, which i s often referred to as a “Program Budget”, has a limited scope and actually presents a listing o f current Government programs that are not always consistent with the Government’s intentions as stated in i t s annual or bi-annual Economic and Social Programs. c. WEAKNESSES OF THE CONVENTIONAL BUDGETARY MAIN PROCESS 2.14 The current budgetary legislation is complex, whereas agencies and mechanisms o f control are fragmented and ineffective. There i s no law below the constitution to clearly establish, in one document, the roles and responsibilities o f all government agencies, relations between agencies, or establishing the authority o f the MINFIN in respect o f the public financial management and accountability. In fact, as an attempt to clarify, operationalize, change, or complement the laws, there i s a mass o f directives issued over the years. They sometimes contradict each other and need to be carefully reviewed and consolidated. 2.15 The budget legal and regulatory frameworks are not strictly followed by executing agencies. O n the one hand, the financial operations undertaken by the national o i l company Sonangol contradict the authority o f the MINFIN and the BNA and make the revenues and debt management opaque. On the other hand, there are clear indications that the budget framework i s not strictly followed by all executing agencies; sometimes reports fail to provide a sincere and truthful picture o f the situation. Whatever the legislation, they should be complied with and, in particular, the annual budget law which i s clear and appears annually should be followed by the executive, results and reports audited by the supreme audit institution. They should finally be scrutinized by the National Assembly. 22 As argued by L e Houreou and Taliercio (2002), “consistency between the budget and its execution i s a precondition for transparency, predictability, and accountability. In a country where budget execution (Le., actual expenditure) bears little resemblance to the voted budget (i.e., the intention t o spend by sectors, fimctions, and programs), an MTEF i s not likely t o b e taken seriously by sector ministries, nor by parliamentarians, nor by c i v i l society” (p. 26). - 20 - 2.16 There seems to be outright violation o f the rules and bypassing o f the system without consequences for the violators. The violations are evidenced by (i) arrears created by the executing agencies above and beyond the authorized ceilings and (ii) the discrepancies that exist between the administrative and financiaVfisca1 accounts and reports. The answerability i s not clearly established and there are no known consequences for the violators o f the law. N o sanctions are mentioned in the legislation or known t o have been taken against violators. Furthermore, rules o f ethics and procedures manuals to assist civil servants in the fulfillment o f their functions either do not exist or date back to the colonial time. 2.17 Finally, the internal and extemal controls have not been working effectively. While the intemal audit bodies o f the Executive lack capacity and motivation, and have not been effective, a Supreme Audit Institution (the Tribunal o f Accounts) did not exist until recently. I t i s worth noticing that Government attempts to reform the system are in the right direction and, with assistance, the intemal and external controls can become functional. By appointment o f the president o f the tribunal, the supreme audit institution has already started i t s preliminary work. This institution needs help and support o f donors to fulfill i t s crucial function. A stronger and more effective scrutiny by the parliament will further reduce the accountability gap. D. THENON-CONVENTIONAL SPENDING SYSTEM 2.18 The non-conventional system i s based o n complex and largely informal compensation mechanisms between Sonangol and MINFIN. These include quasi - fiscal activities and the servicing o f o i l backed loans o n behalf o f the Government that are performed outside o f the formal budgetary framework. Their existence blurs lines o f accountability and transparency in public expenditure management in Angola. 2.19 The volume o f expenditures executed through the non-conventional system i s significant. According to the final 2002 Budget Execution Report, expenditures by Sonangol (excluding subsidies and debt service) outside the SIGFE represented almost 11% o f total executed Government expenditures in the period. In addition, Sonangol handled in 2002 US$ 1.2 billion to service oil-backed loans o n behalf o f the Treasury. In 2002, budgeted price subsidies totaled some US$300 m i l l i o n dollars, o f which US$220 million (equivalent to 2 percent o f GDP) were in arrears by the end o f the year. Most o f these price subsidy arrears (89.8 percent) were due to Sonangol Holding (through Sonangol Distribuidora) for the provision o f petroleum products at below-market prices. The revised 2003 Budget incorporated some US$ 257 m i l l i o n to cover for goods and services directly paid for by Sonangol. 2.20 The aimed expedience i s a also source o f tension. Historically, Sonangol assumed a prominent role in the execution o f expenditures o n the basis purely o f expediency due to its having the resources and skills to perform the activities concemed. However, these activities are not the responsibility o f a commercial o i l and gas company but l i e rather with Government. Sonangol’s continued perfonnance o f these quasi-fiscal responsibilities not only blurs and complicates Sonangol’s commercial operations and aspirations but also takes up the time o f its o w n managers who would be better utilized in -21 - furthering Sonangol’s own operational objectives. It also results in tension between Sonangol, MINFIN and the BNA. 2.21 Sonangol activities on behalf o f the Government are largely unbudgeted and are recovered against tax as offsets. This i s a purely compensation mechanism in which no actual cash transaction occurs, preventing recording in the CUT. Inpractice, DNI agrees on final taxes due, but a separate MINFIN Department [GEREI] negotiates the value o f the offsets. A MINFIN Committee finally agrees with Sonangol the net final balance o f taxes payable. The valuation and settlement o f these complex adjustments against final audited taxes due for the year i s a time consuming process involving several decision stages that require political approval. 2.22 The non-conventional spending mechanism creates a worrying burden to Sonangol. The importance o f transparency to Sonangol in managing i t s expanding commitments has never been more important in view o f its ever increasing range and size o f capital commitments. Measurement o f the size and timing o f payments and required funding for future major commitments requires transparent and accurately recorded transactions with sophisticated planning, budgetary and cost control techniques in order to avoid major cash flow problems. At present, there appears to be a danger o f over stretching o f resources when the foregoing activities are added to the range o f quasi-fiscal and concessionary activities which Sonangol performs o n behalf o f the Government. Ideally, therefore, the “non-conventional” activities should be transferred back to MINFIN and the BNA as soon as possible and the Concessionary responsibilities should be transferred to the Ministry o f Petroleum thereby freeing up skills and resources within Sonangol. E. REFORM A TWO-PRONGED STRATEGY 2.23 The existence o f a dual system o f expenditure execution calls for a two-pronged reform strategy. Having identified the co-existence o f a conventional and a “non- conventional” system o f expenditure execution as the most basic financial management issue in Angola, the PEMFAR proposes accordingly a two-pronged reform strategy: 0 T o strengthen the conventional system, by expanding and enhancing the credibility and effectiveness o f the formal budget preparation, execution, accounting, reporting and control systems. 0 T o “ring-fence” and phase-out the %on-conventional” budget execution mechanisms, while establishing a carefully designed “path towards normalization” whereby Sonangol would relinquish i t s Treasury-like activities over a reasonable period o f time. 2.24 The proposed reform strategy recognizes weaknesses in the conventional system and recommends a gradual return to normalization. The first prong o f the strategy acknowledges that the conventional budgetary mechanisms also possess weaknesses which need to be removed. The presence o f such weaknesses facilitates the emergence and creates incentives for the maintenance o f “non-conventional” practices. The second prong recognizes the high price o f expediency - which becomes both an excuse and a conduit for corrupt practices - and proposes a transitional period in which the “non- - 22 - conventional” mechanisms would be brought increasingly under the control o f the Ministry o f Finance and “agencies o f restraint” such as the Tribunal o f Accounts. 2.25 While the implementation o f neither o f the two prongs would be easy, reforms associated with the second prong are likely to be more politically difficult. For one, most o f the reforms under the first prong are within the sphere o f influence o f the Ministry o f Finance, whose Public Finance Modemization Program (PMFP) may become the centerpiece o f the change process. In addition, the phasing-out and eventual elimination o f the “non-conventional” system i s likely to face obstruction from powerful vested interests which currently insulated from the influence o f the genuine public financial management institutions. 2.26 The transition process should ultimately lead to the following arrival point: 0 Full compliance with the Organic Budget L a w and 0 The elimination o f the non-conventional mechanism o f budget execution. 0 Full compliance with the Organic L a w o f the BNA, restoring i t s role as Angola’s foreign exchange authority. 0 Elimination o f Sonangol’s dual role, and transfer o f its Concessionaire functions to the Ministry o f Petroleum. 2.27 The remainder o f the PEMFAR develops further the storyline presented in this chapter. Part 11, “Practices” (Chapters 3-6), describes the strengths and weaknesses o f the formal budgetary mechanisms, and prepares the ground for policy recommendations for the first prong o f the reform strategy. Chapter 7 discusses the main aspects o f the MINFIN-Sonangol relationship, which l i e s at the core o f the “non-conventional” practices, and proposes a four-phase “ring-fencing” program that forms the basis for the second prong o f the strategy. All comes together in Chapters 8 and 9, in which a basic reform program i s proposed o n the basis o f existing Government tools and initiatives, such as the PMFP. - 23 - 3. BUDGET PREPARATION I n Angola, the most critical budget preparation issues are the coverage and the credibility of the OGE, the State Budget. I n a sound public expenditure management system, a well-functioning budget preparation stage performs two critical functions, by: (i) ensuring consistency between planned aggregate expenditures and macroeconomic targets (aggregate fiscal discipline, or “level one”; and (ii) reflecting strategic policy priorities in planned expenditure allocations (“level two’?. This chapter focuses mostly on “level one” issues; expenditure policy issues w i l l be taken up again in Chapter 8. A. PREPARATION PROCESS IN ANGOLA OVERVIEW OF THEBUDGET T h e Budget Preparation Cycle 3.1 Typical stages o f a budget preparation cycle involve23 : (i) the realization o f macroeconomic projections and the preparation o f a macroeconomic framework for the budget year24; (ii) the allocation o f total budgetary resources among line ministriedspending agencies; (iii) the circulation o f instructions to line ministriedspending agencies by the agency in charge o f budget preparation; (iv) the submission o f bids by l i n e ministries to the agency in charge o f budget preparation; (v) the conduct o f negotiations between the Ministry o f Finance and the line ministriedspending agencies; and (vi) endorsement o f budget proposal by the Cabinet, prior to submission to Parliament for approval. 3.2 The budget preparation cycle in Angola largely follows a similar sequence o f steps, summarized in B o x 3.1 below. Within the legal framework provided by the Organic Budget L a w (or Lei Quadro do Orqamento), an annual budget law i s enacted that presenting the revenue and expenditure estimates for a given fiscal year25. It starts when the Economic Consistency Team26 prepares a proposal for a macroeconomic scenario containing quantitative objectives and targets, and some assumptions on key variables. With that scenario, the Tax Directorate (DNI) makes revenue projections, and based on them, the Budget Directorate (DNO) establishes expenditure ceilings for line ministriedspending agencies. The following step in this chain i s the distribution o f budget preparation guidelines for the following year’s OGE, with the guiding principles for budgetary policy (Orientaqaes de Politica Orgamental), prepared by DNO. Both the See Potter, and Diamond (1999). 24 In countries w i t h multi-year budget planning, the macroeconomic framework can span three o r more years. 25 Angola’s fiscal year runs f r o m January 1 to December 3 1, thus coinciding with the calendar year. 26 Grupo de ConsistEncia Macroeconomica (GCM) - composed by the Minister o f Planning, the Minister o f Finance, the Governor o f the Central Bank, BNA, and the Minister o f Labor and Social Security. - 24 - macroeconomic scenario and the budget guidelines require approval by the Council o f Ministers. By mid-year, line ministriedspending agencies prepare budget requests, which are reviewed by and negotiated with MINFIN. B o x 3.2 summarizes the budget preparation process f r o m the standpoint o f line ministries/spending agencies. I . Box 3.1. Timetable for budget preparation: . - JanuaryMarch Macroeconomic scenario ApriVMay - preparation and distribution o f budget preparation guidelines (“budget call . circular” to line ministrieshpending agencies and Provincial Governments) 1 June/July - Line ministrieshpending agencies prepare budget requests August - Review and negotiation between MINFIN and line ministrieshpending agencies 27 1 By September 30 - Budget proposal presented to the Council o f Ministers . 1 By October 3 1 - Budget proposal endorsed by the Council o f M i n i s t e r s and submitted to the National Assembly. 28 1 By Mid-December - Annual Budget approved by the National Assembly . Box 3.2. Budget Preparation from the Standpoint of Spending Agencies Typically, line ministries, spending agencies and provincial governments receive, around April or May, guidelines for budget preparation (including expenditure ceilings): recurrent expenditures guidelines f r o m MINFIN, and capital spending instructions f r o m MINPLAN (guidelines for PIP include ceiling expenditures for each sector and instructions for resource appropriations). At line ministries, the General Secretariats prepare proposals for recurrent expenditures; those for capital spending are the responsibility o f the Planning Departments. Budget proposals are handed to MINFIN by June each year and negotiations take place around July. For capital expenditures, higher priority is given t o projects already underway; remaining funds can then be allocated for n e w projects. 3.3 Budget ceilings provided to the line ministriedspending agencies in principle are derived f r o m macroeconomic and revenue forecasts (the k e y variables used in the macroeconomic forecasts are listed in Table 3.1 ). 3.4 However, in practice, macroeconomic projections are frequently made available t o DNO with a lag, w h i c h means that the budget elaboration process starts without a proper macroeconomic framework in place. 3.5 In order to establish expenditure ceilings, DNO prepares cost estimates, based on the expenditure actually executed in the previous year, the number o f employees per spending agency and the number o f spending agencies per line ministry. 27 The draft proposal o f the Annual Budget, “ante-project0 de Orqamento”, with recurrent and capital expenditures as prepared by MINFIN. There are no legislative l i m i t s o n expenditure, deficit and borrowing. 28 The Budget approval process underwent some modifications since last year with the introduction o f new methodologies and routines for approval, Le., the whole process n o w takes 6 weeks (divided into 30 specific stages) for conclusion, once the Annual Budget documents are submitted to the National Assembly. - 25 - 3.6 Once the “Orientaqbes de Politica Orqamental” are approved, MINFIN and MI”LAN launch the process o f preparation o f the annual budget and o f drafting o f the Public Investment Program (PIP). A budget call circular i s issued, including the expenditure ceilings by line ministriedspending agencies and provincial government^^^. B. THE CREDIBILITY OF THE OGE 3.7 In order to improve the credibility o f the OGE, the Government needs to make progress in two crucial fronts: (i) the realism o f macroeconomic assumptions and projections, including revenue and expenditure projections; and (ii) the reliability and verifiability o f the cost information provided by the Budget Units, during the preparation o f the OGE, particularly with respect to personnel costs in politically-sensitive areas o f the public administration. Realism o f Macroeconomic Assumptions and Projections 3.8 At the root o f public management issues - such as overoptimistic budget formulation, inadequate cash and debt management or the absence o f a medium-term approach to budgeting - lies inadequate macroeconomic forecasting. With the exception o f overall growth and o i l sector growth that have been accurately forecast, there are important discrepancies between actuals and the OGE assumptions. The fiscal deficit, for example, has been consistently underestimated over the last 4 years (see Chart 3.1) - the same happening with the case o f inflation. Chart 3.1 : Overall Fiscal Balance on a Commitment Basic (% o f GDP): OGE vs. Actuals 1 %6 - Overall Fiscal balance (%of GDP) comitment basis 4 2 0 -2 -4 -6 -8 -10 3.9 The “over-optimistic bias” in the OGE exacerbates reputational risks associated with the Government’s own disinflation program: while i t might be argued that the underestimation o f inflation and deficit i s indeed intended as an announcement o f the Government’s intention to tighten fiscal spending, the fact that those assumption have been proved incorrect in the same direction indicate a degree o f inconsistency that might 29 Revisions o f the draft budget are s t i l l frequent; one reason is, as pointed out above, that budget preparation often starts without the fundamental economic assumptions being fully established. - 26 - have unintended effects. In particular, i t might impel economic agents to require the application of more stringent fiscal restraint measures for a longer period of time before they believe on the Government’s commitment to stabilization and act accordingly3’. Reliability and Verifiability of Cost Information 3.10 The lack o f a centralized payroll system undermines the credibility o f the wage bill estimates (25.8% o f the total Revised 2003 OGE). This i s further complicated by the fact that the MINFIN and MAPESS registries do not include staffs from the Ministries of Defense and Interior. 3.1 1 Because o f difficulties encountered by MAPESS in i t s listing o f all administrative staff, MINFIN has found more efficient to work on the l i s t o f personnel prepared b y each budget unit (UO) for budgetary purposes at the beginning o f the year which, unfortunately, might not yet coincide with MAPESS’ own lists. 3.12 In addition, control over the whole administration o f staff has not been efficient so far, leading to higher real personnel costs than estimated at the time o f budget preparation. Indeed, the UOs’ cash need previsions sent to DNT every month might differ from the six-month cash commitments they obtain automatically from the SIGFE system on the basis o f their annual budget allocation, thus creating additional discrepancies at the time o f payment. 3.13 A review o f budget execution for the first six months of FY03 shows that personnel costs estimates for the year increased from IRO 1,235 million at the time o f budget preparation to IRO 1,749 million as of June 30, 2003, that i s a total increase o f 42%. Yet, for civil servants initial budget figures had been under estimated by 23%, while personnel expenses for the defense, security and public order were revised up by 162%. As the payment o f salaries i s first priority for the Treasury, and because of the high share o f personnel costs in total current expenditures (43%), the impact o f under estimating personnel costs might be one o f the main factors creating scarcity o f cash in the Treasury. This situation, in turn, may lead to arrears in the payment o f other expenditures already budgeted and most probably already committed. Recommendations for Reform 3.14 The process o f generating credible macroeconomic assumptions and projections should not be seen as a pro forma exercise. Past experience - through, for example, the use o f time-series techniques to assess the behavior over time o f key variables, when applicable - should be used to produce more realistic macroeconomic assumptions for budget preparation. 3.15 A review o f public administration costs o f personnel should be undertaken. One o f i t s objectives would be to identify the reasons explaining the discrepancies between budget estimates and reality. This should facilitate a more accurate estimation o f current expenditures during the budget preparation stage and lead to a progressive reduction in the amount o f arrears. 30 Perhaps the main indicator that the Government’s disinflation policies have become more credible i s a sustained increase in the demand for Kwanzas and the accompanying de-dollarization o f the economy. - 27 - 3.16 I t i s critically important to enhance institutional coordination between MAPESS and MINFIN, on the one hand, and the Ministries o f Defense and Interior, o n the other. The personnel information systems at MAPESS and MINFIN should be unified, and a move towards a single cadastre o f all public sector personnel should be conducted. c. THECOVERAGE OF THE OGE 3.17 Important aspects o f Angola’s Budge Law pertaining to the principles o f unity and universality o f the budget are critically and consistently violated. Areas in which OGE coverage i s faulty or non-existent include: (i)expenditures camed out by Sonangol quasi-fiscal spending by BNA; (iii) o n behalf o f the Treasury; (ii) B N A ’ s operational deficit; (iv) donor grants and counterpart finds; (v) own revenues by autonomous institutions, autonomous funds, and provinces; and (v) payments to budget units located abroad (embassies and diplomatic missions). 3.18 Although the numbers vary considerably in that respect, the Government’s 2002 Report o n Budgetary and Financial Execution o f the OGE3’ indicates that approximately 28.4% o f total expenditures (or 14.8% o f GDP) were executed outside o f the formal budgetary system, the SIGFE. Given that, presumably, the entirety o f the OGE i s loaded in the SIGFE as the fiscal year starts, this should give a reasonable approximation o f the degree o f coverage o f the OGE in 2002 - a little more than 70% o f total Government expenditures captured in the 2002 Execution Report. 3.19 Defective coverage has a triple impact on Angola’s fiscal and public financial management. First, i t denotes insufficient transparency and unclear accountability, which creates an environment conducive to fraud and corruption. Second, i t usually leads to an underestimation o f the “true” fiscal deficit, thereby complicating an assessment o f the country’s fiscal stance. Third, i t further undermines the credibility o f the OGE. 3.20 Quasi-Jiscal spending by the BNA32 has mostly taken the form of: (i) costs related to the process o f liquidation o f the C A P (Caixa de Agricultura e Pesca) bank; (ii) salary supplements to certain categories o f c i v i l servants; and (iii) payments in foreign exchange for Government’s liabilities contracted abroad. In total, these expenditures reached U S $ 107.1 m i l l i o n in 2001 (1.1% o f GDP) and U S $ - 92.3 million in 2002 (- 0.8% o f GDP). In addition, BNA’s operational deficit, estimated at U S $ 247 m i l l i o n in 2002 (2.2% o f GDP) has not been included in the OGE. 3.21 A recent report by the I M F ’ s Fiscal Affairs Department33- which counted with the participation o f the Bank’s PEMFAR team - analyzed extensively the other expenditure categories for which coverage i s insufficient. That report’s main findings are highlighted as follows: 0 The OGE has not been able to capture the totality o f donor grants and counterpart funds. Data from OECD’s Development Assistance Committee suggest that during 1995-1999 Angola ODA receipts were actually above the 31 See DNCMI N FN I (2002). 32 See Alvesson Torrez (2003). 33 See Tollini et a1 (2003). - 28 - average for Sub-Saharan Africa. The 2004 OGE estimates donor grants as US$ 14.6 million, or 0.2 % o f the total Budget, but this i s very likely to underestimate the true amounts. New mechanisms for the recording o f donor grants were established by Decree 80/2001. In practice, only those grants for which there are associated Government counterpart funds end up recorded, as they have immediate fiscal implications for the OGE. e O w n revenues by Angolan embassies or diplomatic missions abroad are not systematically recorded in the OGE. The usual control mechanism involves a monthly submission, by such entities, o f their accounts to DNC, including their own revenues as well as their total expenses. [The total share o f embassies and diplomatic missions in the 2004 OGE amounts to only -%. O w n revenues by autonomous institutes34 are reportedly fully recorded in the OGE.] Current legislation states that autonomous institutes must surrender 60% o f its own revenues to the Treasury. In both cases o f o w n revenues, reporting to D N C i s not systematically done. This i s particularly relevant to the release o f budget credits to embassies and diplomatic missions, as DNT would as a r u l e transfer them only the balance o f o w n revenues minus total expenditures. Therefore, there is a built- in incentive for those entities to under-report their o w n revenues in order to increase the inflows o f funds from the Treasury. e O w n revenues by provincial governments are also reported ex post - if at all - to DNC. In many instances, local revenues become, in practice, extra-budgetary funds for provincial governments. There occurs here a “compensation mechanism” analogous to that between Sonangol and MINFIN: provincial governments have an incentive to retain their o w n revenues as monthly financial quotas from DNT often reach them with a delay (see Chapter 4 for more details). In addition, the provincial sub-accounts (“provincial CUTS”) o f the Treasury Single Account (CUT) often maintain a positive balance, contrary to the “principle o f zero balance” advocated in such cases. 3.22 The main recommendations from the FAD report - endorsed by the PEMFAR - on coverage issues are: e to enhance interaction with UNDP, which has closer contacts with the various donor agencies, to obtain updated information about the activities and plans o f different donors; e to enforce a recent agreement between MINFIN and MINARS35 to improve the flow o f information o n humanitarian assistance; e to create an individual code for each donor in the source-of-funds classification modality o f the OGE, facilitating identification o f expenditures in accordance with the sponsoring donor agency; 34 Autonomous funds are being subject to a complete evaluation by MINFIN. Those funds also have own revenues (legislation states that at least 25% o f their total spending be covered by their own receipts). During this period, OGE resources to those funds w i l l be limited to salaries and minimal operational costs. 35 Ministry o f Social Assistance and Reinsertion. - 29 - a to increase the capacity in DNO, DNC, and D N T to adequately oversee the process whereby own revenues are calculated, collected, and reported to MINFIN; 0 to expedite the process whereby DNT releases monthly financial quotas to the provinces (refer to discussion in Chapter 4); and 0 to accelerate the process o f roll-out o f the SIGFE to the provinces, allowing real-time connection between the Treasury Single Account (CUT) at BNA and the provincial CUTS, ensuring observance o f the “zero-balance principle” for the latter. The Revised 2003 OGE 3.23 The Revised 2003 Budget - released in August 2003 - stands out as a landmark document: N o t only does i t s justification document (Relat6rio de Fundamentaqgo) provides one the most candid assessments o f Angola’s public financial management woes but i t also took an important step by incorporating the estimated aggregate amounts o f quasi-fiscal expenditures undertaken by the BNA and Sonangol for the remainder o f 2003. Added to the original 2003 OGE were US$ 257 m i l l i o n for goods and services paid by Sonangol on behalf o f the Treasury, and US$ 144 million. The Revised 2003 OGE also included additional US$ 34 million for fuel price subsidies36 for the rest o f 2003. Preliminary information received from Sonangol for up to October 2003 indicate, however, that those numbers were substantially underestimated, especially regarding fuel price subsidies37. 3.24 However, the Revised 2003 OGE represents the first major attempt by the Angolan Government to regularize its extra-budgetary and quasi-fiscal spending, and make them an integral part o f the budget process. An important caveat i s that even if the aggregate amounts for expenditures paid by Sonangol are included in the State Budget, their execution s t i l l takes place outside o f the SIGFE, the formal budgetary system. As such, they are s t i l l part o f the “non-conventional” mechanisms o f budget execution to be discussed in more detail in Chapters 4 and 7. Recommendations f o r Further Reform 3.25 The Government should strive to increase the coverage o f the OGE. The incorporation in the Revised 2003 Budget o f estimated aggregate amounts o f extra- budgetary spending by Sonangol was clearly an important first step38, but i t does not solve the problem o f OGE coverage. The mostly needed next step i s the establishment o f a regular process whereby those transactions become routinely incorporated in the OGE (even if their actual execution still, for a time, bypasses the SIGFE). First, although there exists information o n the actual transactions carried out by Sonangol on behalf o f the Government (see Chapter 7), i t i s recommended that an independent validation 36 Refer to Araujo et al(2003). 37 When the Revised 2003 OGE was released, there was a general expectation - that did not materialize - that fuel price subsidies would be gradually eliminated over the course o f the second h a l f o f 2003. 