Document of The World Bank Report No: ICR00004448 IMPLEMENTATION COMPLETION AND RESULTS REPORT (IBRD-No. 93732-AO) ON A LOAN IN THE AMOUNT OF US$450 MILLION AND A POLICY-BASED GUARANTEE IN THE AMOUNT OF US$200 MILLION TO THE REPUBLIC OF ANGOLA FOR THE ANGOLA FIRST FISCAL MANAGEMENT DEVELOPMENT POLICY FINANCING June 28, 2018 Macroeconomics, Trade and Investment Global Practice Cameroon, Gabon, Equatorial Guinea, Angola and São Tomé and Príncipe CMU Africa Region CURRENCY EQUIVALENTS (Exchange Rate Effective June 12, 2018) Currency Unit = Angolan kwanza (AOA) US$1 = AOA 238.3 FISCAL YEAR January 1 - December 31 ABBREVIATIONS AND ACRONYMS AGT General Tax Administration (Administração Geral Tributária) CPS Country Partnership Strategy DPF Development Policy Financing DPL Development Policy Loan GoA Government of Angola GDP Gross Domestic Product ICR Implementation Completion and Results Report IMF International Monetary Fund INE National Statistical Institute (Instituto Nacional de Estatísticas) M&E Monitoring and Evaluation MoF Ministry of Finance MPDT Ministry of Planning and Territorial Development NDPI National Directorate for Public Investment OPDA Oil Price Differential Account PBG Policy-Based Guarantee PDO Project Development Objective PIM Public Investment Management PIP Public Investment Program SIGFE Integrated Financial Management System (Sistema Integrado de Gestão Financeira do Estado) SIPIP Integrated Public Investment Management System (Sistema Integrado do Programa de Investimento Público) VAT Value Added Tax Regional Vice President Makhtar Diop Senior Global Practice Director: Carlos Felipe Jaramillo Country Director: Elisabeth Huybens Practice Manager: Francisco Galrão Carneiro Project Team Leader: Souleymane Coulibally ICR Team Leader: Rafael Chelles Barroso ICR Author: Eduardo Somensatto The World Bank (P153243) REPUBLIC OF ANGOLA IMPLEMENTATION COMPLETION AND RESULTS REPORT ON THE ANGOLA FIRST FISCAL MANAGEMENT DEVELOPMENT POLICY FINANCING TABLE OF CONTENTS DATA SHEET A. Basic Information B. Key Dates C. Ratings Summary D. Sector and Theme Codes E. Bank Staff F. Results Framework Analysis G. Ratings of Project Performance in ISRs H. Restructuring 1. PROJECT CONTEXT, DEVELOPMENT OBJECTIVES AND DESIGN .....................................................1 2. KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOMES ...................................................5 3. ASSESSMENT OF OUTCOMES.................................................................................................... 11 4. ASSESSMENT OF RISK TO DEVELOPMENT OUTCOME ................................................................ 18 5. ASSESSMENT OF BANK AND BORROWER PERFORMANCE ......................................................... 19 6. LESSONS LEARNED ................................................................................................................... 21 7. COMMENTS ON ISSUES RAISED BY BORROWER/IMPLEMENTING AGENCIES/PARTNERS ............ 23 ANNEX 1: BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION PROCESSES .................... 24 ANNEX 2: BENEFICIARY SURVEY RESULTS ......................................................................................... 26 ANNEX 3: STAKEHOLDER WORKSHOP REPORT AND RESULTS ............................................................ 27 ANNEX 4: SUMMARY OF BORROWER'S ICR AND/OR COMMENTS ON DRAFT ICR ............................... 28 ANNEX 5: COMMENTS OF COFINANCIERS AND OTHER PARTNERS/STAKEHOLDERS............................ 30 ANNEX 6: LIST OF SUPPORTING DOCUMENTS ................................................................................... 31 MAP ................................................................................................................................................ 32 The World Bank (P153243) A. BASIC INFORMATION Program 1 AO-Fiscal Country Angola Program Name: Management Programmatic DPL Program ID: P153243 L/C/TF Number(s) IBRD-85260 ICR Date: 05/31/2018 ICR Type: Core ICR GOVERNMENT OF Financing Instrument: DPL Borrower ANGOLA Original Total US$ 450.00 million Disbursed Amount US$ 450.00 million Commitment Implementing Agencies: Cofinanciers and Other External Partners: B. KEY DATES AO-Fiscal Management Programmatic DPL P153243 Revised / Actual Process Date Process Original Date Date(s) Concept Review: 12/15/2014 Effectiveness: 10/06/2015 12/31/2016 Appraisal: 03/30/2015 Restructuring(s): Approval: 06/30/2015 Mid-term Review: Closing: 12/31/2016 12/31/2016 C. RATINGS SUMMARY C.1 Performance Rating by ICR Overall Program Rating Outcomes Unsatisfactory Risk to Development Outcome Significant Bank Performance Moderately Unsatisfactory Borrower Performance Moderately Unsatisfactory C.2 Detailed Ratings of Bank and Borrower Performance (by ICR) Overall Program Rating Bank Ratings Borrower Ratings Moderately Moderately Quality at Entry Government: Unsatisfactory Unsatisfactory Quality of Moderately Implementing Moderately Supervision: Unsatisfactory Agency/Agencies: Unsatisfactory Overall Bank Moderately Overall Borrower Moderately Performance Unsatisfactory Performance Unsatisfactory C.3 Quality at Entry and Implementation Performance Indicators AO-Fiscal Management Programmatic DPL P153243 i The World Bank (P153243) Implementation QAG Assessments (if Indicators Rating Performance any) Potential Problem Quality at Entry Program at any time No None (QEA) (Yes/No): Problem Program at Quality of No None any time (Yes/No): Supervision (QSA) DO rating before Closing/Inactive status D. SECTOR AND THEME CODES AO-Fiscal Management Programmatic DPL P153243 Original Actual Major Sector Public Administration Other Public Administration 44 44 Central Government (Central Agencies) 56 56 Major Theme/Theme/Sub Theme Economic Policy 23 23 Fiscal Policy 17 17 Fiscal sustainability 2 2 Public Expenditure Policy 2 2 Tax policy 13 13 Macro-financial policies 6 6 External Finance 2 2 Macroeconomic Resilience 2 2 Monetary and Credit Policies 2 2 Private Sector Development 3 3 Jobs 3 3 Job Creation 3 3 Public Sector Management 64 64 Data Development and Capacity Building 3 3 Data production, accessibility and use 3 3 Public Administration 25 25 Transparency, Accountability and Good 25 25 Governance Public Finance Management 36 36 Domestic Revenue Administration 11 11 Public Expenditure Management 25 25 Social Development and Protection 12 12 Social Inclusion 3 3 Other Excluded Groups 3 3 Social Protection 9 9 ii The World Bank (P153243) Social Safety Nets 9 9 E. BANK STAFF AO-Fiscal Management Programmatic DPL P153243 Positions At ICR At Approval Vice President: Makhtar Diop Makhtar Diop Country Director: Elisabeth Huybens Gregor Binkert Practice Francisco Carneiro Mark Roland Thomas Manager/Manager: Task Team Leader: Souleymane Coulibaly Souleymane Coulibaly ICR Team Leader: Rafael Chelles Barroso ICR Primary Author: Eduardo Somensatto F. RESULTS FRAMEWORK ANALYSIS Program Development Objectives (from Program Document) The Program Development Objective is to strengthen the country’s fiscal management to create the fiscal space needed to better protect the poor and vulnerable Revised Program Development Objectives (as approved by original approving authority) Indicator(s) AO-Fiscal Management Programmatic DPL P153243 Original Target Formally Actual Value Achieved Values (from Indicator Baseline Value Revised at Completion or approval Target Values Target Years documents) Indicator 1: Fiscal rules linking the Oil Price Differential Account to the budget cycle are in place by December 31, 2017. Value quantitative or No Yes Qualitative Date achieved 2013 2017 Comments (incl. % Achieved. Rules approved, but were limited. Did not fully specify the achievement) procedures governing the Fund as anticipated. Indicator 2: Non-oil tax revenue as a share of non-oil GDP increases from 13.4% in 2013 to 15.6% by December 31, 2017. Value quantitative or 13.4 percent of 15.6 percent 10.8 percent of non-oil Qualitative non-oil GDP of non-oil GDP GDP Date achieved 2013 2017 2017 Comments (incl. % Not Achieved. Economic crisis reduced tax revenues achievement) Indicator 3: 100 % of new projects in the 2017 Public Investment Plan are subjected to the new Public Investment Management cycle, systems and regulations. iii The World Bank (P153243) Value quantitative or No No Qualitative Date achieved 2013 2017 Comments Not Achieved. Indicator 4: Fuel price subsidies fall from 5.2% of GDP in 2013 to 2.3% by December 31, 2017. Value quantitative or 5.2 percent of 2.3 percent of 1.3 percent of GDP Qualitative GDP GDP Date achieved 2013 2017 2017 Comments (incl. % Achieved. achievement) Indicator 5: The percentage of households protected from negative income shocks via direct cash transfers grows from less than 3% of the population in 2013 to 7% by December 31, 2017. Value quantitative or 3 percent of 7 percent of 1.4 percent of Qualitative population population population Date achieved 2014 2017 2017 Comments (incl. % Not Achieved. achievement) Indicator 6: By December 31, 2017, empirical results from the analysis of the Inquérito de Despesas e Receitas e Emprego da População-2015 (IDRE-2015) household survey appear in one or more government official documents proposing policy actions. Value Qualitative No Yes No Date achieved 2013 2017 2017 Comments (incl. % Not Achieved. achievement) G. RATINGS OF PROJECT PERFORMANCE IN ISRs H. RESTRUCTURING (IF ANY) iv The World Bank (P153243) 1. PROJECT CONTEXT, DEVELOPMENT OBJECTIVES AND DESIGN 1. This Implementation Completion and Results Report (ICR) assesses the results achieved under Angola’s First Fiscal Management Development Policy Financing (DPF). The DPF was composed initially of two Development Policy Loans (DPLs) and one Policy Based Guarantee (PBG). The first DPL and the PBG were approved in June 2015, with loan disbursement in October 2015. This was the first ever DPL and IBRD loan to Angola, and thus was rated high risk. The Government of Angola (GoA) took longer to use the PBG since it was focused on its first Eurobond issuance in late 2015 and its ambitious plan of raising US$1 billion from a US$200 million guarantee. The PBG was extended until June 2017, but the deteriorating macroeconomic environment with an inadequate policy response from the government led to the cancellation of the PBG and the second DPL, since World Bank policy requires countries to have an adequate macroeconomic policy at the approval of DPLs and usage of PBGs. The Bank, however, remained engaged through Investment Project Financing operations and a strong analytical program to prepare for an eventual reengagement after the elections. 