SIERRA LEONE Integrated Public Financial Management Reform Project Report No. 124660 MARCH 28, 2018 © 2018 International Bank for Reconstruction This work is a product of the staff of The World RIGHTS AND PERMISSIONS and Development / The World Bank Bank with external contributions. The findings, The material in this work is subject to copyright. 1818 H Street NW interpretations, and conclusions expressed in Because The World Bank encourages Washington DC 20433 this work do not necessarily reflect the views of dissemination of its knowledge, this work may be Telephone: 202-473-1000 The World Bank, its Board of Executive reproduced, in whole or in part, for Internet: www.worldbank.org Directors, or the governments they represent. noncommercial purposes as long as full attribution to this work is given. Attribution—Please cite the work as follows: The World Bank does not guarantee the World Bank. 2018. Sierra Leone—Integrated accuracy of the data included in this work. The Any queries on rights and licenses, including Public Financial Management Reform Project. boundaries, colors, denominations, and other subsidiary rights, should be addressed to Independent Evaluation Group, Project information shown on any map in this work do World Bank Publications, The World Bank Performance Assessment Report 124660. not imply any judgment on the part of The Group, 1818 H Street NW, Washington, DC Washington, DC: World Bank. World Bank concerning the legal status of any 20433, USA; fax: 202-522-2625; e-mail: territory or the endorsement or acceptance of pubrights@worldbank.org. such boundaries. Report No.: 124660 PROJECT PERFORMANCE ASSESSMENT REPORT SIERRA LEONE INTEGRATED PUBLIC FINANCIAL MANAGEMENT REFORM PROJECT (IDA41460-Trust Fund No.:TF57287, TF56158) March 28, 2018 Human Development and Economic Management Independent Evaluation Group Currency Equivalents (annual averages) Currency Unit = Sierra Leone Leones (Le) 2007 $1.00 Le 2,985 2011 $1.00 Le 4,349 2008 $1.00 Le 2,982 2012 $1.00 Le 4,344 2009 $1.00 Le 3,386 2013 $1.00 Le 4,332 2010 $1.00 Le 3,987 2014 $1.00 Le 4,524 Abbreviations AfDB African Development Bank MDTF multi-donor trust fund AGD Accountant General’s Department MFED Ministry of Finance and Economic DFID U.K. Department for International Development Development NPPA National Public Procurement EU European Union Authority GBAA Government Budgeting and NSA nonstate actors Accountability Act OBI Open Budget Index HRMO Human Resource Management PDO project development objective Office PEFA Public Expenditure Financial IDA International Development Accountability Association PETS Public Expenditure Tracking IEG Independent Evaluation Group Survey IFMIS Integrated Financial Management PFM public financial management Information System PFMICP Public Financial Management IPAU Integrated Project Administration Improvement and Consolidation Unit Project IPFMRP Integrated Public Financial PFMRU Public Financial Management Management Reform Project Reforms Unit IRCBP Institutional Reform and Capacity PIU project implementation unit Building Project PPAR Project Performance Assessment M&E monitoring and evaluation Report MDAs ministries, departments, and agencies All dollar amounts are U.S. dollars unless otherwise indicated. Fiscal Year Government: January 1 – December 31 Director-General, Independent Evaluation Ms. Caroline Heider Director, Human Development and Economic Management Mr. Auguste Tano Kouame Manager, Economic Management and Country Programs Mr. Pablo Fajnzylber Task Manager until December 9, 2017 Mr. Arun Arya Task Manager Mr. Ismail Arslan Contents Principal Ratings ...................................................................................................................... vi Key Staff Responsible.............................................................................................................. vi Preface.................................................................................................................................... viii Summary .................................................................................................................................. ix 1. Background and Context....................................................................................................... 1 Country Background ............................................................................................................. 1 Project Context...................................................................................................................... 1 2. Objectives, Design, and Their Relevance ............................................................................. 1 Objectives ............................................................................................................................. 1 Relevance of Objectives ....................................................................................................... 1 Design ................................................................................................................................... 2 Components ...................................................................................................................... 2 Implementation Arrangements.......................................................................................... 3 Monitoring and Evaluation Design ................................................................................... 4 Relevance of Design ............................................................................................................. 4 Relevance of Project Design to Objectives ....................................................................... 4 Quality of Results Framework .......................................................................................... 7 3. Implementation ..................................................................................................................... 8 Planned versus Actual Expenditure by Component.............................................................. 8 Implementation Experience ................................................................................................ 10 Implementation of Monitoring and Evaluation............................................................... 12 Safeguards Compliance .................................................................................................. 12 Financial Management and Procurement........................................................................ 13 4. Achievement of the Objectives ........................................................................................... 13 Objective 1: Sustainable Improvement in Credibility ........................................................ 13 Outputs ............................................................................................................................ 13 Outcomes ........................................................................................................................ 14 Objective 2: Sustainable Improvement in Control.............................................................. 18 Outputs ............................................................................................................................ 18 This report was prepared by Arun Arya, task manager until December 9, 2017, when Mr. Arya left IEG. The task manager at the time of esubmission was Ismail Arslan. Mr. Arya and Buffy Bailor, consultant, assessed the project in September 2017. The report was peer reviewed by Stefano Migliorisi and panel reviewed by Robert Mark Lacey. Gaby Loibl provided administrative support. iv Outcomes ........................................................................................................................ 19 Objective 3: Sustainable Improvement in Transparency .................................................... 21 Outputs ............................................................................................................................ 21 Outcomes ........................................................................................................................ 21 5. Efficiency ............................................................................................................................ 24 6. Ratings ................................................................................................................................ 24 Outcome .............................................................................................................................. 24 Risk to Development Outcome ........................................................................................... 25 World Bank Performance ................................................................................................... 26 Quality at Entry ............................................................................................................... 26 Quality of Supervision .................................................................................................... 26 Borrower Performance ........................................................................................................ 27 Government Performance ............................................................................................... 27 Implementing Agency Performance ............................................................................... 28 Monitoring and Evaluation ................................................................................................. 29 Design ............................................................................................................................. 29 Implementation ............................................................................................................... 30 Utilization ....................................................................................................................... 31 7. Lessons ................................................................................................................................ 31 Bibliography ........................................................................................................................... 32 Figure Figure 2.1. Theory of Change for the Sierra Leone Integrated Public Financial Management Reform Project .......................................................................................................................... 5 Tables Table 2.1. Project Components and Subcomponents................................................................ 3 Table 2.2. Project Development Objective Level Indicators .................................................... 7 Table 3.1. Project Cost by Component ..................................................................................... 9 Table 3.2. Source of Funds ....................................................................................................... 9 Table 4.1. Aggregate Government Expenditures .................................................................... 15 Table 4.2. Stock of Expenditure Payment Arrears (Le, millions) .......................................... 15 Table 4.3. Aggregate Government Revenues ......................................................................... 16 Table 4.4. Aggregate Budget Deficits and Surpluses ............................................................. 17 Table 4.5. Deviations in Aggregate Expenditure and Domestic Revenues (percentage) ....... 17 Table 4.6. PEFA Indicators for Credibility............................................................................. 18 Table 4.7. PEFA Indicators for Control .................................................................................. 19 v Table 4.8. Internet and Mobile Application Access ............................................................... 22 Table 4.9. Contracts Using Competitive Methods .................................................................. 22 Table 4.10. PEFA Indicators of Transparency........................................................................ 23 Appendixes Appendix A. Basic Data Sheet................................................................................................ 35 Appendix B. Update of Public Expenditure Financial Accountability Assessment for Measuring Sustainability of Outcomes ................................................................................... 38 Appendix C. Broadening Access to Fiscal and Budget Data.................................................. 71 Appendix D. List of Persons Met ........................................................................................... 73 Principal Ratings ICR* ICR Review* PPAR Outcome Moderately satisfactory Moderately satisfactory Unsatisfactory Risk to Development Moderate Significant High Outcome World Bank Moderately satisfactory Moderately satisfactory Moderately unsatisfactory Performance Borrower Performance Moderately satisfactory Moderately satisfactory Moderately unsatisfactory * The Implementation Completion and Results Report (ICR) is a self-evaluation by the responsible World Bank Global Practice. The ICR Review is an intermediate IEG product that seeks to independently validate the findings of the ICR. Key Staff Responsible Task Manager or Division Chief or Sector Project Leader Director Country Director Appraisal Vivek Srivastava Anand Rajaram Ishac Diwan Completion Roberto O. Panzardi Guenter Heidenhof Yusupha Crookes vii IEG Mission: Improving World Bank Group development results through excellence in independent evaluation. About this Report The Independent Evaluation Group (IEG) assesses the programs and activities of the World Bank for two purposes: first, to ensure the integrity of the World Bank’s self-evaluation process and to verify that the World Bank’s work is producing the expected results, and second, to help develop improved directions, policies, and procedures through the dissemination of lessons drawn from experience. As part of this work, IEG annually assesses 20–25 percent of the World Bank’s lending operations through fieldwork. In selecting operations for assessment, preference is given to those that are innovative, large, or complex; those that are relevant to upcoming studies or country evaluations; those for which executive directors or World Bank management have requested assessments; and those that are likely to generate important lessons. To prepare a Project Performance Assessment Report (PPAR), IEG staff examine project files and other documents, visit the borrowing country to discuss the operation with the government, and other in-country stakeholders, interview World Bank staff and other donor agency staff both at headquarters and in local offices as appropriate, and apply other evaluative methods as needed. Each PPAR is subject to technical peer review, internal IEG panel review, and management approval. Once cleared internally, the PPAR is commented on by the responsible World Bank country management unit. The PPAR is also sent to the borrower for review. IEG incorporates both World Bank and borrower comments as appropriate, and the borrowers’ comments are attached to the document that is sent to the World Bank’s Board of Executive Directors. After an assessment report has been sent to the Board, it is disclosed to the public. About the IEG Rating System for Public Sector Evaluations IEG’s use of multiple evaluation methods offers both rigor and a necessary level of flexibility to adapt to lending instrument, project design, or sectoral approach. IEG evaluators all apply the same basic method to arrive at their project ratings. Following is the definition and rating scale used for each evaluation criterion (additional information is available on the IEG website: http://ieg.worldbankgroup.org). Outcome: The extent to which the operation’s major relevant objectives were achieved, or are expected to be achieved, efficiently. The rating has three dimensions: relevance, efficacy, and efficiency. Relevance includes relevance of objectives and relevance of design. Relevance of objectives is the extent to which the project’s objectives are consistent with the country’s current development priorities and with current World Bank country and sectoral assistance strategies and corporate goals (expressed in poverty reduction strategy papers, country assistance strategies, sector strategy papers, and operational policies). Relevance of design is the extent to which the project’s design is consistent with the stated objectives. Efficacy is the extent to which the project’s objectives were achieved, or are expected to be achieved, taking into account their relative importance. Efficiency is the extent to which the project achieved, or is expected to achieve, a return higher than the opportunity cost of capital and benefits at least cost compared with alternatives. The efficiency dimension is not applied to development policy operations, which provide general budget support. Possible ratings for outcome: highly satisfactory, satisfactory, moderately satisfactory, moderately unsatisfactory, unsatisfactory, and highly unsatisfactory. Risk to Development Outcome: The risk, at the time of evaluation, that development outcomes (or expected outcomes) will not be maintained (or realized). Possible ratings for risk to development outcome: high, significant, moderate, negligible to low, and not evaluable. World Bank Performance: The extent to which services provided by the World Bank ensured quality at entry of the operation and supported effective implementation through appropriate supervision (including ensuring adequate transition arrangements for regular operation of supported activities after loan or credit closing, toward the achievement of development outcomes). The rating has two dimensions: quality at entry and quality of supervision. Possible ratings for World Bank performance: highly satisfactory, satisfactory, moderately satisfactory, moderately unsatisfactory, unsatisfactory, and highly unsatisfactory. Borrower Performance: The extent to which the borrower (including the government and implementing agency or agencies) ensured quality of preparation and implementation, and complied with covenants and agreements, toward the achievement of development outcomes. The rating has two dimensions: government performance and implementing agency(ies) performance. Possible ratings for borrower performance: highly satisfactory, satisfactory, moderately satisfactory, moderately unsatisfactory, unsatisfactory, and highly unsatisfactory. Preface This Project Performance Assessment Report (PPAR) assesses the Sierra Leone Integrated Public Financial Management Reform Project (IPFMRP), which was approved on June 4, 2009, and closed on July 31, 2014. The objective of the project was to sustainably improve the credibility, control, and transparency of fiscal and budget management. This report was prepared by Arun Arya, senior evaluation officer of the Independent Evaluation Group (IEG), with support from Buffy Bailor, consultant. After Mr. Arya’s departure from IEG, Ismail Arslan assumed task leadership. The report was peer reviewed by Stefano Migliorisi and panel reviewed by Robert M. Lacey. It benefited from the strategic guidance of Pablo Fajnzylber, manager, Economic Management and Country Programs. The report presents findings based on a review of the project appraisal document, the Implementation Completion and Results Report, relevant laws of Sierra Leone, and the Government of Sierra Leone’s budget documents, policies, strategies, action plans, and progress reports. Information was also obtained from stakeholder interviews during an IEG mission in Sierra Leone in September 2017 and from interviews with World Bank staff. An informal Public Expenditure Financial Accountability (PEFA) update for selected PEFA indicators was carried out in 2017 to determine the sustainability of reform supported by the IPFMRP. IEG acknowledges the cooperation and support provided by staff of the Public Financial Management Reforms Unit of the Government of Sierra Leone; staff of other relevant government institutions, civil society, and the private sector; and World Bank staff in the country office and in Washington, DC. Special thanks are also due to staff from the U.K. Department for International Development and the European Union for providing valuable inputs in assessing the experience and future strategy for public financial management reforms in Sierra Leone. The names of interviewees appear in appendix D. Following standard IEG procedures, a copy of the draft PPAR was sent to relevant government officials and its agencies for their review and feedback. No comments were received from the borrower. Summary This Project Performance Assessment Report (PPAR) reviews the Integrated Public Financial Management Reform Project in Sierra Leone, which was approved on June 4, 2009, and became effective on December 15, 2009. It closed on July 31, 2014. The project cost of $23.44 million was financed by a $4 million grant from the International Development Association (IDA) and $17.44 million in grants from the U.K. Department for International Development (DFID) and the European Union (EU), which were channeled through a multi- donor trust fund administered by IDA. The Government of Sierra Leone made a counterpart contribution of $2 million. The project’s objective was to sustainably improve the credibility, control, and transparency of fiscal and budget management. Five components made up the project: (i) strengthening macrofiscal coordination and budget management, (ii) reinforcing the control system for improved service delivery, (iii) strengthening central finance functions, (iv) assisting oversight by nonstate actors (NSAs), and (v) project management. Project development objectives (PDOs) were highly relevant at both appraisal and closure. Relevance of design, however, is rated modest. Design did not address the binding constraints to achieving the development objectives, including extrabudgetary funds being more than one-third of total public expenditure; nonsubmission to parliament of supplementary budgets, overcentralized payment authorizations, and low Internet density. Further, the design did not take into account the roles of academia, the media, and the private