38 As well as the f i r s t stage in the process o f “ring-fencing’’ the Sonangol-MINFIN relations, to be discussed in Chapter 7. - 30 - exercise be carried out on such information, under the leadership o f DNO, during the process o f preparation o f the OGE. This exercise could benefit from the financial audit o f Sonangol’s 2003 accounts, which should cover those activities as well. Second, formal mechanisms need to be developed that allow D N O to systematically assess and validate such information on a regular basis, until the phasing-out o f such transactions i s completed. Currently, DNO’s capacity for that task i s quite limited. Therefore, a tailor- made capacity-strengthening process o f its staff i s recommended. D. BUDGETING OF CAPITAL EXPENDITUREd9 3.26 The budget i s de facto compiled by two different ministries - MINFIN for recurrent expenditures and MINPLAN for capital spending. Integration o f current and investment spending is not complete. For all practical purposes, Angola i s a typical case of “dual budgeting”, defined as “a dual process o f budget preparation, whereby the responsibility for preparing the investment or development budget i s assigned to an entity different from the entity that prepares the current budget”40.The “dualism” i s replicated at the line ministry: while recurrent expenditures are the responsibility o f a sector ministry’s general secretariat, capital expenditures are planned and monitored by their departments o f planning and studies. 3.27 The Public Investment Program (PIP) is expected to be an annual budgeting exercise that distinguishes projects by sector and province, and which projects are on- going and new. According to the budget preparation guide1ines4l, UOs must present their investment budget proposal to MINPLAN not later than August 22 o f any given year for the next year. In tum, MINPLAN should present a Public Investment Program (PIP) proposal to MINFIN not later than September 15 to be incorporated in the OGE for the subsequent fiscal exercise4*. The PIP i s s t i l l a small component o f the budget and i s estimated to reach 833 m i l l i o n in 2003 or about 17% o f total budget expenditures. I t does not yet provide any information o n projects that have ended in a particular year, nor o n investment amounts o f on-going projects that will be transferred to subsequent years. In addition, there are no mechanisms t o record long-term investment contracts at the moment o f their signature. 3.28 A System o f Public Investment Management and Programming (SIGIP, in Portuguese) was established, with the following main objectives: (i) formulation and implementation o f a public investment management and programming system, (ii) articulation o f the PIP with the Government’s program and annual budget, (iii) articulation and coordination o f public investment at sectoral and provincial level and (iv) improvements in the technical capacity for preparation and execution o f public 39 The capital expenditure budget, in Angola, includes n o t only the Public Investment Program, but also Government’s transfers to public companies (including the so-called “operational subsidies, see chapter 4) and amortization o f financial liabilities, including external debt. 40 Schiavo-Campo and Tommasi (1999, p. 95). The OGE presentation includes both the recurrent and investment budgets, although each i s prepared by a different entity. 41 Manual de Elaboracao da Proposta Orcamental - MEPO. 42 As a matter o f exception, the 2003 PIP proposal was entirely prepared by MINFIN. 43 Angola’s budgetary measurement unit, equivalent t o US$ 1. -31 - investment projects. The SIGIP i s also expected to contain detailed information about the projects included in the PIP (the OGE only describes the project title, i t s classification code, the agency in charge - either a line ministry or a provincial government - and, the amount allocated to i t within the budget year). Project details such as objectives, amount, beneficiaries, and expected results can be found in project registration forms (fichas de cadastro), which are completed manually by the UOs and fumished to MINPLAN. 3.29 The SIGIP, however, i s s t i l l confined to MINPLAN and not used in practice by the spending agencies. Moreover, an interface between the SIGIP and the SIGFE has not been established, which complicates further the integration between the capital and recurrent budgets. 3.30 In the absence o f a multi-year planning framework44, the PIP does not show investment projects over a timeframe longer than the budgeted year. In practice, MINPLAN i s responsible for PIP programming as well as monitoring o f i t s “physical” execution, whereas MINFIN i s in charge o f financial programming for the PIP and o f monitoring o f its financial execution monitoring. Ongoing Reforms and Actions Contemplated in the PMFP 3.31 The PFMP explicitly acknowledges the need to address the dual budgeting issue. In its budgetary management chapter, and under the general goal o f “increasing the comprehensiveness o f budgetary control”, the PMFP includes the following policy measure: “to review and improve the budgetary process for the Public Investment Program and redefine the interface with MINPLAN”. Although not accompanied by a more specific plan o f action , this measure suggests that the Government i s concerned about the dual budgeting issue in Angola. The SIGIP i s expected to start being a real management tool in the 2004 budget cycle. Furthermore, investments will be classified in accordance with the new public chart o f accounts, including projects’ sources o f financing. However, they will not include the details o f the economic classification because o f i t s complexity and the authorities’ concern that the UOs m a y not have the capacity to prepare the budget o n time. MINPLAN i s also undertaking a process o f cleaning-up the PIP’S investment portfolio. As a result, new investments in 2004 will be limited to the execution o f ongoing projects and to the financing o f feasibility studies for new projects. The final objective o f the Government i s to make a complete inventory o f existing commitments with a view to eliminate payments for capital expenditures that were not followed in the past and that added to the arrears problem during the course o f the year. In the subsequent years, new projects will be prepared in accordance with the new Development Planning Law. Recommendations for Further Reform 3.32 The August 2003 FAD mission, supported by the Bank, concluded that the Government should converge to a situation in which a single entity be responsible for the preparation o f the entire budget, including the process o f negotiation with the sectoral agencies that undertake capital spending. In this context, DNO should be responsible for the preparation o f the OGE as a whole, including the PIP. MINPLA”s role would need 44 This will be discussed in detail in Chapter 8. - 32 - to be restructured accordingly, possibly with the adoption o f a narrower focus, mostly restricted t o medium- and long-term planning. Another possibility could be the eventual merger o f MINPLAN and MINFIN. The Government should, therefore, strive to find a sustainable, long-term institutional solution for the dual budgeting issue. 3.33 I t i s important to stress that this i s a not an issue o f having the capital and recurrent budgets prepared under the same roof; more importantly, the two processes need to be truly integrated. This i s why simply merging the entities in charge o f the State budget i s not a panacea. As noted by Schiavo-Campo and Tommasi (op. cit. , p. 96), “(. ..) the real issue o f dual budgeting i s the lack o f integration between investment and current expenditure programming, and not the formally separate processes in themselves (. . .) to misspecify the issue would lead (and often has) to ‘solving’ the problem by a simple merger o f t w o ministries - even while coordination remains just as weak.” 3.34 From a sequencing point o f view, before a longer-term solution for the dual- budgeting issue i s chosen and adopted, attention should be paid to strengthening the investment budgeting process itself. N o t only should the integration between MINPLAN and MINFIN in the budgetary process increase, but also the monitoring o f the execution of the PIP (see Chapter 4) needs to be considerably improved. At a minimum, in the short term, increased collaboration between MINPLAN and MINFIN in the elaboration o f the PIP should take place, and particular attention should be placed on the establishment o f an interface between SIGIP and the SIGFE. 3.35 Furthermore, the design o f the PIP should follow a multi-year approach, one that i s solidly anchored in the PRSP (refer to discussion in Chapter 8). E. THESTATUS OF BUDGET CLASSIFICATION ISSUES U N I T S AND BUDGET The Multiplicity o f Budget Units 3.36 Starting with the 2001 OGE, a number o f dependent agencies (6rg5os dependentes), especially in the education and health sector, acquired the status o f budget units, or UOs. In total, the number o f UOs increased from 156 in 2000 to 114871 in 2001. As a result, central and provincial hospitals, as well as secondary schools and tertiary education institutions, until then dependent o n the Ministries o f Health and Education, respectively, for their funding, became eligible to receive budgetary resources directly from the Ministry o f Finance45.. The main goal o f this change was to allow greater efficiency in the transfer o f resources from the Treasury to those entities, and to increase control over their financial reporting46.. 45 As noted in UNDP/IOM/UNICEFIWHO (2002), “a direct consequence o f this evolution o f the budget management mechanisms was the decline o f the role o f sectoral ministries in the decisions regarding the distribution o f resources within the sectors” (p. 59). The same report also ably notes that t h i s budgetary disempowerment o f the M i n i s t r i e s o f Education and Health was preceded by a more general administrative deconcentration reform process - in w h i c h only the Ministries o f Interior, Justice, and Finance were unaffected - whereby the provincial governments were given greater responsibility for the conduct o f sectoral policies, at the expense o f the line ministries. 46 See M INF N (2001). I - 33 - 3.37 However, as noted in the FAD mission report, this measure unduly diminished the coordinating role o f the line ministries over the implementation o f their o w n sectoral strategies and policies, and overburdened MINFIN with tasks for which i t had no comparative advantage. This has been recently acknowledged by the Government, leading to a reversal o f that change: the Guidelines for 2004 Budget Preparation (Manual de Elaboracao da Proposta Orcamental - MEO) essentially restored the pre-2001 situation and “demoted” a number o f UOs to their previous status as “6rgBos dependentes”. Budget Classification Issues 3.38 According to the Organic Budget Law, Government revenues should be classified by economic category (current and capital revenues) and sources o f funds. Government expenditures should be classified by administrative responsibility (UOs), by economic category (use o f expenditure) and by function (purpose o f expenditures). However, two main budget classification issues have faced the OGE preparation: (i) differences in classification methods as between key guiding documents; and ( ii)the alignment o f Angola’s fiscal statistics with the methodology o f the IMF’s Government Financial Statistics Manual - (GFSM 2001). 3.39 These discrepancies seem to have been solved, for the Government has now presented to the National Assembly the 2004 budget according to the ME0 classification This i s a noteworthy progress, as until 2003 and despite M I N F I N ’ s efforts to conform the State budget classification to these norms, there were different interpretations o f h o w accounts should be classified. The main discrepancy was between the economic classification o f revenues and expenditures according to the new Chart o f Accounts (Plano de Contas do Estado - PCE) and the MEO. - 34 - 4. BUDGET EXECUTION The importance o f a fully-functioning budget execution system for overall public expenditure management should not be underestimated. From the point o f view o f aggregate fiscal discipline, budget execution information should tell policymakers and analysts alike whether fiscal deficit targets have been fulJilled, that is, outturns are within the resource envelope established in the preparation stage. From the standpoint o f public expenditure policy, the budget execution process w i l l determine whether expenditure priorities were implemented as planned. I n the case o f Angola, a key issue is that the process o f expenditure execution is not circumscribed to the formal system; weaknesses in the latter reinforce the problem. This chapter assesses the main issues involved in the process o f budget execution in Angola. A FOR BUDGET THEINSTITUTIONAL FRAMEWORK EXECUTION 4.1 At the central level, budget execution i s responsibility o f the Treasury Directorate (Direcq5o Nacional do Tesouro, DNT), while at the provincial level i t i s responsibility o f the Provincial Directorate o f Finance (DelegaqBes Provinciais de Finanqas, DPF). The Treasury Directorate i s responsible for the policy and financial programming o f budget execution as well as for the public debt management policy. 4.2 A Treasury Single Account (Conta Unica do Tesouro, or CUT) i s hosted in the Central Bank; treasury operations are carried out by the Government’s fiscal agent, the Banco de Poupanqa e Credit0 (BPC), a publicly-owned commercial bank (see below). The C U T i s sub-divided into two sub-accounts: C U T - M N (moeda nacional), for operations in local currency; and C U T - M E (moeda estrangeira), for operations in foreign currency . 47 4.3 Although the Treasury i s legally responsible for the entire budget execution process, the national o i l company Sonangol carries out Treasury-like operations such as external debt payments o n behalf o f the Government (see Chapter 7). 4.4 Additional clarity on the institutional relationship between Treasury and the Central Bank - including the establishment o f criteria for foreign exchange treasury operations with impact o n the Treasury Single Account - has been provided by the signing o f a Protocol between the Treasury and BNA o n September 19, 2002 (see B o x 4.1). 47 O i l signature bonuses have been incorporated into budget and centralized at CUT-ME; however, they are still being captured through Sonangol and will be financing the Public Investment Program. Oil exploration bonuses are up-front payments f r o m o i l companies t o the Government for exploration rights. In the past bonuses were identified as income in the budget but their use was n o t normally registered in the fiscal accounts. - 35 - B o x 4.1 The Protocol Between the Central B a n k and the Ministry o f Finance The Protocol covers the main areas o f interaction between BNA and MINFIN, and includes: (i) rules for movements in the Treasury Single Account (CUT), both in local currency (CUT-MN) and in foreign currency (CUT-ME); (ii) the remuneration o f Central Bank for services provided and for Central Bank Bills (TBCs); ( iii rules for liquidation in local currency o f external debt payments backed by oil; (iv) rules ) concerning the financial flows resulting f r o m foreign exchange receipts and tax payments; (iv) the resolution o f pending issues o n B N A ' s quasi-fiscal debt; and (v) basic principles for public external debt management. The approved version o f the Protocol does not address the so-called "surrender requirement" o n Sonangol for o i l export earnings, established by Decree 30195 and the basic pillar o f the "petroleum account", according to w h i c h Sonangol would sell to BNA the total proceeds from o i l exports. Article I,item 1 o f the Protocol addresses the issue o f the consolidation o f the CUT: "MINFIN will keep at BNA the Treasury Single Account (CUT), which will collect all State revenues, either paid directly, or paid through the banking system.". An addendum t o the Protocol, accompanied by a Memorandum o f Understanding between MINFIN and BNA, was issued in August 2003, offering greater operational detail as to the workings o f the C U T and operationalizing a number o f fiscal measures established by decree in February 2003. T h e Stages o f the Budget Execution Process 4.5 The stages in the expenditure process are regulated by Article 26 o f the Organic Budget L a w and consists o f three standard steps4" a cabimentaqlio (commitment), o r the certification that a sufficient budget allocation i s available to place contracts and orders, by using the Nota de Cabimenta@o4'; a Ziquidaqlio (verification), or the confirmation that the creditor i s entitled to receive payment for the services provided and goods delivered according to the contract that gave rise to the transaction; this leads to the issuing o f a Nota de Saque", and a the actual payment by the designated bank (BPC). B. LAW COMPLIANCE WITH THE BUDGET 4.6 Several issues, among which two are noteworthy, undermine the designed system that can otherwise be a fairly reliable system. Firstly, unavailability o f funds for payment o f confirmed commitments discredit,the commitment system and the certificates. As a 48 Angolan Ministry o f Finance (2001). Seminirio sobre N o v o Sistema de Execucgo Orcamental e financeira do OGE. Luanda. 49 Nota de CabimentapTo - document that identifies the nature o f the expenditure (according t o the budget classification) and the amount to be paid. 50 Nota de Saque - payment order issued by l i n e ministrylspending agency, which the B P C w i l l use t o pay the creditor. - 36 - result o f unavailability o f funds for payment5’, arrears are created which in tum have further negative impact such as escalation o f prices for the government. Secondly, the lack o f trust in the Nota de CabimentaCGo has produced an undesirable practice which consist o f change in the sequence o f the three steps established by the organic law; the payment order i s issued prior to the delivery and the verification that should take place under the Ziquida@io. This practice and the absence o f complementary controls to ensure delivery o f goods under such procedures i s a potential financial risk that can be exploited by some and must stop immediately. 4.7 In addition, the verification stage o f budget execution has been almost completely ignored, and payments have been made without physical verification and certification that goods have been delivered and services rendered. That explicitly violates the Organic Budget Law52. With the implementation o f the updated version o f the SIGFE from January 1, 2004, measures were introduced to strengthen enforcement o f the verification stage. They include the creation o f a Nota de Liqiiidaqiio that has to be issued by the UO and signed by its manager before a Nota de Saque can be issued. I t i s s t i l l too early to evaluate the impact of those measures on the actual budget execution practices. c. CASH PLANNINGAND MANAGEMENT 4.8 From a legal standpoint, budget execution (expenditures) should be made through “budget credits” (appropriations divided between initial credits and additional credits) effected o n the basis o f one-twelfth (duodkimos) o f the respective budgetary appropriations. Additional budgetary credits are allowed o n the basis o f changes that may have occurred after the approval o f the Annual Budget Law. The current system i s as follows. At the beginning o f the year, the Ministry o f Finance holds back 20% o f the total o f the budget, as a “financial reserve” (cativaqiio), for adjustments during the annual execution. The remaining 80% are distributed to spending agencies according to the one- twelfth rule. Spending agencies might request access to the remaining 20% as long as well justified and o n a punctual basis. Expenditure provisions can be transferred from one line item to another (virement) - except salaries and investments - and are allowed on condition that such adjustments do not exceed the authorized budget ceiling (upon request by line ministriedspending agencies and after approval by DNO). During the execution, i t i s the Budget Directorate’s task to monitor and control the public investments in collaboration with provincial and sectoral levels. 4.9 In practice, the Treasury Directorate prepares a Monthly Cash Plan (MCP) with a view to match financing capacity and financial resource needs. The M C P has to be approved by the Financial Programming Commission (CPF)53,54, in charge o f the 5’ The causes for unavailability o f hnds are most l i k e l y a combination o f unrealistic budget estimates for both expenditures and revenues, inadequate planning, and lack o f firm control over extra-budgetary as w e l l as payments for transactions outside the formal system 52 Refer to Tollini et al. (2003) for a more detailed discussion. 53 Comissgo de ProgramapTo Financeira - composed by the Minister o f Finance, the Minister o f Planning, the Governor o f the Central Bank and the Minister o f Labor and Social Security. The Protocol between BNA and MINFIN sets up a joint team to oversee the M o n t h l y Cash Plan execution. 54 Actually, the Comissiio also approves a quarterly financial programming plan (which has t o take into account the “request for financial funds”, Necessidades de recursosfinanceiros, prepared by line ministries - 37 - financial programming, and by the Standing Commission o f the Council o f Ministers. In practical terms, although entitled to one-twelfth (duodecimos), spending agencies receive what i s approved in the M o n t h l y Cash Plan o n the basis o f i t s level o f execution in the previous month. In short, they receive a monthly Quota Financeira (lump sum), which m a y or not correspond to a one-twelfth (Box 4.2 discusses the practical workings o f the system). The o n l y exceptions are the Ministries o f Health and Education, which are entitled to receive the entire one-twelfth. L i n e ministriedspending agencies are not allowed to carry over any remaining balance o f Quota Financeira to the following month, i.e., it should b e entirely spent within the month. Otherwise, remaining funds revert back t o the Treasury. Box 4.2. The Duoddcimo System in Practice The one-twelfth, or duoddcimo, principle, i s used mostly as a reference point rather than as an effective financial programming tool. I f concretely applied, the monthly Quota Financeira (QF) for a budget unit i would be exactly one-twelfth o f the annual OGE allocation (a) for that unit, as follows: The number 0.8 indicates that MINFIN i s entitled t o retain 20% o f the total allocation. DNT s t i l l uses the above formula to prevent UOs f r o m spending i t s yearly OGE allocation before the end o f the fiscal year. If revenues were perfectly predictable and entered the Treasury in a smooth fashion, financial programming should mirror the monthly duodecimos. That not being the case, the duoddcimo works as an upper limit for budget execution. According to the Treasury Operations Department o f DNT, the actual formula takes into account the amounts already executed in previous months (e) and the number o f months already passed (m): - eit) 12-m(ait QF,, = x 0.8 f=l 12-m in a quarterly basis). L i n e ministries are then informed by MINFIN o f the amount they are entitled to receive for the period followed by the reception o f the monthly quotafinanceira. - 38 - I Box 4.3. Budget Execution from the Standpoint of Spending Agencies Typically, line ministries/spending agencies receive quotas financeiras in response t o their quarterly I “requests for funds”. Although these quotas are expected to be allocated to the spending agencies o n a monthly basis, as the Monthly Cash Plan i s implemented, the process quite often does n o t follow a regular periodicity. The quotas normally do not correspond to the one-twelfth appropriation. When expenditure i s committed (Nota de CabimentapTo, NC), a payment order (Ordem de Saque, 05‘)i s immediately issued and sent to B P C f o r payment (reportedly, payments m a y take some days t o be processed at the bank, causing losses due to exchange-rate fluctuations; capital expenditures are particularly affected by such fluctuations). In an attempt to avoid accumulation o f arrears, line m i n i s t r i e s make an effort to commit expenditures only after receiving the monthly quota financeira (reportedly, there i s n o carryover o f quota Jinanceira balances f r o m month t o month). Changes in line budget appropriations (virement) are permitted upon request to DNO, as long as the total resource envelope i s respected. F o r public companies, quotas are processed through the Treasury’s Department in charge o f overseeing State-Owned Enterprises Dept (CPED); CPED remits the payment order to BPC, which then credits the company’s account in a commercial bank. Once a year, public companies are expected t o send a report o n their accounts t o both DNC and CPED. Expenditure Processing Arrangements 4.10 Due to i t s extensive temtorial coverage (46 branches connected online, in 14 out o f the 18 province^)^^'^^, BPC has been the State’s fiscal agent (banco operador) since October 2001. The way fiscal expenditure processing takes place in Angola can be succinctly described as follows. AAer the Monthly Cash Plan i s approved, funds are transferred from the C U T to a mirror account, or “sub-CUT” at BPC through a transfer order (Ordem de Transferhcia) issued by the Treasury. Since the Treasury monitors the sub-CUT daily operations in real time - a computer software was commissioned by MINFIN and installed at BPC to allow for real-time link between the Treasury and the B P C - transfer of hnds from C U T to sub-CUT occurs only when cash management liquidity needs occur. BPC reports daily o n the balance o f sub-CUT to BNA and MINFIN; since March 2003 BPC transfers to the C U T the entire revenue collected the previous day. 4.1 1 L i n e ministries/spending agencies send payment orders (Ordens de Saque) directly to BPC for execution; withdrawals from their accounts at BPC, which are related to the sub-CUT as transit accounts, are made only upon presentation o f such payment orders. The fact that line ministries/spending agencies are not yet linked online with the 55 In the remaining 4 Provinces (Lunda Norte, Lunda Sul, M o x i c o and Kwanza Norte) budgetary payments are s t i l l being done by another public bank, Banco Comircio e Zndzistria (BCI) (which has been performing this role after many years). 56 Arrangements o f this nature are n o t uncommon. See The W o r l d B a n k (1998): “A common variation i s that the retail banking operations are delegated t o a fiscal agent (normally an authorized commercial bank) by the Central Bank. This model is put in place in those cases where the Central B a n k does not have an adequate network of provincialiregional branches o r does n o t have the capacity to handle the large volume o f transactions that are associated with government payments and receipts. In these cases, the fiscal agent makes payment o n behalf o f the Treasury, the Central B a n k recoups a l l payments made by the agent for government operations and the fiscal agent makes daily deposits o f a l l government revenues to the T S A [Treasury Single Account] in the Central Bank” (pp. 169-171). - 39 - Treasury, D N C and BPC constitutes a bottleneck in the system (requiring that information flows take place via diskettes and print outs). Consolidation of the Treasury Single Account (CUT) 4.12 The mechanism o f the CUT, however, i s far from being consolidated. The main issues impairing i t s consolidation are related to some o f the operations carried out by Sonangol, and to the service o f debt based on oil-backed loans. 4.13 In principle, profit oil, Sonangol's taxes, the Kwanza-equivalent the dollar-value o f taxes paid by the foreign o i l companies, and all other taxes transit through the CUT- MN. However, since the C U T records only actual cash transactions, taxes and profit o i l from Sonangol are often not recorded, due to the compensation mechanisms described in Chapter 7 . This clearly contradicts Article I,item 1 o f the Protocol (see B o x 4.1). 4.14 As for the debt service operations, o i l signature bonuses and proceeds from extemal loans, they should all transit through the C U T - M E with payments being made through the C U T - M E by the Central Bank. However, for contractual reasons, external debt service of oil-backed loans is still made through dedicated escrow accounts, thus bypassing the CUT-ME. Oil-backed debt service does not go through the C U T as i t does not involve an actual cash transaction undertaken by the Treasury. 4.15 In summary, in addition to i t s complexity, this cash management system i s not achieving i t s intended purpose and does not allow adequate planning by OUs nor prevents over commitments. Most probably due to the complexity and difficulties in information flow, the estimates do not reach the destination before the beginning o f the period. The complying OUs wait for the plans before commitment and therefore can't plan ahead o f time and therefore are penalized. Those who are not compliant undermine the system and create arrears. In all circumstances the existing system does not allow the recording o f long t e r m commitments and contracts beyond the expenditure cycle are only recorded for the payable amounts and that only when the payment order i s to be issued. The unexecuted amount o f the contract i s not recorded anywhere and the amounts are not known. D. REVENUES MANAGEMENT 4.16 Although the revenue side o f the budget was not included in the scope o f the exercise, a brief analysis o f the overall revenue management i s presented in this Section. The fiduciary framework for o i l revenue management are discussed in Chapter 7 . 4.17 The principle o f consolidated fund, single revenue account, collection and management o f govemment assets and management and payment o f liabilities through the formal budget system has been formally accepted and imbedded in the legislation. The MINFIN i s making efforts and has made strides in achieving some noticeable results in cash management and rationalization o f the government banking operations. 4.18 As regards the management o f the state revenue, several observations are worth mentioning: 0 Firstly, assessment o f the government revenue, the first step in generation, collection, and, more generally, the preparation o f a meaningful budget, i s - 40 - inadequate and the practice seems somewhat ad-hoc. Due to lack o f capacity, DNI relies heavily on self-assessments and others’ data banks and systems and tries t o cope by checking the big numbers and big tax payers. Secondly, i t i s acknowledged that the collection o f revenue by the banks and transfer o f collected revenue to the single account i s not efficient nor transparent. Once the funds reach the single account, DNT seems to have good control (on line access) over them. 0 Finally, the PEMFAR team did not obtain any evidence, or was not told about control mechanisms and reconciliation o f revenue data with deposits made. However, the practice o f reconciliation o f the deposits data provided by the banks and the credits to the government revenue account seems satisfactory. The upstream accounts o f the revenue (the banks) are not audited and no reconciliation i s done to confirm that all collected revenues end up in the government revenue account. N o t all financing i s recorded and loans are contractedguaranteed by the Government without being recorded. 