2. The ICR covers the whole program and is based on the accomplishments toward the six results indicators identified in the DPF policy matrix. As of December 2017, four of the six results indicators were only partially achieved, while the other two were formally met, however the indicator on subsidies showed some retrocession as a result of partial policy reversal. The overall rating for the operation is deemed as unsatisfactory. Progress has been less than expected due to: (i) implementation weaknesses brought about by the ambition of some reforms, poor government capacity, and insufficient technical assistance from the Bank side; (ii) declining government support to the reforms; and (iii) the challenges of managing a prolonged fiscal crisis. Following recent improvements in economic management, with progress made on all areas of the previous DPF, the Bank is now preparing a new programmatic DPF series. 1.1 Context at Appraisal 3. The DPF supported the government’s effort to deal with a major fiscal crisis, following the sharp drop in the price of oil in 2014. The Bank was already re-engaging with the government on policy based discussions, when the country suffered a considerable external shock. As the largest sector of the economy, petroleum accounted for essentially all foreign exchange earnings and financed nearly 80 percent of government’s budget up to 2013. 4. Falling oil prices impacted gross domestic product (GDP) growth, fiscal, and external accounts. GDP growth declined from 4.1 percent in 2014 to 0.9 percent in 2015, compared with a 6.6 percent growth rate assumed in the revised 2015 budget. The projected budget deficit in 2015 was 18 percentage points of GDP higher than in 2013. A similar shift occurred in the current account, which shifted from a surplus of 6 percent of GDP in 2013 to a deficit of 8.9 percent of GDP in 2015. The external financing needs of the country surged, and despite comfortable foreign exchange reserves, the government had to rely on external financing, despite some initial currency devaluation. 5. The government responded to the crisis by introducing a set of adjustment measures, including a large fiscal contraction. At the time of the operation’s appraisal, the government introduced a revised 2015 budget that proposed sharp reductions in expenditures. There were cuts on public investments, purchase of goods and services, and fuel subsidies, but also measures to increase non-oil revenues. The government planned to expand social protection transfers to mitigate the impact of the adjustment Page 1 of 32 The World Bank (P153243) measures on the most vulnerable. The Bank team assessed the macroeconomic framework at the time of appraisal as adequate for a DPL based on the consistency of the policy responses laid out by the government. The proposed programmatic DPF was consistent with the third component of the Country Partnership Strategy (CPS), which was to help form adequate human and institutional capacity for sound economic management. 6. The Bank engaged with the GoA in a DPF to smooth the fiscal consolidation required by the falling oil prices, to support fiscal reforms aimed at improving macroeconomic management in the medium- term, and ultimately to protect the most vulnerable population from the costs of the adjustment. Table 1. Key Macroeconomic Indicators Leading up to Appraisal 2012 2013 2014 2015 GDP Growth (real, %) 8.6 5.0 4.0 0.9 Current Revenues (% of GDP) 37.0 39.1 35.5 21.3 Current Expenditures (% of GDP) 23.9 30.9 31.7 23.4 Public Investment (% of GDP) 6.9 12.7 15.3 4.6 Budget Deficit (% of GDP) 4.1 0.7 −6.6 −3.2 Inflation (Luanda CPI, %) 9.0 7.7 7.5 14.3 Current Account (% of GDP) 10.8 6.0 −2.6 −8.9 1.2 Original Project Development Objectives (PDO) and Key Indicators 7. The Program Development Objective was to strengthen the country’s fiscal management to create the fiscal space needed to better protect the poor and vulnerable. There were six results indicators (RI), which are reported below with their respective baseline and target.  RI#1 was to have “Fiscal rules linking the Oil Price Differential Account to the budget cycle in place by December 31, 2017�.  RI#2 was to increase “Non-oil tax revenue as a share of non-oil GDP from 13.4 percent in 2013 to 15.6 percent by 2017�.  RI#3 was to have “100 percent of new projects in the 2017 Public Investment Program (PIP) be subjected to the new Public Investment Management (PIM) cycle, systems, and regulations�.  RI#4 anticipated a reduction of “Fuel price subsidies from 5.2 percent in 2013 to 2.3 percent by December 2017�.  RI#5 aimed at increasing “The percentage of households protected from negative income shocks, via direct cash transfers, from 3 percent of the population to 7 percent by December 2017�.  RI#6 established that “By December 31, 2017, empirical results from the analysis of the IDRE- 2015 household survey were to be published in one or more government official documents proposing policy actions�. 1.3 Revised PDO and Key Indicators, and Reasons/justification 8. There were no revisions to the program development objectives and key indicators. Page 2 of 32 The World Bank (P153243) 1.4 Original Policy Areas Supported by the Program 9. The operation covered three main policy areas or pillars: (i) Introducing fiscal rules for the utilization of oil funds and modernizing tax administration and tax policy to increase non-oil revenue collection; (ii) Increasing value for money by enhancing the efficiency and efficacy of public investment management; and (iii) moving from untargeted fuel price subsidies to targeted cash-transfers and evidence based policy-making. Policy Area 1 - Introducing fiscal rules for the utilization of oil funds and modernizing tax administration and tax policy to increase non-oil revenue collection 10. This policy area supported efforts to increase non-oil tax revenue in line with the government program to reduce the dependence on oil revenue and promote the fiscal consolidation foreseen in the revised 2015 budget and beyond. The program supported the merger of the three tax agencies in one single revenue authority (General Tax Administration, AGT) and the revision of key tax laws in the first DPL, while the second DPL foresaw support to the implementation of AGT and the new tax laws. The request to be specific on the tax reforms supported came from the Operations Committee meeting. 11. This policy area also foresaw support, through the second DPL, to the government’s efforts to revise its main oil-funded fiscal stabilization fund – the Oil Price Differential Account (OPDA) – to include more transparent, formal and automatic rules for inflows and outflows in and out the fund. Policy Area 2 - Increasing value for money by enhancing the efficiency and efficacy of public investment management 12. Angola has had a large Public Investment Program (PIP) due to the abundance of oil resources and the need to rebuild the country after the civil war. Public investment efficiency was not very high as acknowledged by the government. The falling oil prices made the government cut the PIP budget significantly, which raised the urgency to improve the efficiency of the PIP. 13. The first phase of the program, supported by the DPF, concentrated on three government initiatives. The first initiative was a clean-up of the PIP to eliminate projects that were not ready for approval, but safeguarding social projects. The second initiative focused on improving the monitoring of PIP through the introduction of indicators and the adoption of a manual specific to riskier and high-profile projects. The last initiative was to begin the integration of the information systems of the budget (Integrated Financial Management System, SIGFE) and the PIP (Integrated Public Investment Management System, SIPIP) by creating a joint committee and implementing specific measures. The second phase, to be supported by the second DPL, included the revision of the PIM cycle with supporting regulations and manuals approved and the actual connection between SIPIP and SIGFE. Policy Area 3 – Moving from untargeted fuel price subsidies to targeted cash-transfers and evidence based policy-making 14. The actions in this policy area built on three key policy initiatives that were already underway: the gradual reduction of fuel subsidies through increases of domestic prices; the expansion of the nascent cash transfer program to the poor; and the creation of conditions and tools for evidence-based policies. Page 3 of 32 The World Bank (P153243) 15. Angola had for long provided fuel subsidies to its citizens by pricing domestic fuel below international prices. The size of the subsidy bill in the budget averaged 6 percent of GDP between 2010 and 2014. The program supported a phased approach to the removal of subsidies with an increase in domestic fuel prices in the first DPL, and the complete removal of most subsidies and introduction of an automatic fuel price adjustments through the second DPL. 16. The program also supported compensatory measures on social protection. It relied on the transfer program, Cartão Kikuia, that provides cash (in the form of a bank card) to vulnerable families for the purchase of essential goods, such as food, school supplies, and agricultural inputs, in stores belonging to the program. The first DPL supported its expansion to 50,000 families, while the second DPL aimed to support its extension to an additional 90,000 families. The second DPL also intended to consolidate all beneficiaries of social protection programs in a single register as a first step to program rationalization. 17. The third component focused on enabling evidence-based policy by improving the statistical information system. The first DPL supported Angola to improve its low statistical capacity through the approval of the National Statistical Strategy 2015-25 by the Council of Ministers. The second DPL was supposed to support the country in carrying out the household survey and a poverty map to improve targeting of the Cartão Kikuia, a measure specifically asked by the Operations Committee meeting. 1.5 Revised Policy Areas 18. There were no revisions of the policy areas during the program. 1.6 Other Significant Changes 19. The most significant development was the decision by the World Bank not to proceed with the second operation under the programmatic approach and to cancel the PBG due to the inadequacy of the macroeconomic framework. A series of supervision missions for the first operation, which also served as preparation of the second operation, monitored the implementation of the program and the adequacy of the macroeconomic framework. During loan preparation, it was clear that Angola was undergoing a crisis caused by falling oil prices that would affect external and internal equilibrium, although the duration and impact of the crisis proved to be longer than anticipated. 