sector, thereby limiting the potential increase in demand for good governance. Weaknesses in the results framework limited the ability to measure the full effect of activities on development objectives. To assess efficacy, the PPAR goes beyond the original indicators. It uses relevant and up-to- date data on the status of applicable Public Expenditure and Financial Accountability (PEFA) indicators and Open Budget Index scores together with additional information provided by the government on budget and fiscal performance during 2014, 2015, and 2016. There are three main conclusions: First, the objective of sustainably improving the credibility of fiscal and budget management has not been achieved. Aggregate expenditure outturns and aggregate domestic revenue outturns were significantly higher than the original approved budget, both during the project duration and subsequently. Expenditure payment arrears rose from below 1 percent during project implementation to more than 17 percent in 2016. Large extrabudgetary funds persist, and supplementary budgets are not presented to parliament despite huge deviations in expenditure and revenue outturns. These factors undermine the credibility of budget and fiscal management. Second, the objective of sustainably improving control of fiscal and budget management was only partially achieved. Control function improved due to the use of the commitment control system, which temporarily contained public expenditure under budget ceilings and cash availability. However, according to the 2014 PEFA assessment, predictability of the availability of funds for committed expenditures, payroll controls, and the internal audit function was weak and rated D+ at project closure. No improvement and some deterioration x occurred compared with the 2007 baseline. Similarly, the scores for recording and management of cash balances, debt, and guarantees and the effectiveness of internal controls for nonsalary expenditure remained unchanged at C+ and C, respectively. Persistent overcentralization of payment authorizations at the level of the Ministry of Finance adversely affects efficiency, transparency, and accountability. Third, the objective of sustainably improving the transparency of fiscal and budget management was not achieved. Public access to key fiscal and budget documents saw virtually no increase. The number of procurement contracts subject to open competition fell. Performance declined or remained static in terms of transparency of intergovernmental fiscal relations, transparency of taxpayer obligations and liabilities, and availability of information on resources received by service delivery units. Little evidence was found that the project constituted an efficient use of World Bank resources. Calculations of economic rate of return or comparative cost estimate took place either at appraisal or at completion. A reported 23 percent reduction in fiscal losses in 2013 resulted from the streamlining of administrative guidelines on tax exemptions; 365 retired staff were removed from the payroll following validation of personnel data. However, efficiency was adversely affected by delays caused by retroactive financing and suspension of a key contract as well as by tardiness in implementation. Moreover, the assessment of the project’s efficiency is negatively affected by the limited improvements in the status of the most relevant PDO indicators. Given the high relevance of objectives, modest relevance of design, modest efficacy of all the three objectives, and a modest efficiency, outcome is rated unsatisfactory. The risk to the limited development outcome is rated high. A large number of extrabudgetary funds, nonsubmission to parliament of supplementary budgets, and overcentralization of payment authorizations persist. External partner support to public financial management (PFM) reform has been curtailed since both DFID and the EU withdrew from the follow-on PFM Improvement and Consolidation Project in February 2017. This followed an independent assessment of the implementation of the Integrated Financial Management Information System (IFMIS). The two partners began some parallel activities as bilateral support, which do not fully compensate their withdrawal of support to the sector. Funding for continuing and deepening the project-supported reforms is lacking. The government finds it difficult to finance the maintenance and licensing cost of the IFMIS. The World Bank suspended the system’s service provider, and project funds cannot be used to pay for it. Several staff trained by the project left. The World Bank’s performance is rated moderately unsatisfactory. Design did not acknowledge or address several binding constraints that prevented the project from achieving its objectives. The support to NSA narrowly focused on civil society and did not include support to academia, the private sector, or the media, thereby reducing its potential impact. Selection of indicators in the results framework was not comprehensive, making it difficult to measure the impact of all project activities on the PDOs. The project design overestimated government institutional capacity and political willingness in managing such a complex project and reforms, and overcoming potential shortcomings. xi While the supervision team provided timely feedback and implementation support, its focus on development impact was insufficient. Several weaknesses—no scanning of payment vouchers, lack of focus on the Human Resource Management and Performance Budgeting modules of IFMIS, inadequate use and compatibility of various IFMIS systems, and weak implementation of the NSA component—went unaddressed. The government’s performance is also rated moderately unsatisfactory. The two ministers of the Ministry of Finance and Economic Development provided adequate leadership and demonstrated overall government commitment during the project life in strengthening PFM across ministries, departments, and agencies and local councils. Capacity building initiatives funded by the project were successfully implemented. The government-appointed appropriate project staff maintained adequate implementation arrangements and generally provided adequate funding. However, little effort was made to address the binding constraints to achieving project objectives. Moreover, the government could have been more forceful on improving public procurement performance. Despite training and other capacity building activities supported by the project, procurement outcomes fell well below expectations. The implementing agency performance is rated moderately satisfactory. The project management team was appropriately staffed to implement the project and to use technical assistance effectively for procurement management. The teamimplemented agreed work plans and provided timely progress reports. Fiduciary requirements were met, including maintenance of financial records, submission of financial statements, and the commissioning of external audits. However, extended delays slowed the finalizing the 2012–13 procurement plan. While component managers were technically competent, they were not made fully aware about the procurement plans associated with their components. The major lessons from this project follow. • In the absence of a conducive PFM policy environment, there are clear limits to what can be achieved through investment project financing alone. In this case, while most of the project’s targeted outputs materialized, its development objectives were largely unachieved. The investment project financing design did not, and perhaps could not be expected to, address the binding policy constraints to achieving the outcomes. In the absence of PFM policy reforms to address constraints, the link between supported activities and intended outcomes was weak. The project could have achieved better results if the policy constraints to enhanced credibility, control, and transparency of fiscal and budget management had been addressed in a parallel manner, including with World Bank support through other lending and nonlending instruments (for example, development policy financing), supported and preceded by Advisory Services and Analytics. • Effective support for improving the demand for good governance can benefit from broadening support beyond civil society organizations to include academia, the media, and the private sector. This project included only civil society organizations. Although it had good outreach with communities, the communities lacked the necessary capacity to analyze economic issues and budget documents. xii • In the context of low Internet density, effective public dissemination of state documents calls for combining online publication with alternative means of diffusion. Where there is very low Internet density, as in Sierra Leone, online dissemination is unlikely to reach more than a small part of the population. Alternative methods of dissemination would have been more productive in attaining the goal of greater transparency. In this case, the problem was exacerbated by the lack of publicity on the availability of documents on the web, thereby limiting outreach even to those with Internet access. • Effective and sustainable World Bank leadership of multi-donor support to PFM reforms requires a continuous effort by staff to consult with external partners. In this case, this effort appears insufficient. DFID and the EU were reportedly disappointed that the reform strategy was tailored almost exclusively to address World Bank priorities. Little room was left for other development partners to contribute, which led them to withdraw funding from the follow-on project. A more partnership-oriented approach could have secured their continued participation. • Effective World Bank support for designing and installing information technology systems requires tailoring solutions to address borrower capacity limitations. For example, it is crucial to ensure all equipment critical to achieving project outcomes is in place and that staff can operate it. Where the installation and implementation of a management information system involves a consultancy, it is desirable to avoid the high costs and dependency frequently associated with time- based contracts. Rather, contracts should be structured around system implementation and allow for building the client’s capacity on a sustainable basis. Auguste Tano Kouame Director Human Development and Economic Management Independent Evaluation Group 1. Background and Context Country Background Sierra Leone is eligible for International Development Association (IDA) lending and designated a fragility, conflict, and violence–affected country. Improved governance remains an essential condition for preserving social peace. Main challenges for Sierra Leone are its legacy of civil war preceded by decades of mismanagement, its culture of patronage politics, and its susceptibility to negative influences associated with the diamond trade. The World Bank Group engages and operates in the country under its Joint Country Strategy with the African Development Bank (AfDB), with the current portfolio aligned with the government’s Agenda for Prosperity and the Post-Ebola Recovery Program (World Bank and AfDB 2010). Project Context Sierra Leone’s poverty reduction strategy of 2008–12 set the context for the Integrated Public Financial Management Reform Project (IPFMRP) at appraisal. The strategy wasbased on a foundation of good governance, which included making public financial management (PFM) stronger (IMF 2008a). The effective functioning of PFM institutions and systems at central and local levels was considered a high priority because of the importance of PFM to effectively implementing a poverty reduction strategy, reducing opportunities for corruption, and attracting budget support from development partners. The 2008 IPFMRP of the Ministry of Finance and Economic Development (MFED) stressed an urgent need to address the lack of budget credibility and predictability, fiscal management challenges, weaknesses in expenditure control (including payroll), and low levels of transparency. The IPFMRP was approved on June 4, 2009, and became effective on December 15, 2009. Expected to close on July 31, 2013, it was extended twice and closed on July 31, 2014. The total project cost was $23.44 million. It was financed by a $4 million grant from IDA combined with grants of $17.44 million from the U.K. Department for International Development (DFID) and the European Union (EU), which were channeled through a multi- donor trust fund (MDTF) administered by IDA. Sierra Leone contributed counterpart funding of $2 million. 2. Objectives, Design, and Their Relevance Objectives The project’s objective was to sustainably improve the credibility, control, and transparency of fiscal and budget management. Relevance of Objectives The objectives of the project were relevant at appraisal and at completion. The project built on the progress made by the government in strengthening its PFM framework and systems by implementing most recommendations made by the Country 2 Financial Accountability Assessment of 2002 (World Bank 2002). This process was supported by IDA through the Institutional Reform and Capacity Building Project (IRCBP) and by technical assistance from EU, DFID, and the AfDB, which improved the timeliness of financial statements and audit reports. However, budget credibility deteriorated sharply in 2007, an election year. The 2007 Public Expenditure Financial Accountability (PEFA) Report and other analyses, such as the report from the International Monetary Fund entitled Sierra Leone: Implementing Public Financial Management Reforms, indicated areas of major weakness remained. They identified lack of budget credibility and predictability, fiscal management challenges, weaknesses in expenditure control (including payroll), and low levels of transparency as major challenges. The project’s objectives were relevant to addressing them. The objectives were also relevant to the World Bank Group’s 2010–13 country assistance strategy, in particular, its goals to improve strategic orientation of the budget, transparency of the resource envelope and allocation, accountability of spending units, transparency and accountability of procurement, and staffing (World Bank and AfDB 2010). The project’s objectives were also consistent with the government’s 2013–18 poverty reduction strategy, which focused on governance, more decentralization, and stronger PFM (Sierra Leone 2013). The relevance of objectives is rated high. Design COMPONENTS The project consisted of 5 components and 17 subcomponents (table 2.1). 3 Table 2.1. Project Components and Subcomponents Components Subcomponents 1. Strengthening macro-fiscal coordination and 1.1 Macro-fiscal management and budget budget management formulation (Appraisal: $2.1 million; Actual: $3.3 million) 1.2 Budget execution, establishment of a predictable budget execution process 1.3 Debt management 2. Reinforcing the control system for improved 2.1 Legal framework service delivery 2.2 Public procurement (Appraisal: $3 million; Actual: $4 million) 2.3 Accounting, recording, and reporting 2.4 Payroll strengthening 2.5 Other aspects of internal control 2.6 Reinforcement of parliamentary oversight 3. Strengthening central finance functions 3.1 Financial management information systems (Appraisal: $12.2 million; Actual: $8.6 million) 3.2 Information and communication technology 3.3 PFM capacity building 3.4 Salary support for MFED civil service staff 3.5 Salary support for MFED consultants 4. Assisting NSA’s oversight 4.1 Support for the engagement on PFM reforms (Appraisal $1 million; Actual $1.1 million) 4.2 Demand-driven subprojects 5. Project management 5.1 Coordination, administration, procurement, (Appraisal $1.9 million; Actual $1.9 million) financial management, monitoring, and evaluation Source: World Bank 2009. Note: MFED = Ministry of Finance and Economic Development; NSA = nonstate actor; PFM = public financial management. IMPLEMENTATION ARRANGEMENTS DFID and EU resources were pooled with IDA resources in an IDA-administered MDTF. Under this arrangement, development partners supported all aspects of the project using one set of arrangements for disbursement, reporting, financial management, and procurement. Oversight and monitoring of the reform program was undertaken by a PFM Oversight Committee, chaired by the MFED’s financial secretary and including stakeholders from across the government. Within the MFED, a director for PFM reforms was appointed with responsibility for developing and implementing the reform agenda, coordinating across components, and overseeing daily project operations. The director reported to the financial secretary and was supported by staff drawn from the Public Financial Management Reform Unit (PFMRU), which was initially supported by the IRCBP and later integrated into the MFED. For project administration, the director was supported by an appropriately staffed Integrated Project Administration Unit (IPAU), which included a project manager, procurement specialist, accountant, coordination officer for nonstate actors (NSAs), and support staff. The unit’s tasks included general project management, administration of subgrant agreements with NSAs, management of bank accounts, fiduciary management, and monitoring and evaluation (M&E). The unit assisted the PFMRU director in managing the centralized training capacity development program and all operating expenses associated with the project as well as 4 preparing annual work plans and budgets, getting approval for these from the PFM Oversight Committee, and reporting on project progress. Fiduciary responsibility rested with the MFED (the implementing agency). Various agencies, such as the National Public Procurement Authority (NPPA), parliament, and several ministries, departments, and agencies (MDAs) outside the MFED as well as departments and directorates within the MFED. These included the Accountant General’s Department (AGD), Budget Bureau, Debt Management Unit, and other MDAs responsible for implementing the technical aspects of the project with support from the IPAU for procurement, financial management, and M&E. Since activities within subcomponents cut across MDAs, various PFM oversight subcommittees were responsible for the technical aspects of implementing relevant subcomponents, and their activities were coordinated by the PFMRU director. For each subcomponent, a lead person was assigned from the MFED or relevant MDA. All procurements and payments were made by the IPAU, and no resources were passed on to the beneficiary agencies except for the subgrants to NSAs. MONITORING AND EVALUATION DESIGN The government had adopted the PEFA framework for the overall monitoring of PFM performance. The independent 2007 PEFA assessment acted as the baseline, with future targets set against it. The PEFA indicators guided the development of a common framework of objectives, indicators, and targets to phase and monitor the project. It was expected that the independent PEFA assessment would be repeated every two years. The PEFA assessments were in fact conducted in 2010 and 2014, and provided a sufficient basis to assess the project and PFM outcomes at the country level. Relevance of Design RELEVANCE OF PROJECT DESIGN TO OBJECTIVES A clear statement of objectives is logically linked with intermediate outcomes, outputs, and activities. The project was focused on the core PFM reform priorities of the government’s IPFMRP, which aimed to achieve fiscal discipline; strategic, efficient, and effective allocation and use of funds; and value for money and probity in the use of public funds. The theory of change identified by the IEG team is presented in figure 2.1. The theory of change shows how different components and their activities contribute to achieving intermediate outcomes, and how the intermediate outcomes would contribute to project outcomes. 