4.19 In summary, fragmentation o f the banking operations (on the collection side) and financial operators has increased the complexity o f the system. Controls need to be rationalized and streamlined, in line with the implementation o f the updated version o f the SIGFE (see Chapter 6). E. PUBLIC EXPENDITURE MANAGEMENT 4.20 In addition to the difficulties that the execution o f an unrealistic budget represents, at that stage, the PEM process i s also confronted with three problems related to weak intemal controls within the expenditure cycle. At present the budget execution i s decentralized and each spending entity (UO) has the authority to execute i t s budget, commit, and issue payment orders. As much as a very positive and welcome development for expediting the budget execution and making the UOs responsible vis-a-vis their budget, this arrangement, poses challenges for intemal controls. 4.21 Angolan P F M system is, at the origin, based o n a reference model which relies heavily on independent ex-ante controls, thus by design ex-post controls are weak. The Angolan departure from that reference model5’ has taken place without the necessary accompanying measures. The elimination o f ex-ante control should be accompanied by the strengthening o f the ex-post controls (by INF, for instance). The commitment i s therefore not verified independently from the UOs before hand, thus posing a risk that the UOs might not respect the quarterly Financial Plans and over commit the government (sometimes done outside the system58). j7The so-called Continental European model (Belgian, French, Portuguese, etc. ) based on which the Angolan model was established. Under the reference model a financial controller, representing the MINFIN, verifies and approves a l l commitments, without which the government i s not committed. 58 . in particular contracts with more than 12 month execution time cannot be fully entered in the system. Only the protion executable within the FY i s committed and the remainder i s n o t tracked in the SIGFE. -41 - 4.22 In addition, the D N C i s dependent o n financial information received from UOs which would be prone to delays, errors, and omission o f information such as the commitments created outside the system. Any control exercised by the D N C can’t be considered effective and independent and accounts and reports produced by i t are reflection o f information received from UOs. 4.23 The flaws observed in the budget execution l i e in the way the budget i s executed in practice and not in the Organic Law. These deviations in execution are probably the rational reaction to two issues (i) lack o f ownership o f the budget by the UOs, and (ii) the cash rationing and very short term and unpredictable management through this rationing. 4.24 The budget i s unrealistic and therefore not easy to comply with in the first place; the cash rationing and management from month to month (sometimes information becomes available well into the month) makes planning impractical for the UOs. Because o f this, until recently, some UOs found i t legitimate to commit above and beyond the authorized commitment ceilings and quarterly plans, and sometimes even above the approved annual budget. The first results o f the Government’s PMFP, however, have made such practices more difficult. 4.25 Shortage o f cash results in a selective payment o f invoices and lack o f transparency as well as creation o f arrears. Worse yet, this undermines the normal sequence o f actions crucial to ensure that the government has received goods and services for payments made. In fact it has been widely acknowledged that the N C B does not reassure the suppliers o f quick payment. Due to this lack o f trust, with consent o f officials, the suppliers wait for payment before delivering the goods and service. This change o f sequence undermines the adequate control o f delivery which under this new arrangement i s to take place after payment (documents are filed after payment and no follow-up action takes place). Anecdotal examples indicate that in some instances the goods are never delivered59. This i s further aggravated by loose public procurement processes, as discussed in detail in the Country Procurement Assessment Report and reported in Annex 2. 4.26 An important part o f the payroll i s that o f the Army, Police and Interior. These are not included in the MINFIN system and their payment i s made only through lump-sum amounts. As for the others, payment o f salaries are made by provinces and there are indications (i.e. unpaid teachers) that the funds are sometimes used for other purposes. 4.27 As payroll expenditures are not well controlled, a census and audit needs to be done followed by the creation o f a uniquelsingle database for the government payroll. I t i s possible to device mechanisms (communicating /interfacing databases) to protect the confidential information without jeopardizing accountability while at the same time i t i s also necessary to develop controls to deal with the reliability o f the non-Defense payrolls. 59A variation of this practice i s what the UOs introduce commitment and immediately after the payment order in order to speed the payment to the suppliers. - 42 - F. AND PRICE SUBSIDY ARREARS PAYMENT ARREARS 4.28 Expenditure control has been based on cash rationing rather than on refraining from placing orders for which funds may not be available. Line ministriedspending agencies, which are the originators o f Nota de Cabimentqa"o, are in theory not supposed to issue them for values above the duodecimo, however several possible practices explain the arrears. First, the Quota Financeira i s usually below the amount o f the duodecimo. Information from line ministries indicate that this i s particularly the case with capital spending, as contracts (often associated with c i v i l works) span several months and sometimes beyond the financial year. Second, the limit i s ignored, the commitment step i s delayed or ignored, or, some commitments are not included in the SIGFE. 4.29 All three steps o f expenditure (Commitment, Liquidation and issuance o f Payment Order, and Payment) are supposed to take place at the right time and with the right sequence and commitments made within the framework o f the budget are expected to be recorded in the SIGFE. However, large unexplained or unrecorded expenditures imply that many transactions occur outside the realms o f the formal rules. Legislation o n sanctions against managers o f public entities who have incurred non-authorized expenditures is not available, Le. there are no practical sanctions that MINFIN can apply to discipline l i n e ministries. Therefore, expenditure controls are not binding at the commitment stage and expenditure approvals are not always backed by cash availability, leading to payment arrears. 4.30 This has also discredited the Nota de CabimentaCa"o and has created a sense o f mistrust by suppliers who would rather deliver the goods and services after payment. Therefore, in order to move to commitment controls, the Government will need to overcome the currently l o w credibility o f i t s Nota de Cabimentaqa"o, which i s likely to have been generated by a history o f arrears. This can be done by automation o f the issuance o f this note by the system and requires careful financial planning and payment o f all o f them. 4.31 On the other hand, budget execution numbers for 2002 show that subsidy arrears are s t i l l very high in Angola6'. In that year, budgeted price subsidies totaled some US$300 million dollars, o f which US$220 million (equivalent to 2 percent o f GDP) were in arrears by the end o f the year. Most o f these price subsidy arrears (89.8 percent) were due to Sonangol Holding (through Sonangol Distribuidora) for the provision o f petroleum products at below-market prices. Substantial arrears have also been accumulated in respect o f the electricity companies (EDEL61and ENE) and the municipal water supply company in Luanda (EPAL). Chart 1 below shows the composition o f price subsidy arrears in 2002. 6o See Araujo et a1 (2003) for a detailed discussion. [Final budget execution numbers for 2003 were not available at the time o f the writing]. 61 EDEL stands for Empresa de Distribuiya'o de Electricidade and ENE for Empresa Nacional de Electricidade. - 43 - Chart 4.1: Price Subsidy Arrears, Company by Company (kz). Fiscal Year 2002 0 2000 4000 6000 8000 10000 12000 4.32 The fact that the State incurs arrears to the service-providing companies helps weaken their financial position and triggers the accumulation o f inter-enterprise arrears in the energy sector (see Figure 4.1 for an illustration). Indeed, i t i s quite often that EDEL cannot honor i t s debt to ENE for the electricity i t receives, and ENE cannot pay Sonangol for the fuels required for electricity generation. As o f June 2002, cross-arrears between EDEL/ENE and ENE/Sonangol were more than K z 1 bn, or US$24 million dollars. Figure 4.1: Cross Arrears: A Chain Reaction EDEL Arrears with’ ENE ’ Arrears with ’ SonanPJl 4.33 In addition, as will be noted in Chapter 7 and in Annex 4, the fact that the Treasury has a history o f subsidy arrears towards Sonangol i s part o f the reason why the mechanism o f the “Petroleum Account” broke down in late 2000, leading to the existing system o f tax and profit oil offsets. - 44 - G. EXECUTION OF THE PUBLIC INVESTMENT PROGRAM 4.34 A s noted in Chapter 3, there are two dimensions to the execution o f the Public Investment Program in Angola: the “financial” dimension, carried out by the Ministry o f Finance, and the “physical” one, carried out by line ministries such as the Ministry o f Public Works, but monitored by the Ministry o f Planning. While the financial dimensional refers to actual payments made, the physical dimension relates to the actual rate of implementation o f the programmed investment projects within the PIP. 4.35 Data o n PIP execution are not usually timely or particularly reliable, especially given the fact that there i s no interface established as yet between the SIGIP and the SIGFE (see Chapter 3). The main source o f information for this section i s MINPLAN’s “2002 Public Investment Program Execution Report”62. Financial execution information i s disaggregated as follows: (i) budgeted amounts; (ii) amounts contemplated in the financial programming; and (iii) amounts actually released to investment projects. The total amount budgeted for the PIP in the Revised 2002 OGE was approximately K z 14 billion (US$ 170 million). Financial programming exceeded the budgeted PIP by about 14%, while actual releases fell short o f programming by around 6%. Chart 4.2 below shows PIP execution numbers for 2002, further disaggregated between centrally-executed and province-executed projects. The bulk o f the difference between budgeting and programming stems from the portion o f the PIP executed by the provincial governments. According to MINPLAN’s report, this difference i s mostly accounted for by unexpected rehabilitation-related expenditures triggered by the peace process, which started in earnest after April 2002. Chart 4.2: Execution o f the 2002 PIP: The “Financial” Dimension (Kz million) 18000 16000 1 14000 12000 10000 0 Budget 8000 W Programming 6000 4000 2000 0 PIP Central PIP Provinces Total Source: M I N P L A N 4.36 While by law the execution o f capital expenditures should follow the same three- stage process (commitment, verification, and payment) o f recurrent spending, in practice the phenomenon o f advance payments i s particularly pronounced with respect to the PIP. MINPLAN’s report alludes to two possible explanations: (i) the lack o f confidence o n the 62 M I N P L A N (2003). “Relatbrio de ExecuG5o do Programa de Investimento PLiblico 2002”. Luanda: March. Final execution data for 2003 were n o t available at the time o f writing. - 45 - Nota de CabimentaqLTo on the part o f private suppliers, as discussed in section A, favoring the practice o f large, upfront payments before the commencement o f works; and (ii)the large concentration o f the PIP in infrastructure projects, in which i t i s common practice to mobilize the material means only after the payment o f the first tranche o f a works contract i s effected. 4.37 Although i t i s difficult to disentangle the dominant reason for this phenomenon, it i s clear that there i s a very large disconnect between financial execution and the estimated value o f what was indeed physically im~lemented~~. MINPLAN estimates that the value o f physical execution represented only 19% o f total actual releases. 4.38 The results o f the 2002 PIP execution process suggest that the non-observance o f the appropriate sequence o f the budget execution process - as established in the Organic Budget L a w - leads to sub-optimal outcomes in terms o f achieving the objectives o f the PIP. The fact that the PIP does not follow a multi-year exercise exacerbates the problem, as i t i s unclear whether there i s any “catching-up” over time between physical and financial execution. I t also indicates that there are serious fiduciary risks associated with the execution o f the capital budget, and that they should be addressed in the process o f designing multi-year investment programs (see Chapter 8 for further discussion). H. DEBT ISSUES MANAGEMENT 4.39 Angola’s debt management process, in practice, i s scattered among three agencies64: 0 The recently-established Debt Unit (Gabinete da Divida) at BNA, which i s in charge o f compiling and managing external debt statistics. This function is contemplated in Article 17 o f the Organic L a w o f the Central Bank. 0 The Treasury’s External Debt Department, which i s more directly involved in the borrowing process itself, as i t is involved in loan commitments, repayments and renegotiations. 0 The national o i l company Sonangol, which acts as an underwriter and provider o f collateral for oil-guaranteed loans6’. Sonangol is also a main source o f information o n the amortization o f such loans. 4.40 To facilitate coordination among these entities, the Council o f Ministers passed o n March 2, 2001 a decree (Decree 7/01) establishing a Public Debt Management Committee (Comite“ de Gestiio da Divida Pziblica, or CGDP). The CDGP reports to the Council o f Ministers and i s composed by the Minister o f Finance (Head), the Minister o f Planning, and the Governor o f BNA. The operational work o f the Committee i s carried out by an Executive Commission, headed by the National Treasury Director, and encompassing representatives from MINPLAN, BNA, and the Ministry o f Foreign Affairs. At a technical level, the work o f the Commission i s supported by BNA’s Debt 63 The valuation methodology employed by M I N P L A N for such estimates i s not discussed in MINPLAN’s report. 64 Aide Memoire o f j o i n t I M F - W o r l d Bank mission t o Angola on external debt management issues, December 2003. 65 See Chapter 7 for more details o n the role o f Sonangol. - 46 - Unit and MINFIN’s External Debt Department (EDD). I t i s interesting to note that Sonangol, which carries much o f the burden o f loan negotiations and contracting, i s not formally represented at any level. 4.41 The Executive Commission i s viewed as a transitional arrangement, until a new Public Debt Framework L a w i s approved by the National Assembly. This law would provide for a new institutional set-up for debt management in Angola66. A joint I M F - World Bank external debt mission visited Angola in December 2003, and proposed the creation o f a Public Debt Management Operations Unit (PDMOU), assisted by an international specialist. The P D M O U would accumulate information collection and forecasting functions, and would also monitor debt-service payments. The P D M O U would not necessarily become a formal “unit” as such, but would rather operate as a working group, comprised o f representatives from BNA, DNT, Sonangol, and BPC (as the latter i s involved at the tail end o f non-oil backed loan repayments). A technical expert from the Bolivian Central Bank, identified by the IMF, is providing technical assistance to the Angolan Government in the implementation o f the recommendations o f the December 2003 mission. The proposal for the new institutional setup i s s t i l l under study by MINFJN. 4.42 With support from an IDA credit, an UNCTAD-designed debt management information system, the DMFAS, was installed at BNA. However, in practice, B N A ’ s Debt Unit compiles external debt information in two databases, DMFAS itself and an in- house debt database (the “old system”). Different sets o f numbers are produced by each o f those systems. The Fund-Bank mission recommended the phasing-out o f the “old system” by June 2004 and the utilization o f D M F A S as the sole debt management information system. The D M F A S interface with MJNFIN s t i l l needs to be established as well. 66 The Bank-supported EMTA credit has also earmarked funds to assist in this process. - 47 - 5. BUDGET ACCOUNTING AND REPORTING “At the heart o f any fiscal and financial information system is the accounting system, which maintains the basic records o f government transactions and, thus, outturn expenditures ’’67. I n Angola, this function is performed by the National Directorate o f Accounting (DNC), a key component o f the Sistema Integrado de Gestio Financeira do Estado (SIGFE), which is, in turn, the backbone o f Angola’s reporting system. A major process of reform is underway in this area, as part o f the Public Finance Modernization Program (PFMP). This chapter describes Angola’s public accounting system, assesses progresses achieved under the PFMP, and identifies main areas for further reform. A. PUBLIC ACCOUNTINGSYSTEM ANGOLA’S 5.1 The public accounting system in Angola i s still cash-based and single-entry, but a transition towards a “double entry” accounting system (mdodo das partidas dobradas) has begun with the roll-out o f the new version o f the SIGFE (see below) that ties up the whole accounting stage o f the budget cycle. All accounting is centralized at the National Accounting Directorate (DirecpTo Nacional da Contabilidade, DNC), but as mentioned before, UOs are obligated to maintain their o w n budgetary accounts and submit monthly budget execution repots to the DNC. 5.2 Two new documents, a public accountability manual and a new chart o f accounts, were prepared in 2001 and were implemented early in 2004 and have facilitated the introduction o f an accounting system based o n double entries. In addition, there i s a plan to produce an Accounts Glossary (Funqio das Contas, Le., the explanation o f the accounts). Every other week spending agencies send to the Accounting Directorate data on the expenditure commitment stage (cabimentaGio or the so called ‘tfloating” debt) that occurred in the two previous weeks. In principle, this recording practice should facilitate the implementation o f an accrual accounting system and improve the quality and reliability o f the reports. 5.3 The full implementation o f the SIGFE and activation o f all o f its automatic controls should be completed as soon as possible. With a robust system in place the functions o f staff need to be changed. The w o r k flow and organization will require a thorough review and reorganization after the full implementation o f the new SIGFE. Such review and redefinition should include independent controls over data entry, classification, and adjustments to the accounts entries. The Government’s Financial and Fiscal Reports 5.4 The system o f information collection and report generation by the D N C i s totally dependent on the information provided by the UOs, mostly through statistical reports and 67 Potter and Diamond (1999). - 48 - vulnerable to error or manipulation at the source. Given the number o f UOs (487 for the 2003 fiscal year), regular and relatively timely production o f consolidated reports i s understandably a cumbersome and difficult task (never achieved so far). The full installation and use o f the maximum capacity o f the SIGFE will certainly improve the controls and automate the production o f reports. It i s crucial though that the internal controls are strengthened on data manipulation and reporting. The SIGFE i s a powerful tool in terms o f report production and, provided the quality and timeliness o f data improves, it would be able to provide reliable outputs. However, currently there are discrepancies between the budget classification and the chart o f accounts used in the SIGFE. Furthermore, payment orders are being issued without commitments already produced for those expenses. There are also partial commitments produced leaving portions o f legally created commitments outside the system and without any monitoring and tracking. These issues, and others need to be addressed in parallel. 5.5 The reconciliation between the Treasury information (banking) and information from the UOs does not regularly take place; thus some financial reports m a y be based on unreliable infomation. In addition, the reconciliation between the administrative and fiscal accounts does not take place routinely and end-year adjustments lack transparency, un-reconciled items remain in the books or arbitrarily allocated. In general, there i s an excessive reliance on the banking system information rather than o n the government’s own accounts. Considering these and the shortcomings in reconciliation o f revenues and expenditures, described earlier, the financial reports are unreliable. However, o n the positive side, the linkage o f the treasury to the bank accounts allows a real time monitoring o f balances. 5.6 The Accounting Directorate produces budget execution reports. I t i s also supposed to submits quarterly accounts to the National Assembly and the final annual Government accounts, Conta Geral do Estado are due to the National Assembly by March 3 1 - only 3 months following the end o f the fiscal year. In addition, i t analyses the accounts o f the Autonomous Funds (the accounts o f State-owned enterprises are responsibility o f the Public Enterprises Department at the Treasury). However, non o f the above takes place with the regularity needed to generate a steady f l o w o f budget execution data to the Ministry o f Finance and other stakeholders. 5.7 For all o f the above reasons, the financial reports are not complete and do not reflect all major government financial transactions. There i s a relatively high risk that the financial reports are not reliable and do not sincerely reflect the financial transactions. These r i s k s will be reduced by the full implementation o f SIGFE but also with strengthening o f internal and independent controls over classification and production o f reports. Non-Financial Assets a n d Liabilities 5.8 Non-financial assets and liabilities management i s under the responsibility o f the National Directorate for the Management o f Non-Financial Assets o f the State (DirecqGo Nacional do Patrimonio do Estado, DNPE). A key concern in this area is to ensure that quality information on the non-financial assets o f the State i s generated in order to increase coverage and accuracy o f the General State Accounts (Conta Geral do Estado, CGE). Indeed, lack o f timely and accurate data o n public debt and o n State physical - 49 - assets inventory prevents the production o f reliable assets and liabilities accounting” (contabilidade patrimonial). The key inputs from DNPE to the CGE are information on the stock o f State non-financial assets (balanqo patrimonial) as well as o n vearly changes in that stock (balanqo das variaq6espatrimoniais). 5.9 Although the DNPE i s supposed to maintain an inventory o f all government non- financial assets at both central and provincial levels, i t s records, maintained mostly by hand (especially in the provinces), are incomplete and out o f date. A proposal to computerize the records and introduce a more sophisticated accounting system (including asset depreciation), which will ultimately be integrated with SIGFE, is being implemented6*. B. MAIN ISSUES PUBLIC ACCOUNTING SYSTEM IDENTIFIED I N ANGOLA’S 5.10 A s in the case o f budget preparation and execution, important flaws encountered in the public accounting system do not l i e in the Organic Budget L a w itself, but rather in the compliance with it: 0 Contrary to Article 41, which stipulates that all budgetary operations have to be recorded according to the .double entry system, public accounting i s s t i l l cash- based, single entry. The single-entry system only records services, supplies and works paid. Consequently, as the state commitments are not being recorded it i s difficult for the D N O to exactly determine whether forthcoming revenues would cover total expenditures needs. 0 Likewise, Article 49 stipulates that fixed assets should be recorded. Although the DNPE is supposed to maintain an inventory o f all government non-financial assets at both central and provincial levels, its records, are for the time being incomplete and out o f date. 0 According to Articles 53 to 57, closed accounts (Conta Geral do Estado) must include all government bodies, services, autonomous funds and the Social security. They should be presented in statements showing budgeted and realized revenues and expenditures, the balance o f financial assets, the balance o f non- financial assets and the changes occurred in non-financial assets. In addition a balance sheet o f the State should be prepared, clearly showing government short- t e r m and long-term assets and liabilities. The present accounting information system does not allow the closing o f accounts as stipulated in the Organic Law. However, one should recognize that the production o f such statements are out o f reach o f countries at the development level as Angola. 0 Compensatory and inter-agencies transactions are not recorded when they are not realized in cash. This creates a serious gap in the presentation o f the country fiscal accounts. I t i s particularly serious when the non-cash compensatory transactions are important, as in the case o f state subsidies to Sonangol being compensated, in principle, by Sonangol debt payment o n behalf o f the State. 68 Support from the EMTA credit w i l l be provided to this activity. - 50 - e The system does not permit to record simultaneously the debit and credit operations when the proceeds o f an existing loan are used to pay directly a supplier. This complicates tremendously the analysis o f the status o f foreign debt and double counting often creates confusion in the accounts. e The new accounting manual and chart o f accounts, originally scheduled for introduction in 2002, have not been adopted or implemented, even though the two documents are available. One o f the structural problems stems from two different charts o f accounts proposed by MINFIN: the structure o f Plano de contas do Estado (PCE) consists in 6 algorisms, while the instructions for 2004 budget preparation (Manual de Elaborapio de Proposta Orqamental - MEO) request 8 a~gorisms~~. 5.11 The Government i s well-aware o f the severe weaknesses affecting i t s accounting system. Box 5.1 enumerates the main problems as identified by the National Accounting Directorate (DNC) itself: Box 5.1. Main Weaknesses in Angola’s Public Accounting System, as seen by DNC 0 Non-centralization of key information in DNC Excessive use o f DNCS staff time for typing and incorporating in the accounting system information “manua1ly”provided by the UOs, especially in the provinces. Inexistence of non-financial asset accounting (contabilidade patrimoniaQ.Inexistence of a formal recording process for the verijkation stage (liqiiidapio) o f budget execution. 0 Delays in the generation o f information by the UOs. 0 Insuflcient knowledge o f key legal and policy-related information. 0 Insuflcient institutional and human-resource capacity. C. INTEGRATED ANGOLA’S INFORMATION MANAGEMENT FINANCIAL SYSTEM 5.12 The Angolan I F M I S (Sistema Integrado de Gest6o Financeira, SIGFE, in Portuguese) has been officially in place since 1999 (Decree 13/99) and has undergone an upgrade during the second h a l f o f 2001, using bilateral assistance (Decree73/01). The model was adopted in 1992 and developed under a technical assistance project jointly executed by the Government, IMF and UNDP. The system was originally based on three modules: budgeting, treasury, and accounting, and the 2001 review focused o n the integration of the three modules. A new process o f upgrading is underway, with significant technical and coverage improvements which has already resulted in the incorporation o f a new module covering non-financial asset accounting (see Chart 5.1). 69 T h i s issue i s in the process o f b e i n g solved as part o f the SIGFE r e f o r m s underway. - 51 - Chart 5.1: The SIGFE at a Glance Modules of the SIGFE: Non-Financial Accounting System Entities Involved: v 1. Treasury Directorate 2. Provincial Directorates of Finance 3. The Central Bank 4. Savings and Credit Bank (BPC) 5. Planning and Studies Departments 6. Administration and Budget Departments Sub-Svstems of the SIGFE: Tax Collection Public Debt Tax Administration Procurement Non-Financial Asset Accounting Human Resources Social Security Strategic Information SIGFE Outcomes: 9 Better control of budgetary, financial, and “patrimonial” execution; 9 Establish the recording of the verification stage of budget execution; 9 Accelerate financial programming; 9 Adoption of double-entry accounting; 9 Use of public accounting as a reliable and timely source of information; 9 Standardize working methods and routines; 9 Link all national territory and budget units abroad; 9 Operation of the system in real time; 9 Improve control over external and internal debt; 9 Consolidate the process of preparation of the quarterly budget execution - 52 - 5.13 As previously pointed out, BPC became the banco pagador for the State in 2001. The reporting mechanism underwent several adjustments. I n addition, the Government accelerated the implementation o f a full on-line system between the Ministry o f Finance (Treasury, Tax and Accounting Directorates), the BPC and the line ministriedspending agencies. A real-time link between the Treasury and BPC i s almost fully operational, which allows the Treasury to regularly check i t s available balance in the “sub-CUT” (Portuguese acronym for the Treasury account in BPC) that mirrors the Treasury Single Account (CUT) in BNA. The next step i s to move as soon as feasible to a full online system linking the C U T and the sub-CUT, and therefore, the Treasury, BNA, and BPC. With the introduction o f a commercial bank, the Treasury Directorate expects to increase efficiency by avoiding delays in the payments operations. 5.14 SIGFE’s legislation clearly defines the entities involved in the system, the rules, procedures and the instruments o f financial and budgetary execution, and the norms o f accounting, control and reporting o f state accounts. In addition, it establishes that financial programming i s the main instrument for public resource management, under the supervision o f the Financial Programming Commission. Quarterly financial programming should be approved by the Standing Commission o f the Council o f Ministers. Currently, the following entities are part o f the SIGFE. 