20. The macroeconomic policy response was adequate at first, but gradually changed course throughout the preparation of the second operation. The policy response from the government throughout the preparation of the first DPL was geared to reestablishing adequate macroeconomic framework, according to the team’s professional judgment seconded by the Bank internal review process, after several requests to strengthen the macroeconomic section and to provide additional information. The government approved a revised 2015 budget promoting large expenditure cuts, raised fuel prices, devalued the currency, and raised interest rates. After the approval and disbursement of the first DPL, the government proceeded with the reforms. Fuel prices were raised on January 1st, 2016 and fuel subsidies were virtually eliminated. The exchange rate was devalued by 15 percent in the last day of 2015, and small monthly devaluations (around 2 percent) were carried out until April 2016. The benchmark policy rate was hiked from 11 to 16 percent in 2016. Lastly, the government approved a new and comprehensive plan to address the effects of the crisis on February 2016. Policy response, however, changed course gradually in mid-2016. The exchange rate was pegged again, restrictions on foreign exchange were imposed, the benchmark interest rate was kept unchanged, turning real interest rates negative, and the Page 4 of 32 The World Bank (P153243) emphasis on fiscal consolidation was eased, giving more priority on the need to boost the economy in the run-up to the 2017 presidential elections. After several months observing the changing government response and engaging with the counterparts, it became clear that the government response would not change soon. The macroeconomic framework then was judged by the Bank as not adequate for a DPF, and the Bank informed the authorities on November 13, 2017 of the decision to cancel the second loan and the PBG. 2. KEY FACTORS AFFECTING IMPLEMENTATION AND OUTCOMES 2.1 Program Performance 21. All prior actions from DPL1 were completed prior to Board presentation and progress was made in achieving the triggers of the second operation. Although the second DPL was dropped, five triggers of DPL2 (#1, #2, #3, #7, and #10) were completed and progress towards completion of the other triggers were made. Two out of the six results indicators were achieved, however indicator #4 retroceded. More information on the prior actions of DPL1, triggers of DPL2, and the linkage with the results indicators are given in section 3.2. Table 2. Basic Data for the Angola DPF Disbursed Approval Effectiveness Revised DPF Closing Date Release Amount Date Date Closing Date AO DPF1 450 30-Jun-2015 27-Jul-2015 31-Dec-2016 - Regular AO 30-Jun-2015 - - Cancelled Guarantee AO DPF2 31-Dec-2017 Cancelled Table 3. Prior Action and Triggers Prior Actions under DPF1 Triggers for DPF2 Prior action #1: Integrated its tax and customs Trigger # 1: Adopt a Presidential Decree introducing administration including through the establishment, fiscal rules governing the flow in and flow out of the institutionalization and provision of a governance Oil Price Differential Account structure for the AGT as evidenced by the: (a) Presidential Decree No. 324/14 published in the Diário da República No. 220 dated December 15, 2014; and (ii) Ministry of Finance Order n. 5513/14 published in the Diário da República No. 149 dated December 24, 2014. Page 5 of 32 The World Bank (P153243) Prior Actions under DPF1 Triggers for DPF2 Prior action #2: Streamlined its tax policy to support Trigger # 2: Implement institutional and operational the modernization of tax administration in order to measures to make the AGT fully functional by ensure compliance by taxpayers, simplify procedures, December 30, 2015: (1) AGT organizational structure expand the tax base, reduce distortions and enhance completed and management of all departments fully equity as evidenced by the: (a) Industrial Tax Code staffed; (2) New strategic plan and implementation (Law No. 19/14, of 22 October) published in the Diário plan for AGT approved by both the AGT Board and da República No. 192 dated October 22, 2014; (b) Minister of Finance; (3) New dedicated Code of Ethics Personal Income Tax Code (Law No.18/14) published completed and applied to entire staff at AGT; and (4) in the Diário da República No. 192 dated October 22, New management unit reporting directly to the chair 2014; (c) Tax on Invested Capital Code (Presidential of the AGT Board established to coordinate Decree Law No. 2/14) published in the Diário da implementation of strategic cultural and operational República No. 190 dated October 20, 2014; (iv) New integration between tax and customs administrations General Tax Code (Law 21/14, of 22 October); (v) Tax as well as change management. Enforcement Code (Law 20/14) published in the Diário da República No. 192 dated October 22, 2014; and (vi) Procedures Code (Law No. 22/14) published in the Diário da República No. 217 dated December 5, 205 Prior action #3: Enacted a revised budget law that, Trigger # 3: By December 30, 2015, continue major among other things, adjusts the 2015 PIP downwards operational business process reengineering and staff by 50 percent to reflect declining oil prices, as training to support implementation of new tax laws: evidenced by the Revised 2015 Budget Law published (1) complete the pilot phase of updating registration in the Diário da República No. 49 dated April 9, 2015. database for the group of corporate sector; (2) expand the taxpayer base through a mix of new operational initiatives, e.g., create mobile teams and task force to reach hard-to-tax groups of taxpayers and those located outside of Luanda; (3) develop new tax forms as commensurate to the revised tax laws; (4) introduce e-filing, make it compulsory to large taxpayers; (5) update and improve strategic plan and manual for tax audit; and (5) continue training on new tax codes. Prior action #4: Strengthened the PIP by restructuring Trigger # 4: The GoA, has adopted: (i) a new it, approving a manual for monitoring risky projects; Presidential decree that establishes the principles, and developing performance indicators to monitor objectives, processes and methodologies for the and measure project performance as evidenced by the development and management of the Public letter from the Minister of Planning and Territorial Investment System, PIP and projects within it; and (ii) Development dated May 11, 2015 attaching the (a) binding annual operational procedures for public report on the cleaning up of the 2014 PIP; (b) manual investment management of preparation and execution of field visits for structural projects issued by the National Directorate of Public Investment on December 2014; and (c) table of performance indicators issued by the National Directorate of Public Investment on December 2014 Page 6 of 32 The World Bank (P153243) Prior Actions under DPF1 Triggers for DPF2 Prior action # 5: (a) Created a joint committee (the Trigger #5: The GoA implements a revised public Joint Committee) comprised of representatives of investment process through the development and MPDT and MoF chaired by the Minister of Finance systematic use of manuals and templates for with mandate to design and implement the necessary preparation of projects, application of prioritization actions to attain full integration as evidenced by the criteria, better implementation and monitoring. Ministerial Decree signed by the Minister of Finance and the Minister of Planning and Territorial Development in May 2015. (b) Initiated the integration of the SIPIP and SIGFE through the following measures: (i) creation of an electronic process for internal budget readjustments of the PIP, including budget units, MPDT and MoF; (ii) merging of MPDT and SIGFE mailing lists for automated alerts for the purpose of: (A) withholding or releasing (as the case may require) budget funds for new projects; (B) activation of additional budget credits; and (C) activation of budget ceilings; and (iii) allowing MPDT to access SIGFE for purposes of extracting PIP budget execution reports at project level; and (D) allowing MoF to access SIPIP for purposes of preparing the annual PIP budget proposal as evidenced by the Minutes of the Joint Committee meeting dated May 11, 2015 Prior action #6: Increased retail fuel prices in order to Trigger # 6: Regulation for linking information systems start the phasing out of inefficient and untargeted fuel and databases related to planning, budgeting, PIM, subsidies which jointly resulted in the following treasury, contracts in relevant Ministries and increases in fuel prices: (a) gasoline and diesel: 50%; institutions enacted promoting effective, open (b) liquefied petroleum gas: 21.6%; (c) kerosene: collaboration and limiting duplication. 34.6%; (d) heavy and light fuel oil: 100%; and (e) asphalt: 18.8% as evidenced by Executive Decree No. 289/14 published in the Diário da República No. 179 dated September 26, 2014 Executive Decree No. 405/14 published in the Diário da República No. 405/14 dated December 24, 2014. Prior action #7: Expanded social protection by Trigger #7: Further implementation of the subsidy completing the registration of 50,000 families in the reform: Borrower’s pilot Cartão Kikuia cash transfer program,  Presidential decree establishing a relation as evidenced by the letter from the Minister of between fuel (gasoline and diesel) price increases Commerce dated April 2, 2015, describing the process and public transport fares increases used to target municipalities and register  Introduction of an automatic upward fuel price beneficiaries. adjustment (gasoline and diesel) Prior action #8: Strengthened evidence-based policy Trigger #8: Register 90,000 more families, for a total making by approving the National Strategy of of 140,000 beneficiaries by the end of 2015. Statistical Development for 2015- 2025 (ENDE 2015- Trigger #9: Develop and implement a Cadastro Único 2025), as evidenced by the publication of the said of all social programs Strategy in the Diário da República No. 64 dated May 7, 2015. Page 7 of 32 The World Bank (P153243) Prior Actions under DPF1 Triggers for DPF2 Prior action #9: Initiated monitoring of its social Trigger #10: IDRE-2015 is carried out and the programs by concluding the preparatory work to anonymized data is made available to the public for launch the 2015 Survey Expenditure, Revenue and analysis. Employment of the Population (IDRE-2015), including Trigger #11: Prepare a poverty map combining the the: (a) design of a national sample frame based on data from the 2014 National Census and the 2008-09 the 2014 census ensuring national and provincial IBEP survey, and use the poverty map to assess the representativeness; (b) design and programming of targeting of the Cartão Kikuia resources, as well as the the electronic questionnaire for the IDRE-2015 survey; allocation of the main other poverty reduction and (c) finalization of the IDRE-2015 sample, as programs. evidenced by the letter from the General Director of INE dated May 7, 2015 attaching the preparatory work for the IDRE-2015. 