5 Figure 2.1. Theory of Change for the Sierra Leone Integrated Public Financial Management Reform Project Inputs and Activities Intermediate Outcomes and Outcomes World Bank Group interventions Outputs Strengthening of public financial that support public financial Achievement of results management management with regard to the identified activities Improved Strengthening • Macro-fiscal aggregates on track credibility of fiscal macrofiscal • Core strategic budget/MTEF process adopted coordination and and budget • Strengthened policy framework and data for budget aid management management management • Budget executed in line with approved budget • Strengthened legal, policy and institutional framework for debt management • Revised legal and regulatory framework reflecting new business processes adopted and in use. Improved control of Reinforcing the • MDA procurement practice strengthened fiscal and budget control system for Comprehensiveness, timeliness and quality management improved service of accounting, recording and reporting delivery enhanced • Strengthened payroll management • Internal audit functioning and reports utilized • Strengthened parliamentary oversight of public finance Strengthening central finance • Secured IFMIS functioning across central functions government and strengthened FM in local Improved councils transparency of • Effective ICT support function established and fiscal and budget operational • Increase in trained permanent PFM staff management Assisting nonstate actors’ across central government oversight Nonstate actors exercise increased scrutiny of over public finances Impacts Fiscal discipline; strategic, efficient and effective allocation and use of funds; and value for money and probity in the use of public funds Source: Integrated Public Financial Management Reform Project Appraisal Document (2009) and Independent Evaluation Group mission. Note: FM = financial management; ICT = information and communication technology; IFMIS = Integrated Financial Management Information System; MDA = ministries, departments, and agencies; MTEF = Medium-Term Expenditure Framework; PFM = public financial management. However, design did not account for several binding constraints to achieving the PDOs. First, a large number of extrabudgetary funds bypass parliamentary scrutiny. The 2014 PEFA report states that extrabudgetary funds were 36 percent of the total public 6 expenditure in 2012 (Coffey 2014). Sierra Leone had about 130 autonomous government agencies managing public funds outside the purview of budget law. These included donor- funded projects administered by project implementation units (PIUs), subvented agencies, and departmental revenues not collected by the National Revenue Authority and not brought into government accounts. Some of the major ones were Sierra Leone Road Transport Authority, National Social Security and Insurance Trust, National Agricultural Response Program, National Telecommunications, and the Sierra Leone Broadcasting Corporation. With a large portion of public expenditure outside the budget law’s remit, budget credibility is highly compromised. Second, supplementary budgets involving important deviations in both revenue and expenditure from the original budgets were not submitted to parliament. They relied exclusively on executive orders, which undermined the parliamentary process, adversely affected fiscal credibility and control, violating provisions of the Constitution of Sierra Leone of 1991 and the Government Budgeting and Accountability Act (GBAA) of 2005.1,2 Through 2016, supplementary budgets were presented to parliament in only one year (2014) as a prior action for the World Bank’s development policy financing—the Sierra Leone Emergency Economic and Fiscal Support Operation. Third, approval of payments during project implementation is overcentralized. All payments require approval of the MFED.3 This requirement continued after project closure and was still prevalent at the time of the IEG mission. It leads to payment delays and adversely affects PFM efficiency. With effective ex ante control over expenditure by commitment control and cash management systems (provided by the Integrated Financial Management Information System [IFMIS]), the MFED could ensure that expenditure was within cash and budget limits, so that the ex post control over expenditure at the payment level was an extra, redundant layer. The MDAs could be authorized to issue payments if they conformed to budget and cash limits. Moreover, the decision-making process surrounding these approvals was not incorporated into the IFMIS, with a consequent lack of accountability and transparency for the most important expenditure decisions taken at the political level. Fourth, the government’s plan to enhance transparency was by way of publishing relevant fiscal, budgetary, and procurement documents on its websites. However, Internet users in Sierra Leone were only 0.3 percent of total population at the baseline in 2008, making such a process insufficient for reaching the public at large. The number of Internet users rose to only 4 percent by project end, but no course correction was made. In recent years, the topics of budget transparency and open data increasingly received more attention by PFM professionals. Based on the most recent studies, to achieve a genuine transparency, it is important not only that governments publish data on websites but also that data they disclose are meaningful and easily accessible by citizens, including through mobile application (Dener and Min 2013). This new trend may be relevant but not fully tested yet for low-income countries. Appendix C presents successful examples of using mobile apps for disseminating fiscal and budget data. In Sierra Leone, mobile cellular subscriptions in 2008 were 18.2 percent of the population, increasing to 34.8 percent in 2012 and 76.7 percent in 2014. This increase could have provided an opportunity for the government to disseminate 7 budget and fiscal information to people by developing a mobile application over the IFMIS database. However, the IFMIS design did not include a web interface. In 2016, mobile cellular subscriptions increased to 97.6 percent, but no efforts were made to develop a mobile application for fiscal information. Support to NSAs was limited to civil society organizations and nongovernment organizations. While they were effective in community outreach, they lacked PFM capacity and a strong voice to demand good governance from the central government. The design excluded academia and the private sector from the NSAs to be supported, which could have been more effective in this regard. QUALITY OF RESULTS FRAMEWORK The results framework was explicitly aligned with PEFA indicators, and the PDO was measured by several indicators (table 2.2). Table 2.2. Project Development Objective Level Indicators PDO Indicators Credibility (i) Variance between total actual primary expenditure (outturn) and total originally budgeted primary expenditures (percent). (ii) The share of actual to budgeted spending on pro-poor spending priorities compared with the ratio of actual to budget expenditures for all other discretionary primary expenditures. (iii) Domestic expenditure arrears (total from all years) as percent of total expenditure for year (excluding interest and donor-financed project expenditure). Control (i) Submission of audited public accounts by auditor general to parliament within 12 months. Transparency (i) Financial and procurement documents listed in PI-10 and in accordance with requirements of the Government Budgeting and Accountability Act and Public Procurement Act, published in the Sierra Leone Gazette or on Ministry of Finance and Economic Development or related institution websites. (ii) Public contracts signed by local councils and 20 largest budget heads (ministries, departments, and agencies) using competitive bidding according to the National Procurement Law and regulations (percent), disaggregated by local councils and 20 largest budget heads. Source: World Bank 2009. Note: PDO = project development objective. The results framework included indicators designed to assess the achievements of the PDO along three dimensions—credibility, control, and transparency. The credibility-related indicators looked at the reduction in variance between total actual primary expenditures and total originally budgeted primary expenditures, reduction of payment arrears, and increase in the share of actual budgeted spending on pro-poor expenditures. These were important indicators to assess the credibility of budget and fiscal management. However, the project design could have been improved further. To assess the impact on the credibility of budget and fiscal management, for instance, it is important to consider changes in the composition of expenditure outturn compared with the original approved budget and aggregate revenue outturns compared with the original approved budget, measured by PEFA indicators PI-2 and 8 PI-3, respectively. The project activities of component 1, Strengthening Macro-Fiscal Coordination and Budget Management, were designed to contribute to these outcomes. One PDO and 12 intermediate results indicators were designed to assess the achievement of the control dimension on fiscal and budget management. The PDO indicator for assessing the control dimension was the timeliness of the submission of audited public accounts by the auditor general to parliament. As in the case of the credibility dimension, to assess the control dimension on budget and fiscal management, it is important to consider some relevant PEFA indicators. These were the changes in the predictability and availability of funds for commitment of expenditures (PI-16); recording and management of cash balances, debt, and guarantees (PI-17); effectiveness of payroll controls (PI-18); effectiveness of internal controls for nonsalary expenditure (PI-20); and internal audit effectiveness (PI-21). Two PDO indicators were used to assess achievement of the transparency dimension of fiscal and budget management. One measured the increase in the number of financial and procurement documents listed in PEFA indicator (PI-10). The second measured the increase in percentage of contracts in compliance with the provisions of procurement legislation and regulations. To assess the impact on the transparency of budget and fiscal management, in addition to the PDO indicators included in the results framework, it would have been useful to consider the change in the extent of unreported government operations (PI-7); transparency of intergovernmental fiscal relations (PI-8); transparency of taxpayer obligations and liabilities (PI-13); and availability of information on resources received by service delivery units (PI-23). In addition, the Open Budget Index (OBI),4 which is a comparative measure of budget transparency, could have been considered while assessing the impact of project activities on the PDO. The relevance of design is rated modest. 3. Implementation Planned versus Actual Expenditure by Component The project costs are broken out by component in table 3.1. 9 Table 3.1. Project Cost by Component Appraisal Project Cost by Estimate Revised Disbursed Component ($, millions) ($, millions) ($, millions) Strengthening macro- 2.15 3.70 3.63 fiscal coordination and budget management Reinforcing the control 3.04 4.48 4.40 system for improved service delivery Strengthening central 15.32 9.94 9.32 finance functions Assisting nonstate actors’ 1.00 1.24 1.21 oversight Project management 1.93 2.15 2.11 Total financing 23.44 21.49 20.66 required Source: World Bank 2015. At appraisal, out of the project cost of $23.44 million, the maximum allocation of $15.32 million was for component 3, Strengthening of Central Finance Functions. Most of it was spent on rolling out IFMIS. However, the allocation was significantly reduced to $9.94 million on actual implementation as the result of the cancelation of salary support to the MFED’s civil service staff and consultants. Allocation of the remaining components was revised upward, and all components spent more than their original allocation. Table 3.2 shows the project cost by source of funds. About 75 percent was financed by the IDA-administered MDTF, which was funded by DFID and EU. A slight downward revision was nade in the cost to $21.49 million, in which IDA’s share of $4 million remained intact. The government’s share was halved, and MDTF’s share was reduced by 5 percent. Table 3.2. Source of Funds Appraisal Estimate Revised Estimate Percentage of Source ($, millions) ($, millions) Appraisal Government 2.00 1.00 50 IDA 4.00 4.00 100 IDA-administered 17.44 16.49 95 MDTF (funded by DFID and EU) Total 23.44 21.49 87 Source: World Bank 2015. Note: EU = European Union; DFID = U.K. Department for International Development; IDA = International Development Association; MDTF = multi-donor trust fund. 10 Implementation Experience The project was approved on June 4, 2009, became effective on December 15, 2009, and was expected to close on July 31, 2013. The four level 2 restructurings did not involve changing the PDO.5 A midterm review occurred on June 18, 2012. The closing date was extended twice for a total of one year: (i) from July 31, 2013, to March 31, 2014 (8 months); and (ii) from March 31, 2014, to July 31, 2014 (4 months) to account for implementation delays and allow time to complete project activities, including IFMIS rollout, approval of the GBAA and subordinate fiscal rules and debt regulations, the commissioning of an IFMIS audit by an independent reviewer, and preparation of a MFED IFMIS strategy plan inclusive of an information and communication technology security policy. The project closed on July 31, 2014. During implementation, the World Bank became concerned about: (i) the lack of clarity in the roles and relationship between the PFMRU director and the IPAU; (ii) nonalignment of functional responsibilities resting with MFED component managers and accountability for results resting with PFMRU; and (iii) the risk of weakening the project’s internal control mechanisms by consolidating fiduciary, project implementation, and coordination responsibilities under one unit. This problem was subsequently resolved when the government agreed to move away from the PIU model to pilot the mainstreaming and integration of project management fiduciary systems into governmental systems. The IPAU continued as an implementation unit for other externally funded projects in the MFED. A major weakness in IFMIS implementation was identified by the IEG mission. While MDAs submitted their payment requests to the MFED through a workflow on IFMIS, the vouchers and supporting documents could not be submitted in the system because scanners were not procured and supplied to the MDAs. This led to manual submissions of vouchers and supporting documents for every payment request by MDAs to MFED. These documents were sent daily by MDAs to the AGD. Although payment requests were processed online, this process was preceded by manual scrutiny of vouchers and supporting documents by the staff of AGD. This led to delays, subjectivity, inefficiency, and lack of transparency. There were also serious technical limitations in the design of the system’s architecture.6 No business continuity and disaster recovery plans have been made to enable continued activities of the IFMIS in the event of technical failure. The server components at a remote backup facility are unlikely to offer an adequate environment for IFMIS technical operations in case of server loss at the data center. No wide area network links connect MDAs to the facility. Given a disaster at the primary data center, it would not be possible to transfer operations to the facility. Moreover, the facility is unmanned, not operational, and incomplete, and physical access is inadequately secured. An independent assessment of IFMIS implementation by a consulting firm was commissioned by the government in October 2016 to identify any operational, functional, and system design deficiencies. The main findings are as follows: 11 • Key modules, such as Performance Budgeting and Civil Service Management, were not in use, despite being installed after government approval. The government is not deriving full value from the IFMIS. • The inability to use these modules results from the fact that applications were not tested, accepted, deployed, and operationalized as planned. • Nonusage of these modules subsequently led key user departments, such as the Budget Bureau and the Human Resource Management Office (HRMO), to consider acquiring separate applications to use in their key processes. This would create a risk of suboptimal discrete systems that would not be fully integrated. • For modules in use, mainly Financial Accountability and Purchasing Accountability, not all licenses are fully used. There are 86 unused licenses in Financial Accountability and 90 in Purchasing Accountability modules. There is also an indication that licenses for the Civil Service Payroll application may be overused (32 licenses). • Complete and seamless integration has been made more difficult, if not impossible, by the fact that the Performance Budgeting and Financial Accountability modules are running a lower version of the application while the Civil Service Management is running a higher and more recent one. The sole provider of the IFMIS system, FreeBalance Inc. of Canada, was suspended by the World Bank for future contracts on January 29, 2013, based on investigations by the Integrity Vice Presidency. IFMIS had been rolled out to 12 MDAs. The suspension prevented the government from authorizing any further payments to the provider from project funds or from entering into a new contract. This negatively affected IFMIS implementation in the last two years of the project. To date, the government has used its own funds to make payments to FreeBalance of about $932,000, mostly for annual product support and maintenance license fees. The IEG mission found that the PETRA Financials accounting software,7 introduced in 19 local councils, was not operating because of lack of maintenance and technical support. Personnel of the consulting company, which had installed the system, were not allowed by their governments to visit countries suffering from the Ebola outbreak. For local councils, the government wants to replace the software with IFMIS. Project support to the HRMO was focused on Payroll Management and neglected Human Resource Management modules. Although the system had both, only the Payroll Management module was activated. The project primarily served the needs of AGD and did not give adequate attention to the needs of HRMO. The lack of integration between Human Resource Management and Payroll Management affected the integrity of the payroll. Tensions arose between higher-paid project contract staff and regular civil servants. These tensions were exacerbated by the project-supported measures under subcomponents 3.4 and 3.5 (Salary Support for MFED Civil Service Staff and Consultants) to mainstream the externally funded staff and consultants into the civil service. The World Bank was obliged to cancel funding under both subcomponents because the government failed to meet disbursement conditions.8 Subcomponent 3.4 was formally canceled, although subcomponent 12 3.5 was not. To avoid disruption of project activities, local technical assistants were funded by the government during the life of the project, and a holistic human resource strategy covering a wider group of MDAs as well as MFED was developed. The implementation of NSA oversight was focused on local authorities. Oversight of the central government was largely ignored, although it was included in project design. During implementation, project management—including the project oversight committee, PFMRU director, and NSA director—took no steps to hold the central government accountable due to lack of political will. The NSA Secretariat did not have its own office but was housed within MFED, which adversely affected independence from the central government. The space provided for the NSA Secretariat allowed barely enough room for two officials to work. Such facilities did not permit periodic meetings with civil society organizations. Although the NSA Secretariat published the Citizens’ Budget and focused on improving its content, there was little focus on dissemination. Because of lack of planning and resource allocation, only 200 copies were printed and distributed. IMPLEMENTATION OF MONITORING AND EVALUATION The project’s results framework was consistent with the PEFA framework, which enabled PFMRU and IPAU to measure outcomes of PFM reforms and project objectives coherently. A focal person at IPAU was appointed to collect, compile, and analyze data received from the managers of the different project components. This arrangement helped ensure the timely compilation of M&E data for inclusion in the quarterly progress reports shared with the World Bank. Both original and revised indicators were closely monitored and progress assessed based on periodic targets throughout the project’s life. The results framework was assessed as part of the June 2012 midterm review, and proposed revisions were formally adopted and implemented, albeit with considerable delay. According to the April 2013 restructuring memorandum, the changes to the results framework allowed adding and modifying outcome indicators that had proved difficult to monitor, such as Public Expenditure Tracking Surveys (PETS) or certain PEFA indicators. However, the results framework was not adjusted to reflect the cancelation of subcomponent 3.4 (Salary Support for MFED Civil Service Staff), which accounted for $3 million, or about 15 percent, of the original project cost. The M&E framework of the project appraisal document was used for most of the project life. The revised M&E framework, which was discussed at a midterm review in June 2012, was used for about six months in 2014. It is unclear to what extent these progress reports were shared and discussed with other external partners not part of the MDTF but engaged in PFM and other public sector reforms. SAFEGUARDS COMPLIANCE There were no environmental safeguards issues for this category C project, and the World Bank’s environmental safeguards policies were not triggered. The project did not identify any social safeguards issues at appraisal. Hence, its compliance was not verified at completion. The project was designed to contribute to the 13 government’s higher-level objective of improving strategic allocation of resources and quality of expenditures required for the implementation of the Poverty Reduction Strategy Paper. The achievement of the PDO was based on the key indicator related to the share of actual to budgeted spending on pro-poor expenditure priorities; however, this was not achieved. Gender aspects were given due attention by the government, especially in the implementation of capacity building initiatives with 22 percent of the beneficiaries being women. Support to NSAs was intended to lead to broadened participation of civil society in its interactions with the state. FINANCIAL MANAGEMENT AND PROCUREMENT Overall financial management performance was rated satisfactory by the supervision team throughout the project life. Close oversight by the World Bank’s financial management specialist; provision of appropriate training to the project accountant in processing withdrawal applications; and the task team’s close coordination with the relevant development partners helped ensure timely transfer of MDTF funds to the government. No disbursement issue was mentioned. Project audits were carried out as stipulated and interim financial reports were prepared in a timely manner, but with some qualifications. During the early part of the project, procurement responsibility was vested in a central project administration unit within MFED under the direct supervision of the PFMRU director and the overall purview of the financial secretary. These arrangements were interrupted when MFED proposed an IPAU that resulted in some changes in procurement arrangements. The World Bank raised its concerns, and the government agreed to mainstream the functions of project administration, including procurement, within the government’s organizational structure. The MFED-appointed procurement officer from within the civil service had little experience with World Bank procurement requirements and was provided training. This led to delays, but procurement nonetheless complied with related operational policies. 