5.15 The Treasury Directorate (DNZJ is the backbone o f the SIGFE in charge o f supervising, ruling and coordinating the system and is also responsible for the management and control o f the Conta Unica do Tesouro, CUT, hosted at the Central Bank, as well as o f the sub-CUT at BPC (the CUT’S mirror account). Other entities with responsibilities in the SIGFE are: 0 the Provincial Directorates o f Finance (Delegaq6o Provincial de Finanqas, DPF) o f each province, responsible for provincial-level entries into the system; 0 the Central Bank (Banco Nacional de Angola, BNA), the banker o f the State and host o f the Conta Unica do Tesouro, CUZ the Banco de Poupanqa e Cre‘dito, BPC, the State-owned commercial bank in charge o f the Government budgetary expenditure and collection o f revenues, the fiscal agent, through the sub-CUT nourished by the Conta Unica do Tesouro; 0 the Planning and Studies Departments (GEP) at the line ministriedspending agencies, either at the central or local level, in charge o f identifying the budgetary needs and responsible for its distribution among i t s subordinate agencies; 0 the Administration and Budget Departments (DNAGO) o f line ministriedspending agencies, in charge o f budgetary and financial execution. - 53 - D. THERECENT PUSH FOR REFORMS 5.16 The ongoing process o f reforms o n the SIGFE and the accounting system - carried out in the context o f the Public Finance Modernization Program (PMFP), see Chapter 9 - benefited from the advice o f the August 2003 FAD mission, supported by the PEMFAR team7’. However, the main motivation came from MINFIN itself, and D N C has played a key role in spearheading those reforrhs. 5.17 The ongoing reform process - which has a major bearing on D N C - includes the following objectives: (i) to control budgetary, financial, and “patrimonial” execution; (ii) to formally establish the recording o f the verification stage o f budget execution; (iii) to accelerate financial programming; (iv) to adopt double-entry accounting; (v) to allow the use o f public accounting as a reliable and timely source o f information; (vi) to standardize working methods and routines; (vii) to link all national territory and budget units abroad as part o f the SIGFE; (viii) to allow the system to operate in real time; (ix) to improve control over external and internal debt; and (x) to consolidate the process o f preparation (and submission to the Tribunal o f Accounts and the National Assembly) the quarterly budget execution reports as well as the CGE. 5.18 During 2004, the SIGFE reform process will focus mostly o n the introduction o f non-financial asset accounting (contabilidade patrimonial); the integration o f budget execution with financial and “patrimonial” execution; strengthening of financial programming; training o f staff o n the use o f the new version o f the SIGFE software; and the establishment o f a centralized database. 5.19 In order to be h l l y responsive to this new environment, during 2004 D N C will undergo a process o f restructuring o f i t s organic functions as well as o f its human resources profile. A key component o f this process will be a capacity building effort, which will include training o n operational public accounting; accounting for decision- making; and adapting and streamlining the administrative flows and processes in the Government. 5.20 Part o f the reform process includes the establishment o f an interface between the SIGFE and the various “subsystems”, which include: (i) tax subsystem; ( ii)salary bill subsystem; (iii) non-financial asset subsystem; (iv) debt management subsystem; and (v) procurement subsystem. World Bank support through the EMTA credit has been sought in the process o f establishing most o f these subsystems. E. AREAS MAIN FOR FURTHER REFORM 5.21 A series o f additional measures should give the authorities the necessary tools to manage the State accounts in a more efficient and transparent way. These measures include: 0 At present, D N T financial programming does not take into account information available from the SIGFE and rather makes i t s financial projections from its o w n information sources. Under these circumstances no control o f budget execution can really be made. This situation should be corrected. ’ O See Tollini et al. (2003). - 54 - 0 Revise the SIGFE so that i t does record all UOs’ verification (liquida@o) o f services rendered or equipment or materials received. 0 Introduction o f the accounting manual and the new chart o f accounts before fiscal year 2004 (January 2004). The conflict between the two charts o f accounts (DNO and PCE) should be solved and the 2004 budget should be presented to the National Assembly based on one chart o f account only. 0 All State transactions would be made according to accrual accounting, including those concerning the State budget, all financial transactions and all fixed assets records. 0 All P C E transactions should be recorded with specific assignment to the accounts to be credited or debited. 0 Non-financial assets should be recorded. 0 Public enterprises accounting and reporting should be made part o f a regular assessment process. - 55 - 6. AUDIT AND CONTROL SYSTEMS Despite the fact that Angola is one o f the only countries in the Southern African Development Community that has a Supreme Audit Institution, neither the internal nor the external functions are performed satisfactorily in the country. Recent literature on public expenditure management has emphasized the importance o f adequate institutional arrangements for the quality o f fiscal outcomes. I n particular, cross-country analysis suggests that less transparency is associated with higher fiscal disequilibria7’. Ministry- level control and audit agencies are a key element in ensuring transparency and enforcing accountability in fiscal management. And so are Supreme Audit Institutions (SAIs) with respect to external controls and audits. This chapter identiJies Angola’s main weaknesses in this area and proposes a reform approach to strengthen the country’s public resource management processes. A. INTERNAL CONTROL, ACCOUNTABILITY, AND FIDUCIARY FINANCIAL RISK 6.1 In Angola, the intemal audit function i s mostly carried out by the National Inspectorate o f Finance (INF). In turn, the external audit function i s the responsibility o f the recently established Tribunal o f Accounts. The Tribunal i s the Supreme Audit Institution (SAI) whose sole function i s to audit State accounts, pronounce o n them, and impose fines. 6.2 Both the internal and external audit functions in Angola have a potentially very important role in the implementation o f the reform strategy recommended in the PEMFAR. Currently, however, i t can be said that neither function i s performing satisfactorily in Angola. Despite i t s long history, INF i s too weak to exert any real impact in terms o f enforcing compliance with the Organic Budget Law72. At the same time, the Tribunal o f Accounts i s s t i l l very incipient and its strength and independence have yet to be put to test. 6.3 I t i s important to note that, both in the Organic Budget L a w and in the PMFP, the terms “internal control” and “internal audit” are used interchangeably or, in many instances, the expression “intemal control” is used when it is actually meant “intemal audit”. As pointed out in Havens (1999), internal (or management) controls refer to policies and procedures to ensure the proper functioning o f entities to which they apply, while internal audit refers to the mandate to report to the top leadership o f an organization o n findings related to the functioning o f the management control systems in place. Within that definition, “intemal audit i s part o f an organization’s management control structure” (Havens, op. cit., p. 219). I NF i s more concerned with internal audit 71 See The World Bank (1998). 72 Cf. the mission report for the IMF’s Fiscal Affairs Department mission t o Angola in August 2003 (Tollini e t al, 2003). - 56 - proper, although i t i s mandated to perform some non-audit internal control activities as 6.4 Therefore, i t can be said that internal control i s a collective term describing the whole system o f internal checks and balances - including authorization limits, reviews, counterchecks, procurement regulations, and internal audit - put in place by an organization or government in an effort to protect i t s assets and income from loss, and protect the integrity and accuracy o f i t s records74. 6.5 For all these reasons, the internal control system constitutes a major link in the financial accountability chain, and i t s quality an important indicator o f the l e v e l o f fiduciary r i s k in a country.. Furthermore, the executive, not only should execute the budget as desired, i t has the responsibility to provide timely and accurate reports to the parliament o n the execution o f the budget. Therefore, intemal controls should expand to cover accounting and reporting. 6.6 Institutionally, internal controls are at two levels. First, within various budget units (UOs) and maintained by the accounting officer responsible for the execution o f and reporting on the budget for each UO. Second, at the country level, maintained by the Minister o f Finance as the responsible for the overall PFM (budget execution and reporting). I t is also the responsibility o f all parties mentioned above to maintain adequate controls to ensure the safeguard and maintenance o f state’s assets and patrimony. Finally, i t i s the responsibility o f the MINFIN (DNT) to ensure safeguard and adequate management o f government funds and their use for purposes intended and in the public interest. 6.7 This requires that (i) the rules and regulations are clear and written directives are issued and procedures manuals are available to staff, (ii) training i s provided so that the staff i s knowledgeable and well informed o f those rules, and (iii) control are in place to deter and detect irregular practices. Expenditure Controls 6.8 The system o f expenditure control in Angola consists o f three main components: (i) a priori controls on expenditures by UOs at “Commitment”, “liquidation” and “Ordonancement”, (ii) mechanical controls exercised (or to be in place in near future) by SIGFE computerized system, (iii) a-posteriori control available t o the MINFIN through INF. 6.9 Regarding (i) - these a-priori controls are in theory at commitment as well as prior to the issuance o f payment orders (Liquidations step). Different services apparently follow the quarterly financial plans sent to the DNT and the system o f one twelfth to 73 Other elements o f the overall intemal control system at the level o f the M i n i s t r y o f Finance include the accounting controls exerted by the National Accounting Directorate and SIGFE’s automated controls over the commitment and payment stages o f the expenditure cycle. 74 Such a system reduces opportunities for fraud, and reduces the possibility o f error. The emphasis i s o n risk elimination or minimization in a l l significant areas. A common characteristic o f a good intemal control system i s that nobody i s able to initiate, authorize, execute, and record a transaction without the intervention o f another person. Such a system makes frauds and mistakes difficult, because they require collusion by at least two persons. - 57 - contain commitments at the level o f budget credit. There are also separation o f duties and approval processes in place to ensure that the expenditures are for the goods and services required. However, unfortunately, procedures manuals are not available to staff nor i s there a functioning internal audit within UOs to provide assurance to the accounting officer that the formal rules are being followed. There seems to be relatively high risk that informal rules and exceptional measures are frequently used. In particular the following areas require attention and action for strengthening o f the internal controls: 0 Payroll - it seems that the decentralized treatment o f the payroll and lack o f robust linkage (unique database) among the UOs, MINFIN, and MAPESS has made the system vulnerable to manipulation and error, resulting in undue payments. 0 Classification o f expenditures - there i s anecdotal examples which indicate a relatively high risk that the expenditures can be classified inaccurately not reflecting the real nature o f the expenditure. 0 Payments for undelivered goods - it is widely acknowledged that payment orders are being issued before the delivery o f related goods (refer to discussion in Chapter 4 on the non-observance o f the verification stage o f the budget execution process). This breakdown o f the internal control sheds doubt about the delivery o f goods even after payment and i s a high risk area for consideration and strengthening. The reason given for such practice is the distrust o f suppliers and the reputation o f the government as bad payer (this i s inconsistent with the existence o f arrears, as suppliers arrears exist). 0 Commitment control (also see Chapter 4 , o n payment arrears) - given the fact that the commitment i s done by a large number o f UOs it i s crucial that a centralized system controls and manages the commitments as to ensure the commitments within the approved budget (provided the budget i s realistic). I t i s unclear at this stage whether the SIGFE can reject unplanned commitments. But, as mentioned before, indications are that this automatic control i s not taking place. 6.10 Regarding (ii), the mechanical controls by computerization under the SIGFE are a positive step towards commitment and payment control which, once fully operational, will help the UOs and MINFIN to manage the resources better and dispose information very timely. These mechanical controls include verification o f budget availability and issuance o f unique commitment numbers and commitment certificates, and issuance o f payment orders only for issued commitment certificates. This system, if well managed, can provide the assurance to all parties that the government commitments will be paid and to change the practice o f premature issuance o f payment orders, thus improving controls on delivery o f goods. 6.11 An unresolved issue i s the data against which the commitments are approved. The question o f what data i s downloaded into the SIGFE for budgetary control needs to be carefully reviewed and answered. Such data needs to be reliable and have direct relevance to the reality and availability o f cash for payment o f government commitments, once they become due. At present there are different explanations for commitment - 58 - approval Le. (i)the approved budget and the system o f 1/12, (ii) quarterly Financial Programs prepared and submitted by the UOs, (iii) quarterly Financial Programs as prepared by the DNT and approved by the two committees, or (iv) the monthly cash projections. 6.12 Related to this issue are the contracts for large works and goods going beyond the duodecimo or the fiscal year. As the system may not allow commitment within the system, those commitments take place outside the SIGFE and i t maybe that payment for those contracts drain the treasury and result i s arrears on the commitments duly executed. This issue needs careful study and resolution. A change in budget (to a multi-year budget) and modification o f the system i s necessary to stop commitments outside the system and to control commitments within the cash availability rather than financial plans made by the UOs. B. INTERNAL AUDIT: OVERVIEW Legal and InstitutionalFramework 6.13 The primary function o f a governmental internal audit department i s to verify whether government transactions and activities are complying with the law, as w e l l as with the internal control checks and balances, and the procurement regulations. Internal audit in the Angolan Government is the responsibility o f the National Inspectorate o f Finance (Inspecqgo Nacional de Finanqas, INF). INF i s the technical support agency within MINFIN in charge o f carrying out financial control activities o f all budget units across the public administration as w e l l as o f the State-owned enterprises. I n collaboration with the National Tax Directorate (DNI), INF can also conduct financial control activities (mostly restricted to tax issues) o f private companies. 6.14 INF was created in 1958, twenty-seven years before Independence, and to a great extent i t s work program still reflects legislation passed during colonial times. According to the Organic Budget L a w (Law 9/97 - Lei Quadro do OGE), the internal control and audit functions are to be “institutionalized” within MINFIN, requiring complementary regulation through a Council o f Ministers Decree, which was never passed75. The k e y legal document defining the responsibilities o f I N F i s MINFIN’s o w n Charter (Estatuto Organico, Council o f Ministers’ Decree 4/98). On the basis o f Decree 4/98, INF has initiated work on the preparation o f draft regulations for itself, which represents the first major effort in conducting a legal and institutional reform o f MINFIN’s internal audit agency since i t s creation in 1958. Auditing Standards 6.15 INF does not have an audit manual, does not publish an annual report o n i t s activities, and practices audit standards adapted from those prevailing in Portugal. A 75 The internal control and audit functions in Angola are not clearly defined. In principle, the m a i n internal control and audit agency in Angola should b e the General Inspectorate o f State Administration (IGAE, in Portuguese). IGAE was established by L a w 9/92, but the accompanying regulations have never being prepared. IGAE should, in principle, coordinate a l l previously established intemal control agencies, notably INF and the inspectorates established in the level o f the sectoral m i n i s t r i e s . L a w 2/92, however, allows some flexibility for the autonomous functioning o f INF. - 59 - review o f i t s audit questionnaire shows that, out o f 41 questions, 25 (61%) deal with timeliness o f accounting and reporting; one deals with timeliness and correctness o f deposits; 3 deal with ensuring that expenditures are within budgetary limits; only 2 deal with checking that expenditures are properly authorized; one deals with legality o f salaries and benefits; one with the legality o f budget management commission; one with ensuring that reconciliation’s are prepared in a timely manner; one with ensuring confirmation o f balances; one with the legality o f the reporting format; one with prevention o f waste; one with proper inventory; one with orderly filing; and two with correct bookkeeping and accounting. Although the audit program covers most o f the traditional areas o f internal control and accounting, there i s clearly heavy emphasis on timeliness o f reporting, and legality o f various activities, rather than on ensuring that built-in automatic intemal checks and balances are operating, as they should. 6.16 In practice, only few controls are carried out annually by INF. This i s mostly due n addition, the Internal Regulations and to i t s limited human and material resources76. I Terms o f Reference for INF, that were prepared in 1958 are not adapted to the modem intemal control role INF should play. Reporting and Follow-up Mechanisms 6.17 The usual reporting procedure i s as follows: Audit reports, once prepared, are submitted to the National Inspector, who then transmits them to the Minister o f Finance. The Minister, in tum, sends a copy o f the report to: (i) the line Minister for the area being to the National Directorate o f Accounting (DNC); and (iii) audited; (ii) to the Tribunal of Accounts, when there are clear problems o f public financial management. 6.18 According to rNF, the follow-up role lies with DNC; in practice, however, there are no concrete follow-up mechanisms o n the cases brought to light by the audits, once the audit reports are issued. The absence o f regular follow-up procedures further undermines the credibility o f the internal control and audit functions in Angola, as i t facilitates the creation o f an environment o f impunity for wrongdoers. c. REFORMING THE INTERNAL CONTROL AND AUDIT FUNCTIONS 6.19 The severe weaknesses identified in the performance o f the internal control and auditing functions in Angola essentially imply that the quality o f public expenditure i s subject to a permanent risk. I n fact, INF i s insufficiently capacitated and in need o f a business plan and a risk management strategy. In addition, I N F ’ s findings and recommendations are not acted upon. Finally, even the automated controls through SIGFE do not offer much respite, since they : (i) are not utilized at their full potential; and (ii)are constrained by the system’s limited coverage. 6.20 A s noted, the required regulations for the internal control function in Angola have not been passed, despite the fact that the Organic Budget L a w was approved in 1997. 76 At its Luanda headquarters, INF i s organized in three departments, dealing respectively with: the public administration (i.e. the ministries); public enterprises; and private enterprises. INF also has branch offices in the provinces. I t has a total o f 60 staff members, o f who about h a l f are in headquarters, and the other h a l f are in the provinces. O f the 60 auditors, only six have university degrees at the time o f the P E M F A R writing. - 60 - Therefore, there i s s t i l l a great deal o f uncertainty as to the institutional framework within which I N F i s supposed to operate. This uncertainty needs to be removed as soon as possible, if INF i s to properly perform the tasks mandated to i t by the existing legislation. Ongoing Reforms and Actions Contemplatedin the PMFP 6.21 The PMFP objective with respect to intemal control77 i s to “restructure the Control System with a view to improving the process o f evaluation o f public resource management”. To meet this objective, the PMFP aims to: 0 Develop a concept o f a systemic model for the internal control function. 0 Adjust the functioning o f INF to the new model. 6.22 Concretely, the PMFP proposes a sequence o f reform actions, starting with an assessment o f altemative internal control models. Once a model i s chosen, intemal control guidelines - covering administrative and financial aspects - would be established. The PMFP also refers to possible changes in legislation, although i t does not offer details as to what such changes should be. Recommendationsfor Further Reform 6.23 The actions identified in the PMFP should be translated into a clear implementation action plan. They are conceptually correct; however, they are presented in fairly general terms which do not lend themselves easily to operationalization. Ultimately, intemal control and audit reform will hinge upon the choice o f the systemic model referred to above. 6.24 The Bank-supported E M T A credit has earmarked resources to assist the internal control/audit component o f the PMFP. M o r e specifically, EMTA will support: (i) the design of an appropriate institutional framework for INF, thus assisting in the development o f the “systemic model” envisaged by the PMFP; (ii) the technical work underpinning the revision o f the relevant legislation; (iii) the preparation o f manuals o f guidelines; (iv) the design o f curriculum o f the career o f INF inspector as well as o f a training program for INF’s technical staff. 6.25 Other major areas in need o f deeper reform - beyond what is currently contemplated in the PMFP - include: 0 Intemal audit work should be enhanced and be treated as a high-priority matter, both at the level o f INF and the UOs, to improve the quality o f expenditures and the reliability o f the reports generated by the system. 0 Awareness should be raised within the c i v i l service as a whole o f the importance o f the INF, and its credibility should be enhanced through concrete examples o f successful follow-up actions. 0 Rules o f ethics for public sector staff should be established and training geared towards changing individual behavior should be implemented. 77 “Intemal control” i s often used instead o f “intemal audit” in the Government’s documents. -61 - 6.26 The ultimate test o f INF’s strength and relevance would be i t s participation in exerting preventive controls and conducting regularity audits - in coordination with the Tribunal o f Accounts - on the compensation mechanisms between MINFIN and Sonangol, and investigating other potential sources o f leakages from the formal system o f expenditure execution. T o perform at that level, however, INF’s capacity should be enhanced and the Inspectorate’s terms o f reference should be clarified, as proposed above, particularly vis-a-vis other agencies such as the IGAE. D. AUDIT FUNCTION: THE EXTERNAL OVERVIEW Legal and InstitutionalFramework 6.27 The Tribunal o f Accounts (Tribunal de Contas) was institutionalized by the Constitution o f the Republic o f Angola78, I t i s regulated by i t s Organic L a w 5/96 o f April 12, 1996 and i s organized by Decree 23/01 o f April 12, 2001. Decree 24/01 o f April 12, 2001 establishes fees for the services rendered by the Tribunal. The Tribunal o f Accounts i s also a member o f the International Organization o f Supreme Audit Institutions (INTOSAI), the African Organization o f Supreme Audit Institutions (AFROSAI), and the Southern African Development Community Organization o f Supreme Audit Institutions (SADCOSAI). The membership to these three organizations guarantees the Tribunal access to I N T O S A I auditing standards, which are acceptable to the World Bank for audits o f government accounts. 6.28 The Tribunal o f Accounts i s an extemal body, with its de jure independence assured by law, performing the following main functions: (i) controlling and auditing the Government accounts and financial statements, particularly the General State Accounts (Conta Geral do Estado, CGE); (ii) undertaking ex ante control (fiscalizag8o preventiva) o f contracts above $350,000 and (iii)controlling a posteriori (fiscalizaqtio sucessiva), i.e. regularity auditing, of spending agencies. According to i t s Organic L a w (Articles 58 and 60), the Tribunal’s report and final position o n the General State Account must be presented to the National Assembly and published in the official gazette (Diario da Repfiblica). 6.29 The President and the other six judges o f the Tribunal o f Accounts are appointed by the President o f the Republic. They are selected through a competitive process, based o n their qualifications and experience, by a jury consisting o f the [President o f the Supreme Court], a member o f the Superior Council o f Judicial Magistrates, and a Professor from the L a w Faculty o f the Agostinho Net0 University. 6.30 As part of the judicial system, the Tribunal essentially follows the model o f the Supreme Audit Institutions in Italy, France, Portugal, Spain, and most South American countries, in contrast with the Anglosaxon system, where the Auditor General model prevails. The closest SA1 to Angola’s in Southern Africa (in terms o f objective and structure) i s the Tribunal Administrativo in Mozambique. However, the latter deals with a broad range o f administrative and tax issues, in addition to the State accounts. 78 Law 23/92 o f September 16, 1992, No. 3, Article 125 - 62 - 6.31 The Tribunal o f Accounts has only become effective recently, as i t s first President was appointed on April 4, 2001. I t s de facto independence from the executive has yet to be tested. If confirmed by actual practice, such independence would allow the Tribunal to play a major role in the implementation o f the two-pronged strategy for public financial reform proposed by the P E M F A R (see Chapter 2 as well as Section D below). 6.32 According to its Organic Law, the Tribunal i s empowered to control and conduct any analysis, review and supervision o f all personnel, services and entities that are part o f the central and local administration, all institutes and associations, public enterprises, other enterprises where the State has majority o f capital, as w e l l as other entities, private or public, that execute financial transactions involving State resources. In principle, the Tribunal is, therefore, entitled to control the various existing UOs. Auditing Standards and Coverage 6.33 The Tribunal i s required by law to express opinions o n the following: a State financial activities in the areas o f assets and liabilities, receipts and expenditures, cash flow, and public debt; a Compliance with the Organic Budget L a w and i t s supporting legislation; 0 Direct and indirect State responsibilities, including guarantees; a Inventory o f State assets; a Subventions, subsidies, loans, and other forms o f State aid; and a Implementation o f action programs, investments, and financing o f State-owned enterprises, including employment and application o f subsidies to autonomous funds. 6.34 In addition, the Tribunal’s program o f ex ante controls i s designed to: (i)confirm the legality and regularity o f receipts and payments; and (ii)ensure that contracts offer the most advantageous terms when they are signed. Organization and Staffing 6.35 In addition to the Office o f the President and his staff, the Tribunal is jurisdictionally divided into two chambers, dealing respectively with ex ante controls and ex post audits. In terms o f detailed operations, i t i s divided into t w o Directorates, namely: Directorate o f Technical Services (DTS), which carries out all the audits, and the Directorate o f Administrative Services (DAS) responsible for finance, human resource management, transport, public relations, documentation, and information technology. The DTS, headed by an Auditor-General (“Contadoria Geral”) i s further sub-dividedinto five departments, the first two o f which are responsible for ex ante controls in various jurisdictions and institutions, while the remaining departments are responsible for ex post audit in various jurisdictions. 6.36 Capacity i s an issue for the Tribunal. The [enabling Act] provides for the Tribunal to have 170 employees: 7 judges and one deputy chief prosecutor; 31 directors, division chiefs and other managerial staff; 54 professionally qualified auditors and senior staff; 15 middle level technicians, and the rest administrative and support staff. Only 70 employees have been hired by the Tribunal, primarily due to inadequate office space. O f - 63 - the 70, only six are qualified to carry out audits, and only one o f them can audit a computerized accounting system. Five more auditing staff are currently being trained in Portugal. Salaries are said to be totally inadequate to attract the right kind o f staff; no provincial office has yet been established, even though the Tribunal i s supposed to have them. Quality and Timeliness o f Reporting 6.37 The Tribunal has yet to issue i t s first audit report o n the General State Accounts (CGE). The CGE for the fiscal year ended December 31, 2002 will be the first to be audited by the Tribunal. Preparations have started for this task. Indeed, since January 2003, all UOs have been requested to submit to the Tribunal a detailed l i s t o f all their transactions, being revenues or expenditures. These l i s t s will be compared to the financial statements submitted by MINFIN. A special working group has been set up to carry out this work. Technical assistance has been provided by Portugal and INTOSAI, which w i l l furnish a software program to the Tribunal that will facilitate the task. Although i t i s too early to evaluate the Tribunal’s quality and efficiency in conducting this work, at least all necessary measures seem to have been taken to minimize the risks o f arriving at an unsatisfactory end-product. Government Follow-up to the Tribunal’s Recommendations 6.38 In Angola, legislation grants the Tribunal o f Accounts the power to impose penalties and fines on institutions and individuals for a variety o f infractions, including, for example, failure to submit financial statements in time, or to provide information required by the auditor. This i s potentially one o f the strongest advantages o f the Angola model. However, this follow-up power s t i l l needs to be tested in practice as most SAIs suffer from the weakness that their recommendations are often ignored by the government for years. 