2.2 Major Factors Affecting Implementation 22. There were three broad factors affecting implementation: (a) declining political commitment to enact ambitious policy reforms during a crisis; (b) weak institutional capacity coupled with insufficient technical assistance in some areas to sustain demanding reforms; and (c) the restrictive fiscal conditions. Political Commitment and Ambitiousness 23. Implementation of the overall program was hampered by the inability to pursue ambitious policy reforms. In most areas, the program called for the introduction of significantly new policies, systems, and procedures. Those would have been challenging to implement even during normal times. But, they became more difficult at a time when the authorities had to deal with sharply deteriorating fiscal conditions. The reforms were delayed, in many instances, because of the need to address more immediate and pressing issues of crisis management. The change of leadership in the Ministry of Finance also contributed to the declining political support to reforms, evidencing that the support and buy-in for reforms was less broad-based than initially thought. In addition, as the presidential elections approached the support to such reforms dimmed, which led to the dismissal of the minister of finance, a great reform champion. Some reforms also called for a coordinated effort among several ministers such as the PIP and social protection programs that could only be accomplished with significant political support at the highest level. The experience showed that forging alliances to agree on more centralized guidance takes considerable time and strong political leadership. 24. An ambitious program was driven in part by the existing policy dialogue and the high expectations set by all stakeholders. The Bank and the government were in a process of re-engagement before the crisis, carrying out policy dialogue and discussing a possible operation. With the onset of the crisis, negotiations deepened and accelerated. To obtain broad support within the Bank for the first policy based loan to Angola, there was a need to show commitment to a strong program. The Ministry of Finance (MoF), the main counterpart, also wished to promote fundamental reforms. There was an agreement to design an operation with far reaching measures, both to show reform commitment and meet the high expectations set by stakeholders. 25. Progress was greater where goals were more consistent with the political priorities of fiscal crisis management and required less coordination outside of the MoF purview. This included the two Page 8 of 32 The World Bank (P153243) broad fiscal consolidation measures: the reduction in fuel price subsidies and improved tax collection. These policies were aligned with the broader fiscal objectives and were more successfully implemented. Institutional Capacity and Administrative Constraints 26. A related factor affecting implementation was weak institutional capacity to carry out the demanding technical and administrative reforms expected under the program. For example, the Ministry of Planning and Territorial Development (MPTD), responsible for PIP, did not have the administrative wherewithal to deal with a comprehensive set of reforms: it lacked the expertise to review and implement an improved PIP cycle, its database had gaps and it lacked the convening power to have all spending ministries contributing with the date needed and learning the new PIM cycle. In additional, it did not receive sufficient technical assistance from the Bank or other development partners. Similar difficulties were faced in the implementation of the poverty map and the creation of the single register, despite receiving significant technical assistance from the Bank team. 27. Coordination among government agencies also proved difficult to achieve and constituted a hindrance to the successful implementation of the program. Government actions such as the expansion of the kikuia card, the creation of a single registry of social program, and the revision of the PIP cycle required coordinated action across different agencies and levels of government, often across an extensive geographical coverage. It was recognized early on that administrative weakness could delay the timetable for the expansion of the program. 28. The experience showed that building institutions requires continuous engagement and supervision, along with improving administrative capacity. The areas of relative success, such as with the reforms of AGT and the initial stages of ENDE, received continued, extensive and on-site support from large private consultancy firms and concentrated on upgrading human capital. The scope of the proposed measures under the AGT program was far reaching and benefitted from expert advice and international experience. The operational measures needed to build an effective new AGT entailed organizational restructuring and considerable training. The Bank remained closely engaged with the client until mid- 2016. . Fiscal Management 29. Fiscal conditions affected many expected outcomes. Shrinking government revenues limited the available resources to several executing agencies. This was most acute in the social policy areas. The National Statistical Institute (INE) did not receive the funds needed to finish the census and to carry out planned surveys. Allocations to the Cartão Kikuia were reduced in 2016, leading to irregular payments to beneficiaries and an indefinite postponement of any planned expansion. The sharp contraction in funding was also the predominant concern for the PIP. The conduct of fiscal policy in late 2016 and 2017 also undercut the objective of controlling budget subsidies for fuels, by allowing the subsidies to re-emerge implicitly, financed by the state-owned oil company – Sonangol, by not making use of the automatic fuel price adjustment mechanism. 30. On the case of the PBG, its cancelation was a mix of low capacity and waning political commitment to the reforms. After the approval of the PBG, the debt management office’s (DMO) capacity was dedicated fully to the issuance of Angola’s first Eurobond, which happened after considerable delays on November 2015. The Bank had advised to do the PBG first to create confidence in the market and lay Page 9 of 32 The World Bank (P153243) a better foundation for the bond. After that, the DMO turned its capacity to the syndicated loan backed by the PBG aiming to raise USD 1billion. This ration was not deemed realistic by the Bank, but the DMO initially agreed with the underwriter’s position. The operation took longer than expected since the government and the creditors could not agree on the terms of the operation, while at the same time the political commitment to reforms slowed in Angola, making investors warier of lending to the country. In the end, the underwriter did scale the transaction back to a $500, but when it became clear that the macroeconomic environment would not improve in the short term, the WB pulled off the guarantee and the operation was called off. 2.3 Monitoring and Evaluation (M&E) Design, Implementation and Utilization 31. The design of the M&E system was rather straightforward. Six results indicators were chosen to capture the broad developments under each pillar. Half of the indicators had specific numerical targets that were possible to identify and follow, despite lack of methodological references such as in results indicator #5. The other half were general documentation and publications certifying that measures had been taken. Some of the qualitative indicators, however, sounded like a restatement of the prior action. The data for most of the indicators could be found in policy or government public documents. 32. Regarding implementation, the MoF was to monitor and evaluate the performance of the operation. To facilitate the coordination among the many agencies and ministries, the government had agreed to appoint a dedicated technical counterpart team to monitor implementation. But the committee was never formally identified and installed. Instead, different groups responsible for each of the areas of the program became responsible for separately monitoring progress. There was a general lack of capacity among implementing agencies to carry the data collection and evaluate programs. Bank teams followed developments closely and reported periodically on their progress, also identifying the measures that needed attention for the successful completion of the program. 33. The utilization of the M&E system was limited. The requirements for keeping abreast of the six broad indicators were rather simple: they were goals to be achieved by the end of 2017. Most of the numerical indicators were reported regularly in government documents, while qualitative indicators were monitored during Bank missions. However, indicators were not used for decision making, except for the indicators on non-oil revenue and subsidies which were used for fiscal policy decisions. 2.4 Expected Next Phase/Follow-up Operation 34. A new government came into office in late 2017. The authorities adopted a macroeconomic framework designed to address growing economic imbalances, resuming most of the reforms supported by the DPF and taking on board other policy advice previously received from the Bank. The new administration also sought continued support from the international community. As a result, the Bank is resuming policy dialogue and is now preparing a new DPF series. Page 10 of 32 The World Bank (P153243) 3. ASSESSMENT OF OUTCOMES 3.1 Relevance of Objectives, Design and Implementation Overall Relevance Rating - Modest Objectives – Substantial 35. The objective of the DPF remains highly relevant for the current development priorities. The crisis underscored the need to strengthen fiscal management, while improving social protection programs. The country still needs to increase non-oil revenues and better target subsidies. The need to improve public sector investment and promote empirically based policy-making remains. The objectives are still consistent with prevailing broad national strategies, identified in the documents that guide economic policy in Angola. The objectives are also in line with the World Bank Group FY14-16 CPS for Angola. The proposed programmatic DPF fitted under the third major component of the CPS, which was to help form adequate human and institutional capacity for sound economic management, develop skills needed in a diversifying economy, and build resilience to potential shocks from the global economy or climate change. 36. The substantial rating stems in part from the rationale for setting ambitious objectives. While objectives turned out to be overambitious, the objective of the reforms were highly relevant for Angola. The goals of the DPF were broadened during the preparation process to include measures on social protection to mitigate the adverse effects of the macroeconomic adjustment on the vulnerable population. Both the Bank and the authorities had reasons to use the DPF as a conduit for promoting major reforms. The government agreed to ambitious targets in part to demonstrate commitment for important measures, and to signal to markets the direction of policy. The Bank, under its approval process, needed a strong program to garnish international support for extending the first policy based loan to Angola. Relevance of Design - Modest 37. The design of the program correctly focused on the areas of fiscal management and social protection. The high dependence on petroleum mandated prudence and the adoption of stabilizing rules for managing oil revenues. The reduction of un-targeted subsidies was a priority at a time of fiscal contraction; so was the prioritization and rationalization of the PIP. Even the attempt to expand the cash transfers, though challenging, was laudable at design. One of the important elements of the design was the anticipated introduction of better management practices, with greater reliance on data and expertise. 