4. Achievement of the Objectives To assess the efficacy of this project, the IEG team went beyond the PDO indicators in the results framework and examined all the relevant indicators within PEFA and the OBI. To assess the sustainability of project outcomes, the team carried out an informal examination of the status of relevant PEFA indicators in 2017, based on additional information requested and received from the authorities on budget and fiscal performance during 2014, 2015, and 2016. The results are presented in appendix B. This section discusses main conclusions on project efficacy. Objective 1: Sustainable Improvement in Credibility OUTPUTS The following outputs were produced: 14 • A budget framework paper was developed as part of the budget process, and MDAs prepared their budget in accordance with the Medium-Term Expenditure Framework guidelines. • A new macroeconomic forecasting model (the Sierra Leone Integrated Macroeconomic Model) was introduced and training conducted for key forecasting staff within MFED, the National Revenue Authority, and Statistics Sierra Leone. • Budget committees established in MDAs and local councils were trained on strategic planning and forecasting. • A medium-term debt strategy was developed, incorporating policy on public debts and guarantees. • A debt sustainability analysis was prepared and updated regularly. • The Cash Management Committee was established in MFED. OUTCOMES Three PDO-level indicators measured outcomes under this objective. Variance between total actual primary expenditure (outturn) and total originally budgeted primary expenditures (percentage). This indicator was adopted after the midterm review in March 2013. The target was to keep the deviation below 10 percent. At project closure, it was 7.9 percent. Although the target was achieved, according to the 2014 PEFA report, the deviation from the budget was more than 15 percent in all the three years under assessment. Year 2013 was not part of the PEFA assessment, and the variance in that year was reported to be 7.9 percent, which was on target. Expenditure deviations beyond 2013 were examined by the IEG team as an informal PEFA assessment. Table B.1 in appendix B shows the expenditure deviations in years 2014–16 in the context of measuring PEFA indicator PI-1. For most of the project’s duration, the deviation has been higher than 15 percent, reflecting poor budget credibility. The IEG mission examined the deviation after project closure and found the deviation to be above 10 percent for two out of the three years (table 4.1). In 2016, it was more than 25 percent. 15 Table 4.1. Aggregate Government Expenditures Budget Actual Expenditure Deviation Year (Le, millions) (Le, millions) (Le, millions) (Percentage) 2010 1,138,339.52 1,488,282.83 349,943.31 30.7 2011 1,371,167.82 1,594,971.9 223,804.08 16.3 2012 1,754,034.6 2,169,121.4 415,086.80 23.7 2013 2,047,307.13 2,208,411.52 161,104.39 7.9 2014 2,628,240.20 3,018,169.80 389,929.60 14.8 2015 3,002,337.20 3,266,211.90 263,874. 70 8.8 2016 3,401,788.60 4,263,274.10 861,485.50 25.3 Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: Excluding debt servicing and donor-funded projects. Similar to the aggregate expenditure outturns, a consistently high variance is seen in the composition of expenditures. See table B.4 for the extent of the variance in expenditure composition by program, administrative, or functional classification for three years, excluding contingency items during 2014–16 for measuring PEFA indicator PI- 2.1. Le = Leone. The share of actual to budgeted spending on pro-poor spending priorities compared with the ratio of actual to budget expenditures for all other discretionary primary expenditures. At the baseline, actual and budgeted pro-poor expenditure was 57.5 percent, whereas the actual and budgeted expenditure in other categories was 69 percent. The former should be greater than the latter by project closure. However, at closure, actual pro-poor expenditure was 114.1 percent of the budget; for other categories, it was 130 percent of the budget. Thus, the target was not achieved. At baseline, both pro-poor and other expenditure fell short of the budget by 30–35 percent. At closure, both exceeded the budget by 14– 30 percent. Again, this reflects poorly on budget credibility. Domestic expenditure arrears (total from all years) as a percentage of total annual expenditure (excluding interest and donor-financed project expenditure). The ratio was 9.4 percent at baseline, and the target was 8 percent. The status of expenditure arrears in 2011–12 was reviewed by the PEFA assessment and that between 2014 and 2016 was assessed by the IEG team in its informal PEFA assessment in 2017. See table B.9 for expenditure arrears while measuring PEFA indicator PI-22. Table 4.2 notes the stock of expenditure payment arrears. Table 4.2. Stock of Expenditure Payment Arrears (Le, millions) Category 2011 2012 2013 2014 2015 2016 Arrearsa 13,187 11,312 10,419 — 9,442 455,050 Expenditureb 1,714,743 2,201,984 2,208,412 1,700,916 1,751,727 2,617,892 Percentage 0.77 0.51 0.47 0.00 0.54 17.38 Source: Public Debt Management Division Audited Accounts—Ministry of Finance and Economic Development Budget Bureau estimates. Note: Le = Leone. a. Stock of arrears: Checks payable and checks on hold at the Bank of Sierra Leone and the Accountant General Department. b. Expenditure: Total primary expenditure including nonsalary, noninterest recurring expenditure, domestic capital expenditure, and contingency expenditure. The stock of arrears was under 1 percent of expenditure during the last three years of the project period (2011–13), and it was also decreasing. The project fully achieved the 16 target. Even after closure, the stock of expenditure arrears was kept below 1 percent in 2014 and 2015. However, it increased sharply to above 17 percent in 2016, which can be partially explained by the austerity measures undertaken by the government. The IEG team also examined the extent of deviation in the actual domestic revenue collection compared with original approved budget (PEFA indicator PI-3). The findings are presented in table 4.3. Table 4.3. Aggregate Government Revenues Budget Actual Revenue Deviation Year (Le, millions) (Le, millions) (Le, millions) (Percentage) 2010 844,110 1,007,626 163,516 19.3 2011 1,156,570 1,462,100 305,530 26.4 2012 1,606,932 1,869,205 262,273 16.3 2013 2,066,076 2,318,243 252,167 12.2 2014 3,313,749 3,403,155 89,406 2.7 2015 3,231,200 3,381,705 150,495 4.7 2016 3,832,242 3,615,440 −216,802 −5.6 Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: Information regarding years 2010, 2011, and 2012 are based on Public Expenditure and Financial Accountability reports and those related with 2014, 2015, and 2016 are based findings from the Independent Evaluation Group mission. Le = Leone. Revenue outturns have been significantly higher than the original budget projections throughout the project implementation period. They continued to be high even after the project period but by a lesser margin. In 2016, the revenue outturns were lower than the budget projection. See table B.7 for assessing the aggregate revenue outturn compared with the original approved budget while measuring PEFA indicator PI-3. While revenue outturns were greater than the original approved budget, they were less than the expenditure outturns, leading to deficits in 2010, 2011, and 2012. Budget surpluses appeared in 2013, 2014, and 2015, but the deficit returned in 2016 (see table 4.4). 17 Table 4.4. Aggregate Budget Deficits and Surpluses Revenue Deviation Budget Actual (Le, (Percentage) Yeara (Le, millions) (Le, millions) millions) 2010 1,007,626 1,488,283 −480,657 −32.3 2011 1,462,100 1,594,972 −132,872 −8.3 2012 1,869,205 2,169,121 −299,916 −13.8 2013 2,318,243 2,208,411 109,832 5.0 2014 3,403,155 3,018,170 384,985 12.8 2015 3,381,705 3,266,212 115,493 3.5 2016 3,615,440 4,263,274 −647,834 −15.2 Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: a. Information for 2010, 2011, and 2012 is based on Public Expenditure and Financial Accountability reports. Information for 2014, 2015, and 2016 is based on findings of Independent Evaluation Group mission. Deviations in expenditure and revenue outturns were significantly large during the project implementation period, leading to low budget credibility. Deviations in expenditure outturns continued to be high even after project implementation, while the deviation in the revenue outturns fell significantly. However, the budget credibility was hit by very high deviations in expenditure outturns after the project implementation period. Table 4.5. Deviations in Aggregate Expenditure and Domestic Revenues (percentage) Deviation in Aggregate Expenditure in Deviation in Domestic Revenue Collection Respect to Original Approved Budget in Respect to Original Budget Estimates Yeara (PI-1) (PI-3) 2010 24.1 19.4 2011 20.4 26.4 2012 14.2 16.3 2013 7.9 12.2 2014 17.5 2.7 2015 27.1 4.7 2016 31.2 −5.7 Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: a. Information for 2010, 2011, and 2012 is based on Public Expenditure and Financial Accountability reports. Information for 2014, 2015, and 2016 is based on findings of Independent Evaluation Group mission. PI = PEFA indicator. In sum, the budget as approved by parliament lacks credibility. Actual expenditure and revenues vary significantly from the originally approved budget (table 9), and the variance is not presented for parliamentary approval, although required by the existing legal framework. Instead, the changes are approved at the executive level. Expenditure payment arrears were low during project implementation, but rose subsequently. Combined with the presence of many extrabudgetary funds, these factors lead to low credibility of budgetary and fiscal management (table 4.6). 18 Table 4.6. PEFA Indicators for Credibility PEFA Score Year PI 2007 2010 2014 2017a PI-2. Composition of C C D+ D expenditure outturn compared with original approved budget PI-3. Aggregate revenue B C D C outturn compared with original approved budget Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: PEFA = Public Expenditure and Financial Accountability; PI = PEFA indicator. a. PEFA score 2017 is estimated based on an informal PEFA assessment carried out by the Independent Evaluation Group team during its mission. Please refer to appendix B for details. The project’s contribution to the achievement of this objective is rated modest. Objective 2: Sustainable Improvement in Control OUTPUTS The following outputs were produced: • A draft PFM bill was approved by the cabinet. • A draft bill to revise the National Public Procurement Act of 2004 was approved by the cabinet. • A Public Debts Management Law was enacted by parliament in 2011. • Procurement was institutionalized in MDAs through the establishment of procurement cadres. • The websites of MFED and NPPA were revamped to publish key fiscal, financial, and procurement information. • The oversight and scrutiny of public finances by parliament committees was enhanced by organizing reviews regionally, clearing backlogs of audited public accounts, and making their reviews open to public. • The IFMIS (Treasury, Payroll, and Purchase modules) was rolled out to 12 key MDAs. • The PETRA Financials accounting package was introduced to all 19 local councils. • Yearly financial statements for local councils are now produced in a timely manner. 19 OUTCOMES One PDO-level indicator that measured this objective was “submission of audited Public Accounts by the auditor general to Parliament within 12 months.” This target was achieved but determined in such a way that no improvement is seen over the baseline. The PEFA report of 2010 gave a C score to PEFA indicator PI-26.2, which implied that the auditor general was already submitting its audit reports to the legislature within 12 months of the end of the period covered. The bar was not raised to 8 months or 4 months, which would have achieved a PEFA rating of B or A, respectively, by closure. In addition to the PDO indicator, relevant PEFA indicators were considered by the IEG team (see table 4.7). They show little improvement compared with the baseline. Table 4.7. PEFA Indicators for Control PEFA Score Year PEFA Indicator 2007 2010 2014 2017a PI-16. Predictability in the availability of C+ C+ D+ D+ funds for commitment of expenditures PI-17. Recording and management of cash C+ C+ C+ C balances, debt and guarantees PI-18. Effectiveness of payroll controls D+ D+ D+ D PI-20. Effectiveness of internal controls for C+ C+ C C nonsalary expenditure PI-21. Effectiveness of internal audit D+ D+ D+ D Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: PEFA = Public Expenditure and Financial Accountability; PI = PEFA indicator. a. PEFA score 2017 is estimated based on an informal PEFA assessment carried out by the Independent Evaluation Group team during its mission. Please refer to appendix B for details. The IEG mission observed the control function in the ministries of Education and Science and Technology; Health and Sanitation; and Works, Housing, and Infrastructure. Some improvement is seen due to the use of the commitment control system in the IFMIS. MDAs cannot issue local purchase orders unless authorized by the IFMIS based on budget and cash availability. MDAs are required to presubmit PETS forms, including details of the cash required for each quarter. The MFED reviews its revenue collections and cash availability before authorizing cash release for each quarter. Expenditure commitment control procedures thus operate through budget ceilings set within IFMIS, which limits commitments to actual cash availability. They do not comprehensively cover all expenditures such as those that are off budget. Evidence of an expenditure arrears build-up appears in recent years. A strong degree of centralization of payment authorizations over and above commitment control and cash management functions persists. Although commitment control and cash management functions under the IFMIS work well, all payments require the approval of the MFED. In 2016, payments above Le 500 million accounted for 86.8 percent of all payments, requiring the approval of the MFED. Payments between Le 7.5 million and Le 499 million accounted for 13.03 percent of all payments, requiring approval by the Minister of State. Payments of less than Le 7.5 million, accounting for 0.17 percent of all 20 payments, were approved by the AGD. The decision-making process was kept out of the IFMIS workflow, which ended at the AGD level. The payment vouchers were submitted manually to the ministers’ offices—a flaw in the design of the IFMIS. Among payment authorizations, 99.83 percent was kept out of the IFMIS control mechanism on the premise that information technology systems should be installed only at the bureaucratic level, not at the political level. Because of its decision-making authority, the political level also needs to operate under the purview of the IFMIS. Without it, there is a lack of accountability and transparency for the most important decisions made at the political level. These procedures also lead to delays and adversely affect the efficiency of PFM. In sum, overcentralization of payment approval undermines the project’s objective of enhanced budget control through the installation of the IFMIS. A workflow submission to MFED for approval of payments within the respective thresholds of each authority could have been included in the IFMIS, but was not. Predictability in the availability of funds for commitment of expenditures is an important dimension of efficient public expenditure management. While strong central control over the forecasting and disbursement of available funds can be effective in ensuring that total budget ceilings are not exceeded, it has deleterious management and operational consequences for MDAs and government suppliers. This is particularly the case for goods and services, which are not statutory payments (such as salaries) and receive a lower priority. Cash management is not fully transparent. Although the cash position is centrally monitored on a weekly basis by MFED, this does not always translate into a high level of predictability and reliability at the MDA operational level. Frequently, a significant number of in-year budget adjustments are made and not always carried out in a transparent manner. Serious weaknesses persist in payroll control and in the internal audit function. Various payroll verification exercises were undertaken during 2011–13, but the integrity of payroll remains significantly undermined by incomplete personnel records and the personnel database, and the inconsistencies between them. The internal audit function is operational for the most important central government entities. Of the total number of audits conducted in a year by the internal audit team, about 20 percent are system audits and about 80 percent are transaction audits. Internal audit reports are issued regularly for most audited entities and distributed to the audited entity, the MFED, and the Supreme Audit Institution. However, the audits do not meet international auditing standards, and their recommendations are frequently ignored (Coffey 2014). The quality of parliamentary oversight is compromised by the fact that supplementary budgets are not presented to the legislature. In-year adjustments made by the executive, do not, therefore, receive parliament’s approval. This violates both the constitution and budget law. According to the 2014 PEFA report, Sierra Leone has about 130 autonomous government agencies, which are asked to report to the Treasury and Other Government Accounts Service in the AGD.9 In 2012, these funds spent Le 810,293 million, or 36 percent of the total expenditure of Le 2,201,984 million.10 As per section 107 of the Public Financial Management Act of 2016, any new subvented agency, other entity in the central government, or social security fund may not be established or incorporated except by an act of parliament. 21 This would certainly regulate future establishment of new extrabudgetary funds. However, it has no bearing on those extrabudgetary funds in existence at the time of passing this act. In 2012, about 36 percent of total expenditure passed through extrabudgetary funds without parliamentary scrutiny. Overall, the project’s contribution to this objective is rated modest. Objective 3: Sustainable Improvement in Transparency OUTPUTS The following outputs were produced: • An NSA Secretariat was established in MFED, and a network of NSAs was also created. • More than 170 NSAs were trained on financial management. • The Citizens’ Budget is published annually. • Demand-driven grants are awarded to 15 NSAs nationwide. OUTCOMES Two PDO indicators measured the achievement of this objective. Financial and procurement documents listed in PEFA indicator PI-10 and in accordance with GBAA and NPPA requirements are published in the Sierra Leone Gazette or on the websites of MFED or related institutions. At the baseline in 2006, only three out of six required fiscal documents were published. Annual budget documentation, year-end financial statements, and external audit reports were made available to the public through publication on the government’s website. However, the information and documents related to in-year budget execution, contract awards above $100,000 equivalent, and resources available to primary service units were not made available. The target was set at six, but by project closure, no improvement was registered. The budget and fiscal documents published on the government’s website likely hardly reached the public because of low Internet density. At the baseline, the percentage of the population using the Internet was 0.3 percent (table 4.8). During implementation, it increased, but was still only 4 percent at closure. This severely limited access to budget and fiscal documents. A new global trend in budget transparency is making budget-related information and data easily accessible by citizens using mobile communication devices. This recent undertaking may be relevant but not fully tested yet for low-income countries (see appendix C for successful examples of using mobile apps to disseminate fiscal and budget data). In Sierra Leone, mobile subscribers made up 97.6 percent of the population in 2016. The growth in number of people having mobile subscriptions based on World Development Indicators is shown in table 4.8. However, IFMIS data were not optimized for mobile phones. 22 Table 4.8. Internet and Mobile Application Access Technology 2008 2009 2010 2011 2012 2013 2014 2015 2016 Individuals using the Internet 0.3 0.3 0.6 0.9 2.5 4.0 6.1 6.3 11.8 (percent of population) Mobile cellular subscriptions 18.2 20.6 34.8 36.4 35.0 65.7 76.7 89.5 97.6 (per 100 people) Source: World Development Indicators database, access November 21, 2017. The proportion of public contracts signed by local councils and the 20 largest MDAs using competitive bidding should have been 95 percent per project target,11 according to the National Procurement Law and regulations. Achievement by the project in 2013 was 89.3 percent, and the target was partially achieved. This is reflected by a PEFA PI-19.1 score of A in 2014.12 In 2010, Sierra Leone had received a score of B, which meant that only 50– 75 percent of contracts were awarded on the basis of open competition. There was, therefore a notable improvement. However, over the period 2010–17, there was a downward trend in the total number of contracts above the threshold that were subject to open competition (table 4.9). Table 4.9. Contracts Using Competitive Methods Contracts Above Contracts Threshold Contracts Above Awarded Through Awarded with Total Contracts Threshold Open Competition Open Competition Year (no.) (no.) (no.) (percent) 2010 2,439 353 317 89.8 2011 1,063 160 126 78.8 2012 1,630 241 153 63.5 2014 — — — 30.0 2015 — — — 40.0 2016 — — — 50.0 2017 — — — 65.0 Source: Coffey 2014 for 2010–12; National Public Procurement Authority data for 2014–17. Note: — = not available. The reasons cited for the use of less competitive methods are late disbursement of funds, which leads to tight deadlines and time frames resulting in the use of less competitive methods. Ceiling restrictions and the subsequent small budget allocations and release of funds mean that the entities do not have enough funds for large quantities requiring open competition. Some development partners have their own procurement procedures that may differ from local regulations in terms of the threshold. The capacity of procurement officers to follow procedures is limited, resulting in instances of splitting procurement packages to be below thresholds. In response to national emergencies, use of restricted methods is considered best practice. Regardless of reasons, the reduction in the use of open competition shows lack of transparency. 