6.39 The relationship and collaboration between the Tribunal and the National Assembly’s Commission o n Economy and Finance has yet to start in earnest. The first test will come in March 2004, when the Tribunal i s due to submit i t s first report to the National Assembly. E. AREAS MAIN FOR REFORM 6.40 The Tribunal o f Accounts’ financial autonomy falls somewhat short o f the degree o f independence provided by the law. Although the Ministry o f Finance (MINFIN) normally accepts the budget proposals submitted by the Tribunal, subsequent cash rationing by the MINFIN affects the Tribunal as much as i t affects line ministries. According to the Tribunal, while its annual budget i s part o f the public sector budget, i t is a sovereign entity, as are the National Assembly and the Government. Greater predictability in budgetary receipts would clearly enhance the operational efficiency o f the Tribunal. 6.41 I t i s important to phase out the current practice o f ex-ante controls carried out by the Tribunal while internal control mechanisms are strengthened. The assignment o f ex- ante control activities to the Tribunal o f Accounts i s compatible with the Declaration o f - 64 - L i m a on the main principles o f control o f public finances (IX I N T O S A I Congress, October 1977). Section 2 o f the Declaration stresses that “ex-ante controls by a superior Tribunal presents the benefit o f avoiding prejudicial acts but that, at the same time, might create an excessive amount o f administrative burden to the supreme Tribunal, as well as creating a certain confusion in the distribution o f responsibilities according to public rights”. However, in many countries ex-ante controls by the Tribunal o f Account have been eliminated, as they might create a potential conflict o f interest. Such controls may also dilute the Tribunal’s independence, by putting it in a position o f having to audit its own work. Preventive audits are usually a function o f internal audit and control system, not o f external audit. I t i s recommended that these ex-ante controls by the Tribunal o f Accounts be phased out and eventually eliminated, to the extent that internal controls are strengthened. 6.42 Adequate staffing and appropriate compensation levels are critical to ensure the de facto independence o f the Tribunal. I t is unclear how well the Tribunal can apply auditing standards, given the limited numbers and experience o f its staff. The Tribunal has submitted to the National Assembly a proposal to increase the number o f i t s staff as well as the levels o f i t s salaries. The request i s n o w with the President o f the Republic. 6.43 The time allowed for submission o f the CGE to the Tribunal should be reduced to six months. According to Article 58 o f the Organic Budget Law, the Government must submit the CGE for any given year to the Tribunal o f Accounts up to December 3 1 o f the following year. This lengthy interval raises questions about the relevance and usefulness o f the CGE. In particular, this timetable prevents the National Assembly from fully appreciating the budget execution performance for a given year and using it as an input into the approval process for the OGE for the next year. This hampers the learning process and makes it difficult to incorporate lessons from the previous fiscal exercise into the design o f the immediately subsequent one. 6.44 The costs o f including the largest public companies in the Tribunal’s work program should be carefully assessed and incorporated into the Tribunal’s financial needs assessment, and borne by the OGE. Although the Tribunal’s responsibilities include the control of public enterprises’ financial statements, they are not part o f the ongoing and 2004 work program o f the institution. Given the economic and financial importance o f some key public enterprises in Angola, they should not be left out o f the Tribunal’s work program from 2005 onwards. Capacity issues can be temporarily addressed through outsourcing, that is, by hiring internationally-reputable audit f i r m s that could perform regularity audits in selected public enterprises o n behalf o f the Tribunal79. 6.45 In consonance with the two-pronged reform strategy proposed in the PEMFAR, the Tribunal should be empowered to conduct strategic regularity audits o n the compensation mechanisms between MINFIN and Sonangol (see Chapters 2, 4, and 7). This would be an important element o f the “ring-fencing” o f the so-called non- conventional mechanisms o f budget execution discussed elsewhere in this report. By 79 This would not exempt the Tribunal f r o m overseeing the work o f the contractor and ensuring quality control. Outsourcing could also generate capacity constrains o f a different nature: “In any event, the SA1 must equip i t s e l f with staff who are sufficiently skilled in this type o f auditing [especially financial auditing] to assess the quality of the work, even if the w o r k i s to be outsourced.” (Havens, op. cit., p. 209). - 65 - doing so, the Tribunal o f Accounts w o u l d be truly engaged in the process o f strengthening the formal mechanisms o f budget execution and phasing out the informal ones. This involvement could become the most important test o f the Tribunal’s relevance and independence. 6.46 The ultimate focus o f the Tribunal o f Accounts should l i e o n a strategy - in close coordination with INF and MINFIN more generallys0 - to strengthen management controls at the level o f the UOs and k e y public enterprises, rather than trying to detect and eliminate each and every managerial and financial irregularity in Angola’s public resource management processes. *O Part o f this coordination could be done by incorporating reform actions related to the Tribunal into the work program o f MINFIN’s PMFP, to help ensure that sufficient budgetary resources w o u l d be allocated in the future t o the Tribunal’s reform efforts. - 66 - 7. THE FIDUCIARY FRAMEWORK FOR OIL REVENUE MANAGEMENT IN ANGOLA At the center o f Angola’s fiduciary framework for o i l revenue management the special relationship between the Ministry o f Finance, the Central Bank, and the national o i l company, Sonangol. A transition towards a fully functional relationship between these three key institutions is proposed. This ‘path towards normalization” is the basis for the second p r o n g of the reform strategy proposed in the PEMFAR and should involve: (i) the establishment o f clear reporting mechanisms from Sonangol to both the Ministry o f Finance and the Central Bank on a l l transactions it carries out on behalf of the Treasury; (ii) strengthened internal control mechanisms (both within the Ministry o f Finance and Sonangol) to ensure that such transactions are conducted in accordance with legitimate fiscal policy objectives and acceptable public expenditure management practices; and (iii) improved accounting procedures, whereby the National Accounting Directorate at MINFIN obtains from Sonangol a l l the information it needs to prepare the yearly General State Accounts as well as Quarterly Budget Execution Reports. A. BACKGROUND 7.1 Oil and gas revenues are central to Angola’s economic, social and indeed political performance. O i l revenues already account for 40 % o f GDP, 80 % o f government revenues and 95 % o f foreign exchange earnings. These revenues are expected to grow rapidly as exploitation o f major deepwater offshore o i l and natural gas reserves gets underway. 7.2 While the record to date has not been encouraging, these revenues, if properly managed, have the potential to drive Angola’s economic and social recovery. Over the past few months, the Government has shown increasing signs o f commitment to establishing a fiduciary framework for the management o f petroleum revenues designed to correct past misuse or abuse and deliver the positive developmental impacts expected o f the sector. 7.3 The key agencies involved in petroleum revenue management are MINFIN; BNA; MINPET; and Sonangol In theory, the roles and responsibilities o f these agencies are as follows: 0 MINFIN: overall economic and financial policy, budgetary management, tax collection, audit and reconciliation, and oversight o f Sonangol as the Government’s representative. 0 BNA: depository for taxes and other petroleum payments, clearing house for financial flows, debt and foreign exchange management. 0 MINPET: sector policy and oversight. - 67 - 0 Sonangol: Government's representative in the sector as concessionaire, cost audits, collection o f selected payments due Government and commercial participant in the sector. 7.4 In practice, expediency considerations have eroded many o f the individual agency responsibilities and transferred them to Sonangol. Better staffed, better funded and with critical control over o i l production, Sonangol has either assumed these responsibilities at the Governments' request, or has taken them over o n i t s own initiative to address agency failings which might otherwise adversely impact Sonangol's commercial performance. 7.5 This Chapter examines three key dimensions o f the fiduciary framework for the management o f Angola's o i l revenues, namely: a) taxation; b) Treasury-like operations (including debt and foreign exchange management); and c) sector governance (regulatory roles and oversight o f Sonangol). Performance in each area i s critiqued, with particular emphasis o n the influence o f expediency considerations, and a series o f reform steps, or a path towards normalization o f the roles o f the several agencies involved i s recommended". B. TAXATION PETROLEUM 7.6 Petroleum revenue flows start with petroleum taxation. For purposes o f this Chapter, taxation i s taken to include all forms o f tax and other payment required under law or contractual provisions. The principal components o f petroleum taxation thus defined in Angola are as follows: 0 Bonuses. These are lump sum cash payments, the most important o f which i s paid on award o f a contract to explore and produce. Bonus payments may also be required o n declaration o f a commercial discovery and o n commencement o f production. Other bonuses, recently introduced, go to social funds. Bonuses are bid by would-be foreign investors and are paid to the national o i l company, Sonangol. 0 Royalties. Royalties are levied as a percentage o f gross production from each concession or contract area. Existing royalties fall in the range o f 16.67 percent to 20 percent. Royalties and the taxes listed below are paid to the Ministry o f Finance. 0 Petroleum Income Tax. The Petroleum Income Tax (PIT) i s levied at a 65.75 per cent rate o n normal profits from older concession areas, and at a 50 percent rate on profit o i l from newer Production Sharing Agreement (PSA) areas. *' This chapter draws o n the recently completed Diagnostic Study o f the O i l Sector w h i c h was funded jointly by the Angolan Government and the W o r l d Bank. The a i m o f the Diagnostic Study was t o review the upstream petroleum sector (Le., crude o i l production activity) as i t affected the public finances o f the country. Specific objectives were to assist the Government in enhancing the transparency o f revenue flows f r o m o i l production, and to build capacity for managing those flows. - 68 - 0 Petroleum Transactions Tax. The Petroleum Transactions Tax (PTT) applies only to the Cabinda Concession Area and i s designed to ensure that the Government receives an increased share o f Concession Area profits when o i l prices are very high. 0 Profit Oil. Investors operating under PSAs, including Sonangol, are obliged to pay the Government, through Sonangol, a share o f their production after recovery o f cost. This Profit O i l share i s negotiated contract by contract, and increases as function o f either daily or cumulative production. 0 Payments to Provincial Government. These are payments made to the governments o f Cabinda and Zaire out o f PIT. Payments are made directly by the o i l companies, and the Ministry o f Finance adjusts i t s reported P I T revenue accordingly. 7.7 Revenues from all these sources are currently o n the order o f U S D 4.7 billion annually, and, as suggested above, can be expected to grow sharply over the next few years.82 7.8 The Government has for some time had in place audit procedures designed to ensure that taxes assessed and paid are accurate. The first o f these i s a “value for money” audit conducted on behalf o f Sonangol. This audit verifies or challenges, as appropriate, the costs claimed by industry as recoverable against required tax or profit o i l payments. The second audit, prepared for the National Tax Directorate (DNI) in the MINFIN, i s a tax audit, verifying or challenging the taxpayer’s calculation o f payments due and payments made. These audits are conducted by internationally qualified and recognized firms, selected by transparent, intemational competitive tender. Sonangol is covered by both audits and treated the same as i t s international partners. As a result o f these two audits, and the original U S dollar accounts and invoices o n which they are based, the Government i s regularly provided with a very reliable set o f U S dollar numbers for tax payments due, payments made, and revenues received, except in relation to Sonangol. 7.9 Reliable numbers on tax collection and payments are indispensable to o i l revenue management and in this respect Angola i s well positioned. Several issues have yet to be addressed, however. The first o f these relates to Sonangol’s payment, o r apparent non- payment o f taxes. 7.10 Reports on payments made versus payments due show Sonangol’s payments to be irregular, delayed and frequently well below what i s calculated by the auditors as due. This is often interpreted as deliberate o n the part o f Sonangol, the charge being made that payments are withheld to cover inefficiencies or misuse o f funds - which i s not necessarily the case. As suggested in section A, Sonangol, at the Govemment’s request, undertakes a wide range o f activities or tasks o n behalf o f the Government (see discussion o f quasi-fiscal expenditures under the next heading). In the past, the Government has sometimes failed to reimburse Sonangol, or has been delayed in doing 82 Volume will certainly increase dramatically. The pace o f revenue growth will depend additionally on the future path o f o i l prices. - 69 - so, making it very difficult for Sonangol to meet i t s own obligations or otherwise behave in a proper commercial manner. Sonangol’s reaction has been to reduce the tax and profit o i l payments it owes Government by the amount o f the costs i t has incurred o n Government’s behalf. 7.1 1 The consequences o f this practice for effective revenue management are adverse. Total payments to Angola from the petroleum sector are underreported and final net payments are made uncertain and delayed as a result o f contentious and often protracted negotiations between Sonangol and the MINFIN, over the allowable tax offsets. Disputes arise because in the past there has not been clarity on which activities qualify for offset treatment, and because expenditures under qualifying categories have not been audited. 7.12 The second outstanding issue relates to the conversion o f U S dollar numbers and accounts into Kwanzas. The O i l Revenue Diagnostic Study revealed a number o f discrepancies in the Kwanza reporting o f o i l revenues among the several agencies involved (MINFIN, BNA and Sonangol). While a number o f outside observers have interpreted these findings as evidence o f abuse, the Study itself attributes them primarily to institutional failings. Translations o f U S dollar revenue figures into Kwanzas have been made at different times and different exchange rates by the different agencies, resulting in confusion and inevitably in some discrepancies between payments reported as made and revenues reported as received which do not exist in the dollar figures. In response, the Government will continue the reporting in U S dollar figures i t has recently begun, and has assigned a high priority to the improved reconciliation o f Kwanza reporting among Government agencies. 7.13 The final issue is transparency - the public reporting o f petroleum revenues. Transparency i s important for a number o f reasons: a) i t fosters public and democratic debate on revenue use; b) i t acts to increase accountability in the executive and legislative branches o f government at all levels, reducing the opportunities for corruption and the potential for waste o f public funds; and c) increasingly demanded by not only citizens within the country, but also by international financial institutions (public and private), donor organizations and international c i v i l society, transparency can be expected to attract increased finance and investment. While in the past, the Government was not prepared to go public with i t s o i l revenues, i t n o w is, and implementation difficulties appear to have more to do with institutional capacity than lack o f will. Revenue numbers have appeared on the MINFIN’s website , and the Government has also authorized their release in the latest IMF Article I V Staff Report. Ease o f access to both sources i s a problem, however, and audiences for both are limited. c. TREASURY-LIKE OPERATIONS BY SONANGOL Overview 7.14 Sonangol’s activities on behalf o f the Treasury are at the core o f the non- conventional mechanisms o f budget execution as previously referred to. These include quasi - fiscal activities (including fuel price subsidies) and the servicing o f o i l backed loans o n behalf o f the Government. - 70 - 7.15 All those transactions are performed outside o f the formal budgetary framework, the SIGFE. An ever-present risk therefore exists that their total executed amount will exceed the fiscal targets o f the OGE. This i s clearly inimical to aggregate fiscal discipline and macroeconomic stability. In addition, the fact that such expenditures are not subject to the automatic controls o f the SIGFE also implies that lines o f accountability are blurred. Further, their existence gives rise to complex and largely informal compensation mechanisms between Sonangol and MINFIN. 7.16 Because Sonangol controls the o i l volumes required to back loans to Angola, and because Sonangol i s familiar with o i l and financial markets and has the requisite commercial expertise, Sonangol i s the principal agency charged with arranging loans o n behalf o f the Government, largely bypassing MINFIN and the BNA, notwithstanding the existence o f a Debt Management Commission (refer to Chapter 4) in which all three agencies nominally have a role to play. In addition to the debt i t arranges o n the Government’s behalf, Sonangol raises debt o n i t s o w n account. Since Sonangol i s wholly state owned, this i s essentially part o f the national debt along with Government debt83. 7.17 The “expediency” proved financially costly to the Government, as i t led to the accumulation o f tax arrears by Sonangol. During the late Nineties, a formal compensation mechanism existed through the so-called “petroleum account” at BNA, whereby the latter would act as a “clearing house” for the financial flows - in Kwanzas and in U.S. dollars - between MINFIN, BNA and Sonangol. This mechanism broke-off as the three partners did not fully comply with the terms o f the legislation. Sonangol stopped paying i t s “cash call” obligations through the “petroleum account” before the end o f 2000, and started accumulating tax arrears vis-&vis the Treasury that amounted to 4.6% o f G D P in 2001. 7.18 Additionally, the Central Bank has been excluded from the compensation mechanisms and has become partially unable to monitor the country’s foreign exchange. A key legal and fiduciary concern emerged as a result o f the demise o f the “petroleum account” in 2001: since then, the BNA i s excluded from the compensation mechanisms, thereby preventing i t from fulfilling its role as the Government’s official Manager o f all foreign exchange. This violates the spirit o f the Organic L a w o f the Central Bank. 7.19 From a fiscal point o f view, Sonangol’s Treasury-like operations are not negligible and have reached a sizable magnitude. According t o the final 2002 Budget Execution Report, expenditures by Sonangol (excluding subsidies and debt service) outside the SIGFE represented almost 11% o f total executed Government expenditures in the period. In addition, Sonangol handled in 2002 U S $ 1.2 billion to service oil-backed loans o n behalf o f the Treasury. In 2002, budgeted price subsidies totaled some US$300 m i l l i o n dollars, o f which US$220 m i l l i o n (equivalent to 2 percent o f GDP) were in arrears by the end o f the year. Most o f these price subsidy arrears (89.8 percent) were due to Sonangol Holding (through Sonangol Distribuidora) for the provision o f petroleum products at below-market prices. 83 These practices, together with an evident lack o f coordination among the several agencies involved, have made i t very difficult to access accurate o r complete data o n Angola’s debt, a state o f affairs confirmed by the O i l Diagnostic exercise and by W o r l d Bank and IMF missions over the past several years. L a c k o f transparency in turn has locked the country into oil-backed loans as the only source o f loan finance, resulting in high prices and short tenures, again underscoring the need to move to greater transparency. - 71 - 7.20 The “petroleum account” was discontinued in 2002 and replaced by largely informal compensation mechanisms involving Sonangol and MINFIN, but again excluding BNA. Such mechanisms take the form o f tax and profit o i l offsets, whereby Sonangol “pays i t s e l f out” for the expenses i t incurs on behalf o f the Treasury. One o f the main implications - as discussed in Chapter 4 - is that the Treasury Single Account (CUT) can n o longer capture the totality o f fiscal transactions - which violates the Protocol between BNA and MINFIN. 7.2 1 Independently o f how organized these non-conventional mechanisms can become, they violate the r u l e o f the law. While during 2002 the compensation process seemed to be conducted in a rather haphazard way, since 2003 i t became somewhat more organized and predictable. This in itself presents the additional danger o f creating an artificial “functionality” that could lead to the perpetuation o f a mechanism that violates basic legislation and ultimately weakens the Ministry o f Finance as the chief fiscal institution in Angola. 7.22 In the remainder o f this section, we will discuss in more detail the workings and the eventual collapse o f the “petroleum account”, followed by an analysis o f the currently prevailing compensation mechanisms between MINFIN and Sonangol. The Petroleum Account Mechanism 7.23 Under the previously existing petroleum account mechanism, a transit account was established at the BNA in which all credit and debit transactions between Sonangol and the Treasury and between Sonangol and its foreign partners were recorded. BNA acted as a clearing house for the financial flows involved, which were essentially foreign exchange flows. This ensured that the BNA would be the focal point for foreign exchange management. 7.24 Under Decrees 15/8gS4and 30/9585, the mechanics o f financial flows seemed to be clearly defined, albeit complex: Sonangol would s e l l to BNA the total proceeds from o i l exports. In turn, BNA credits the company with the corresponding Kwanzas. From those, the company would pay i t s main obligations, Le., taxes to the Treasury and cash call to operators. If any surplus remained, i t would be transferred to a Sonangol’s account in a commercial bank or, if there was a deficit, funds would be transferred from Sonangol’s commercial bank accounts to BNA. Figure 7.1 illustrates the workings o f this mechanism. 84 Sonangol should give to the OGE the revenues received in i t s role as a concessionaire, being authorized to retain up to 10% o f these proceeds t o pay contractors supervision 85 Companies under general foreign exchange regime should sell to BNA the total foreign currency revenue resulting f r o m exports o f crude o i l and should be credited by the corresponding amount in Kwanzas - Article 1, nr. 2 and Article 3, nr. 1 and 3. - 72 - . Figure 7.1 : Financial Flows under Decree 30/95 Sonangol - Commercial ‘ Sonangol - Oil Sector Operations Foreign Cash Call BNA - - - - -- - - - - - - -1 --- --------- Operators (Petroleum Account) Pays Kz for external debt I I I O $ u : U I gg Committed oil ........................................................................... j $; ! & 2 a‘ E - 0 t Trustees: 1 Cabinda Trust & Soyo Creditors ........................................................................... 7.25 In practice, the mechanism broke down because the monies owed.to Sonangol from the transit account were either not transferred in full or not quickly enough to allow Sonangol to meet either i t s tax obligations or the cash calls o f its foreign joint venture partners (Sonangol’s payment obligations as an equity participant in joint venture operations). Sonangol’s response to this state o f affairs was to retain its foreign exchange earnings in i t s o w n accounts and, bypassing the BNA, pay i t s tax and cash call obligations directly, undermining the B N A ’ s foreign exchange management responsibilities An Informal Compensation Mechanism 7.26 The accumulation o f arrears against each other (tax arrears on the side o f Sonangol, and fuel-price subsidy arrears o n the side o f the Treasury) seems to have been a major motivation for the introduction o f some sort o f compensation mechanism (albeit informal) to replace the “petroleum account” - but without the participation o f BNA. 7.27 The new mechanism works largely through tax and profit o i l offsets: Sonangol’s expenditures on behalf o f the Treasury are deducted from their obligations towards the latter. During 2002, this mechanism appears to have compensated only partially for Sonangol’s treasury-like spending - s t i l l allowing for the accumulation o f fuel-price subsidies o n the part o f the Treasury (as seen in Chapter 4, Chart 4.1). Chart 7.1 below shows the mechanism operated during 2002. - 73 - Chart 7.1: The Informal Compensation Mechanism at Work in 2002 (Kz million) +Production Tax 8000 7000 Transaction Tax 6000 5000 Income Tax 4000 --“-Profit Oil 3000 2000 +Tax on Distribution 1000 Activities 0 -Compensated Quasi Fiscal Spending -+- Non-Compensated Source : Sonangol . 7.28 Quasi-fiscal spending (excluding external debt service) was compensated during 2002, mostly, but not exclusively, against profit o i l (see Chapter 7.1)86. Total non-debt quasi-fiscal spending by Sonangol during this year i s estimated to have reached U S $ 597 million. The timing for both spending and offsetting in the course o f the year gives rise to a number o f observations: N o t all quasi-fiscal spending by Sonangol was compensated in 2002 through offsets. Quasi-fiscal spending not compensated in 2002 was offset against Sonangol’s obligations to the Treasury during 2003. Quasi-fiscal spending by Sonangol started l o w in 2002, and increased substantially from March onwards, peaking in November. This pattern reflects both an increased demand for peace-process-related spending (e.g., quartering o f former UNITA rebels) and rising fuel-price subsidies, in the absence o f price adjustments for petroleum products, in the context o f a highly inflationary economy. However, the compensation process really kicked-off from M a y onwards, and mostly relied o n profit-oil offsets (some minor recourse to tax offsets took place throughout 2002 as well). I t i s interesting to note that profit-oil-based compensation was substantially “frontloaded” in June and July, which is likely to have been a factor in the cash-flow difficulties affecting the Treasury in the fourth quarter o f 2002. 86 I t appears that most o f the debt service conducted through Sonangol was compensated by tax offsets during 2002. - 74 - 7.29 I t could be argued that, to a large extent, 2002 was an exceptional year, and that the fiscal phenomena observed over the course o f that year reflected the new demands (e.g. demobilization-related spending) arising from the peace process. During 2003, however, an informal compensation mechanism continued to operate, arguably on a more predictable and systematic basis. 7.30 The ‘‘informal rules o f the game” with respect to the compensation mechanisms adopted in the course o f 2003 are as follows. Quasi-fiscal spending i s n o w fully compensated against profit oil, while external debt service conducted by Sonangol i s compensated against tax offsets. Although the total value o f offsets for 2003 were not made available to the P E M F A R team, i t i s estimated that profit-oil offsets have been more than sufficient to cover Sonangol’s non-debt treasury-like operations. In turn, external debt service costs handled by Sonangol are deducted monthly from i t s taxes due. At any given month, the following theoretical sequence o f events has been applied: 0 If debt service due (handled by Sonangol) = X 0 And taxes due by Sonangol = Y < X 0 Then the Treasury issues a payment order in favor o f Sonangol equivalent to X. 0 Finally, the difference (Y-X) i s rolled over to the following month. 7.31 The estimated total external service handled by Sonangol o n behalf o f the Treasury in the way described above i s approximately [US$ 1.2 billion]. 7.32 In addition, as noted in Chapter 3, the revised 2003 Budget incorporated some U S $ 257 m i l l i o n to cover for goods and services directly paid for by Sonangol, introducing a further degree o f predictability (for MINFIN) with respect to the behavior o f the profit o i l offsetss7. However, preliminary Sonangol numbers for up to October 2003 already indicate a total o f approximately US$ 815 m i l l i o n (of which U S $ 421 m i l l i o n as fuel price subsidies) total Treasury obligations against Sonangol. Therefore, the budgeted amount for Sonangol’s spending was far below the accrued actuals up to October 2003. Although an important step, the budgeting o f Sonangol’s quasi-fiscal spending i s not sufficient in i t s e l f to ensure that the non-conventional execution system would not undermine aggregate fiscal discipline. D. SECTOR GOVERNANCE PETROLEUM 7.33 Sector governance has a bearing on petroleum revenue management in two respects: a) through certain aspects o f sector regulation undertaken by Sonangol in i t s Concessionaire capacity; and b) through oversight o f Sonangol’s investment operations. 