38. One of the shortcomings was that some of the proposed reforms were limited in scope, while others were overly ambitious. The size and reach of the targeted cash transfer program was inadequate to protect a large segment of the poor and vulnerable from the adverse effects of the fiscal adjustment program. The implementation of a single registry of social programs beneficiaries in a country with many dispersed programs was highly ambitious, given the time taken by more advanced countries to achieve this objective. The same can be said for conducting a household survey and a poverty map for a country that was in the process of conducting the first census after the end of the civil war (2002). Page 11 of 32 The World Bank (P153243) 39. The design of the program benefitted from the Bank’s accumulated knowledge in Angola and its involvement in certain areas. The World Bank had prepared a review of the PIM and PIP, and was well informed of its complexities. There had been evaluations of the tax administration and of possible tax measures to improve non-oil revenues. The International Monetary Fund (IMF) had analyzed in depth Angola’s fuel subsidies. Finally, international experience was used in the design of an expanded coverage under the incipient direct transfer program (Cartão Kikuia). The fiduciary management of the public sector was weak and prone to manipulation, so the Bank obtained assurances in the handling of the funds through dedicated accounts. Relevance of Implementation - Modest 40. The implementation of the program did not match the ambition of its design. As noted earlier, there were many areas where implementation was successful, particularly in cases where policy requirements were consistent with the fiscal needs of the program, where there was unabating political commitment, adequate technical assistance and low need for inter-government coordination. Those covered primarily the reduction in fuel subsidies and the increases in non-oil revenues. The execution of the program fell short in many areas, especially where the expected measures called for fundamental managerial and administrative reforms, such as with the PIP. Also, when the policies required strong political support to overcome vested interest, such as with the consolidation of the social sector programs. Greater progress in those areas will require time and more concerted efforts, as mentioned earlier. 3.2 Achievement of Program Development Objectives Efficacy Rating – Modest 41. The efficacy rating is based on the achievement of the results indicators’ targets. When strictly measured by the results indicators, achievement was sparse; only a third of the final targets were met, and results indicator #4 was not met in substance, despite being formally achieved. But when considered from the perspective of the higher order objectives of the DPF, there has been more progress. The operation launched some initiatives that are proceeding, though much slower than expected. It also helped to build institutional capacity that could serve as the basis for future progress towards the targets. Still, the achievement of progress is rated as modest since there were major shortcomings in achieving the goals. 42. The achievements in each policy area are described below. The indicators related to improved fiscal management (RI #1 and #3) were partially achieved, since fiscal rules linking the OPDA to the budget cycle were enacted. Indicators on the creation of fiscal space (RI #2 and #4) were not achieved since non- oil revenues fell short of the target and the subsidies reform was partially reversed. Finally, the objective of better protecting the poor was not achieved either as indicators #5 and #6 were also missed. The deviations can be attributed in some cases to political and institutional constraints. In other cases, the goals were too ambitious and faced considerable organizational challenges, such as with the reforms in the PIP and the social sector program. Page 12 of 32 The World Bank (P153243) Policy Area 1 - Introducing fiscal rules for the utilization of oil funds and modernizing tax administration and tax policy to increase non-oil revenue collection Results indicator #1 - “Fiscal rules linking the Oil Price Differential Account to the budget cycle are in place by December 31, 2017.� Achieved 43. A presidential decree governing the management of the fund was issued in 2015 and remains in effect. This policy area had no prior action in DPL1, but Trigger #1 (from DPL2) called for the adoption of a presidential decree introducing fiscal rules governing the flows of the Oil Price Differential Account. The decree, published on June 2015, laid out the inflow and withdraw rules. It specified the basis for calculating deposits of excess revenues generated by higher prices of oil than those used for budget planning. It also identified the four conditions when resources could be withdrawn or used (for example whenever the debt to GDP exceeded 60 percent or at times of high debt service payments). Although the decree fulfilled the trigger for the planned second operation, and established some links between the oil price and the budget cycle, the Bank team felt that the rules could be improved. The major concern was that the President retained considerable discretion in triggering the use of the funds, albeit within the broad parameters of the decree, when the outflow mechanism should be more automatic. Resources could be accumulated even when the country was running a budget deficit. The Bank team provided technical assistance by preparing a policy note with simulations and examples from other resource-rich countries on how Angola could improve the Oil Price Differential Account (OPDA). This policy note did not find much traction in 2017, but it is now being used as the rationale behind the upcoming revision of the OPDA and the Sovereign Wealth Fund. Nevertheless, the constitution of a fiscal stabilization fund was an improvement to fiscal management and proved useful when the resources were tapped in 2016 and 2017, avoiding even larger expenditure cuts and budget deficits. Therefore, the result was not only formally achieved, but also in substance as the OPDA served the purpose it was created for. Results Indicator #2 - “Non-oil tax revenue, as a share of non-oil GDP, increases from 13.4 percent in 2013 to 15.6 percent by 2017.� Actual result = 10.8 percent of GDP. Not Achieved. 44. The target was reached in 2015, but has declined since then, to 10.8 percent of GDP in 2017. The primary reason for the shortfall was the depth of the ensuing recession. Economic activity declined sharply, as did imports, one of the main source of revenues. Many enterprises suffered losses, delayed payments, and even closed. Even the state-owned oil company accumulated tax arrears. These factors overwhelmed the benefits derived from the tax reforms supported as prior actions to the first DPL and triggers to the second DPL, which were all achieved. 45. The administrative and institutional reforms in tax administration continued throughout the program. The prior actions of the first DPL1 were followed through with the implementation of the triggers for the second operation. The new organizational structure of AGT was completed. A new managerial team was installed and became responsible for the integration of the three agencies. The new management team reengineered business’ process, issued strategic and annual plans to improve revenue collection, and reduced compliance costs. Staff was trained to implement new tax laws, improve taxpayer services, and expand the database. The modernization of AGT can be considered one of the successes of the operation, even though the results indicator was not achieved signaling some problem with the choice of results indicator. The most promising development for improving tax revenues was the decision by the government to introduce a value added tax (VAT) in 2019. During the operation, AGT has received technical assistance support in the form of tax expert visits from the IMF’s Fiscal Affairs Department., Page 13 of 32 The World Bank (P153243) which has led to the recent request for support from the Bank and the IMF for the implementation of the VAT. During the operation, the Bank and the government also engaged in a RAS proposal, which was signed in late 2017 to support tax simplification among other things. Policy Area 2 – Increasing value for money by enhancing the efficiency and efficacy of public investment management Results Indicator #3 - “100 percent of all new projects the 2017 Public Investment Plan are subjected to the new PIM cycle, systems, and regulations.� Not Achieved. 46. Although there was progress in PIM, the results indicator was not achieved due to its broad and ambitious undertaking. The prior actions and triggers required within two years a significant revamp of the whole public investment management system. There was progress such as improvements in SIPIP, a more inclusive database of projects, and increased dialogue between planning and budgeting. The government drafted new manuals - guiding appraisal, costing, selection, and monitoring of public investment, and a new Presidential decree establishing the principles, objectives, processes, and methodologies for the management of the PIP, but those were never formally approved and implemented. Thus, the 2017 PIP was formulated still under the old regulations and the results indicator target was missed. This policy area suffered from the inability of the Bank or other development partners to provide more intensive technical assistance. The IMF did one “experimental� Public Investment Management Assessment mission in March 2015, that was not followed up. The government used resources from the African Development Bank to hire one external consultant, but these resources were not enough and the work suffered from delays and discontinuities. The new government moved the PIP unit from the Ministry of Planning to the Ministry of Finance, under the purview of the state secretary for budget, which should help coordination and steer the reforms going forward. Overall, the government took some steps to improve PIM by adopting the prior actions of DPL1 and making progress towards the triggers 4, 5, and 6 of DPL2, but the target of the results indicator was not met because the draft regulations were never put in place so the PIP is still being formulated and approved under the old rules. Policy Area 3 - Moving from untargeted fuel price subsidies to targeted cash-transfers and evidence based policy-making Results indicator #4 – “Fuel price subsidies fall from 5.2 percent of GDP in 2013 to 2.3 percent by December 2017�. Actual result = 1.3 percent of GDP. Achieved. 