23 During the IEG mission’s stakeholder consultations, the Sierra Leone Chamber of Agribusiness Development expressed concerns about the transparency of public procurement processes. The chamber stated that the government was not receptive to the idea of openness regarding either bidding documents or the award of contracts. Contract agreements are not accessible, and the procurement process is not transparent. Those who are awarded some of the contracts are not qualified. Road and bridge constructions were prime examples. In addition to the PDO indicators, three other changes in the relevant PEFA indicators were analyzed by the IEG team. These show stable or declining performance compared with the baseline, except for transparency of intergovernmental fiscal relations and taxpayer obligations and liabilities (table 4.10). Table 4.10. PEFA Indicators of Transparency PEFA Score Year PEFA Indicator 2007 2010 2014 2017a PI-7. Extent of unreported government Not rated Not rated D D operations PI-8. Transparency of intergovernmental B A B A fiscal relations PI-13. Transparency of taxpayer obligations C+ D+ C B and liabilities PI-23. Availability of information on A A C D resources received by service delivery units Source: Ministry of Finance and Economic Development; Budget Bureau / Integrated Financial Management Information System. Note: PEFA = Public Expenditure and Financial Accountability; PI = PEFA indicator. a. PEFA score 2017 is based on an informal assessment carried by IEG team during its mission. Please refer to appendix B for details. Sierra Leone’s OBI score and ranking were also considered. Sierra Leone received a score of 52 percent on budget transparency in 2015, implying that the government provided limited budget information to the public. Nevertheless, this score was higher than the global average of 45 percent and those of other countries in the Region, including Ghana (51 percent), Liberia (38 percent), São Tomé and Príncipe (29 percent), Nigeria (24 percent), and Equatorial Guinea (4 percent). Sierra Leone’s score increased substantially from 39 percent in 2012 to 52 percent in 2015. The OBI score partially reflects the fact that, since 2012, the government has increased the availability of budget information by publishing the Citizens’ Budget and improving the comprehensiveness of the enacted budget, in-year reports, and year-end report. Nonetheless, no progress was seen in making the prebudget statement and mid-year review available to the public. Moreover, publication on the government website meant that the material was inaccessible to a large percentage of the population because of low Internet density. Only 200 hard copies of the Citizens’ Budget were printed. Transparency is also compromised by the high level of unreported government expenditure, especially from the large number of autonomous government agencies. Extrabudgetary expenditures need to be minimized and brought below 10 percent of total 24 expenditure per international standards. There is a need for preparing an inventory of extrabudgetary funds, followed by a policy decision to retain only those funds with objectives that cannot be achieved if operating under current budget laws. The rest could be slowly phased out by amending the legal and regulatory framework. Unless this is done, the transparency of budget and fiscal management will continue to be seriously compromised. Overall, the project’s contribution to this objective is rated modest. 5. Efficiency At appraisal, the benefits of the project were expected to flow from improved budget credibility, better internal controls, enhanced oversight, and increased transparency in fiscal and budget management. But, identifying and quantifying the direct and indirect financial, economic, and social benefits and attributing outcomes to project interventions was considered difficult, and an economic rate of return was not calculated. Some efficiency gains were identified at completion. First, the streamlining of administrative guidelines on tax exemptions and the establishment of clear criteria for eligibility for individuals and companies led to a reported 23 percent reduction of fiscal losses in 2013.13 Second, the validation of personnel data using the IFMIS Personnel Management module resulted in the removal of 365 staff from the payroll because of their retirement status. Third, savings likely resulted from a least-cost approach in public procurement and from alignment of fees for training activities to national and international standards, although these were not quantified. However, efficiency was negatively affected by (i) delays related to the unplanned retroactive financing of the contract related to IFMIS development; (ii) suspension of the IFMIS service provider contract, which negatively affected MFED’s ability to get appropriate technical support and additional software licenses; (iii) persistent delays in making and implementing decisions and in formalizing revisions resulting from agreed modifications; and (iv) the second extension of the closing date that led to the inability of government to use the remaining MDTF funds of about $950,000.14 Moreover, the assessment of the project’s efficiency is negatively affected by limited improvements in the status of most relevant indicators of the PDOs. Efficiency is rated modest. 6. Ratings Outcome The project’s objectives were relevant to country context at appraisal and closure, to the World Bank’s country assistance strategy, and to the government’s IPFMRP and poverty reduction strategy (World Bank and AfDB 2010). However, the relevance of design is rated modest because it did not address the binding constraints to public sector reform, including a large number of extrabudgetary funds, overcentralization of payment authorizations, and nonsubmission of supplementary budgets for parliamentary approval. Design was also 25 compromised by the lack of a strong and comprehensive results framework. Efficacy in achieving all three project objectives—credibility, control, and transparency of fiscal and budget management—is rated modest. Efficiency is also rated modest. There were no quantified analyses of value for money and several important operational and administrative inefficiencies. Taken together, these amount to major shortcomings, and outcome is rated unsatisfactory. Risk to Development Outcome Several issues pose a high risk to the sustainability of the limited development outcomes. First, many extrabudgetary funds, overcentralization of payment authorizations, and nonsubmission of supplementary budgets to parliament persist more than three years after project closure. Unless these binding constraints are addressed, the project’s objectives cannot be substantially achieved. Second, both key development partners—DFID and the EU—expressed their dissatisfaction with the project’s performance after the independent assessment of IFMIS. Both withdrew their funding from the follow-on Public Financial Management Improvement Consolidation Project (PFMICP). They also expressed dissatisfaction at the tight tailoring of the PFM reform strategy to World Bank priorities, leaving little room for them to contribute to the thinking on reforms. This, together with the late release of AfDB financing, led to a funding shortfall for almost all PFMICP activities, which undermines the sustainability of reforms supported by the project under review. Although the World Bank participates in formulating a new PFM reform strategy for 2018–21, other external partners have initiated parallel technical support to the Ministry of Finance that enhances the risk of fragmentation and inconsistencies. Third, after more than 10 years of IFMIS implementation, the service provider has still not established a local office to provide in-country local support or transferred knowledge to a local firm. This poses a substantial risk to the sustainability of the IFMIS. Fourth, because external partner funding was withdrawn from the PFMICP follow-up, the government does not have sufficient funds to pay for the maintenance, further rollout, or upgrade of IFMIS software. The problem is exacerbated by the more recent further debarment of Free Balance Inc. by the World Bank for six months, in relation to a sanctionable misconduct under the IFMIS project for Liberia. This debarment precludes the company from participating in World Bank-financed projects. Hence, the burden of maintaining and further rolling out IFMIS will be on the government. In view of serious resource constraints, this poses a major risk to sustainability. If the annual maintenance fee to the service provider is not paid, the government will lose the right to use the software. Fifth, some key staff trained on IFMIS and PETRA systems have been transferred to other departments and agencies, while others have moved to the private sector where they receive higher compensation. The risk to development outcome is rated high. 26 World Bank Performance QUALITY AT ENTRY The project built on the progress made by the government in strengthening its PFM framework and systems by implementing most of the recommendations made by the Country Financial Accountability Assessment of 2002, supported by IDA through IRCBP and technical assistance from the DFID, EU, and AfDB (World Bank 2002). The project had solid analytical underpinnings from the 2007 PEFA assessment and other analyses of PFM systems conducted by the International Monetary Fund. The project design was also informed by the government’s 2008 IPFMRP. The project rightly focused on the government’s agenda and priorities, and the World Bank team correctly identified the need for appropriate institutional arrangements for managing the reform process. The project design applied several lessons drawn from experience in PFM and public sector management reforms in other countries and the PFM reforms undertaken in Sierra Leone.15 The task team was composed of the right skills mix during preparation and coordinated well with other development partners. The project had a clear theory of change with a clear statement of objectives logically linked with intermediate outcomes, outputs, and activities. However, the project design did not attempt to address the binding constraints to achieving project outcomes and overestimated government institutional capacity and political willingness in managing such a complex project. The project lacked a strong and comprehensive results framework. Although it was explicitly aligned with PEFA indicators, it included certain indicators that were not affected by project activities or were insignificant since the target was already achieved at the baseline. The remaining indicators though relevant, did not fully measure the impact of the project activities on the PDO. Quality at entry is rated moderately unsatisfactory. QUALITY OF SUPERVISION The supervision team was responsive to the requests of the project team and in the timely provision of quality feedback and implementation support. Supervision missions were undertaken at least twice each year from December 2009 to July 2014. This allowed the task team to provide implementation support to relevant stakeholders. The World Bank’s team and its approach was proactive, demonstrating flexibility in addressing issues. The midterm review in June 2012 assessed progress, highlighted issues, and proposed remedial actions that were agreed on and led to more focused attention on resolving priority issues. The World Bank’s task team was composed of qualified technical experts who possessed experience in the various key project areas (that is, accounting and budgeting, internal controls, IFMIS and information and communication technology, procurement, parliament, civil society and NSAs, and project management). As such, the implementation 27 support missions yielded detailed and constructive recommendations for reaching or meeting practical and beneficial results. The World Bank closely coordinated with other external partners to ensure that relevant funds were mobilized and transferred to the government in accordance with agreed modalities on the use of these funds. The World Bank provided the needed fiduciary oversight on the use of both the IDA and MDTF grants. Supervision tended to focus more on process than development impact. • The supervision team did not detect the absence of scanning payment vouchers. The voucher and supporting documents could not be submitted in the IFMIS, thus requiring manual submissions for every payment request by MDAs. This adversely affected efficiency and transparency of budget management. • The Human Resource Management and Performance Budgeting modules were not activated. The integrity of the payroll was adversely affected. Moreover, the Performance Budgeting module was not in use, despite having been approved by the government and installed. Without using the implemented applications, the government was not deriving full value from the investment made in the IFMIS. • Issues concerning the compatibility of various IFMIS modules remained unaddressed. Not all licenses were fully used, even for Financial Accountability and Purchasing Accountability, which were in use. Moreover, the Performance Budgeting and Financial Accountability modules were running a lower version of the application, while the Civil Service Management was running a more recent one, making seamless integration difficult, if not impossible. • The NSA oversight activities were weakly implemented. The supervision team did not address the fact that the central government was not held accountable by NSA, the office accommodations for the staff implementing the component was inadequate, and the dissemination strategy was ineffective. The quality of World Bank supervision is rated moderately unsatisfactory. Together, these leads to an overall rating of World Bank performance of moderately unsatisfactory. Borrower Performance GOVERNMENT PERFORMANCE The two ministers of MFED provided adequate leadership and demonstrated commitment by strengthening PFM across MDAs and local councils. Capacity building initiatives funded by the project were successfully implemented. Key officials from the NPPA, Revenue and Tax Policy Division, Budget Bureau, Economic Policy and Research Unit, Internal Audit Department, NSA Secretariat, PFMRU, Information and Communication Technologies Division, Public Debts Management Division, Multilateral Projects Division, and AGD attended relevant courses within Sierra Leone and outside of the country during the 28 project period. Approximately 290 officials across MDAs and local councils were also trained at the Institute of Public Administration and Management, University of Sierra Leone, in the areas of public procurement, internal audit, HRM, and administration. The government appointed appropriate project staff, maintained adequate implementation arrangements, and provided sufficient counterpart funding for most of the project life. The government used the IFMIS for better control of public expenditure. The AGD exercised more effective oversight of MDAs. Appropriate diagnostics like the PEFA assessment were used to identify weaknesses in PFM and set its priorities. The government collaborated well with all external partners and used their support to improve different aspects of the PFM. The government used project support in revising the GBAA, financial management regulations, NPPA, and the local council financial administration regulations. Draft bills were approved by the cabinet during implementation and were enacted after project completion. However, the government could have been more forceful about improving public procurement performance. Despite training and other capacity building activities supported by the project, procurement outcomes fell well below expectations. As noted in DFID comments and in the 2014 annual report of the auditor general, procurement practices remain a problem and have a track record of hindering overall progress of PFM reforms, including PFM project implementation. The government could not meet the disbursement condition associated with subcomponent 3.4, Salary Support for MFED Civil Service Staff, resulting in the nonutilization of funds and the cancelation of the subcomponent from the project. Although the component was included under the Pay and Performance Project, its implementation could have contributed to more immediate results in terms of institutionalization and professionalization of PFM functions. NSA-related activities could not achieve their expected outcomes because the government did not provide the conditions for the related staff to work independently and be held accountable. Academia, media, and the private sector were not engaged, and efforts only narrowly focused on civil society. The Citizens’ Budget was underfunded, undermining its meaningful dissemination. The government took no action to address the three major binding constraints to achievement of project outcomes—the large number of extrabudgetary funds, nonsubmission of supplementary budgets, and overcentralization in payment authorizations. Government performance is rated moderately unsatisfactory. IMPLEMENTING AGENCY PERFORMANCE The project management team, led by the PFMRU director, was appropriately staffed to implement the project and effectively used technical assistance. It could implement agreed-on work plans and provide timely progress reports. Fiduciary requirements were met, 29 including maintenance of financial records, submission of financial statements, and commissioning of external audits. Progress reports based on the results framework were timely prepared and shared with the World Bank and development partners. The project director frankly discussed project achievements and bottlenecks in coordinating the reform agenda with the World Bank and development partners. However, extended delays occurred in finalizing the 2012–13 procurement plan. The transfer of the senior procurement officer to another MDA in May 2013 and his replacement with a senior procurement officer with limited knowledge of the World Bank’s procurement and financial rules hindered the speed of executing procurement-related activities, including the development of timely procurement plans. This resulted in undue delays in the processing of urgent procurement activities toward the end of the project and a sizable undisbursed amount ($0.947 million). While component managers were technically competent, they were not made fully aware of the procurement plans associated with their components, which led to more undue delays in procurement and implementation. Implementing agency performance toward the end of this project weakened as the implementing agency had to cope up with both the implementation of this project and the design of next phase of PFM reforms (Sierra Leone PFM Reform Strategy 2014–17) and the follow-on PFMICP. Implementing agency performance is rated moderately satisfactory. The overall borrower performance rating is moderately unsatisfactory. Monitoring and Evaluation DESIGN The results framework was explicitly aligned with PEFA indicators; however, it included certain indicators that were not affected by project activities or were insignificant since the target was already achieved at baseline. The remaining indicators, though relevant, did not fully measure the impact of the project activities on the PDO. To assess the impact on the credibility of budget and fiscal management, it was important to consider changes in the composition of expenditure outturn compared with the original approved budget and aggregate revenue outturns compared with the original approved budget. They were not included in results framework. The Quality Enhancement Review also suggested that the revenue variance captures revenue underforecasting as well as overforecasting, and should be measured. Also, the definition of “poverty-reducing expenditures” included in the results framework was problematic because opinions could differ about which part of public expenditure reduces poverty and which does not. Similarly, to assess the impact on control of budget and fiscal management, it was important to consider the changes in the predictability of the availability of funds for commitment of expenditures; recording and management of cash balances, debt, and guarantees; effectiveness of payroll controls; effectiveness of internal controls for nonsalary 30 expenditure; and effectiveness of internal audit. Project activities were contributing to these outcomes, but they were not included in the results framework. To assess the impact on the transparency of budget and fiscal management, in addition to the PDO indicators included in the results framework, it was important to consider the change in the extent of unreported government operations; transparency of intergovernmental fiscal relations; transparency of taxpayer obligations and liabilities; and availability of information on resources received by service delivery units. In addition, OBI was a good measure of budget transparency and could have been included in the project’s results framework. IMPLEMENTATION Periodic M&E of the project was undertaken in a satisfactory manner. With regard to M&E data management, a focal person at the IPAU was appointed to collect, compile, and analyze data received from various component managers. This arrangement helped ensure the timely compilation of M&E data for inclusion in the quarterly progress reports shared with the World Bank. Both the original and revised indicators were closely monitored, and progress was assessed based on periodic targets throughout the project’s life. The results framework was assessed as part of the June 2012 midterm review, and proposed revisions were formally adopted and implemented with considerable delay. Changes to the results framework were made to allow adding and modifying outcome indicators that had proved difficult to monitor, such as PETS or certain PEFA indicators. However, the results framework was not adjusted to reflect the cancelation of component 3.4, Salary Support for MFED Civil Service Staff, and component 3.5, Salary Support for Consultants, which together accounted for $4 million or about 20 percent of the original project cost. The M&E framework of the project appraisal document was used for most of the project life. The revised M&E framework discussed at the June 2012 midterm review was used for about 6 months in 2014. The progress reports were shared and discussed with other donors, including those not part of the MDTF but engaged in PFM and public sector reforms and other country stakeholders beyond the project team. At least two PDO indicators, one of which measured the impact of public spending in key sectors, were partly validated by the results of the PETS carried out by government. Data generated from the PETS were instrumental in identifying areas for corrective action with regard to (i) all essential facilities, including medical supplies and educational materials funded during a particular period, reaching the intended beneficiaries; (ii) proper distribution of essential drugs to health centers; (iii) outcome of the payment of the school fees subsidy; (iv) distribution of textbooks to targeted primary schools; (v) procurement and distribution of seed rice to farmers; and (vi) utilization of public resources at service delivery facilities and civil works carried out by contractors in the reconstruction and rehabilitation of schools. At least two surveys were conducted during the project period. 