7.34 Sonangol as Concessionaire. As title-holder o f each o i l exploration and production license, or Concessionaire, one o f Sonangol’s critical roles i s to approve all major procurement contracts. Given Sonangol’s equity participation in a wide range o f o i l field service and supply companies, there exists clear potential for conflict o f interest in i t s exercise o f the Concessionaire role. Awards o f procurement contracts to non- 87 I t i s important to note that the mere introduction o f such amount in the OGE does not mean that the actual spending would take place within the formal budgetary system (that is, through the SIGFE and the CUT). - 75 - competitive suppliers could significantly raise costs, especially where contract provisions may allow for cost recovery with an “uplift”, i.e, greater than 100% recovery. The main adverse impact o f such cost inflation will not be on the foreign investor (who will recover the cost against taxes) nor o n Sonangol, whose subsidiaries will benefit from the contract awards, but o n the Treasury, whose tax revenues will be reduced, possibly significantly. This potential problem was identified in the O i l Diagnostic, and while disputed by Sonangol, i t s existence was confirmed in informal meetings with Sonangol’s international o i l company partners.88 7.35 Oversight ofSonango1. As a wholly owned state enterprise, Sonangol both generates and consumes public funds on a massive scale. Sonangol’s investment requirements will escalate rapidly in the near t e r m to fund i t s 20% share o f ultra deep water o i l field developments, expected to cost in excess o f U S $ 1 billion each. Other major investments are also under consideration - participation in a LNG export project, and in a grass roots 300,000 barreldday refinery at Lobito. Cleary, these investments have a macroeconomic significance, and should only be made in the context o f a careful budget debate which takes into account not only the merits o f the investments themselves and the use o f public funds in particular, but also competing budgetary demands, e.g., for social and other infrastructure. Unfortunately, the state i s not well equipped to exercise the necessary oversight. MINFIN, the state’s representative o n Sonangol’s Board, does not have the skills or resources to critically examine or oversee Sonangol’s performance and plans, their macroeconomic relevance notwithstanding. E. SUMMARY AND RECOMMENDATIONS 7.36 Implementing an appropriate fiduciary framework for petroleum revenue management is enormously important for Angola. Some progress has been made in this regard, but a significant agenda remains. The paragraphs below summarize the observations made in this Chapter and make a number o f recommendations for the transition from present circumstances to full reform. 7.37 Taxation. Reasonably robust audit procedures are already in place that provide the Government with credible U S dollar numbers o n tax payments made and received, at least as far as foreign taxpayers are concerned. Total tax payments and receipts are clouded by tax offset procedures followed by Sonangol to compensate for a range o f quasi-fiscal activities performed o n behalf o f the Government. Clarity further deteriorates when U S dollar figures are translated into Kwanzas by different agencies at different rates without coordination. Finally, while the Government has taken steps towards increased transparency o f tax revenues, public access is still relatively restricted. 7.38 Recommendations for the transition include: a) negotiated and public clarity on which quasi-fiscal activities will be undertaken by Sonangol and qualify for tax offsets; b) explicit audit o f these activities. This i s expected under the ongoing independent audit of Sonangol; c) adoption o f U S dollar accounting for o i l revenues; d) capacity building at 88 Sonangol retains 10% o f the value o f the Government’s o i l entitlement f r o m production sharing contracts which i t sells on behalf o f the GOA. This fee i s intended t o cover not only the cost o f theoil marketing activity, but also the cost o f other regulatory obligations assigned to Sonangol as Concessionaire, as well as part o f the quasi-fiscal operations discussed earlier in t h i s Chapter. - 76 - each o f the key agencies (MINFIN, BNA, Sonangol), with an emphasis on improved inter-agency coordination. This might be achieved by setting up a Petroleum Revenue management Unit in MINFIN with representation from the other agencies involved. 7.39 Budgetary Management. Quasi-fiscal expenditures, principally by Sonangol, occurring outside the normal budget process, have undermined fiscal discipline and also led to misconceptions o n the use o f funds. A second concern i s the absence in the budgetary process o f any mechanism designed to deal with o i l revenue volatility. 7.40 The Government (MINFIN)has moved to incorporate at least some o f Sonangol’s quasi-fiscal operations into the budget, but needs to go much farther. Agreement on allowable offsets and planned audits will help. Ultimately the quasi-fiscal activities should be moved out o f Sonangol. Although this cannot happen immediately, in the meantime all-quasi fiscal activities could be “ring-fenced” within Sonangol for better monitoring and control and ease o f eventual transfer. The phases o f the ring-fencing process i s described in Box 7.1. - 77 - Box 7.1. Phases o f the “Ring-Fencing” Process First phase First step - ensure that the value o f these activities i s estimated in advance and included in the annual budget. (This has started with the 2003 Revised Budget). Second step - agree what quasi activities are to be identified and measured within Sonangol for recovery (that i s subject to the profit o i l and tax offsets). The improved measurement o f the associated costs would include the introduction o f time writing techniques and cost allocation systems. Third step - the foregoing transactions would be included as a separate task within the proposed independent audit o f Sonangol. Second phase Submission by Sonangol o f an independently audited analysis o f unbudgeted expedient type costs, in accordance with a Cabinet Minute would simplify and accelerate the approval process within MINFIN. The present independent audit o f taxes paid should be extended to include an audit o f the calculation o f net taxes due to GOA after netting o f f offsets and the value o f quasi-fiscal activities. This w o u l d result in the whole cycle o f tax flows being audited f r o m the taxes paid by Sonangol to the final receipt o f these taxes by MINFIN after the deduction o f the offsets and quasi-fiscal activities. Thus an independently audited full reconciliation between taxes originally due and n e t taxes finally received by MINFIN would be available. Thirdphase Another important aspect would be that approval o f Sonangol’s tax retums should be strengthened by centralizing both the approval o f gross taxes due and the value o f offsets t o arrive at the net taxes due and paid within M I N F I N ’ s T a x Directorate. The foregoing should be formalized in a n e w procedure approved by Cabinet. Fourth phase Formal reporting mechanisms f r o m Sonangol t o MINFIN, and from Sonangol to BNA, need t o be established, particularly regarding: Extemal debt service. tax and profit o i l offsets and o i l exports undertaken by Sonangol. 7.41 Some consideration should be given to the application o f fiscal rules to the budget process that would set aside a portion o f excess funds resulting from unexpected o i l price increases. 7.42 Budgetary management, and indeed all aspects o f petroleum revenue management, would be greatly helped by use o f a petroleum sector financial model such as that developed in the context o f the Oil Diagnostic Study (see Annex 3). 7.43 Financial Management. There i s an urgent need for greater transparency with respect to both debt and foreign exchange transactions. The ongoing independent audit o f Sonangol should provide some o f this, but i t will take a concerted comprehensive approach to get the required results. A capacity building program in the BNA should be - 78 - pursued in order to guarantee a safe return o f all debt and foreign exchange operations currently carried out by Sonangol can be safely restored to that institution. 7.44 Sector Governance. Certain aspects o f Sonangol’s role as Concessionaire, notably i t s approval o f all major procurement contracts create a significant potential for conflict o f interest and loss o f revenues to the Treasury. Further, the GOA lacks the capacity to oversee effectively Sonangol’s revenues and investment programs although their scale is enormous and has clear macroeconomic relevance. 7.45 Tempered by expedience and by the need to build capacity in oversight agencies, consideration should be given to transferring Sonangol’s Concessionaire roles to the Ministry o f Petroleum, where they are traditionally found in international practice. Ring- fencing these activities first within Sonangol would help prepare the transfer (see Box 1 above). The urgency o f proper oversight o f Sonangol i s such that the GOA/MINFIN might be well-advised to engage qualified consultants sooner rather than later to assist in performing this function. 7.46 A “Path to Normalization ”. “Normalization” refers in this context to a situation in which the “non-conventional” or “expedient” activities would have been eliminated over an agreed time period. This would represent the arrival point o f a transition process involving the key institutions - MINFIN, BNA, and MrNPET. However attention must be paid as part o f the transition process to upgrade the resources, skills, training, internal procedures and reporting and communication within and among these institutions in order for them to be ready to take back ownership o f the Treasury-like activities. These changes would involve improvements to salary structure, employee career planning and benefits, all o f which would require commitment to institutional reform. The main milestones o f the transition process are summarized in Chart 7.2. Chart 7.2: Schematic View o f the Transition Current Situation The Transition The Arrival Point Ring-fencing o f Coexistence o f a formal expedient mechanisms. Full compliance with budget execution system Elimination o f fuel price Organic Budget Law. with “expedient” subsidies. Full compliance w i t h execution mechanisms. Assessment and o f Organic L a w o f the Compensation ongoing quasi-fiscal Central Bank. mechanisms between activities by Sonangol, Transfer o f Sonangol and MINFIN. followed by transfer t o concessionaire role to Weak oversight appropriate institutions. MINPET. increasing fiduciary risk. Implementation o f a debt BNA not fully exerting management strategy, its role as foreign reducing recourse t o oil- exchange authority. backed loans. Ongoing reforms at Capacity-building at Sonangol, MINFIN, and MINFIN, MINPET, BNA. BNA. - 79 - 7.47 The transition process should ultimately lead to the following: 0 Full compliance with the Organic Budget L a w and 0 The elimination o f the non-conventional mechanisms o f budget execution. 0 Full compliance with the Organic Law o f the BNA, restoring i t s role as Angola’s foreign exchange authority. 0 Elimination o f Sonangol’s dual role, and transfer o f i t s Concessionaire functions to the Ministryo f Petroleum. 7.48 I t i s important to stress that different time scales will be involved in the process o f reaching each different component o f the “arrival point”. In particular, The Government should adopt a phasing-out strategy specific to each category o f treasury-like operation. 0 The gradual elimination o f fuel price subsidies through periodic price adjustments would lead to substantial savings for the State. I n parallel, appropriate social safety nets should be designed and implemented. 0 The Government should carefully examine the rationale for the various other categories o f quasi-fiscal activities carried out by Sonangol o n behalf o f the, Treasury. Those for which justifications are weak should be simply eliminated. A plan should be put in place in which justified quasi-fiscal activities would be transferred to the appropriate organizations within the State. 0 Only sustained fiscal adjustment over the medium-term will eliminate the deficit- financing motivation for the contracting o f external loans. Fiscal adjustment should also give multilateral and bilateral creditors the confidence required to increase Angola’s access to concessional external financing. A clear debt management strategy should also be designed in which the recourse to oil-backed financing would be made less necessary over a realistic timeframe. 7.49 A proposed three-year reform program - indeed the second prong o f the reform strategy proposed in the P E M F A R - i s presented in Chapter 9, Table 9.2. - 80 - 8. LINKINGBUDGETINGAND POLICY Currently, Angola’s Public Investment Program (PIP) is not prepared on the basis o f a multi-year methodology, nor is the State Budget (the OGE) solidly grounded on a medium-term policy framework. However, both on the revenue and on the expenditure side there have been progress in the development o f policy tools that may prove to be the seeds for a successful medium-term approach to budgeting and policy. This chapter notes that the expected increase in oil revenues as well as Angola’s reconstruction needs make it a l l the more necessary for the country to implement a medium-term approach to budgeting, as well as to strengthen the links between budgeting and policy. A. THENEED FOR A MEDIUM-TERM TO BUDGETING APPROACH 8.1 The start o f production in the giant oilfield Girassol in 2002 - which added approximately 200,000 barrels to the country’s daily o i l production - inaugurated a new era for Angola’s o i l sector, one in which production will be dominated by newer, deeper- water oilfields. N o t only the composition o f o i l production is changing, but its volume as well. As noted before, total daily o i l production i s expected to rise from the current 1 m i l l i o n barrels to more than 2 million in 2008. 8.2 The particular nature o f the contracts (production sharing agreements, or PSAs) established between the Government, through Sonangol, in its concessionaire role, and the o i l companies89 may further complicate the already intricate task o f forecasting revenues from oilg0. The PSAs - the dominant arrangement in deep- and ultra-deep- water concession blocks - essentially imply that the behavior o f cost o i l and profit o i l do not follow a linear pattern over time, which means that simple spreadsheet-based linear models - such as the one currently used by the National Tax Directorate - will not properly capture the ensuing revenue fluctuations. 8.3 Over the next few years, Angola’s revenues from o i l will be subject to three additional sources o f variation, in addition to crude price volatility: (i) the changing composition o f production; (ii)the increasing volume o f extraction; and (iii)the increased importance o f PSAs and i t s implications for the behavior o f the State’s profit oil. Thus, ~~~ 89 Refer to Chapter 7 for a fuller discussion o n the nature o f PSAs as w e l l as o f Sonangol’s role in Angola’s ublic financial management. Witness the discussion o n the effective rate o f o i l taxation in the IMF’s Staff Report for the 2003 Article I V consultation (Intemational Monetary Fund (2003)). The share o f o i l receipts in G D P declined in 2002, despite rising o i l prices and production, presumably because cost o i l deductions increased as well. The lack o f a systematic analysis o f the behavior o f o i l tax and profit o i l receipts meant that t o some extent the authorities were taken by surprise and could n o t make timely adjustments in i t s expenditure patterns t o reflect the behavior o f revenues. - 81 - more than desirable, i t is very necessary that Angola adopts an approach to budgeting that goes beyond the one-year framework o f the OGE, to avoid unnecessary disruptions in cash management and to correctly forecast the Government’s fiscal stance. In particular, the Government should take advantage o f tailor-made instruments, such as the O i l Sector Financial M o d e l (to be discussed in Section B below), to improve the quality and reliability o f o i l revenue forecasts, over a multi-year period. 8.4 The need for a multi-year approach to budgeting i s even greater when it comes to establish sound public expenditure policies. The link between the latter - especially at the sectoral level - and the budget process i s weak in many developing countries, and Angola i s no exception. As briefly discussed in Chapter 1, the OGE has been conceived to a large extent as a wartime budget, and the composition o f Government expenditures essentially reflect that. Public expenditure policies - including fuel price and utility tariff subsidies - are established in a largely ad hoc manner, and without the underpinning o f a medium- t e r m spending plan or strategy. Therefore, not only the l i n k s between policies and budgeting are weak, but the foundation for sound public expenditure policymaking i s very shaky as well. The adoption o f a medium-term approach to budgeting, therefore, has to be accompanied by a medium-term approach to public expenditure policy as well. 8.5 The adoption o f such medium-term approaches should not imply the outright implementation o f a Medium-Term Expenditure Framework (MTEF). As pointed out by recent literature (see L e Houerou and Taliercio (2002)), MTEFs cannot work if a solid budgetary management process - particularly as regards budget execution - i s not in place”. Insofar as Angola’s public financial management remains characterized by a dichotomy between formal and “non-conventional” mechanisms o f budget execution, an M T E F i s clearly not an option. But this does not prevent the Government from moving towards the conditions required for a successful MTEF92. Indeed, some o f the seeds for a successful medium-term approach to budgeting and policy have already been planted and will be discussed in Section C. B. EXISTINGOBSTACLES TO LINKING BUDGETINGAND POLICY The Absence o f a Multi-Year Public Investment Program 8.6 The Public Investment Program, by i t s o w n nature, should be the starting point for a better integration between policies and their funding in a medium-term context. Indeed, the PIP i s composed o f projects with maturities longer than one year. However, the PIP has been prepared as an annual tool, which in practice subverts i t s multi-year nature. Although i t s presentation in the OGE has considerably improved over time, the PIP continues to be little more than a l i s t o f projects to be carried out over the course o f a 91 “The MTEF should also rest upon a solid budget foundation, which w o u l d encompass many elements, though chief among them is budget execution that complies with the adopted budget. Consistency between the budget and i t s execution i s a precondition for transparency, predictability, and accountability. In a country where budget execution (Le., actual expenditure) bears l i t t l e resemblance to the voted budget (Le., the intention to spend by sectors, functions, and programs), an MTEF i s not l i k e l y t o be taken seriously by sector ministries, nor by parliamentarians, nor by c i v i l society” (op. cit., p. 26). 92 One o f the activities foreseen as part o f the Bank-supported EMTA credit is the conduct o f a study o n the preconditions for a successful MTEF in Angola. - 82 - given fiscal year. The PIP does not provide any information on projects that have ended in a particular year, nor investment amounts o f on-going projects that will be transferred to subsequent years. I t also does not provide an overview o f public investments over a timeframe longer than the budgeted year. 8.7 The W o r l d Bank and the IMF have recommended that the PIP preparation and execution processes need to be thoroughly reviewed, including the individual analysis o f each projects, including their merits. Projects should be analyzed not only o n financial and physical terms, but also in terms o f benefits accrued to the economy if those expenditures are continued or terminated. Currently, there i s very little evidence that projects are selected for inclusion in the PIP on the basis o f a strategic approach. 8.8 A process o f “cleansing” (saneamento) o f the PIP started in 2004. Further improvements have been made in terms o f i t s presentation in the OGE, which i s now made within the organic, functional, and territorial standpoints. The PIP classification system i s n o w closer to the one proposed in the Manual de Elaboraqgo do Orqamento (MEO) 2004 (see Chapter 4). The 2004 PIP contains only either ongoing projects or the costs o f feasibility studies for newly-proposed ones. MINF’LAN has been taking stock o f all projects approved between 2001 and 2003 with a view to phase them out or terminate them. The ultimate stated objective o f the “cleansing” i s the adoption o f a multi-year budgeting perspective for the PIP. 8.9 The elaboration o f all pieces o f the OGE should be under the responsibility o f a single entity, in this case the National Budget Directorate at MINFIN. MINPLAN’s current responsibility for the PIP would then be transferred to DNO, but this would not necessarily imply that MINPLAN and MINFIN would merge: MINPLAN could still retain its role in medium-term planning, including the preparation and monitoring o f the poverty reduction strategy. In the short term, attention should be paid to the establishment o f an adequate interface between the SIGIP and the SIGFE, as well as to the integration o f the teams in charge o f the PIP and the recurrent budget. The Lack o f a Medium-Term Policy Foundation for the OGE 8.10 The OGE i s often referred to as a “Program Budget”, when i t actually presents a listing o f current Government programs. The activities under such listing are not always consistent with the Government’s intentions as stated in i t s annual or bi-annual Economic and Social Programs (Programa Econ6mico e Social, PES). For example, MINPLAN launched the PES 2003-2004 in October 2002, aiming to cover two fiscal years. While the original 2003 OGE was prepared with the PES in view, the revised 2003 OGE as well as the 2004 OGE considerably depart from it. The OGE’s justification chapter (Relat6no de Fundamentaggo) thus becomes the ultimate policy basis for the expenditure allocations for the corresponding year. However, the justification chapter i s prepared exclusively in the confines o f MINFIN, with little consultation with line ministries, although the latter have an opportunity to react to i t when it i s submitted to the Council o f Ministers for approval. - 83 - c. THESEEDS FOR A SUCCESSFUL MEDIUM-TERM TO BUDGETING APPROACH AND POLICY T h e Revenue Side: T h e Oil Sector Financial M o d e l 8.1 1 As part o f the work on the Oil Sector Diagnostic a “Financial Model” was developed with the following purposes94: (i) to forecast tax and other petroleum revenues accruing to the Angolan State; (ii) to provide the Government with a tool to monitor and manage o i l revenues; (iii) to provide five-year financial forecasts on o i l operations o n a monthly and quarterly basis; and (iv) to assist the Government in the reconciliation o f financial flows between MINTIN, BNA, and Sonangol. 8.12 Table 8.1. below summarizes the main exogenous (input) and endogenous (output) variables o f the model: Table 8.1 : Financial Model: M a i n Exogenous and Endogenous Variables M a i n Exogenous Variables (Inputs) M a i n Endogenous Variables (Results) Crude o i l prices Total Government’s o i l revenues (taxes, bonuses, profit oil, plus Sonangol’s revenues as exploration partner) Exploration cost profiles Total Government’s o i l revenues (taxes, bonuses, profit oil, minus Sonangol’s revenues as exploration partner) Alternative production profiles Sonangol’s after tax net financial flows (revenues minus costs and taxes) Fiscal instruments (taxes, bonuses, profit-oil IOCs’ after tax net financial flows (revenues minus sharing, royalties) costs and taxes) Cost-recovery ceilings Royalties due Behavior o f the depreciation o f the Petroleum Petroleum income taxes due Income Tax Sonangol’s equity share Petroleum transaction taxes due Carrying o f Sonangol during exploration phase Profit o i l due 8.13 The Financial Model, if well-implemented and properly utilized, i s expected to generate benefits in a number o f areas, all o f which facilitating the establishment o f a medium-term approach to o i l revenue management and substantially increasing the accuracy and reliability o f o i l revenue forecasts. These benefits include: (i)improved financial planning and reporting; (ii) enhanced revenue and cash management; (iii) improved monitoring o f strategic developments in the o i l sector, including future explorations and investments; and (iv) strengthened Government’s ability to negotiate future contracts, including the forecasting o f future revenues. 8.14 The National Tax Directorate should become the model’s “institutional home”. Establishing such “home” will require the implementation o f a capacity building effort and the design o f an appropriate incentive system to ensure that the model i s manned by 93 KPMG (2003). 94This subsection draws extensively o n a presentation made by KPMG to the Angolan Government and Bank and Fund staff in January 2003. See Annex 3 for a detailed exposition o f the model. - 84- skilled and motivated staff. This can be developed with support from the Bank through earmarked EMTA credit resources for the institutionalization o f the model. The Expenditure Side: The Post-Conflict Rehabilitation and Reconstruction Program 8.15 The Post-Conflict Rehabilitation and Reconstruction Program (PCRRP), prepared by the Ministry o f Planning with assistance from the World Bank, i s an important step towards devising a multi-year PIP. The PCRRP was conceived in the context o f the peace process, which allowed for a more systematic approach to rehabilitating and rebuilding social and economic infrastructure damaged or destroyed during the conflict. Naturally, the PCRRP, which covers the period [2003-20051 focus o n the most immediate post- conflict investment needs, and as such should be understood as a subset o f a potential multi-year PIE'. 8.16 The PCRRP comprises capital expenditure outlays o n rural development, health services, education, reinsertion & community development, transport, roads & bridges, electricity, water supply, and urban infrastructure & services, totaling U S $ -billion over i t s duration. Part o f those resources are also to be used for capacity building activities as well as for the management o f the program. The expected distribution o f expenditures is described in Chart 8.1 below. Chart 8.1: Composition o f the PCRRP, 2003-2005 Rural dev. IIHealth 0 Education I7 Reinsertion H Transport E4 Roads Electricity I7Water H Urban Cap. Bldg. 0 Prog. Mgmt. I Source: W o r l d Bank and MINPLAN staff calculations. 8.17 The PCRRP was originally developed to be submitted to the donor community for support, in the context o f a post-war donor conference or roundtable. Since the conference did not take place in the intended period, the Government decided to fully cover the PCRRP costs for 2003, as part o f the Public Investment Program. Table 8.1. shows the expected weight o f the PCRRP as a component o f the PIE'. - 85 - Table 8.2: The Share o f the PCRRP in the total Public Investment Program 2003 2004 2005 2006 2007 Estimated capital spending (US$ 1,066 1,076 1,164 1,399 1,744 million) O f which: PCRRP (US$ million) 600 600 600 700 370 Source: Government o f Angola (2003). Estratigia de Combate 2 Pobreza. Draft for Discussion. 8.18 F r o m an operational point o f view, a major risk will be implementation capacity. Nevertheless, a detailed analysis o f this issue i s beyond the scope o f the PEMFAR. I t suffices here to recall the large hiatus between “physical” and financial execution o f the PIP discussed in Chapter 4. With the substantial increase in capital expenditures to be brought about by the PCRRP, such concerns will be magnified. Nevertheless, this program can be viewed as a first effort at adopting a multi-year approach to a relevant subset o f the PIP. Thus, it i s fair to say that the PCRFV subset o f the PIP has already assumed a more medium-term character. The PCRRP will also be a key component o f the Government’s Poverty Reduction Strategy. The Poverty Reduction Strategy 8.19 The Government circulated in January 2004 to the donor community a near-final version o f i t s Poverty Reduction Strategy (Estrattgia de Combate a Pobreza, or ECP). The ECP, if successhlly finalized and implemented, has the potential to become the first medium-term tool for public expenditure policymaking by the Government since Angola opted to become a market economy. 8.20 The ECP proposes to implement a three-year public expenditure program focusing on ten priority areas, namely: (i) social reinsertion; ( ii)demining; (iii) food security and rural development; (iv) HIV/AIDS; (v) education; (vi) health; (vii) basic infrastructure; (viii) employment and vocational training; (ix) governance; and (x) macroeconomic management. The estimated costs o f the ECP for the 2003-2006 period total U S $ 3.17 billion. The ECP costs are presented as part o f a multi-year fiscal framework, which should be the basis for any future effort at establishing an MTEF. 8.21 Chart 8.2 below describes the composition o f expenditures proposed by the ECP, which i s broadly consistent with the one for the PCRRP. - 86 - Chart 8.2: Poverty Reduction Strategy (ECP): Priority Areas in Terms o f Multi- Year Allocations, 2003 - 2006 OSocial reinsertion Demining 0 Food securitylrural dev. 0 HlVlAlDS Education Health Infrastructure 0 Employment Governance Macro Mgmt. OAssociated recurrent costs Source: ,trategia de Combate a Pobreza (MINPLAN) Recommendations for Further Reform 8.22 In the final analysis, a successful adoption o f a medium-term approach to budgeting and policy needs to rest o n three main pillarsg5: 0 Expenditure control mechanisms, particularly with respect to budget execution, need to be strengthened; 0 The budget process needs to be guided by public expenditure policy, not the reverse; and 0 Revenue forecasting capacity needs to be enhanced. 8.23 The first pillar i s at the core o f the proposed two-pronged PEMFAR strategy for Angola - particularly in what concerns the strengthening o f the formal budgetary execution system. And much progress can be made on the third if the Financial Model i s adequately implemented. The second one - public expenditure policy driving the budget process - requires a sound articulation between short- and medium-term policy and planning instruments, as well as improved coordination between MINFIN and MINPLAN. 8.24 Chart 8.3 suggests a possible model o f articulation. The model advocates the establishment o f clear l i n k s between four basic medium-term programming instruments - the PIP, sectoral/provincial plans, the PRSP, and the M T E F - with the two basic yearly 95 Cf. Schiavo-Campo and Tommasi (1999), p. 88. - 87 - operational plans - the Economic and Social Programs (PES) (often bi-annual) and the State Budget (OGE). The model proposed here follows closely that o f Mozambique. The main difference i s that in Mozambique all six instruments are under the responsibility o f a single entity, the Ministry o f Finance and Planning. In Angola, MINPLAN i s in charge o f the PIP, the sectoral/provincial plans, and the PES, while MINFIN i s responsible for the OGE. The PRSP process has also been assigned to MINPLAN, while an eventual M T E F would likely be part o f MINFIN’s work program. 8.25 While merging MINFIN and MINPLAN might facilitate such articulationg6, this would be neither a necessary nor a sufficient condition for a successful integration o f the medium- and short-term processes. As noted by Schiavo-Campo and Tommasi (op. cit., p. 96) for the particular case o f “dual budgeting”, “the real issue (...) i s the lack o f integration between investment and current expenditure programming, and not the formally separated processes in themselves. This i s important, because to misspecify the issue would lead (and often has) to ‘solving’ the problem by a simple merger o f two ministries - even while coordination remains just as weak. A former minister becomes a deputy minister, organizational ‘boxes’ are reshuffled, a few people are promoted and others demoted. But the dual budgeting problem remains alive and w e l l within the bosom o f the umbrella ministry”. 