47. Fuel subsidies, as measured by budgeted expenditures showed a decline to 1.3 percent of GDP in 2017. The prior actions and triggers, which were mostly completed, called for the gradual elimination of fuel subsidies, reducing this expenditure in the budget. The government eliminated all relevant fuel subsidies by the end of 2015. It also introduced an automatic mechanism for fuel price adjustments on January 1st, 2016. However, the mechanism was never used and after exchange rate devaluations and an increase in international oil prices, fuel price subsidies re-emerged as domestic prices were below the international price according to the formula set by the automatic adjustment mechanism. Subsidies are now paid indirectly by Sonangol, as they are not registered and paid through the budget. Eventually, the government had to inject US$10 billion (or close to 10 percent of GDP) into Sonangol to cover the costs of those implicit subsidies in late 2017. Even though the target was formally achieved, the objective of removing subsidies was not. The new government, however, has resumed the fuel subsidy reform agenda Page 14 of 32 The World Bank (P153243) and has already announced plans for changes in electricity and water prices, and is re-analyzing the mechanism for fuel price subsidies. 48. The DPF also called for the removal of subsidies in public transportation but that was dropped during the preparation of the second DPL. Given the social and demographic profile of public transportation users in Angola, and its small impact in the budget, it was judged that the elimination of subsidies to public transport would hurt the most vulnerable population, while saving little fiscal resources. Also, cooking gas and lightning oil remained subsidized, given their pro-poor effect and low budgetary impact. Results Indicator #5 - “The percentage of households protected from negative income shocks, via direct cash transfers, grows from 3 percent of the population to 7 percent by December 2017.� Actual result = 1.4 percent of population. Not achieved. 49. The expansion of the country’s main transfer program has been much slower than expected. This results indicator was supposed to be achieved by the prior actions and triggers supporting the expansion of the Cartão Kikuia program, however, only 40,000 families have been added since 2015. There were 90,000 families (or 1.4 percent of the households) registered in the program by the end of 2017. This would give a coverage of 1.4 percent of households vis-à-vis a target of 7 percent of the population, although the DPF program document did not clearly identify the population that should serve as basis for the calculation. 50. There were two main reasons for the delay, including limited financial resources and organization capacity. The authorities curtailed the funding for the responsible implementing agencies, and reduced the budget allocations for the transfers. Starting in 2016, payments became erratic and the government accumulated large arrears under the program. Positive developments included the geographical expansion of the coverage and the community based support system. 51. There were major challenges in the administration of the social protection programs and the Bank offered support. There is an evaluation of the program in process, with a view to enhance operational effectiveness and the “quality� of targeting. Progress was made with the transfer of the Kikuia Card from the Ministry of Commerce to the Ministry of Social Action and then with its merger with the ministry of Families to form the new MASFAMU, which was assigned the responsibility for the coordination of the programs. But the ministry does not have the information systems and technical capacity to carry out the responsibility. There is still a need for the approval of the Presidential Decree governing the single registry, and establishing the operational modalities, roles and responsibilities, and the mechanisms for collaboration among the agencies. The Bank has a long history of supporting this sector in Angola though its local development projects with FAS and technical assistance funded by trust funds to social sector ministers. The Bank provided technical assistance and nudged the reforms by bringing expertise from Brazil on the single registry, consultants to do a poverty map, among other activities. Nonetheless, progress fell short of the results indicator target. Page 15 of 32 The World Bank (P153243) Results Indicator #6 - “By December 31, 2017, empirical results from the analysis of the Inquérito de Despesas e Receitas e Emprego da População-2015 (IDRE-2015) household survey appear in one or more government official documents proposing policy actions.� Not Achieved 52. The lack of financing and organizational resources led the authorities to postpone the IDRE- 2015 household survey until 2018. The prior actions and triggers under this results indicator called for the development, release, and analysis of household surveys and granular statistical data to inform policy making with evidence. During loan preparation, it was agreed that the government would fund IDREA, however, the shrinking government resources led the government to withdraw the funds to IDREA and even caused delays to the processing, treatment, and release of the 2014 census results. The survey is now being prepared under a recently effective World Bank Technical Assistance loan, the Statistics Project. The pilot survey took place during Nov-Dec 2017. The main data collection has started in early 2018 and will end January 2019.The empirical results from the analysis of the survey will not be published until later in 2019. In summary, there has been progress and the target is likely to be achieved in 2019. 53. Throughout the supervision of DPL1 and preparation of DPL2, the Bank and other development partners provided technical assistance to help Angola achieve the results indicators targets. Table 4 links the TA received to each prior action. In addition to the reports and projects mentioned below, it should be noted that for RI#2, AGT counted with support from external private consultants such as McKinsey, that were working side by side on a daily basis with the client. Resources from the African Development Bank also allowed the government to hire a PIM consultant that supported the revision of the PIM process, with the corresponding supporting documents. In some cases, however, this TA was not enough as in the case of statistics, in which the bottleneck was lack of resources or in the case of PIM, in which more constant and wide-reaching TA was called for. Table 4. Results Indicators and Technical Assistance Result Indicator Technical Assistance (TA) provided and final report/ closing date RI #1 Angola’s Oil Price Differential Account: Simulations and Suggestions for Improvements (P159791) – May 2017 RI #2 Technical Assistance Report from the Fiscal Affairs Department of the IMF RI #3 Strengthening the Management of Public Investment – IMF TA Report – November 2015 RI #4 Angola: Fuel Price subsidy – The Way Forward – Technical Assistance Report IMF Country Report 15/28 – February 2015 RI #5 Support to a Systemic Approach to Social Protection and Poverty Reduction (P156589) – June 2017 RI #6 Angola Statistics Project (P157671) was initiated in October of 2015 and was approved by the Board on December 2016. Page 16 of 32 The World Bank (P153243) 3.3 Justification of Overall Outcome Rating Rating: Unsatisfactory 54. The overall outcome rating is based on the combination of a relevant DPF objective that fell short in design and achieving its objectives. The substantial relevance of the objectives, which reflect the ambition of the undertaking, was tempered by shortcomings in the design and implementation of the operation. Although the operation supported the right and most pressing reforms for Angola, its design had shortcomings – be it the over-ambitiousness of most targets, over-reliance in one reform champion, and the lack of clarity in some triggers and results indicators. The operation faced shortcomings also in its efficacy – evidenced by the small achievement of the results indicator, which led the overall rating to be unsatisfactory. 3.4 Overarching Themes, Other Outcomes and Impacts (a) Poverty Impacts, Gender Aspects, and Social Development 55. The DPF had a strong focus on poverty and social development issues. Its primary objective was to create the fiscal space needed to better protect the poor and vulnerable. It addressed important policies that would benefit the poor, such as the expansion of the cash transfer program. In fact, the entire third pillar was dedicated to creating the conditions for policies and programs that targeted and benefited the poor. The program began to build the institutional structure to improve targeting and service delivery. The Bank prepared a Poverty and Social Impact for Angola, to inform the analysis of the incidence of subsidy reforms and the extension of the social protection program. Overall, the poor benefited from the program for example by the extension of the Cartão Kikuia, but to a lesser extent than it was envisioned. (b) Institutional Change/Strengthening 56. Unquestionably, one contribution of the DPF was to promote institutional change and strengthen institutional capacity. This is taking place slowly in most policy areas. The adoption of rules to shield the budget from variations in oil revenues reflected the attempt to improve the conduct of fiscal policy. The tax reforms should spur greater participation and compliance, perhaps creating a shift in cultural attitudes toward taxes. The apparatus that is being built to modernize the PIP should eventually create an entirely new institutional structure for dealing with allocative decisions. The entire national statistical strategy is designed to encourage evidence-based policy making, which would mark a major shift in the administration of social policies. (c) Other Unintended Outcomes and Impacts Not Applicable. Page 17 of 32 The World Bank (P153243) 3.5 Summary of Findings of Beneficiary Survey and/or Stakeholder Workshops 4. ASSESSMENT OF RISK TO DEVELOPMENT OUTCOME Ratings: Significant 57. The risk that the program’s development outcomes are not maintained or fully realized in the foreseeable future is significant. The assessment is based on two broad factors: the expected time and effort that will be needed to close the current gap in the outcomes; and the degree of political support that will be required to implement the reforms. Both capture the core sources of risk to the program, which are deficient administrative capacity, inadequate resources, and shifting political priorities. 