31 UTILIZATION The government used the M&E reports to measure project progress and PFM reforms more effectively, and the World Bank used them to recommend more focused corrective actions for removing bottlenecks and accelerating implementation. Data were also useful in tracking and monitoring the impact of spending. PETS data allowed MFED to identify areas of improvement in budget formulation and execution. Two PETS reports published in October 2010 and September 2011 were both instrumental in the identification of mitigation measures to correct systematic failures in financial programming and procurement of essential goods on service delivery in key sectors such as health and education. The PETS exercise was not carried out in 2014; however, based on the 2014–17 PFM reform strategy, the government is aware of the usefulness of PETS. It proposes to mainstream these data in a broader public expenditure management system to track service delivery in a continuous and systematic way. Overall, project M&E is rated modest. 7. Lessons The World Bank’s support for PFM reforms in Sierra Leone through the implementation of this project provides potentially useful lessons for the design of similar operations in other fragility, conflict, and violence–affected countries, especially in the Africa Region. • In the absence of a conducive PFM policy environment, there are clear limits to what can be achieved through investment project financing alone. In this case, while most of the project’s targeted outputs materialized, its development objectives were largely unachieved. The investment project financing design did not, and perhaps could not be expected to, address the binding policy constraints to achieving the outcomes. In the absence of PFM policy reforms to address constraints, the link between supported activities and intended outcomes was weak. The project could have achieved better results if the policy constraints to enhanced credibility, control, and transparency of fiscal and budget management had been addressed in a parallel manner, including with World Bank support through other lending and nonlending instruments (for example, development policy financing), supported and preceded by Advisory Services and Analytics. • Effective support for improving the demand for good governance can benefit from broadening support beyond civil society organizations to include academia, the media, and the private sector. This project included only civil society organizations. Although it had good outreach with communities, the communities lacked the necessary capacity to analyze economic issues and budget documents. • In the context of low Internet density, effective public dissemination of state documents calls for combining online publication with alternative means of diffusion. Where there is very low Internet density, as in Sierra Leone, online dissemination is unlikely to reach more than a small part of the population. 32 Alternative methods of dissemination would have been more productive in attaining the goal of greater transparency. In this case, the problem was exacerbated by the lack of publicity on the availability of documents on the web, thereby limiting outreach even to those with Internet access. • Effective and sustainable World Bank leadership of multi-donor support to PFM reforms requires a continuous effort by staff to consult with external partners. In this case, this effort appears insufficient. DFID and the EU were reportedly disappointed that the reform strategy was tailored almost exclusively to address World Bank priorities. Little room was left for other development partners to contribute, which led them to withdraw funding from the follow-on project. A more partnership-oriented approach could have secured their continued participation. • Effective World Bank support for designing and installing information technology systems requires tailoring solutions to address borrower capacity limitations. For example, it is crucial to ensure all equipment critical to achieving project outcomes is in place and that staff can operate it. Where the installation and implementation of a management information system involves a consultancy, it is desirable to avoid the high costs and dependency frequently associated with time- based contracts. Rather, contracts should be structured around system implementation and allow for building the client’s capacity on a sustainable basis. Bibliography Coffey. 2014. Public Expenditure and Financial Accountability (PEFA) Assessment in Sierra Leone Central Government Systems 2010–2012. Reading, UK: Coffey International Development Ltd. Dener, Cem, and Saw Young Min. 2013. Financial Management Information Systems and Open Budget Data: Do Governments Report on Where the Money Goes? Washington, DC: World Bank. https://openknowledge.worldbank.org/handle/10986/15897. IMF (International Monetary Fund). 2008a. Sierra Leone: Implementing Public Financial Reforms. Washington, DC: IMF. ———. 2008b. Sierra Leone: Poverty Reduction Strategy Paper—Progress Report 2005–2007. Washington, DC: IMF. ———. 2016. Sierra Leone: Sixth Review Under the Extended Credit Facility Arrangement, Financing Assurances Review and Request for Waiver for Nonobservance of Performance Criteria-Press Release; and Staff Report. Washington, DC: IMF. Sierra Leone, Government of. 2013. “The Agenda for Prosperity: Road to Middle Income Status.” Sierra Leone’s Third Generation Poverty Reduction Strategy Paper (2013–2018). Freetown: Government of Sierra Leone. World Bank. 2002. Sierra Leone—Country Financial Accountability Assessment: An Assessment of the Central Government. Washington, DC: World Bank. ———. 2008. Sierra Leone—Second Government Reform and Growth Credit Project. Washington, DC: World Bank Group. http://documents.worldbank.org/curated/en/526341468166758167/Sierra-Leone-Second- Government-Reform-and-Growth-Credit-Project. ———. 2009. “Sierra Leone—Integrated Public Financial Management Reform Project.” Project Appraisal Document. Washington, DC: World Bank. 33 http://documents.worldbank.org/curated/en/602731468336617795/Sierra-Leone-Integrated-Public- Financial-Management-Reform-Project. ———. 2015. “Sierra Leone—Public Financial Management Reform Project.” Implementation Completion and Results Report ICR2830. Washington, DC: World Bank Group. http://documents.worldbank.org/curated/en/838821468144869431/Sierra-Leone-Public-Financial- Management-Reform-Project World Bank and AfDB (African Development Bank). 2010 . Sierra Leone—Joint Country Assistance Strategy for FY10–FY13. Washington, DC: World Bank. http://documents.worldbank.org/curated/en/969721468303585015/PDF/522970CAS0CORR1Box3456 18B01PUBLIC10.PDF. 1 Section 27 (2) of the 2005 Government Budgeting and Accountability Act provides that no payment shall be made in excess of the amount granted in appropriation. 2 Section 112 (3) of the 1991 Constitution requires supplementary estimates to be presented to Parliament for approval. Section 112 (4) requires that approved supplementary estimates be presented in Parliament the following year as a supplementary appropriation bill. Section 114 (2c) grants authority to the president to approve expenditures that are not part of the Appropriations Act (in emergency cases). 3 While payments below Le 7.5 million ($1,000) can be approved at the bureaucratic level in the Accountant General’s Department, payments between Le 7.5 million and Le 499 million ($67,000) require approval by the Minister of State and those above Le 499 million require approval by the Minister of Finance and Economic Development. 4 The Open Budget Index assigns countries covered by the Open Budget Survey a transparency score on a 100- point scale using 109 of 140 survey questions. These questions focus specifically on whether the government provides the public with timely access to comprehensive information contained in eight key budget documents in accordance with international good practice standards. 5 The restructurings occurred in (i) December 2010 to allow for the retroactive financing of expenditures incurred during the predecessor project supporting development of the IFMIS; (ii) June 2011 to accommodate “additional” European Union resources (€2 million) and reflect the reallocation of these additional funds across various expenditure categories; (iii) April 2013 to delete Subcomponent 3.4 Salary Support for Ministry of Finance and Economic Development Civil Service Staff and revise the allocation of grant proceeds accordingly; and (iv) March 2014 to reallocate International Development Association and multi-donor trust funds grant proceeds and revise the total grant amount. 6 The architecture is not designed for redundancy, scalability, high availability, business continuity, and resilience. Most information and communication technology components were designed in unified mode with single points of failure and with little space for expansion. There is a high risk that business may be disrupted because of component failure during operations. 7 PETRA Financials is a user friendly and robust financial management system with inherent controls, designed specifically for use by government agencies. With the capability of budgetary control, expenditures and revenue recording, and capture of a multidimensional chart of accounts, PETRA Financials is considered a powerful tool for local government’s financial accountability and control. 8 For any disbursement to be made under component 3.4, a policy for establishing sustainable human resources capacity in the Ministry of Finance and Economic Development in form and substance satisfactory to the Association should have been in place, and a Financing and Implementation Plan consistent with the policy for each fiscal year, in form and substance satisfactory to the International Development Association should have been agreed. 9 The agencies that did not report in 2012 included: Strategy and Policy Unit, National Social Security and Insurance Trust, Sierra Leone Road Transport Authority, National Agricultural Response Program, National Telecommunications, Sierra Leone Roads Authority, Sierra Leone Broadcasting Corporation, and Extractive Industry Transparency Initiative. 34 10 This is based on the Public Expenditure Financial Accountability assessment of 2014. The Independent Evaluation Group mission tried to obtain the current status of extrabudgetary funds from the government, but no comprehensive information was available with any agency. There is a need to do a thorough study to comprehensively identify extrabudgetary funds and remove them in a time-bound manner. 11 Measured in terms of budget heads. 12 Over 75 percent of contracts above the threshold are awarded through open competition. 13 According to the Revenue and Tax Department. 14 About $0.95 million of the multi-donor trust fund allocation was not used due primarily by the suspension of Free Balance Inc., and cancellation of procurement of information and communication technologies equipment during the final months of the project. 15 A key lesson of public financial management (PFM) reforms is that in low-capacity environments such as Sierra Leone there should be an incremental approach, starting with basics (for example, strengthening credible budget and financial accountability). Another lesson is that supporting PFM reforms should seek to stimulate not only the supply of better PFM systems, but also the demand for better management of public resources. Appendix A. Basic Data Sheet INTEGRATED PUBLIC FINANCIAL MANAGEMENT REFORM PROJECT (IDA41460-Trust Fund No.: TF57287, TF56158) Table A.1. Key Project Data ($, millions) Appraisal Actual or Actual as Percentage of Category Estimate Current Estimate Appraisal Estimate Total project costs 20.90 20.66 98.85 IDA 4.00 4.16 104.00 Co-financing 14.90 16.49 110.67 (MDTF) Table A.2. Cumulative Estimated and Actual Disbursements Category FY10 FY11 FY12 FY13 FY14 Appraisal estimate 1.09 2.00 2.90 3.70 4.00 ($, millions) Actual ($, 0.3 1.23 1.71 2.72 4.16 millions) Actual as percent 27.5 61.5 59.0 73.5 1.04 of appraisal Date of final disbursement: 04/17/2014 Table A.3. Project Dates Category Original Actual Initiating memorandum 10/15/2007 12/17/2007 Negotiations 08/28/2008 04/21/2009 Board approval 06/04/2009 06/04/2009 Signing 10/19/2009 10/19/2009 Effectiveness 12/15/2009 12/15/2009 Closing date 07/31/2013 07/31/2014 36 Table A.4. Staff Time and Cost Staff Time and Cost (World Bank budget only) Staff weeks $, thousands (including Stage of Project Cycle (number) travel and consultant costs) Lending/Supervision/ICR FY09 33.71 210,092 FY10 24.86 198,202 FY11 44.96 259,905 FY12 26.70 150,759 FY13 29.33 161,389 FY14 3.58 75,848 Total 162.87 1,056,195 Table A.5. Other Project Data Borrower/Executing Agency: Follow-on Operations Credit Amount Operation Number ($, millions) Board Date Public Financial Management Improvement 5350-SL 28.50 27-November- and Consolidation Project & 6078-SL 2013 Table A.6. Task Team Members Responsibility or Names Title Unit Specialty Lending Vivek Srivastava Senior Public Sector AFTPR Task Team Leader, Public Specialist Sector Sahr Kpundeh Senior Public Sector AFTPR Public Sector Specialist Manush Hrsitov Senior Counsel LEGAF Legal Anton Leis Garcia ST Consultant LEGLA Legal Tony Bennett DFID Consultant DFID Financial Management Oluwe Pratt FM Specialist AFTFM Financial Management Tsri Apronti Procurement Specialist AFTPC Procurement Rajiv Sondhi Senior Finance Officer LOAFC Disbursements Motoki Hayakawa Public Sector Specialist AFTPR Public Sector Supervision/ICR Vivek Srivastava Senior Public Sector GGODR Task Team Leader, Public Specialist Sector Roberto Panzardi Senior Public Sector GGODR Task Team Leader, Public Specialist Sector Ramesh Siva Lead ICT Specialist GTIDR ICT 37 Ismaila Ceesay Lead FM Specialist GGODR Financial Management Tsri Apronti Procurement Specialist AFTPC Procurement Joyce Olubukola FM Specialist AFTFM Financial Management Agunbiade Reynaldo Castro ST Consultant GGODR Operations Fatu Karim-Turay Office Assistant AFMSL Project Support Salieu Jalloh Office Assistant AFMSL Project Support Macmillan Operations Specialist AFTPR Operations Anyanwu Albert Mugera Senior Public Sector AFTPR Public Sector Specialist Sahr Kpundeh Senior Public Sector AFTPR Public Sector Specialist Tony Bennett DFID Consultant DFID Financial Management Mustaqphya Katta Office Assistant AFMSL Project Support Lydie Ahodehou Program Assistant AFTPR Project Support Asli Gurkan Governance Specialist GSURR Governance Young Kyu Kang Senior Public Specialist GGODR Public Sector Shawkat M.Q. Senior Procurement AFTPC Procurement Hasan Specialist Adu-Guyamfie Procurement Specialist AFTPC Procurement Abunyewa Mustapha Sundifu Team Assistant AFMSL Project Support Katta Dan Nicolau Senior FM Specialist AFTPR Financial Management Christopher Gabelle Senior Governance GGODR Governance Specialist Sydney A. Olorunfe FM Specialist GGODR Financial Management Godwin Anders Jensen M&E Specialist GPSOS Monitoring and Evaluation Adama Davida Team Assistant AFMSL Project Support Ginorlei Appendix B. Update of Public Expenditure Financial Accountability Assessment for Measuring Sustainability of Outcomes 1. The Independent Evaluation Group (IEG) mission conducted an informal update of the Public Expenditure Financial Accountability (PEFA) assessment to estimate the sustainability of project outcomes after the project ended. Relevant budget data were sought from the government on relevant PEFA indicators for years 2014, 2015, and 2016. Following are the findings from IEG’s informal 2017 PEFA updates in respect to three major objectives: impact on credibility of budget and fiscal management; impact on control over budget and fiscal management; and impact on the transparency of budget and fiscal management. Impact on Credibility of Budget and Fiscal Management PI-1. Aggregate Expenditure Outturn Compared with Original Approved Budget (i) The difference between actual primary expenditure and the originally budgeted primary expenditure (that is, excluding debt service charges, but also excluding externally financed project expenditure; table B.1). Table B.1. Aggregate Government Expenditure Budget Actual Difference Year (Le, millions) (Le, millions) (Le, millions) (Percentage) 2014 2,628,240.20 3,018,169.80 574,158.50 114.85 2015 3,002,337.20 3,266,211.90 861,293.0 108.80 2016 3,401,788.60 4,263,274.10 1,304,418.60 125.30 Source: Ministry of Finance and Economic Development Budget Bureau. Score: C. Aggregate expenditure outturn was between 85 percent and 115 percent of the approved aggregate budgeted expenditure in at least two of the last three year (table B.2). Table B.2. Aggregate Government Expenditure: PEFA Rating PEFA Score Year PI 2017 2014 PI-1 Aggregate Overall Score: C Overall Score: D Expenditure Outturn • Dimension (i) Score: C • Dimension (i) Score D Note: PEFA = Public Expenditure Financial Accountability; PI = PEFA indicator. PI-2. Expenditure Composition Outturn 2. This indicator measures the extent to which reallocations between the main budget categories during execution have contributed to variance in expenditure composition. It contains three dimensions and uses the M1 (WL) method for aggregating dimension scores. 39 PI-2.1. Expenditure composition outturn by function (ii) Extent of the variance in expenditure composition by program, administrative or functional classification during the last three years, excluding contingency items. 3. The budgeted and actual expenditure data (excluding interest and donor-funded projects) and the variances in PI-1 are shown in table B.4. Table B.4. Variation in Aggregate Government Expenditure (percent) Contingency Share of the Year Total Expenditure Deviation (PI-1) Total Expenditure Variance Budget 2014 114.83 17.5 3.95 2015 108.79 27.1 2.90 2016 125.32 31.2 2.25 Source: Ministry of Finance and Economic Development Budget Bureau. Note: Excluding debt servicing and donor-funded projects. PI = Public Expenditure Financial Accountability indicator. 4. These variances have been derived from the 20 largest voted expenditures in each year with the rest grouped together to form a twenty-first category. Detailed calculations are presented below in the annex to indicator 2. Score: D. Variance in expenditure composition by program, administrative or functional classification has exceeded 15 percent in two of the last three years PI-2.2. Expenditure composition outturn by economic type (iii) Extent of the variance in expenditure composition by economic type classification during the last three years, excluding contingency items. Table B.5. Results Matrix Composition Variance Year (percent) 2014 5.7 2015 14.6 2016 18.5 5. These variances have been derived from the 20 largest voted expenditures in each year. Detailed calculations are presented below in the annex to indicator 2. Score: C. The expenditure variance by economic classification was less than 15 percent in at least two of the last three years. PI-2.3. Expenditure from contingency reserves 40 (iv) The average amount of expenditure actually charged to the contingency vote over the past three years. 6. As can be seen in the table B.4 above, the contingency vote actually charged for the past three years has been 3.95 percent, 2.90 percent and 2.25 percent for 2014, 2015 and 2016, respectively. This results in an average of 3.035 percent for the review period. This is a fall in the score of the 2014 PEFA which gave it and A. It is also worth noting that the miscellaneous budget line is routinely overspent. Score: B. Actual expenditure charged to a contingency vote was averaging between 3 percent and 6 percent, inclusive, of the original budget (table B.6). Table B.6. PEFA Ratings PEFA Score Year PI 2017 2014 PI-2. Expenditure Overall Score: D Overall Score: D+ Composition Outturn • Dimension (i) Score: D • Dimension (i) Score D • Dimension (ii) Score: C • Dimension (ii) Score A • Dimension (iii) Score: B Note: PEFA = Public Expenditure Financial Accountability; PI = PEFA indicator. PI-3. Aggregate Revenue Outturn Compared with Original Approved Budget 7. The indicator focuses on both domestic and external revenue, which comprises taxes, social contributions, grants, and other revenues including those from natural resources, which may include transfers from a revenue stabilization fund or a sovereign wealth fund where these are included in the budget. External financing through borrowing is not included in the assessment of this indicator. This means that grants from development partners will be included in the revenue data used for the indicator rating, but borrowing on concessionary terms from development partners will not. 8. Revenue outturn can deviate from the originally approved budget for reasons unrelated to the accuracy of forecasts, such as a major macroeconomic shock. For this reason, the scoring calibration allows for one outlier year to be excluded. The focus is on significant deviations from the forecast that occur in two or more of the three years covered by the assessment. PI-3.1. Aggregate revenue outturn 9. This dimension measures the extent to which revenue outturns deviate from the originally approved budget. Outturn and budgeted revenue data for 2014, 2015 and 2016 are presented in table B.7. 