8.26 The importance o f the PRSP process in strengthening the l i n k s between policy and budgeting should not be underestimated. A sufficiently consultative process would increase the odds that the line ministries would take ownership o f the policies envisaged in the poverty reduction strategy, and would reduce the degree o f MINTIN discretion in devising public expenditure allocation across sectors and provinces. Therefore, the ultimate instrument for articulating policies and financing will be Angola’s ECP. A s the public financial management process gets strengthened, and the “seeds” for a medium- term approach to budgeting continue to be cultivated, the Government should consider a careful move towards a Medium-Term Expenditure Framework - as one o f the ECP pillars. 96 See discussion on FAD’s recommendations - supported by the Bank’s PEMFAR team - in Section B, in the particular case o f the PIP-OGE connection. r, 0 .- I r a - 89 - 9. THE PUBLIC FINANCE MODERNIZATION PROGRAM The Government’s Public Finance Modernization Program (PMFP), initiated in 2002, has the potential to be the main vehiclefor the implementation o f the two-pronged reform strategy proposed in the PEMFAR. The PMFP - which counts on World Bank support through the Economic Management Technical Assistance (EMTA) credit - is already fairly well-placed to push ahead with reforms associated with the f i r s t prong o f the reform strategy, namely, “Strengthening the Formal Expenditure Execution System ”. The main challenge for the PMFP lies precisely on the second prong - “Ring-Fencing and Phasing-Out the Non-conventional Execution System”. Not only some o f the key institutional actors currently fall outside the scope o f the PMFP, but also the implementation o f the needed reforms may face stiff political resistance from influential quarters in Angola. Therefore, strong political support at the highest levels o f government is required for the PMFP to succeed in this task. The reform action plan proposed by the PEMFAR aims to strengthen, not to replace, the PMFP. Accordingly, this chapter argues that a solidly implemented PMFP may prove to be the main tool for the Ministry of Finance to establish fuller control over public financial management in Angola. A. MODERNIZATION THEPUBLIC FINANCE PROGRAM 9.1 In 2002, the Government took the initiative to put together a wide-ranging Public Finance Modernization Program, PMFP, with three components: A Public Finance Management component, which aim is to strengthen the public financial system, i.e., budgetary, financial, non-financial assets, accounting, internal control systems, public pricing, and insurance markets. 0 An Information Technology component; and 0 A Training and Capacity Building component. 9.2 The overall objective o f this program i s to improve institutional capacity within the Ministry o f Finance. The PMFP contains i t s own diagnostic o f the weaknesses o f Angola’s public finance management system (see Box 9.1)The PMFP originally covered the period 2002/2004 - but i t i s very likely to require an extension - and includes activities that are also part o f other ongoing reforms packages, such as the Customs Reform process. - 90 - Box 9.1. Weaknesses in the Angola financial management system as seen by the Government in the PMFP The P M F P contains its own diagnostic o f the existing weaknesses of Angola’s public finance management system, many o f which reflect previous discussions with Bank and Fund staff: Most weaknesses were found in the following areas: 0 In the mechanisms and procedures for budgeting, managing, use and registration o f State financial resources; 0 In the lack o f appropriate integration o f the PIP in the regular budget preparation and execution processes 0 In the mechanisms and procedures for the recording and monitoring o f public debt, donor grants and compensation funds; 0 In the mechanisms and procedures for recording and monitoring o f non-financial State assets; In the mechanisms and instruments for monitoring and evaluating public enterprises; 0 In the mechanisms and procedures regarding internal control; 0 In the existing Customs Administration procedures; 0 In the insufficient automation o f all management processes involved in the SIGFE 9.3 In order to monitor the development and the evolution o f these actions, a coordinating body, “CoordenaqZo de Programa”, was set up within the Ministry o f Finance in July 2002 (Despacho Intemo 37/GMF/02). The coordinating body i s also in charge o f needs assessment, coordination o f technical assistance, identification and hiring o f consultants, and periodic progress evaluation. The PMGFP contains a breakdown o f the program by components and sub-components; the objectives and targets envisaged; the specific actions to be implemented; an implementation calendar and cost estimates for 2003/2004. B. PROGRESS IN PROGRAM IMPLEMENTATION 9.4 Implementation o f the PMFP received a substantial impetus in the last few months o f 2003, when the reforms related to the new version o f the SIGFE - including those recommended by the FAD-WB mission o f August 2003 - started being implemented. M a i n areas o f progress include: 0 A strengthened commitment stage o f the budget cycle. N o w budget units are limited in their ability to commit funds by both the annual amount budgeted in the OGE and by the quarterly amount authorized by financial programming. 0 A strengthened verification stage o f the budget cycle. Previously unobserved to a large extent, the verification stage i s n o w solidly incorporated into the SIGFE. Payments to suppliers can only be made after a “nota de 1iquidaqZo” i s issued in the system, attesting that the corresponding good was delivered or service rendered. 0 All central government budget units in Luanda are linked online to the SIGFE. 0 An improved integration among the following subsystems o f the SIGFE: (i) budget; (ii) treasury; (iii) accounting; and (iv) non-financial asset management (“patrimonio”). - 91 - 9.5 As noted, many o f the PMFP activities - including public procurement reform - will b e supported by EMTA. MINFIN has been diligently working to put together the terms o f reference for these activities. C. MAIN AHEAD CHALLENGES 9.6 The main challenges involved in meeting the PMFP goals will be political and institutional rather than merely technical. The main issue i s how to engage organizations which are not part o f the PMFP implementation but o f which activities l i e at the core o f the PMFP goals. This i s particularly the case o f Sonangol and, to a much lesser extent, the Tribunal o f Accounts. 9.7 As will become clear in the next section, the PMFP i s already quite advanced in terms o f the main priority areas for reform within the first prong o f the proposed strategy, namely, strengthening the formal system o f expenditure execution. However, the PMFP i s clearly silent with respect to the second one: ring-fencing and phasing-out the “non- conventional” spending mechanisms. 9.8 Difficulties are greater with the second prong as political economy considerations are bound to play a major role. The process o f dismantling the “non-conventional” processes over a period o f time will involve dealing with powerful vested interests and traditional sources o f rent-seeking. In addition, i t will also require the establishment o f an appropriate institutional hierarchy, one in which the Ministry o f Finance would restore its undisputed role as the prime public financial management institution in Angola. 9.9 From a technical standpoint, the PMFP could become the instrument whereby MINFIN would increasingly extend its control over all public financial management activities in the country, including those currently undertaken by Sonangol. Therefore, the PEMFAR recommends that the necessary politico-institutional conditions be created that the PMFP can also incorporate the priority actions under the second prong o f the proposed strategy. 9.10 External audit activities undertaken by the Tribunal o f Accounts should also become part o f the PMFP. This should not be seen as a w a y o f weakening the Tribunal vis-a-vis MINFIN. O n the contrary: the ability o f the Tribunal to meet i t s goals would only be strengthened if they are mainstreamed onto the PMFP. The institutional implication is that the PMFP becomes the Government o f Angola’s - rather than only MINFIN’s - chief device for public financial management reform in Angola9?. D. PRIORITY AREAS FOR REFORM 9.11 The three-year action plan presented in Tables 9 . l a and 9 . l b highlights the key areas o f reform already being proposed in the PMFP and proposes an additional set o f measures required to meet the goals o f the two-pronged reform strategy proposed in the PEMFAR. I t must be stressed that the proposed action plan aims to strengthen and complement, by no means replace, the PMFP. 97 I t should be stressed that an additional challenge i s that for the most part the Bank-supported EMTA project does not finance activities related to the second prong o f the proposed strategy. - 92 - 9.12 Finally, the adoption o f specific legislation should f o r m the basis for undertaking the reforms under the second pillar o f the strategy. Currently, the legislative pieces (laws and decrees) are fairly scattered, and in many cases outdated. This n e w legislation should include a timetable for the measures required in the ring-fencing and phasing-out o f the non-conventional expenditure execution mechanisms. Table 9.1 : Priority Actions: Strengthening the Formal System Priority Action Year 1 Year Year Agency in Alreadyin EMTA 2 3 Charge the PMFP? Funding? Budget Preparation Increase realism o f X MINFIN No No macroeconomic (GEREI) and assumptions, especially Economic targeted inflation and Consistency overall balance Group Adopt O i l Diagnostic’s X MINFIN (DNI, No Yes Financial M o d e l to DNO, GEREI) generate more accurate revenue projections Improve coordination X MINFIN (DNO) No No between MINFIN and and M A P E S S MAPESS o n payroll information Establish a single X MINFIN No No cadastre for all public (Human sector personnel Resources, DNO), M A P E SS, Ministries of Defense and Interior Establish interface X MINFIN (DNC) No Yes between SIGIP and and MINPLAN SIGFE Turn the Public X MINPLAN No Yes Investment Program into a multi-year exercise Budget Execution Ensure consolidation o f X MINFIN, BNA No No the Treasury Single Account (CUT) Use financial X MINFIN (DNT) Yes No programming (Initiated) projections as the ceilings for commitments - 93 - Priority Action Year 1 Year Year Agency in A l r e a d y in EMTA 2 3 Charge the P M F P ? Funding? Introduce a reasonable X MINFIN (DNT) Yes No degree o f flexibility in (Initiated) the cash management process Train key UO (budget X X X MINFIN Yes Yes units) staff o n the modus (especially (Initiated) operandi o f the n e w DNC) version o f the SIGFE Enforce the correct X MINFIN (DNC, Yes No sequencing o f the budget DNT, INF) (Initiated) execution process (cabimentaqlo -> liquidaqiio -> pagamento) , by making full use o f SIGFE’s automated controls Rationalize cash X MINFIN (DNT) Yes No rationing mechanism (Initiated) and improve communication with uos [Implement X X MINFIN (DNT) Partially Yes recommendations on and BNA debt management made by December 2003 IMF- WB mission] Implement X X X MINFIN Partially Yes recommendations on (Procurement public procurement Reform reform made in the Committee) context o f the C P A R Accounting and Reporting Reduce time allowed for X MINFIN (DNC) No No submission o f the CGE and Tribunal o f to Tribunal o f Accounts Accounts to six months after the end o f the fiscal year Establish mechanisms X MINFIN (DNC) Partially No for the production o f timely and complete accounts, including regular follow-ups with UOs, especially those with a history o f non- compliance Provide training on X X X MINFIN(DNC) Yes Yes accounting and reporting (Initiated) practices o n a regular basis for relevant staff in the UOs - 94 - Priority Action Year1 Year Year Agency in Alreadyin EMTA 2 3 Charge the PMFP? Funding? Establish accounting - and X MINFIN Yes Yes financial management (DNPE) (Planned) procedures and routines for non-financial assets Complete adoption o f X MINFIN (DNC) Yes No accounting manual as (Initiated) w e l l as the n e w Chart o f Accounts. Internal Audit Design an appropriate X MINFIN (INF) Yes Yes institutional framework (Planned) for INF, as part o f the development of a concept o f a systemic model for the intemal control function Review relevant X X MINFIN(INF) Yes Yes legislation for the (Planned) internal control and audit functions Operationalize the x MINFIN (INF) Yes Yes internal control and audit (Planned) functions, through: the preparation o f manuals of guidelines the design of curriculum of the career of INF inspector as w e l l as of a training program for I N F ’ s technical staff Establish r u l e s o f ethics X MINFIN (INF) No Possible for staff and provide the (training) associated training External Audit M o v e towards strategic X Tribunal of No No regularity audits o n the Accounts compensation mechanisms between MINFIN and Sonangol Phase out and eventually X X X Tribunal of No No eliminate ex-ante Accounts controls by the Tribunal o f Accounts - 95 - Priority Action Year 1 Year Year Agency in Already in EMTA 2 3 Charge the PMFP? Funding? Hire internationally- X X X Tribunal of No No reputable audit f i r m s that Accounts could perform regularity audits in selected public enterprises o n behalf o f the Tribunal as required 7 Table 9.2: Priority Actions: Ring-Fencing and Phasing-Out the “Non-conventional” System Priority Action Year 1 Year Year Agency in the PMFP? 2 3 Charge b‘Ring-Fencing”:Phase 1 Ensure that the value o f these activities estimated in advance and included in the annual is X MINFIN (DNO), cooperation with Sonangol in No I No budget Agree what quasi -fiscal MINFIN (DNO, No I No activities are to be DNI, DNT), in identified and measured cooperation within Sonangol for with Sonangol recovery (that i s subject to the profit o i l and tax offsets Sonangol a separate task within the independent audits o f Sonangol “Ring-Fencing”: Phase + L Submit an independently Sonangol audited analysis of unbudgeted “non- conventional”- type costs, in accordance with a Cabinet Minute to simplify and accelerate the approval process within MINFIN Extend the scope o f the MINFIN (DNI) tax audits t o include an examination of the calculation o f net taxes due to the Treasury after netting o f f offsets and the value o f quasi-fiscal activities - 96 - Year 1 Year Year Agency in Already in EMTA 2 3 Charge the PMFP? Funding? Strengthen the approval X X MINFIN (DNI) No Possible process for Sonangol’s tax returns by centralizing both the approval o f gross taxes due and the value o f offsets to arrive at the net taxes due and paid within MINFINs Tax Directorate Formalize the foregoing Council of No No in a new procedure Ministers approved by Cabinet “Ring-Fencing”: Phase 4 Establish formal Decision: Limited reporting mechanisms Council of from Sonangol to Ministers MINFIN and to BNA, particularly regarding: Implementation Extemal debt BNA, service MINFIN (DNI, Tax and profit o i l DNC, DNT, offsets INF), and Oil exports Sonangol undertaken by Sonangol. Other Required X MINFIN No No subsidies through periodic price -1 adjustments jiscai . . activities X MINFIN, collaboration in No No undertaken by Sonangol with Sonangol to the appropriate organizations within the MINFIN (DNT) No Possible management strategy and BNA, in should also be designed collaboration in which the recourse to with Sonangol oil-backed financing would be made less necessary over a realistic timeframe - 97 - E. INDICATORS PERFORMANCE 9.13 Success in the implementation o f a revised and expanded PMFP depends critically on a close monitoring o f selected performance indicators. T o the extent possible, these indicators should b e quantifiable. Table 9.2 suggests a number o f possible indicators, together with their respective sources o f information. Table 9.3: RecommendedPerformance Indicators Indicator I Agencies in Charge Data Collection Strategy Percent o f total Government 1 MINFIN Information to be collected through expenditures executed through quarterly and yearly budget execution the SIGFE. reports, plus ad hoc reports generated by DNC. Volume o f resources transiting MINFINIBNA Information t o be collected through: through the Treasury Single DNT’s Department for Treasury Account (CUT), as a percent o f Operations; BNA’s Accounting total Government spending. Department. Payment arrears t o suppliers MINFIN Information t o be collected through DNT and D N C . Intra-public sector arrears. MINFIN Information to b e collected through DNT’s Public Enterprises Department. Difference between practiced and MINFIN Information t o b e collected through cost-recovery prices for MINFIN’s Gabinete de Preqos petroleum products. (GAPREC), in coordination with Sonangol. Oil-backed debt stock as a MINFIN, BNA, Sonangol Information t o b e collected primarily percentage o f total external debt. through D M F A S system at BNA; reconciliation with DNT’s and Sonangol’s own databases t o b e made regularly. Oil-backed debt service as a MINFIN, BNA, Sonangol Information t o b e collected primarily percentage o f total exports. through D M F A S system at BNA; reconciliation with DNT’s and Sonangol’s own databases t o be made regularly. Volume of quasi-fiscal MINFIN, Sonangol Information to be collected through transactions incurred by Sonangol o n behalf o f the Treasury. - 98 - ANNEXES - 99 - ANNEX I Angola - Selected Economic and Financial Indicators, 1999-2003 Indicators 1999 2000 2001 2002 2003 Output, income, and prices ( Annual percentage change, unless otherwise indicated) Real G D P 3.3 3.0 3.2 15.3 4.5 Real G D P per capita 0.7 0.2 0.2 12.0 1.4 GNI per capita (Atlas method) in U S dollars 440 470 530 680 740 Consumer price index (annual average) 248 325 153 109 98 External sector Exports (f.o.b.)(based on current U S dollars) 45.6 53.6 -17.5 27.7 18.3 Oil 45.3 58.5 -18.5 32.0 18.8 Non-oil 47.4 20.2 -8.7 -6.8 13.0 Imports (f.0.b.) (based on current U S dollars) 49.5 -2.2 4.6 16.7 25.7 Export volume 1.4 1.6 -2.6 23.1 8.7 Import volume 55.2 1.4 6.5 14.3 10.7 Terms o f trade (decline - ) 49.0 56.8 -13.8 1.7 -3.4 Real effective exchange rate -18.7 20.1 13.0 1.8 12.3 (depreciation -) Money and credit end period (As percent of beginning period M3) N e t domestic assets -362 -327 54 48 13 Money and quasi-money (M2) 533 3 04 163 158 57 Interest rate (3 month time deposit-in percent) 36 46 56 41 27 Public finances (In percent o f GDP, unless otherwise indicated) Total revenues 46.8 51.7 42.5 39.0 36.2 o f which:oil 41.1 46.2 33.9 29.9 29.1 Grants 4.0 2.3 2.3 0.0 1.3 Total Expenditures 82.4 60.4 46.3 47.9 43.8 Overall balance (accrual basis) -35.6 -8.7 -3.8 -8.9 -7.5 Overall balance (cash basis) -24.7 17.6 -4.9 -1.4 -6.4 External sector Current account balance incl. Transfers -28.1 9.0 -15.1 -5.8 -6.2 Debt-service in percent o f exports o f goods and services 44.4 36.3 41.2 26.4 24.0 ( In m i i l i o n of U S dollars, unless otherwise indicated) Gross domestic product in current prices 6088 8864 9472 11204 13183 Gross official reserves In millions o f US$ 496 1198 732 375 620 In months o f imports o f goods and services 0.8 1.9 1.1 0.5 0.7 Exchange rate- period average (local currency per US$) 2.8 10.0 22.1 43.5 74.6 Oil production (thousands o f barrels per day) 746 748 740 903 970 Price o f Angola's o i l 17.6 27.2 24.2 24.3 26.6 Source: IMF. - 100- ANNEX I1 COUNTRY PROCUREMENT ASSESSMENT REPORT SUMMARY OF FINDINGS AND PROGRESS REPORT9* A. FINDINGS OF THE CPAR In 2001, the Government o f Angola spent some 53 percent o f total public expenditure or about 12 percent o f GDP o n public procurement. However, the public procurement and management system i s not providing optimal benefit to the country’s development needs. A country procurement system” analysis, in the context o f the Country Procurement Assessment Report, reviews usually several factors, such as: (i) the degree to which the government promotes a culture o f accountability; (ii) the status o f procurement staff, including salary structure and capabilities; (iii) the degree to which procurement is free from political interference; (iv) the existence o f clearly written standards; and, (v) the degree o f efficiency, transparency and value for money. In each o f these areas, the Angolan procurement system needs major improvements, and, consequently, i s considered high-risk. Three major problems: transparency, enforcement, and lack o f capacity, are considered the main shortcomings o f the system, as discussed below: a Weaknesses of the legal framework and lack of enforcement W h i l e there are laws pertaining to public procurement matters that are written but not adequately publicized or enforced (Decrees no 7/96 and 22-A/92 regulate the main procurement aspects), some procurement practices have no legal basis but are vigorously enforced loo. Even when they are familiar with its rules, c i v i l servants seem to ignore procurement law without being penalized; and enforcing o f the L a w i s erratic. A national study conducted by a national consultant found that 54.2% o f the members o f personnel in charge o f procurement in the various ministries and at the provincial authorities level were not aware o f the existence o f Decree no 7/961°1. Procurement regulations are not enforced because control mechanisms, either internal (within procurement entities) or external are missing at all levels. In addition, none o f the state-owned enterprises visited had written 98 Prepared by Slahhedine Ben-Halima and Christine D e Mariz. 99 A public procurement system m a y be considered w e l l functioning if it achieves the objectives o f transparency, competition, economy and efficiency, fairness and accountability. The absence o f adequate public advertising, public bid opening, announcement o f award, record keeping, oversight authority, audit or recourse, prevents a clear quantitative insight in the Angolan procurement system. Procurement within the central government i s for the most part carried out without consideration t o cost-saving measures. loo The conviction that the Government should buy o n l y f r o m Angolan suppliers and importers; the legislation or the b e l i e f that Government should buy o n l y f r o m Angolans-at any price-, contributes to excluding foreign bidders f r o m the Angolan market, and consequently increases the cost for the Government. The government Gazette i s published and printed, but i t i s inadequately distributed in the various Ministries. - 101 - internal procurement rules, describing methods and levels o f authorization for spending public funds. The government Gazette does not seem to contribute to transparency as well as to the dissemination o f regulations, as it i s relatively expensive and apparently not widely distributed within the govemment itself. Inefficient and costly procedures and practices Sound public procurement policies and practices are among the essential components o f good governance. Good procurement practices reduce costs and allow for timely results; poor practices lead to waste and delays, and are often the basis for allegations o f corruption and government inefficiency. The Angolan government acquires goods, works and services, but not in a consistent manner or according to rules generally recognized as binding. A number o f procurement procedures and practices are neither economic nor efficient, and are not in compliance with recommended procurement best practices. Although nearly all goods are imported, most o f them are procured through local suppliers. However, if the government awarded contracts to international suppliers, i t would contribute to substantial savings for the Treasury (especially for bulk commodities). The two areas where practices need to change in order to enhance fairness and transparency in the procurement process are: (i)adoption o f open competitive bidding as the primary method o f procurement; and (ii) establishment o f a transparent procurement systemIo2. There is currently no effective mechanism to lodge complaints in relation to the bidding process. The lack o f oversight, transparency, and audit, has created many . opportunities for corruption. W e a k procurement organization and capacity There is a lack o f procurement capacity at all levels, both in Luanda and in the provinces, including poor record keeping within Ministries. This deficit in capacity i s due to the following reasons: (i) very l i t t l e importance i s given to public procurement in general; and (ii) lack o f adequately trained personnel in the full range o f activities linked to procurement functions. I t i s stimulating to note however, that as part o f the imports management process (mainly relating to efficient collection o f customs duties and other import taxes), efforts have recently been made to improve Customs Service staff skills through technical assistance provided by Crown Agents. However, this i s only one side o f phase o f the trade cycle (revenue collection), the other one --public expenditure on acquisition o f goods, works and services-- needs appropriate Government attention to ensure that benefits f i o m improved revenue collection are not eaten away by poor public procurement practices. I O 2 Upfront setting up o f evaluation criteria, bid publication, publication o f contract award, complaint mechanism. - 102 - 0 Weak audit and anti-corruption mechanisms There i s n o oversight mechanism to ensure regular application o f procurement regulations, and there are no procurement audits. Procurement regulations are not enforced because procuring entities internal control mechanisms are missing at all levels, as i s the case for external mechanisms. The Tribunal o f Account staff would be able to exercise audit oversight over government procurement, but lacks proper training to assume its responsibilities and perform i t s duties. Corruption in government procurement i s usually suspected. The two entities whose role i s to determine cases o f corruption and apply sanctions, the Tribunal o f Accounts'03 and the High Authority ~ , just being set up, not fully operational, and Against C ~ r r u p t i o n ' ~are inadequate. 0 Payment Delays breed higher award prices I t i s a common practice for Government Ministries and other departments to delay paymentslo5by more than 90 days (internationally accepted practices in that matter are 30 to 45 days) to suppliers and contractors. Most payment delays are due to poor planning, especially in relation to the budgeting process and the actual funding available to pay against committed. Interest penalties are included in some larger contracts but are neither asked for nor paid. Although allowed by Decree 7/96, price adjustment clauses are not used, and the r i s k o f inflation and the effects o f devaluation o f the local currency are passed on to suppliers and contractors. The expectation o f slow payment has two equally damaging consequences: (i) it imposes on the government's contract partners the burden to pre-finance contract payments for several months and therefore discourages participation o f qualified businesses with limited liquid funds- particularly newly established Angolan businesses; and (ii) it encourages bidders to factor the cost o f late payments to their bid prices, resulting in higher than necessary award prices. This in tum increases the burden o n the Treasury and negates the positive effects o f the revenue enhancement mechanism. Moreover, high construction prices in Angola m a y be explained in part by very slow payments. B. WHAT CAN BE DONE ? Nothing can be done without ownership Without the Government's full support o f the initial changes proposed in the CPAR, the combination o f weak capacity with lack o f control will not allow any newly introduced system to function, n o matter how well ~~ The Tribunal hopes to start i t s operations in 2003 with the audit o f the General State Accounts for fiscal year 2002. The H i g h Authority Against Corruption has never became operational, as i t s president has never been elected. Delayed payment, lack o f national industrial production, h i g h cost o f imports, unfair competition o n the part o f the informal sector, complexity o f the tax system, lack o f appeal systems, and crumbling infrastructure (roads and ports), have a negative impact o n the final cost o f goods and services. - 103 - designed. Until now, the team working on procurement reform has only been a small group o f consultants reporting to the Ministry o f Finance, and there i s n o formally established working group with the appropriate authority to move procurement reform forward. I t i s not clear who will eventually be responsible for making the findings and recommendations o f this report the basis o f a national consensus and commitment to change the existing procurement system. Five recommendations for the short term In the short term (between 2002-2004), improved transparency and integrity, the application o f basic discipline as well as managing the process with due diligence, will result in savings for the country. These savings obtained thanks to a more efficient procurement system could in turn be used by the Government to alleviate the condition o f some vulnerable groups o f the population. The action plan below aims at helping the Government to improve efficiency and value for money o f public procurement. A certain number o f activities could be initiated in parallel, over the next few months. However, the first action to be taken should be the revision o f the legal framework for public procurement, given the well-known challenges in matters o f enforcement due to too many exceptions included in the current Law. This critical initial action to revise the legal framework should be followed immediately by others to improve record keeping; establish a policy directorate for procurement in the Ministry o f Finance; and insure that institutions set up to provide sound controls and accountability, begin to function effectively in order to effectively counter fraud and corruption. 1. Identify a champion to spearhead the Procurement reform process and organize a high-level workshop for the Council o f Ministers to raise ministers and their procurement staff awareness about the benefits for the Government o f a more open biding environment. In addition, the Ministry o f Finance should issue circulars or publish newsletters o n a regular basis intended for the private sector, to ensure that i t i s kept informed o f the practices followed by the government. Information meetings with both the public and the private sectors should be regularly held to complement these circulars. .. 