58. Administrative capacity was one of the major impediment to achieving the outcomes. The challenges are substantial with the development of the institutional arrangement to administer the PIP effectively. It will take a lot of effort to have all agencies complying with a new PIM cycle, systems and regulations, after they are in place. Finally, a whole new administrative capacity must be built to carry out the work planned under policy area 3. This includes the information gathering and analytical work anticipated under the national statistical strategy, as well as the organizational structure required for the expansion of the targeted cash transfer program. 59. The lack of adequate resources could hamper the ability to make further progress towards the development goals. First, continued fiscal restrictions could force the government to postpone allocating the necessary funding for the programs. This risk is highest with the expected expansion of the Cartão Kikuia. The same could apply to PIP, tax modernization, and the statistical strategy. Second, deteriorating macroeconomic conditions could undermine components of the program. This is probable with the attempt to increase non-oil revenues, given the possible erosion of the tax base, or with the PIP, given the pressures to use more expedient allocative measures. Finally, the third element is inadequate human resources. Several components have been delayed because of staffing issues. The experience so far has shown that more and better prepared personnel will be required for the next phase of the tax reform, public investment, and social mitigation programs. 60. Finally, shifting political priorities could divert the focus of the authorities from the development goals of the program. The governance and political risks are present in essentially all policy based operations. Progress and success in many areas will require steadfast political support. But there are always competing demands, or even perhaps the recovery in the price of oil, can weaken political resolve. The most vulnerable components are the rule-based management of the OPDA and the structural reforms of the PIP. In several cases, there are still pending presidential decrees that must be approved for the reforms to proceed to the next level. More broadly, many of the activities will require collaboration and coordination among diverse agencies and levels of government, which may not be easily attained. Page 18 of 32 The World Bank (P153243) 5. ASSESSMENT OF BANK AND BORROWER PERFORMANCE 5.1 Bank Performance (a) Bank Performance in Ensuring Quality at Entry Ratings: Moderately Unsatisfactory 61. The rating is based on a combination of factors, most importantly the degree of initial engagement and the breadth of the program. The Bank was already re-engaging on policy discussions prior the crisis, and was prepared to be responsive to the government’s request for support. Important to note that this was the first major opportunity for policy based lending in Angola. It came at a time of crisis, when support from the international community was critical and needed to be timely. The operation was relevant and strategic for advancing some key policy reforms. The Bank dedicated considerable resources to the design and structuring of the program. During preparation, the Bank emphasized the importance of focusing also on protecting the poor and vulnerable from the consequences of the crisis. Staff had a good understanding of the development challenges facing the country. The design was informed by the findings and recommendations of recent analyses conducted by the World Bank and other institutions. The Bank was also very careful in handling fiduciary responsibilities, quite aware of the risks emanating from weak, often opaque, financial management and procurement systems. 62. The rating is moderately unsatisfactory, because at entry the Bank established very optimistic targets for the development outcomes and underestimated the constraints for achieving them. The Bank and the government were overly optimistic about possible accomplishments under the program. The targets, as measured by the results indicators, anticipated considerable work and efforts, without sufficient technical assistance in some areas. The program also contemplated advancements in policy areas that lacked full political support. It was very ambitious to believe that some of the reforms could advance as quickly as expected, without considerable political encouragement and collaborative efforts from many government officials. Reaching some of the goals may take many more years. For example, strengthening of the public-sector investment program will entail major changes in systems and procedures, and will require participation and contribution from very diverse groups with competing interests. The same applies to the reforms under pillar 3. The expansion of the Cartão Kikuia must reach remote areas and overcome major administrative roadblocks. These simultaneous actions on many fronts that can only be tackled if there is unwavering support at the highest political levels. (b) Quality of Supervision Ratings: Moderately Unsatisfactory 63. The intensity of supervision waned with the passage of time, as the program’s implementation began to slip and the macroeconomic conditions deteriorated. In the early stages of supervision, there was very close engagement, with periodic visits from staff, who accompanied closely the progress in each area. Those missions focused on the pre-established triggers for moving forward with a second operation. The results of the dialogue were periodic reports on the advancements and frequent communications with the authorities on the measures deserving attention. The process peaked in early 2016, with the internal discussions of a project concept note for the second DPL. The Bank had urged for improved macro Page 19 of 32 The World Bank (P153243) management, but those recommendations were not taken. The Bank grew skeptical that the macro framework would be adequate for a DPF and decide to wait for conditions to improve, which led to the decline in the intensity of the policy dialogue and ultimately affected the pace of implementation 64. The Bank, however, continued to support some of the policy reforms during the rest of the implementation period through technical assistance, trust funds, and investment projects. But did not re-engage in policy based lending until a new government moved, in late 2017, to address the macroeconomic imbalances and reinstate an adequate macroeconomic framework. (c) Justification of Rating for Overall Bank Performance Ratings: Moderately Unsatisfactory 65. The overall rating reflects a very proactive start, albeit with shortcomings, with a strong and timely response to a crisis, soon proceeded by a slow decline in the pace of implementation. The quality at entry reflected an accumulated country knowledge and solid analytical background. It was also the first opportunity for the Bank to be actively engaged in policy reforms through policy based lending. The Bank dedicated a substantial portion of the country’s lending allocation to help pursue important reforms. The opportunity was unique and in the eagerness to promote fundamental changes, the Bank established ambitious targets. Those could not be attained within the expected timetable and set basis for unsatisfactory ratings for outcome. More importantly, many of the policy areas did not enjoy the level of political support needed for successful implementation. The focus of the authorities was on crisis management, and the Bank could not re-establish the level of policy engagement that was present during preparation. The continuous drop in oil prices also affected the achievement of the results, as resources to expand the Cartão Kikuia program and evidence-based policy became unavailable. However, had the oil prices held steady, the problems of waning political commitment and lack of technical capacity would have remained and results would not have been fully achieved. Nevertheless, a process has been launched, with a series of reforms that eventually will move forward at slower pace. 5.2 Borrower Performance (a) Government Performance Ratings: Moderately Unsatisfactory 66. The government performance refers mostly to the central role played by the MoF; the overall entity responsible for the operation. The rating is based on the significant shortcomings in execution compared to expectations. The ministry failed to take the necessary steps to ensure quality of implementation and achieve development outcomes. More importantly, the ministry was unable to secure the approval of many decrees and reach important agreements that were pre-conditions for moving forward to the next phases of implementation. 67. The ministry was responsible for many aspects of the program. It was to facilitate the coordination among the many agencies. The government had agreed to appoint a dedicated technical counterpart team to monitor implementation, but the committee was never formally identified and installed. The ministry was also central for ensuring adequate appropriations for the components of the Page 20 of 32 The World Bank (P153243) program. Many areas suffered from limited resources. In addition, the MoF was a co-participant in the integration of data and management information system, which have not proceeded as well as planned. 68. More significantly, the rating reflects the inability to obtain the approval of several decrees critical for the program’s success. The government was also to adopt a new Presidential decree defining the principles, objectives, processes and methodologies for the management of the PIP, along with binding annual operational plans. The MPDT worked on the decree, held consultations, but the draft decree remains to be sanctioned. Another required Presidential decree was for the governance and management of the single registry of social programs, which is also yet to be approved. (b) Implementing Agency or Agencies Performance Ratings: Moderately Unsatisfactory 69. The performance of the implementing agencies is rated moderately unsatisfactory as many agencies failed to implement fully the reforms to achieve the development objective. The MoF as an implementing agency did not act on the fuel price automatic adjustment mechanism and failed to communicate in advance with the Bank on the OPDA decree. INE did not have the resources and capacity to conduct the poverty map and the household survey without significant handholding support from the Bank. The National Directorate for Public Investment (NDPI) had to resort to external consultants to draft the policy reforms, that were delayed and never implemented. The notable exception was AGT, which carried all reforms foreseen in the program in a mostly timely fashion. (c) Justification of Rating for Overall Borrower Performance Ratings: Moderately Unsatisfactory 70. As both government and implementing agencies are rated moderately unsatisfactory, the overall rating is moderately unsatisfactory. 