41 Table B.7. Aggregate Revenue Outturns Budget Actual Actual as Percent of Year (Le, millions) (Le, millions) Budget 2014 3,313,749 3,403,155 103 2015 3,231,200 3,381,705 105 2016 3,832,242 3,615,440 94 Source: Ministry of Finance and Economic Development Budget Bureau. Score: C. Actual revenue was between 92 percent and 116 percent of budgeted revenue in at least two of the last three years. PI-3.2. Revenue composition outturn The revenue composition outturns are shown in table B.8. Table B.8. Revenue Composition Outturns Composition Variance Year (percent) 2014 16.0 2015 10.5 2016 20.8 Score: D. The Revenue Composition Outturn has shown variance, which has been more than 15 percent in the two of the last three years (table B.9). Table B.9. PEFA Ratings PEFA Score Year PI 2017 2014 PI-3 Revenue Composition Overall Score: D+ Overall Score: D Outturn • Dimension (i) Score: C • Dimension (i) Score: D • Dimension (ii) Score: D Note: PEFA = Public Expenditure Financial Accountability; PI = PEFA indicator. PI-22. Expenditure arrears 10. This indicator measures the extent to which there is a stock of arrears, and the extent to which a systemic problem in this regard is being addressed and brought under control. It contains two dimensions and uses the M1 (WL) method for aggregating dimension scores. PI-22.1. Stock of expenditure payment arrears 11. Table B.10 shows the expenditure arrears. 42 Table B.10. Expenditure Arrears Indicator 2014 2015 2016 Arrearsa — 9,442 455,050 Expenditureb 1,700,916 1,751,727 2,617,892 Percentage 0.00 0.54 17.38 Source: Public Debt Management Division/Audited Accounts; Ministry of Finance and Economic Development Budget Bureau estimates; amount in Le million. a. Stock of arrears: Checks payable and checks on hold at the Bank of Sierra Leone and the Accountant General’s Department. b. Expenditure: Total primary expenditure including, nonsalary, noninterest recurring expenditure, domestic capital expenditure and contingency expenditure. 12. It is noted that stock of arrears has was kept below 1 percent in 2104 and 2015 partly due to the prioritizing the eradication of arrears. However, in 2016, this mushroomed to 17.38 percent. This may be explained by the austerity measures undertaken by the government, and thus, huge piles up of checks remain unpaid and also held at the BSL and the Accountant General’s Department (AGD). 13. Discussions with the accountant general noted that arrears cannot be built outside the system because payments are made through local purchase orders and for which you must have a budget allocation. For contract-based payments that do not require local purchase orders, there may be arrears if the contract is not executed. The AGD does not have information on those arrears. However, on separately, a contract management database exists that is maintained by the AGD. So, when a contract comes for payment, the contract amount is recorded and when payment is made on those contracts, it is also recorded, ensuring that the outstanding balances on that particular contract are noted. Score: A. The stock of expenditure arrears is no more than 2 percent of total expenditure in at least two of the last three completed fiscal years. PI-22.2. Expenditure arrears monitoring 14. The amount of stock of arrears is in the first instance generated by the Integrated Financial Management Information System (IFMIS) and then subject to a verification exercise by the AGD carried out annually. For payments that are contract based that do not require local purchase orders, there may be arrears if the contract is not executed. The AGD does not have information on those arrears. 15. In Sierra Leone, arrears are not classified as such unless verified by the AGD. Commitments not honored as a result of nonavailability of cash are rolled over quarter by quarter for payment with the final quarter payments being sufficient to prevent any accumulation of arrears into the following financial year. 16. An age profile of arrears can be determined on a one by one basis from the IFMIS. However, an overall record of the age profile is not available. Current arrears are held on an excel spreadsheet in the Ministry of Finance and Economic Development (MFED) Public Debt Management Division. 43 Score: B. Data on the stock and composition of expenditure arrears is generated quarterly within eight weeks of the end of each quarter (table B.11). Table B.11. PEFA Ratings PI PEFA Score Year 2017 2014 PI-22 Expenditure arrears Overall Score: B Overall Score: B+ • Dimension (i) Score: A • Dimension (i) Score A • Dimension (ii) Score: B • Dimension (ii)Score B Note: PEFA = Public Expenditure Financial Accountability; PI = PEFA indicator. Impact on Control over Budget and Fiscal Management PI-21. Predictability of In-Year Resource Allocation 17. This indicator assesses the extent to which the central Ministry of Finance is able to forecast cash commitments and requirements and to provide reliable information on the availability of funds to budgetary units for service delivery. It contains four dimensions and uses the M2 (AV) method for aggregating dimension scores. 18. MDAs enter into operational expenditure commitments based on these forecasts. Although, the true commitment and ability to spend is determined by the level of funds which materialize and are actually available to meet those commitments. Achieving an effective balance in the level of control to meet both central control requirements and effective operational functioning by the MDAs, can often present significant challenges. 19. PEFA guidance refers to predictability for MDAs in the availability of funds being facilitated by effective cash flow planning, monitoring and management by the Treasury based on regular and reliable forecasts of cash inflows. To be reliable the amount of funds made available to an entity for a specific period should not be reduced compared with those forecasted during that period. It is also recognized that should in-year adjustments to allocations need to be made then the impact on predictability, and on the integrity of original budget allocations, is minimized by specifying in advance an adjustment mechanism that relates any adjustment to the budget priorities in a systematic and transparent manner. If, in practice, adjustments can and do take place informally without clear rules this is likely to impose further unquantifiable delays on new commitments. 20. In Sierra Leone, there is a high level of central control exercised over the forecasting and actual disbursement of available funds by MFED and while this can be effective at a government-wide level in ensuring total budget ceilings are not exceeded—some unsatisfactory management and operational consequences for individual MDAs, and government suppliers, can result. 21. In practice variations in the availability of funds imposes delays on MDAs in incurring new commitments and making related payments when cash flow problems arise. 44 This is particularly so for goods and services which are regarded as expenditures and which receive a lower priority than statutory payments and priority projects. PI-22.2. Cash forecasting and monitoring 22. A Cash Management Committee is established that meets on a weekly basis to monitor the cash position. The government’s cash flow projections are updated twice yearly in line with the timing of the IMF missions for discussions. Score C: The cash position is centrally monitored on a weekly basis although cash flow projections are only updated twice a year. PI-22.3. Information on commitment ceilings 23. The Budget Bureau notifies MDAs of their cash allocations and of their ceilings and anticipated allocations twice yearly based on the approved budget. Previously, quarterly fiscal allocations based on the budget were given. The move to the half yearly notification of allocations is now in alignment with the fact that while quarterly allocations to MDAs was the practice, in fact MFED did not actually revise the forecasts except in the March and September negotiations with the IMF. Score: C. MDAs are notified of the ceilings twice a year (previously quarterly). Although the cash position is centrally monitored on a weekly basis by MFED this does not always appear to translate into a high level of predictability and reliability at the MDA operational level. Informal cash control and allocation mechanisms toward priorities determined at a level above MDAs can be used during periods of cash flow problems. PI-21.4. Significance of in-year budget adjustments 24. Section111 subsection 3b (ii) and section 112 subsection 3and4 of the Sierra Leone constitution sets out the formal process by which supplementary budgets should be submitted to parliament. A presidential warrant (section 114–2c) authorizes extrabudgetary expenditure signed by the President when there is an urgent need in cases of emergencies. 25. The Ministry of Finance does impose reductions on MDAs because of recurring and significant revenue shortfalls, and when extrabudgetary expenditures are authorized. The Budget Bureau has developed procedures to protect priority and statutory expenditures including poverty related expenditures, wages and salaries, debt service payments and other nondiscretionary expenditure. These procedures, including commitment controls have been communicated to MDAs through the Financial Secretary’s Circular,which sets out the rational for changes. Score: D. Significant in-year budget adjustments to budget allocations are frequent, and are partially transparent. 45 PI-13. Debt Management 26. This indicator assesses the management of domestic and foreign debt and guarantees. It seeks to identify whether satisfactory management practices, records, and controls are in place to ensure efficient and effective arrangements. It contains three dimensions and uses the M2 (AV) method for aggregating scores. PI-13.1. Recording and reporting of debt and guarantees 27. External debt management and monitoring is carried out jointly by BSL, the Accountant General’s Department and Public Debt Management Division (MFED). This is done monthly with the BSL having the major responsibility and operational role in maintaining the database. The external debt is recorded in the Commonwealth Secretariat Debt Recording and Management Systems. 28. The monthly reconciliation that takes place is the key reconciliation for the AGD for the update of the fiscal table, and reconciliations are also prepared at the time of IMF missions. The challenges faced under the Integrated Public Financial Management Reform Project (IPFMRP) project was that the IFMIS system does not handle multicurrency transactions and therefore a lot of the debts especially external debts are foreign currency denominated, and this poses a problem in the recording within the system. 29. A quarterly report on the external debt is produced by the Public Debt Department. Domestic debts monitoring is done on a weekly basis, and this is typically for the Treasury bills issued by the BSL. Interest from the Treasury bills is managed by the public Debt Management Unit within MFED and the Central Bank. When payments are due, the BSL will inform the Minister of Finance and Economic Development, and the instruction issued for the account of government to be debited. The debit advice is used to process within IFMIS after the bank has executed. This comes as a posting controlled within the same allocation for all payments, Principal as well. Score: B. Domestic and foreign debt and guaranteed debt records are complete, accurate, and updated quarterly. Most information is reconciled quarterly. Comprehensive management and statistical reports covering debt service, stock, and operations are produced at least annually. PI-13.2. Approval of debt and guarantees 30. The approval of the debt and guarantees of the Sierra Leone government is captured in the Public Debt Management Act 2011. PART II–AUTHORISATION AND BORROWING PURPOSES “Subject to section 118 of the Constitution, the Minister shall have sole authority to borrow money on behalf of Government by concluding loan agreements, issuing Government securities, or entering into supplier’s credit agreements and to issue Government guarantees, both in Sierra Leone and elsewhere and in local and foreign currencies.” It further goes on to say, “The Minister shall have sole authority to sign loan and supplier’s credit agreements for and on behalf of Government.” The government may borrow: 46 • to finance government budget deficits; • to maintain a credit balance on the treasury main account at a level determined by the Minister; • to provide such government loans or credits to local councils, public enterprises and any other entity as may be approved by parliament; • to honor obligations under outstanding government guarantees; • to refinance outstanding debt or repay a loan prior to its date of repayment; • to immediately protect, mitigate or eliminate effects caused by a natural or environmental disaster or any other national emergency; • to replenish the international reserves; • to meet requests by the Bank of Sierra Leone to issue government securities for the sole purpose of supporting monetary policy objectives; and • any other purpose as parliament may, by resolution, approve. Score: A. Primary legislation grants authorization to borrow; issue new debt, and issue loan guarantees on behalf of the central government to a single responsible debt management entity. Documented policies and procedures provide guidance to borrow, issue new debt and undertake debt-related transactions, issue loan guarantees, and monitor debt management transactions by a single debt management entity. Annual borrowing must be approved by the government or legislature. PI-21.1. Consolidation of cash balances 31. The Government of Sierra Leone does not operate a single treasury account which consolidates all government accounts. Currently, the Treasury Accounts, maintained at the BSL managed by the Accountant General’s Department are consolidated and therefore excludes many of the departmental bank accounts, mainly for externally assisted projects and subvented agencies. The treasury single account is only piloted in 10 government subvented agencies. The BSL is challenged by the configuration of their systems to match that of the AGD which is ready. 32. For the accounts that are managed by the AGD, a weekly Revenue Report is produced which captures the daily cash collected and which forms the basis of the information required for the release of funds. The cash and bank reconciliations are done within the AGD. Score: C. A significant number of MDA accounts are not consolidated with the treasury single account. PI-23. Payroll Controls 33. The HRMO’s is in charge of civil servants, however but there are other categories of employees and payrolls outside of this definition for which Government of Sierra Leone provides the funding. The payroll therefore includes all payrolls of central government including all MDAs and autonomous government agencies (AGAs) (for example, police, focus, teachers, military and other subvented agencies). 47 34. The payroll is one of the largest items of the Government of Sierra Leone’s expenditure and accounts for almost 50 percent of domestic revenues. In 2013, it was noted that there were over 22,000 categorized as civil servants but this amount has been trimmed down to about 18,000. The payment of salaries for other quasi-agencies like SLBC, SLRA and SALPOST does not come from the central government budget.1 35. In 2013, a new software system Civil Service Management (CSM) version 6.0 replaced Human Capital Accountability, which was reputed to only concentrate on the payroll aspect and not human resource. It was noted that only the payroll aspect of the software was rolled out to HRMO, though the financial aspect and budgeting aspect is not accessible to them. There is therefore no link between the payroll and human resource with the later informing the former when amendments to personnel is required. They are challenged by the fact that only six (6) licenses for the software are available to them whereas the need is for 10 additional licenses; the six licenses are used by two supervisors and four inputters. 36. There are other issues highlighted which include the software not been upgrade since inception of the software even though it is in need of an upgrade. The HRMO has no timely support from information and communication technology, who are responsible for making changes to the software and this results in some staff not being paid as the HRMO has no control over the addition or removal of staff from the system as the need arises. 37. There are two distinct control measures within the system on recruitment. a) HRMO determines the establishment and b) MFED determining the number of posts. The HRMO only has inputting rights, whereas, the AGD is responsible for payment of the payroll and instructs the transfer of funds to the individual accounts of public servants held in commercial banks. Thus, the personnel and payroll functions are segregated for control purposes. There are plans to roll out the inputting tasks of personnel to the MDAs, currently done by the HRMO. PI-23.1. Integration of payroll and personnel records 38. HRMO ensures that personnel files within the system are properly opened and maintained to allow employees to receive the correct pay and for proper records management. The personnel and payroll functions are segregated for control purposes. The control within the system is the segregation of duties between inputting done by HRMO and payment done by the AGD. 39. The subvented agencies which are funded by the Government of Sierra Leone operate their own accounting systems which have no system interface with IFMIS and other records such as payroll. However, centralized payroll controls exist where the AGD checks the salary payment vouchers which are submitted monthly. The software has not been upgrade since inception of the software even though it is in need of an upgrade. 40. A key control, and the verification required to ensure meaningful integration of the payroll and personnel records, can only be attained when physical checks of employees are established together with reconciliation of the personnel files held by MDAs, and the master file held by HRMO and the corresponding IFMIS-CSM data. 48 Score: D. even though a few payroll verification exercises have been undertaken in the recent past, it still does not mask the fact that the integrity of the payroll is significantly undermined by lack of complete personnel records and personnel database, or by lack of reconciliation between them. Currently the European Union is funding a separate project, (a bio metric head count of all civil servants). The HRMO did not commission this. PI-23.2. Management of payroll changes 41. When changes to personnel records and the payroll are informed to HRMO and the AGD they are normally actioned within a month. It is however difficult to substantiate the delays from MDAs and other bodies in the notifications to AGD. We noted during our meetings with the HRMO changes can take as much as up to three months to rectify resulting in some staff not being paid. Score: C. Personnel records and payroll are updated at least quarterly and require some retroactive adjustments. PI-23.3. Internal control of payroll 42. The Payroll Unit within AGD has the main responsibility to ensure that all government employees are paid on time and at the correct salary scale. In 2013 the Human Capital Accountability system was replaced by CSM and we understand that this has additional system controls within it. 43. The three levels of autonomy previously mandated by the AGD continue to be: • Full self-accounting for Sierra Leone Police (since 2007) and the Ministry of Defence (since late 2009). • A “rolled out status” has been given to ten MDAs of which nine benefit from a lower level of autonomy than the self-accounting MDAs – these MDAs input and process all payroll vouchers up to approval level prior to sending them to the AGD. The AGD then undertakes a higher level of further scrutiny before the printing of checks. • For all other MDAs, AGD receives instructions for payroll input and amendments from HRMO except for teachers whose instructions come from the Ministry of Education, Youth and Sports. 44. The overall comprehensive and multilayered control environment referred to above contributes to the internal controls which operate over changes to personnel records and the payroll. Monthly checks are run on the central system as it captures data. Checks are run against PIN numbers and the establishment list and a report produced for HRMO. This IFMIS difference report is amended / confirmed against the HRMO archive list of personnel files before AGD makes payment. Internal Audit Unit staff in HRMO and other MDAs also have a control role to play over payroll. Score: B. Authority and basis for changes to the personnel records and the payroll are clear and evidenced. 