11. Create a comprehensive procurement framework. Set up a national working group composed o f national consultants'06, lawyers, representatives o f the public sector (especially the Ministries o f Planning, Finance and Public Works), national experts and private sector representatives, to assist an international legal This consultant in drafting o f the new procurement l e g i ~ l a t i o n ' ~ ~ . legislation could The Bank has already established a dialogue and close partnership with the consultants financed by the Norwegian Agency for Development (NORAD). The mission held working sessions with the consultant team and benefited f r o m the wealth o f information o f i t s work. Some o f the mission current findings and recommendations are based o n this information. A s t o the aspects related t o the legal framework and regulatory setting, the mission agreed with the consultant that the w o r k accomplished so far should be completed with the assistance o f an international specialist in procurement law. lo7 Adopt the open competitive bidding as the primary method o f Procurement. Open competitive bidding is rarely used as a procurement method (most often, bidding i s limited t o obtaining three bids f r o m local - 104- be based o n the Brazilian procurement law, considered a good model in Portuguese-speaking countries. Following the enactment o f the Law, the Government should: (i) issue a complete set o f regulations, standard bidding documents for goods, works, and services applicable; ( ii)launch a sustainable, overall procurement training program at various levels, and a n awareness campaign for the public sector, the private sector and the c i v i l society. None o f the proposed changes can succeed without: (iii)extensive training and advertisement; and (iv) an updated anti-corruption law (taking into account the new modem legal framework for procurement). ... 111. Introduce record keeping to establish a transparent procurement system”*. During the first year pilot program, at least five Ministries should start keeping records on all procurement activities above a certain threshold o f contract value, and transmit them to the Ministry o f Finance. This Ministry should gather and analyze data, which will help improving i t s oversight mandate over public sector procurement. While in the short t e r m this can be with proper filing o f hard copy documentation, in the medium term, the design o f a computerized record keeping and reporting system would facilitate this operation. Design record keeping procedures for ministries and provincial authorities, including a determination o f who prepares procurement documentation and where the documents are filed. iv. Establish a directorate at the Ministry o f Finance to undertake procurement policy formulation and procurement analysis, because a w e l l functioning procurement system requires that an authority must be in charge with both procurement policy formulation and “maintenance” o f the procurement framework. 0 Provide for adequate funding, staffing, technical assistance and training to undertake the directorate’s regulatory and administrative functions. v. Establish sound controls (audits) and accountability. I t i s expected that in the long-term: (i) internal controls by the National Inspectorate o f Finance will be fully operational; and (ii) external controls by the Tribunal o f Accounts will also be operational. The key next steps are to : 0 Nominate the President and members o f the High Authority Against Corruption and provide i t w i t h an adequate budget to fulfill i t s mandate; companies), although this i s the procurement method that guarantees economy, efficiency and transparency, in most circumstances. In addition, this method is mandated by Decree 7/96 for larger tenders. Based on experience in other countries, Government could save as m u c h as 20 percent through competitive public procurement. ‘ O s The Mauritius procurement filing system could be used as a n example. - 105 - 0 Remind the target group o f the approaching deadline for production o f audited accounts under resolution 6/01, with copy o f the resolution attached; 0 Prepare a work program for the Tribunal o f Accounts and include the audit o f some pilot ministries in first year; and 0 Provide for legislation prohibiting reappointment of civil servants guilty o f corruption. C. THE WAY FORWARD Three possible models could be adopted for the reorganization o f procurement in Angola. After the govemment has presented its views o n these options, or decided upon another model, the legislation should be drafted taking into account the selected reorganization model. The stated intent to decentralize procurement should be preceded by a study o f local procurement capacity in the provinces, and by an effort to organize such capacity. The table below explains in more detail the three options for the re- organization o f the procurement system. ~ Option 1 Option 2 ODtion 3 Responsibilities: Strengthen procurement Successive creation over Creation of planning o f procurement, capacity within each a period o f three years o f independent contract recruitment of ministry and province, ten free-standing starting with a group o f procurement management management independent consultants to prepare bid documents three - four (3-4) o n a units, six in the provinces agencies, probably and supervise execution, pilot basis. Staff i s paid (two each in north, center ten (10). Staff - not launching of bids, normal civil service and south) and four in civil servants - opening, evaluation and salaries. Supported by Luanda. Staff- civil recruited through signature o f contracts, technical assistance paid servants within the by the government, and M i n i s t r y of Finance - competition. monitoring o f execution, certification o f payments, work only for the recruited through Operations o f these keeping records. ministry or province competition by that agencies and some where they are usually Ministry and paid market- technical assistance employed. oriented salaries f r o m the initially subsidized government budget, Supported by technical by govemment, later assistance paid by financed by fees government. generated by the agencies for their services to client institutions in the public and private sector o n a contract basis. Oversight o f the system, Ministry o f Finance M i n i s t r y o f Finance Ministry o f Finance data analysis and statistics Complaints by and Procuring entity - Courts Procuring entity - Courts Procuring entity - Courts appeal to Audit (a posteriori) Tribunal of Accounts Tribunal o f Accounts Tribunal ofAccounts - 106 - Option 1, i s the option that would be preferred in terms o f sustainability. The expected outcome i s that after a period o f technical assistance support and the expansion o f the system to all the Government departments and provinces, Angola would become self- sufficient and would have the means to carry out i t s procurement processing effectively. Examples o f countries where Option 1 i s practiced are Mozambique, Guinea and Ethiopia. However, this approach can work only if the pilot is expanded to all Government entities, the legal framework, including setting o f oversight and appeal mechanisms and the institutional capacity are established, and the staff in charge o f procurement is organized as a cadre o f specialists with adequate resources to perform their tasks. In other words, this approach requires a Government that i s h l l y committed to procurement reform, and a c i v i l service that fulfills i t s duties appropriately and has the means to do so. Option 2, is the one to use in case progress regarding the setting o f the legal and institutional framework lags behind schedule or proves to be difficult to achieve in a timely fashion. This option will, in the meantime, enable Angola to develop a procurement system that would somehow guarantee that public resources are used as efficiently as possible. Although more expensive than Option 1, because i t includes a technical assistance component, this option can be implemented rather quickly. There could be, however, a problem o f sustainability with Option 2, in that i t may not be possible to pay selected c i v i l service staff above the usual scale o f salaries o n a continuing basis. Option 3, could be considered as the continuation o f or a solution to Option 2, in the sense that members o f the c i v i l service hired on a contractual basis could leave and create self-standing procurement agencies. With this option the Government i s assured o f a high standard quality service provided by independent and accountable companies with clear standards o f ethics; and oversight (through ex-post) performed by independent contract management agencies. Although results are s t i l l being assessed, recent examples o f countries where Option 3 i s practiced are Senegal and Mali. Given the scarcity o f qualified personnel in that area, the problem with this Option i s whether there will be sufficient human resources available in the country to staff these companies. The other issue i s that members o f the c i v i l service could feel ostracized and become reluctant to collaborate with them. D. PROGRESS AS OF MARCH 2004 As o f March 16, 2004, the recommendations (i);(ii); and (iv) mentioned above were successfully implemented or are well advanced. i. Identify a champion to spearhead the Procurement reform process and organize a high-levelworkshop for the Council o f Ministers to raise ministers and their procurement staff awareness about the benefits for the Government o f a more open biding environment. - 107 - J The taskforce or “Grupo de trabalhos” was created by despacho No. 35/03 on April 22, 2003. This taskforce is composed o f the Director o f Patrimonio do Estado (coordinator o f the taskforce); two representatives o f the Ministry o f Finance o f whom one i s the deputy coordinator o f the taskforce; one representative from the Ministry o f Planning ( Ministerio do Planeamento); one representative o f the Ministry o f Public Works ( Ministen0 das Obras Publicas); and two national consultants. The taskforce i s responsible for: (i) coordinating and spearheading the procurement reform process; (ii) make technical recommendations on the existing reports (namely, the Country Procurement Assessment Report- CPAR and the report financed by the Norwegian Agency for Development); (iii)preparing adequate materials (decree and legal documents) for implementingthe procurement reform; (iv) preparing meetings with other partners to discuss the ongoing process; (v) disseminate information to general public about the objectives and results expected o f the reform; (vi) prepare a capacity building program necessary to successfully implement the reform; and (vii) presenting twice per month a progress report to Ministry o f Finance or relevant entities. J The taskforce held i t s first meeting on October 8, 2003 and since then, held working sessions o n a regular basis (at least twice per week). The taskforce mainly (i) discussed the option to be selected for reforming procurement; ( ii)prepared various encounters with other Ministries and representatives o f the private sector, and in particular one which took place o n November, 16 2003; ( iii) prepared the workshop to be held in 2004 which represents the keystone o f the reform process and it i s meant to validate the procurement reform process and action plan. J On November 24, 2003 and parallel to the procurement reform, the Bank’s Procurement team held a procurement seminar o n WB procurement m a i n principles and a clinic on World Bank procurement procedures. Representatives from the private sector and from Ministries (about 20 persons) attended the WB’s presentation which was composed o f two back-to-back presentations: (i) the first presentation was a procurement seminar that meant to present procurement main principles to the private community; and (ii)the second presentation was intended for managers working o n W o r l d Bank projects mainly, the Economic Management Technical Assistance (EMTA), F A S (Fundo de Apoio Social, and Demobilization Program. I t took the form o f a procurement clinic, which was meant to address procurement issues that projects managers m a y encounter o n a daily basis. Also in parallel and in December, the C P A R team kept informed the Norwegian Agency for Development (NORAD) on the status o f the procurement reform (NORAD financed a report on procurement delivered in 2002). ii. Create a comprehensive procurement framework. Set up a national working group composed o f national c~nsultants’~~, lawyers, representatives of the public logThe Bank has already established a dialogue and close partnership with the consultants financed by the Norwegian Agency for Development (NORAD). The mission held working sessions with the consultant team and benefited f r o m the wealth o f information o f its work. Some o f the mission current findings and - 108 - sector (especially the Ministries of Planning, Finance and Public Works), national experts and private sector representatives, to assist an international legal consultant in drafting o f the new procurement legislation”’. 4 From October to December 2003, the taskforce appointed to coordinate the procurement process discuss the different institutional arrangements for procurement. The outcome o f these discussions i s critical, as it will enable the taskforce to draft a new procurement l a w based on the option selected. 4 The institutional model selected by the taskforce is closer to option 1 proposed and described above, i.e., strengthen procurement capacity within procuring entity (mainly, Ministries and provinces). The “sectorial” (for instance, Ministries o f Health, Transportation, public Works) and “local” entities will be, responsible for: (i) preparing the bids, with technical assistance o f consulting firms, if necessary, and in particular for the preparation o f the bidding documents and terms o f reference; (ii) creating evaluation commission when needed; (iii) signing contracts in l i n e with the prevailing rules; and (iv) providing all relevant information about bids to the regulatory body. iv. Establish a directorate at the Ministry o f Finance to undertake procurement policy formulation and procurement analysis. 4 The new procurement L a w will provide for the creation o f a Regulatory body called “Orgao de Regulaczo e Apoio”. So far, there is n o independent institution in charge o f formulating procurement policy. A well functioning procurement system needs an authority in charge with both procurement policy formulation and “maintenance” o f the procurement framework, and to enforce proper application o f the existing and forthcoming procurement legislation. This Orgdo de Regulaciio e Apoio will have several duties and mainly, (i) to approuve rules, methodes for procurement; ( ii) to meet regularly with the procuring entities at the central and decentralized levels; (iii) to gather, centralize and disseminate information and data ob procurement and procurement procedures; (iv) to support and advise procuring entities and foster training programs; (v) to work in close partnership with the SIPIP (Sistema de Informaqzo do Programa de Investimentos Publicos); (vi) disseminate relevant information about bidders and more generally private companies ( mainly, capacity and price); and (vii) to solve procurement conflict as the Orgdo de Regulaciio e Apoio will also play the role o f a dispute resolution entity. recommendations are based o n this information. As t o the aspects related to the legal framework and regulatory setting, the mission agreed with the consultant that the w o r k accomplished so far should b e completed with the assistance o f an international specialist in procurement law. ‘lo Adopt the open competitive bidding as the primary method o f procurement. Open competitive bidding i s rarely used as a procurement method (most often, bidding i s limited t o obtaining three bids f r o m local companies), although this i s the procurement method that guarantees economy, efficiency and transparency, in most circumstances. In addition, this method is mandated by Decree 7/96 for larger tenders. Based o n experience in other countries, Government could save as m u c h as 20 percent through competitive public procurement. - 109 - E. LATEST DEVELOPMENTS AND EXPECTED IMMEDIATE NEXT STEPS b Document for the Minister o f Finance wasprepared The taskforce produced a short document for the Minister o f Finance. This document i s based o n the CPAR and NORAD financed report and present the weaknesses o f the prevailing system and the actions to be taken to improve the current procurement system [see Annex 11. The document focus on the five pillars determined by the CPAR, Le., (i) the legal and regulatory framework; (ii) the institutional framework; (iii) the procedures and practices; (iv) the capacity; (v) the audit, control and appeals measures, and the anti-corruption measures. b Approval by the GOAand transmittal to the World Bank The document produced by the taskforce was delivered to HE. Minister o f Finance in order to be approved [see Annex 21. The document along with a letter from the taskforce was then transmitted to the World Bank on February 25,2004. b Terms o f reference for consultants to be drafted Terms o f reference for intemational lawyer, national lawyer, procurement specialist and procurement institutional reform specialist should be prepared. bDrafting of the new procurement Law, regulations bidding documents and user’s manual Staring beginning 2004 The plan o f work i s organized in t w o phases, in accordance with the terms o f reference, with Phases Iand I1overlapping to a considerable degree. 0 In phase I,Dra. Theresinha Lopes, the national procurement lawyer, and the intemational lawyer to be selected, in consultation with the procurement reform taskforce, will prepare a new public procurement law and comprehensive set o f new procurement regulations to implement the Angola public procurement law. 0 In phase 1 1, the technical procurement specialist, to be selected, will develop the standard bidding documents, which will be used by procuring entities pursuant to these regulations and user’s manual. 0 Also in phase 1 1, the procurement institutional reform specialist will prepare the intemal regulations; financing needed for the creation o f this regulatory body; and terms o f reference for the newly created Orgho de Regulae60 e Apoio. b Workshop to be held in 2004 to validate GOA procurement reform program This full day workshop to be chaired by H.E. the Minister o f Finance w i l l endeavored to discuss the documents prepared. To raise awareness and build consensus among people involved in procurement, several groups will be formed to discuss the procurement action - 110- plan more thoroughly. A final document and action plan will be prepared upon completion o f the workshop. c 4 0 3 - 0 U I B m 0 s- XO 3 6 h h I 3 3 3 I I 0 I $1 r3 I m m - 114- ANNEX111 THE OIL SECTOR FINANCIAL MODEL”’ The Angolan o i l sector, or “aggregate” financial model i s a bespoke policy tool devised specifically for the government o f Angola to assist in managing petroleum revenue flows, and, in an economy so dependent on oil, to assist in the preparation and control o f national budgets. The financial model i s intended as the focal point o f a revenue management system that i s able to monitor, control and reconcile the flow o f revenues into the treasury in a transparent and accountable manner. The revenue management system will be the responsibility o f a dedicated agency resourced by professionals who will employ secure and efficient procedures to generate data and reports, and who will communicate w i t h industry and government at a senior level. 1. The aggregate financial model is structured around a database populated with technical o i l field data and fiscal information provided by foreign o i l companies and Sonangol. The database forms the key input to a series o f pre-tax and post-tax cashflow models that derive the final projections o f government tax take, contractor and operator net cashflows and Sonangols share o f net cashflows. The post-tax calculations include all o f the royalties, taxes and profit o i l payable by investors according to the concession agreements and operating production sharing agreements signed with the government to 2000 (see schematic below). Prepared by David Reading and Peter Macnab (AUPEC). - 115 - Aggregate Financial Model I I I I I ECONOMIC FISCAL I I TAX I ‘OMPAN’ RECONClLAlTlON I RESULTS INPUTS INPUTS I CALCULATIONS I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I I The financial model has been constructed in a spreadsheet format, using Microsoft Excel, so that all calculations remain visible and easily accessible to potential users. Given the capacity constraints identified within government ministries, and the additional training required by users to understand, execute and manage the financial model, i t was not considered appropriate to create an alternative ‘black box’ version o f the model at this stage. The main input parameters to the model are highlighted below. O i l prices Production and Cost profiles Fiscal instruments (tax rates, bonuses, profit sharing tiers, royalties) Cost recovery limits Depreciation schedules Equity (Sonangol and IOC’s) Sonangols carry through the exploration phase Outputs generated by the financial model are too extensive to review in detail here. The quarterly projection reports provide a wealth o f financial output disaggregated by operator, partner, and block, including Sonangol. Aggregate results are also extensive and are summarised in each quarterly projection report. The main summary outputs shown at an aggregate l e v e l in the quarterly projection reports are defined below. - 116 - Total Government T a k e Total Government Take. This comprises all taxes, royalties and bonuses payable, plus 100% o f all profit oil, plus Sonangols post-tax net cashflow (that i s Sonangol Holding or Sonangol P&P’s equity share in any field). Total Government Take is shown to highlight the total govemment take that accrues to the nation as a result o f Angolan o i l activity. Total Government Tax Take Total Government Tax Take i s essentially the Total Government Take less Sonangol’s post-tax net cashflow. Total Government Tax Take i s equal to the total revenue that we believe the govemment should expect to receive from o i l operations. This figure comprises all field taxes and royalties due from all o f the field partners (including Sonangol Holding and Sonangol P&P’s share o f royalties and taxes due to the government if they have an equity share). The figure also includes all field profit o i l due to the government under the PSA regime (governments share o f profit o i l i s administered by Sonangol Holding who lift and sell the o i l o n the governments behalf through i t s trading arm, and remit the sales value to the government less the administration fee equal to 10% o f the fields profit oil, which i s retained by Sonangol Holding). The figure also includes signature and production bonuses and any non-fiscal payments that companies are contractually obliged to pay to the government. Sonangols N e t Cashflow Sonangol’s net cash f l o w i s equal to the share o f post-tax project net cashflow that Sonangol receives as an equity partner in several blocks. Sonangol Holding i s an equity partner in onshore blocks FSRST and Block 0. Sonangol P&P i s an equity partner in blocks 2,3, and 14 and operator o f Block 4 (Kiabo) and block 34. Contractors N e t Cashflow The Contractors net cashflow shows how much wealth the operating groups (comprising all o f the partners but excluding Sonangol) generate from Angolan o i l activity. Royalties Royalties comprise all field royalties due to the government (from all partners including Sonangol). Petroleum Income T a x (PIT) Petroleum Income Tax comprises all field P I T due to the government (from all partners including Sonangol). - 117- Petroleum Transactions Tax (PTT) Petroleum Transaction Tax comprises all field P T T due to the government (fi-om all partners including Sonangol). PTT is only levied on fields in Block 0. Profit Oil Profit O i l comprises all field Profit Oil due to the government fi-om i t s share o f production. Using the Aggregate Financial model The financial model was developed to achieve the following four main objectives To project petroleum revenues and related taxes to the state from o i l activity in Angola To provide a tool to assist the government in monitoring and managing petroleum revenues To provide 5 year financial projections o f petroleum operations o n a quarterly and annual basis To assist the government in reconciling the financial flows between Sonangol, the Bank o f Angola and the Ministry o f Finance. There are a number o f areas in which the government can immediately employ the financial model. One important area is in monitoring loan repayments backed by oil. The financial model can be used to execute altemative price scenarios, that at an aggregate level will provide estimates o f break-even prices, below which i t would be difficult for Angola t o service all o f i t s o i l backed loan commitments. Knowing this would be important in ensuring that future repayments could be met, and that the country did not over-extend i t s e l f in securing loans backed by oil. Initial use o f the model has already focused o n a number o f key areas described below. I t i s our recommendation that hrther work should be progressed by the government in each o f these. The financial model can be used as an effective planning and reporting tool. The model i s a dynamic instrument that should be regularly updated with both current and future production, price, cost and tax information. By employing scenario analysis users will be able to forecast altemative price scenarios, production and cost forecasts and measure their impact on future government revenues. This will be critical in preparing, annually - 118- the National Budget, and developing future policy in key areas o f government spending. As an input to the policy decision making process the financial model will become an indispensable government tool. In addition to modeling price scenarios the government may wish to measure the financial impact o f cost over-runs o n specific projects, declining production trends or the phasing o f future investments by the industry. The financial model will be important in enabling the preparation o f an effective, and transparent reporting mechanism that shows o i l tax revenues, and supports inter-ministerial financial reconciliations. The financial model will be a central platform in the reconciliation process. The Tax Directorate o f the Ministry o f Finance has, over the years, constructed tax spreadsheets that are used to monitor the collection o f tax revenues and to reconcile taxes paid with the independent review o f taxes by third parties such as Ernst and Young. However the process is flawed in that the Tax Directorate i s unable to independently verify tax calculations using i t s own models, and cannot rely on the independent review, as this i s not a standardized audit. I t i s critical that the Tax Directorate develops i t s own expertise, with the model, to confidently project future tax revenues. With the introduction o f a transparent and co-ordinated reporting mechanism these outputs should then be compared with the actual tax flows due by, and received from the o i l companies o n an ongoing basis. The financial model will initially complement the spreadsheets being used by the Tax Directorate. Ultimately i t should be a govemment objective to replace the current flawed system with the financial model and new reporting mechanisms. The financial model will provide valuable guidance in enabling the government to effectively manage the revenues being generated by the petroleum sector, and in particular to verify o i l company’s tax declarations. Overall the model will contribute to improved cash management. The govemment will be closely involved in monitoring and guiding the future strategic direction o f the petroleum industry in Angola. The financial model will provide important information to assist the government in formulating policies relating to investment levels, the pace o f future exploration and development, the extent o f Sonangol equity involvement, the growth in the support and service industries, and o i l related employment. The financial model maybe employed as an important tool to influence macro-economic policy. The model can be modified to provide meaningful estimates o f the share o f the, petroleum industry in the National Budget, and to measure the impact o f the growth o f the o i l sector on the non-oil economy. I. PC Specification The size o f the financial model i s currently over 25 megabytes. This can b e compressed to approximately one third o f the original size. - 119- The most effective storage for the model i s o n a compact disc in a zipped directory. The model can b e downloaded onto a hard drive o r executed f r o m the CD (provided the CD drive o n the PC matches the C D drive stored o n the compact disc). To store and run the financial model the following PC specification i s recommended: At least 3 0 M G hard disk space for the models Personal o r multimedia computer with a pentium 2 processor Office 2000 Microsoft Windows@ 2000 operating system 64 MB memory CD-ROM drive Microsoft Mouse - 120 - BIBLIOGRAPHY Alvesson, M., S. Bhattarai e G. Pastor (2003). “Angola: Sources and Uses o f State O i l Revenue”. International Monetary Fund: Angola - Selected Issues Papers. July. Alvesson, M. e A. Torrez (2003). “Central Bank Operations and Macroeconomic Stabilization in Angola”. International Monetary Fund: Angola - Selected Issues Papers. July. Araujo, J. and H.G. Ramos (2002). “Angola: Public Finance Framework”. World Bank, mimeo, April. Araujo, J.; J.G. Gasha e G. Pastor (2003). “Fiscal Subsidies in Angola”. International Monetary Fund: Angola - Selected Issues Papers. July. Bamett, Steven and Ossowsky, Rolando, 2002, “Operational Aspects o f Fiscal Policy in Oil-Producing Countries” IMF Working Paper 02/177 (Washington: International Monetary Fund) DNC/MINFIN (2002). Relatdrio da Execup?o Orqamental e Financeira do OGE. Eifert, B., A. Gelb and N.B. Tallroth (2002). “The Political Economy o f Fiscal Policy and Economic Management in O i l Exporting Countries”. World Bank Policy Research Working Paper 2899, October Government o f Angola (2003). Estratkgia de Combate Ci Pobreza. Draft for Discussion. Hodges, Anthony (2001). Angola: From Afro-Stalinism to Petro-Diamond Capitalism. Lysaker (Norway), Oxford and Bloomington & Indianapolis: The Fridtjof Nansen Institute & The International African Institute, in association with James Currey & Indiana University Press. KPMG (2002 - 2003). Avaliacgo do Sector Petrolifero Angolano. Relat6rio Inicial (draft). Luanda. Intemational Monetary Fund (2002). Angola - Staff Report for the 2002 Article IV Consultation. Washington, D.C., March. International Monetary Fund (2003). Angola - Staff Report for the 2003 Article IV Consultation. Washington, D.C., July. MINFIN (2001). “Informaq5o sobre a Execuqgo do Programa Economico e Social do Govemo para 2001 - ISemestre”. Moreno, J.M; T. Prietto, and L.C. Nerosky (2001). Evaluation del Sistema de Gestion Financiera Pdblica en Angola: Propuesta de un Sistema de Gestion Financiera Publica - 121 - Inteczrado. Consultant report prepared for the Ministry o f Finance o f Angola and the European Union, July. Potter, B.H. and J. Diamond (1999). Guidelines for Public Expenditure Management. Intemational Monetary Fund. Schiavo-Campo, S. and D. Tomasi (1999). Managing Government Expenditure. Manila, Philippines: Asian Development Bank. Tollini, H.; W. Luiz ; C. Duarte ; J. Araujo ; and H. Ramos (2004). “Angola: Analise e Recomendaqaes na Area da ExecuqZio Orqamental”. Intemational Monetary Fund, Fiscal Affairs Department, January. UNDP/IOM/UNICEF/WHO (2002). Public Financing o f the Social Sectors in Angola. Luanda, August. World Bank (1993). Public Expenditure Issues and Priorities During the Transition to a Market Economy. Washington, D.C. World Bank (1997). Countw Profile o f Financial Accountabilitv (CFPA). Washington, D.C. World Bank (1998). Public Expenditure Management Handbook. Washington, D.C.. World Bank (2002). Country Procurement Assessment Report. Washington, D.C.