6. LESSONS LEARNED 71. Four broad lessons and one PBG-specific lesson emerge from the experience with the first DPF to Angola. First, the need to be realistic in establishing targets and timetables. Second, the need to complement policy reform with institutional building, and calibrate the goals and timing according to the level of capacity available to implement the program. Third, the need to ensure broad ownership and commitment to reforms on the government side, so the program is not severely affected by the departure of key counterparts. The fourth lesson related to the importance of the process of preparing and supervising the operation. In addition, there is a fifth lesson related to the PBG. The lessons are consistent with the identified main impediments to implementation, which were diminished political commitment to enact fundamental policy reforms, and weak institutional capacity to implement the program. The lessons arise from an evaluation of both the design and implementation. 72. One of the main issues identified in the relevance of the design was the overly optimistic targets for the development outcomes. A common tendency is to use a DPF as an opportunity for effecting major changes. This was particularly true here, given the pressures to design a strong program, worth of Page 21 of 32 The World Bank (P153243) international assistance, and to include important actions for garnishing support, such as focusing on social protection. Experience shows that often, in the eagerness to obtain the needed financing, the authorities tend to agree with ambitious reforms and targets to be carried out under optimistic timeline. Development partners accept commitments that are unlikely to be fulfilled. The failure to attain most results indicators under the DPF highlighted the importance of considering the political and institutional constraints to development in Angola. The execution of the program fell short in those areas where the expected measures called for fundamental managerial and administrative reforms, such as with the PIP; and the policies required strong political support to overcome vested interest, such as with the consolidation of the social sector programs or the introduction of automatic adjustments to fuel prices. 73. The second lesson from the operation is the need to identify what can reasonably be accomplished within the limited time and capacity available. Results indicators should be established according to the sector’s administrative capacity and the ability of the Bank and development partners to provide the technical assistance needed by the clients. There are two important factors to consider, after the experience under the program. One is that results were more difficult to reach the larger the gap between administrative capacity and complexity of task. This was true in the case of both the public investment reforms and the national statistical strategy, where available experience did not match the magnitude of the work and effort required. The program anticipated substantial and comprehensive reforms that could take several years to be fully effective. 74. A lesson throughout the ICR is that progress and success require steadfast political support and broad government ownership of the reforms. There was greater progress in areas, such as subsidies and tax reforms, that were more consistent with the overall macroeconomic policies of the governmentIn many areas of the program, the achievements varied according to the degree of focus and assistance provided by the central authorities. The lack of priority support was reflected in the inadequate allocation of resources, in failure to fulfil policy commitments, in being proactively engaged, and taking corrective measures. In some policy areas, such as with the expansion of the targeted cash transfer program, the lack of full support was reflected on poor achievements. As noted before, it will be difficult for some of the reforms to advance as quickly as expected, without much greater political participation and encouragement, as well as more collaborative efforts from all government officials. The commitment at the highest level is crucial, since often has a cascading effect. It permeates down to agency and working level staff. It motivates everyone involved to also be more active and supportive. 75. One lesson is that future policy-based support must be accompanied by a solid understanding of the political commitment to the reforms. This extends beyond the most obvious implication, which is identifying intervention areas that are fully aligned to existing political priorities. After all, the most successful components under the DPF, for example, tax reform, were those most closely linked to the overall macroeconomic stabilization program. The reference here is a broader concept of commitment; one that is not only codified in agreements, but also reflected on the repeated actions and transparent conduct of the participants in fulfilling expectations. The program, under the DPF, was constructed with a sequence of future policy and administrative measures that established expectations. The design anticipated a certain level of commitment to a path of reforms, which did not materialize. While changing political priorities can divert focus temporarily, the understandings to persist with the core reforms, such as with approval of key presidential decrees, must transcend short-term competing demands. The understanding must also include core concepts to be observed. Among them is the collaborative efforts to provide the resources necessary to advance the program, and to share the information necessary to track progress, of both the program and macroeconomic conditions. Page 22 of 32 The World Bank (P153243) 76. The fourth and final lesson is on the importance on the process of preparing and supervising the program. A multi-sectorial program like this DPF requires a multi-sector team as a counterpart on the government side. In addition, multi-sectorial program demands more coordination from the government side. Dedicated government staff with project management skills and a certain level of autonomy is key for the success of the project in terms of timeline and achievement of results. The implementing agency was always overstretched throughout the preparation and supervision. The multi-sector committee, which was supposed to be in charge of monitoring the results was never enacted and staff in the Ministry of Finance lacked autonomy to steer the operation. 77. A specific lesson related to the PBG is on the need to coordinate with a third actor, which is the commercial bank leading the operation and to confirm market appetite for the transaction. The PBG transaction was delayed due to the inability of the debt management office to conduct two large and innovative operations – Eurobonds and a syndicated loan guaranteed by the PBG– at the same time. This relates to the lesson on adequate capacity laid out above. However, in the specific case of the PBG there is a lesson on the need to coordinate with the lead arranger of the PBG transaction. In the case of Angola, the commercial bank coordinating the PBG was the same coordinating the Eurobond issuance, and its view was not always aligned with the World Bank on the sequencing of the transaction. 78. This was the first policy based IBRD operation in Angola, and proved to be a learning process for both the government and the Bank. The experience showed that it will take much longer than expected to achieve some of the goals. The process will be more challenging than anticipated. To overcome the obstacles, there is a need for greater commitment and involvement by the authorities, along with perhaps continued participation and support from the development partners. The process will inevitably entail forging agreements on some of the basic core principles that should guide both the reforms and the relationship with all development partners. Angola faces considerable development challenges, and is still reeling from the crisis. Economic conditions are still fragile and the financing needs remain significant. Tackling the future challenges will require greater and more concerted efforts by all of those involved with the development of Angola. 7. COMMENTS ON ISSUES RAISED BY BORROWER/IMPLEMENTING AGENCIES/PARTNERS (a) Borrower/Implementing agencies (b) Cofinanciers (c) Other partners and stakeholders (e.g. NGOs/private sector/civil society) Page 23 of 32 The World Bank (P153243) ANNEX 1: BANK LENDING AND IMPLEMENTATION SUPPORT/SUPERVISION PROCESSES (a) Task Team members Responsibility/ Names Title Unit Specialty Lending Souleymane Coulibaly Lead Economist GMTA1 TTL Helene Bertaud Lead Counsel LEGES Lead Financial Gianfranco Bertozzi Officer/Debt Capital FABBK Markets & CBP Mazen Bouri Program Leader AFCC1 Ana Maria Carvalho Senior Operations Officer AFMAO Senior Public Sector Kjetil Hansen GGOAC Specialist Elisa Gamberoni Senior Economist GMTE3 Edith Kikoni Senior Economist GMTE3 Sr Financial Management Ousmane Kolie GGOMN Specialist Jose Janeiro Senior Finance Officer WFACS Biagio Mazzi Senior Financial Officer CMD Luz Meza-Bartrina Senior Counsel LEGLE Tuan Minh Le Lead Economist GMTA4 Pedro Olinto Program Leader LCC5C Carolina Renteria Lead Economist GMFDR Supervision Rafael Barroso Senior Economist GMTA1 ICR Team Leader Eduardo Somensatto Consultant GMTA1 ICR Author Page 24 of 32 The World Bank (P153243) (b) Staff Time and Cost Staff Time and Cost (Bank Budget Only) Stage of Project Cycle US$, Thousands (including travel No. of staff weeks and consultant costs) Lending 38.35 432,097.35 Total: 38.35 432,097.35 Supervision/ICR 0.28 32,123.62 Total: 0.28 32,123.62 Page 25 of 32 The World Bank (P153243) ANNEX 2: BENEFICIARY SURVEY RESULTS (if any) Page 26 of 32 The World Bank (P153243) ANNEX 3: STAKEHOLDER WORKSHOP REPORT AND RESULTS (if any) Page 27 of 32 The World Bank (P153243) ANNEX 4: SUMMARY OF BORROWER'S ICR AND/OR COMMENTS ON DRAFT ICR (if any) Dear Rafael, The Government of the Republic of Angola recognizes that the performance of the first Fiscal Management Programatic DPL did not meet the expected results. Nonetheless, the Government would like to reiterate its commitment to poverty reduction and social protection. In terms of concrete actions a new Social Protection program called The Integrated Program for Local Development and Poverty Reduction, was approved on May 7th, 2018 by the Economic Commission of the Council of Ministers, and will benefit 30% of the most vulnerable, and 70% of ex-militants. Best regards Manuel Francisco Pedro Director � Ministério das Finanças • Largo da Mutamba • Luanda • Angola ✆ Telefone: 222 70 60 00 • Ext.: 6102  www.minfin.gov.ao Page 28 of 32 The World Bank (P153243) Page 29 of 32 The World Bank (P153243) ANNEX 5: COMMENTS OF COFINANCIERS AND OTHER PARTNERS/STAKEHOLDERS Page 30 of 32 The World Bank (P153243) ANNEX 6: LIST OF SUPPORTING DOCUMENTS Governo da República de Angola, Decreto Presidencial 83/15 de 4 de maio de 2015 Governo da República de Angola, Lei do Orçamento Geral de 2018. Page 31 of 32 The World Bank (P153243) MAP Page 32 of 32