49 PI-23.4. Payroll audit 45. A few payroll verification exercises have taken place over that past few years. This dimension considers the last three years. A teacher verification exercise is ongoing. The Public Financial Management Improvement Consolidation Project (PFMICP) has conducted a biometric headcount of all civil servants excluding health in 2016. However, this aspect of payroll remains deficient. While there is evidence of some audit and verification exercises operating in respect of payroll these are not regular, systematic nor comprehensive. Score: C. Integrity of payroll continues to be undermined by lack of full verification and reconciliation between the personnel database and payroll records; however sufficient controls exist to ensure integrity of the payroll data of greatest importance PI-25. Internal Controls on Nonsalary Expenditure 46. This indicator measures the effectiveness of general internal controls for nonsalary expenditures. Specific expenditure controls on public service salaries are considered in PI-23. The present indicator contains three dimensions and uses the M2 (AV) method for aggregating dimension scores. PI-25.1. Segregation of duties 47. The regulation through the Act ensures that certain documentations should be in place, which makes provisions for clear segregation of duties. Within the IFMIS application, there are several controls: • You cannot process a transaction in the absence of a budget allocation, which is managed by the budget director. • All payments below SLL60m has to be issued with a local purchase order generated by IFMIS for all MDA’s even those MDA’s which IFMIS have not been rolled out to, and this ensures that you cannot commit the Government of Sierra Leone without a budget allocation. • All contracts above SLL60m, a contract management team from the MOFED is established which deals with such contracts. A certificate is issued by this committee and without which a contract cannot be signed and payment is not made. • Staffs that process local purchase orders within the system are not involved in receipts within the system and processing of vouchers. • A checklist is available for the supporting document that is required to support payment based on the type of payment. • The system does not allow a single person to process a transaction from inception to conclusion thereby enforcing segregation of duties controls. Several staffs have to be involved. • All checks are printed centrally within the AGD, and not from the MDAs to which IFMIS has been rolled out. The AGD is not involved in the procurement and 50 processing of the various MDA’s transactions; however, those documents that have been presented for payment to the AGD are further checked before any payment is made. 48. In summary, the execution of the budget is by allocation. Allocation is the method by which MDA’s acquire goods and services, and payments are made on limits for which you cannot go over by the central bank. The administrative controls set the limits for payments and are based within the AGD by the information and communication technology department. Score: B. Segregation of duties is prescribed throughout the expenditure process. Responsibilities are clearly laid down for most key steps while further details may be needed in a few areas. PI-25.2. Effectiveness of expenditure commitment controls 49. The use of controls within IFMIS used for all payments within the AGD ensures that most government’s payment obligations remain within the limits of annual budget allocations (as revised) and within projected cash availability, thereby avoiding creation of expenditure arrears. 50. Arrears cannot be built outside the system because payments are made through local purchase orders and for which you must have a budget allocation. For contract based payments that do not require local purchase orders, there may be arrears if the contract is not executed. The AGD does not have information on those arrears. However separately, a contract management database exists that is maintained by the AGD. So, when a contract comes for payment, the contract amount is recorded and when payment is made on those contracts, it is also recorded, ensuring that the outstanding balances on that particular contract are noted. 51. For these types of contract which have been signed by MDAs, the AGD will have no knowledge of it until it is brought for processing and payment, as it was initiated out of the system. There is no contract management module within the IFMIS, and what is available is the expenditure module for the execution of the budget. The AGD is requesting the Contract Management module. Score: C. Expenditure commitment control procedures exist which provide partial coverage and are partially effective. PI-25.3. Compliance with payment rules and procedures 52. This dimension assesses the extent of compliance with the payment control rules and procedures based on available evidence. 53. We have noted the Auditor General’s report on the Government of Sierra Leone accounts for 2014 and 2015 and from our discussions with the Audit Service; the issues are common even for the unaudited 2016 accounts. They are summarized in the report as: 51 Several matters that are common to virtually all MDAs, PEs, and councils are as follows: • Significant procurement irregularities. • Payments without supporting documents. • Payments without adequate supporting documents. • Impress accounts without retirement details. • Withholding Tax not deducted and paid to the National Revunue Authority (NRA). • Stores items not taken on ledger charge. • Monthly bank reconciliations not prepared. • No fixed asset accounting or control policies. • Fixed asset registers not maintained. • Fixed asset registers not updated. • No identification codes on assets. • No effective internal audit units. • Performance contracts not presented for audit inspection. 54. “They range from bad practice to outright breaches of legislation and, I may add, all have been reported in previous years. Many are simple to address but for the reasons outlined they are not dealt with in any serious way.” Score: D. The majority of payments is compliant with regular payment procedures; however, the majority of exceptions are not properly authorized and justified. PI-26. Internal Audit 55. This indicator assesses the standards and procedures applied in internal audit. It contains four dimensions and uses the M1 (WL) method for aggregating dimension scores. 56. The Internal Audit Directorate (IAD) of the Government of Sierra Leone was legally recognized through enactment of the Government Budgeting and Accountability Act (GBAA) 2005. The Section 6 of the Act clearly spells out the mandate of the IAD, which covers all MDAs and local councils. Since inception the IAD has developed and strengthened its capacity. The GBAA 2005 has now been repealed by the Public Financial Management Act of 2016. The Public Financial Management Act of 2016 has currently tabled in parliament further strengthens the internal audit function. It is a determined attempt by the government to give full effect to the preamble to secure transparency and accountability in the appropriation, control and management of the finances and other financial resources of Sierra Leone. 57. The aim of Internal Audit is embodied in its mission statement “To ensure that proper internal checks and controls, financial and otherwise exist in Public Financial Accounting and Reporting and that all assets and records are protected and accounted for.” 52 58. Both the IPFMRP and the PFMBESP projects were able to capacitate the IAD through regular technical support and structured training programs. The IPFMRP project in particular enabled the directorate to employ qualified accountants. Currently, there are circa 125 internal auditors in 41 MDAs in central government covering about 69 percent of budget expenditures. There are also 25 internal auditors in all 19 local councils. The IAD in MFED has the harmonization and supervision role of internal audit units across government. It has the statutory authority to co-ordinate and manage internal audit cadre, to set standards and monitor performance. PI-26.1. Coverage of internal audit 59. IA in central government has increased in number and strength as a result of the employment of more internal audit personnel during the past few years 2014–2016. Currently, the internal audit function has been rolled out and has established IA units in 41 MDAs. This covers over 69 percent of the budgeted expenditures of CG. The directorate holds quarterly meetings comprising of two representatives from each IAU. IAUs are divided into ten (10) zones and each zone is headed by a qualified / and or experienced auditor who serves as Coordinator. 60. The Public Financial Management Act of 2016 has clearly mandated the vote controller of a budgetary agency, subvented agency and other entities in the central government, local councils, social security fund, or public enterprise to establish an Audit Committee of the entity that will report to the minister of that entity. The act also mandated an internal audit department, division, or unit of a budgetary agency to report to the Audit Committee of the agency or entity and the Director of Internal Audit, Ministry of Finance. For the period under review a total of 363 audits activities were included in the comprehensive public sector internal audit plan to be completed by 39 MDAs. 61. Within the local councils most of the current crop of internal auditors have no background in internal audit. Some training and capacity was provided however this did not increase their capacity and as a result they are unable to cope with the demands of the job according to the director. The Local Counils were not fully catered for under the PFMICP. Score: C. Internal audit is operational for central government entities representing the majority2 of budgeted expenditures and for central government entities collecting the majority of budgeted government revenue. PI-26.2. Nature of audits and standards applied 62. The staff and heads of internal audit units of central government are qualified accountants. Currently, a total of 17 internal auditors in central government are registered with the Institute of Internal Auditors to write the Certified Internal Audit (CIA) exams. They have also planned to register additional ten (10) staff this year to write the CIA exams. 63. The internal auditors use audit standards, guidelines, and practice notes issued by the Institute of Internal Auditors in planning and executing internal audit plans. There is also an internal audit manual developed and approved. It takes into account the standards, guidelines, 53 and practice notes issued by in the Institute of Internal Auditors. All internal auditors have been trained on its use. For the period under review, the following is noted: • Training of internal auditors on risk-based auditing • Training of recently recruited internal auditors on audit planning, execution, reporting, and follow-up • Validation of the internal audit manual for local council internal auditors • Special investigations • Promoting effective dialogue in internal audit through quarterly meetings • Coaching and mentoring of local councils’ internal auditors • Training of heads of internal audit unit financial management and procurement on AfDB projects • Development of an ERM policy and framework for Sierra Leone Score: B. Internal audit activities are focused on evaluations of the adequacy and effectiveness of internal controls. PI-26.3. Implementation of internal audits and reporting 64. At the time of review the period covered was January to December 2016. The review noted total of 363 audits activities were included in the comprehensive public sector internal audit plan to be completed by 39 MDAs. Actual results showed that only 176 audits representing 48.5 percent of the planned activities were completed as at December 2016, resulting in 187 (51.5 percent) not accomplished. The internal audit prepared and we review an annual report for the period under review, that is, 2016. Score: D. While an annual audit program exists, the majority of programmed audits were not (48.5 percent only) completed, as evidenced by the distribution of their reports to the appropriate parties. The requirement for C rating or higher is not met. PI-26.4. Response to internal audits 65. While the IPFMRP project support increased the quality of the audit reports through capacity building, training and logistics, the PFMICP has been so successful. Under the latter project, only a single activity was undertaken in a one-day seminar for audit committees. The logistics were not available and though the committess were setup, they were not functional due to the lack of resources. 66. There are only limited examples of audit reports where management has responded to audit findings and recommendations. There is therefore no clear evidence of effective and timely follow-up of internal audit findings by management. Score: D. The requirement for C rating or higher are not met. 54 PI-30. External Audit 67. Reliable and extensive external audit is an essential requirement for ensuring accountability and creating transparency in the use of public funds. This indicator examines the characteristics of external audit. It contains four dimensions and uses the M1 method for aggregating dimension scores. 68. The Supreme Audit Institution of Sierra Leone is the Audit Service of Sierra Leone (ASSL) which supersedes the previous auditor general’s Department and was established in 2004. The Audit Services Act 1998 granted the ASSL the responsibility and mandate to carry out the external audit of all central and local government revenue and expenditure as well as institutions which receive more than 50 percent funding from the Government of Sierra Leone. The current legislation that regulates external audit is the 1991 constitution of the Republic of Sierra Leone, the Audit Service Act (2014) and the Public Financial Management Act 2016. 69. The auditor general is mandated to audit the public accounts of sierra leone and all public offices, including the courts, local government administration, the university of Sierra Leone, and other public institutions and statutory corporation and so on. The auditor general is also mandated to carry out other audits such as performance and value for money audits. PI-30.1. Audit coverage and standards 70. This dimension assesses key elements of external audit in terms of the scope and coverage of audit as well as adherence to auditing standards for three years (2014, 2015, and 2016). The scope of audit indicates the entities and sources of funds that are audited in any given year and should include extrabudgetary funds and autonomous agencies. 71. The audit examination of the public accounts for 2014, 2015 and 2016 was risk-based audit. According to the auditor general, “We select, on a test-basis, samples of transactions from the Integrated Financial Management Information System (IFMIS); the governments computerized accounting system, and examine the supporting documentation. We also verify the accuracy of the compilation process used to create the financial statements and review the system of internal control in place. These controls include, but are not limited to, segregation of duties, authorization, and record keeping.” 72. We also examine financial disclosure matters and in that regard, form a view on how well (or not) the disclosure rules of the prescribed accounting framework are followed. As will become clear below, there are without exception, problems with every one of those elements.” 73. The coverage of the audit of the 2013 annual report was 83.7 percent, 2014 annual account of the central government is 85 percent of the government budget and that of the 2015 annual report was 88 percent of Governemnet budget. The 2016 accounts have not yet been tabled in parliament. 74. ASSL complies with the International Organization of Supreme Audit Institutions professional and ethical standards and is very active in its regional body, the African 55 Organization of English-speaking Supreme Audit Institutions (AFROSAI-E).The Auditor General of Sierra Leone is Chairperson of the Governing Board of AFROSAI-E for a three year period in 2014. 75. The review noted that in 2015, two separate performance audits where carried out on the Management of Government Commercial Buses by the Sierra Leone Road Transport Corporation and the Management of Government Residential Quarters by the Ministry of Works. Score: B. Financial reports of central government entities representing most total expenditures and revenues have been audited using ISSAIs or national auditing standards during the last three completed fiscal years. The audits have highlighted any relevant material issues and systemic and control risks. PI-30.2 Submission of audit reports to the legislature 76. The dimension assesses the timeliness of submission of the audit report(s) on budget execution to the legislature, or those charged with governance of the audited entity, as a key element in ensuring timely accountability of the executive to the legislature and the public. 77. The review noted that for 2014 and 2015 the Auditor General met the mandate to submit her report to the parliament, within 12 months of the end if the period covered, while the 2016 accounts are a working progress (See table B.12). 78. Performance audit reports were also noted to have been carried out in 2014 and 2016 and presented to the Parliamnet for the PAC to consider. (See table B.12). 56 Table B.12. Public Accounts Statement of Number of Months Accounts Year Date of Receipt from Date Submitted to to Complete the Ended Accountant General PAC in Parliament Audit 31st December 2016 31st March, 2017 Work in progress Work in progress 31st December 2015 31st March, 2016 Friday, 23rd Friday, 30th December, 2016 September, 2016 31st December 2014 31st March, 2015 Monday,7th Wednesday, 30th December, 2015 September, 2015 Performance Audit Date Completed by Date Reported to Date of PAC Report ASSL Legislature Hearings 2014 - Rehabilitation May, 2014 15th October, 2014 To be confirmed by of Rural Feeder Roads the Clerk of PAC 2014 - Agricultural October, 2014 26th November, 2014 " Mechanization 2014 - Management of September, 2014 26th November, 2014 " Municipal Solid Waste by Local Councils 2015 None 2016 - Management of May, 2016 20th May, 2016 " Government Commercial Buses 2016 - Management of May, 2016 20th May, 2016 " Government Residential Quarters Source: Audit Service Sierra Leone: Work on the 2016 accounts are still in progress and are expected to be laid before parliament in December 2017. Score: C. Audit reports were submitted to the legislature within nine months from receipt of the financial reports by the audit office for the last three completed fiscal years. PI-30.3. External audit follow-up 79. The requirement for audited entities to respond to the Auditor General’s management letter on systems weaknesses and the PAC’s recommendations within 30 days still applies, though there have been significant improvement over the ensuing years since the last PEFA2014. In comparison to 2013, the Auditor General noted in her report that “the rate of implementation of her recommendations appears to have marginally improved from 19 percent in 2013 to 24.2 percent in 2015. It would be pleasing to report that the improvement was real but it is not. It has arisen simply because the Ministry of Foreign Affairs is not included in the 2014 figures as the annual audit was not undertaken in time for inclusion due to logistical constraints arising from the EVD crisis.” 80. Almost without exception our observations and recommendations are not being given the attention they deserve or that parliament, citizens and international donors have a right to expect. For example, Freetown City Council has implemented less than 8 percent of our 57 recommendations and, from the eight entities reviewed, another three have implemented less than 13 percent. In absolute numbers, for the four years 2010 to 2014 there were 953 recommendations of which 231 were implemented, 75 are in progress and 647 were not implemented.” 81. Similarly in her report of 2015 she noted that “It comes out clearly that our observations and recommendations are not being given the attention they deserve, or that parliament, citizens, and international donors have a right to expect. Our assessment for the last five years has exposed a minor increase in the percentage of improvement for five MDA’s and two have regressed by an average of 8 percent.” 82. Ministry of Health and Sanitation has maintained its implementation rate of last year at 25 percent of our recommendations. The Ministry of Defence and the Ministry of Education, Science and Technology made a minor increase of 4 percent while the Office of the President, Ministry of Finance and Economic Development (MFED) and Freetown City Council, made an average increase of 15 percent on recommendations implemented (31.3 percent in 2014 to 46 percent in 2015). Overall only 28.8 percent of recommendations have been implemented by the eight selected entities. In absolute numbers, for the five years 2011 to 2015 there were 959 recommendations for the selected entities of which 276 were implemented, 65 are works-in-progress and 618 were not implemented. There is a clear pattern of repeated observations across all the audit entities we reviewed. 83. It is worth noting that the Internal Audit Department carried out a follow-up review on the implementation of the recommendations in the annual report of the Auditor General for the year ended 31st December 2013. The review is undertaken in line with the multidonor budgetary support progress assessment framework that indicates that follow-up on the implementation of the Auditor General`s Report on the Accounts of Sierra Leone should be a priority. According to the indicator, under the theme accountability and oversight with a focus on external oversight, requires both the Audit Service and Internal Audit to follow up the implementation by vote controllers of 60 percent of the recommendations made in the Auditor General`s report on the 2013 public accounts. 84. A joint review was undertaken and the results were as follows: 85. A total of 197 audit recommendations were captured from the 20 MDAs on the follow-up review. Out of the 197 audit recommendations, 61 recommendations representing 31 percent of the total were noted to have been fully implemented, 29 recommendations representing 15 percent of the total recommendations were partly implemented while 107 (one hundred and seven) representing 54 percent are yet to be implemented. 86. Thus, the ASSL could only confirm 31 percent of the recommendations had been fully implemented. It also noted that 24 recommendations from 8 MDAs were left out in Internal Audit’s follow-up report because of the noninclusion of these recommendations contained in the ASSL table of common issues. Score: B. ASSL now has statistics of, and is monitoring, the auditee’s responses to audit queries set out in the MDAs’ management letters which indicates that this has shown significant improvement since PEFA 2014. The evidence of systematic follow-up is less clear. 58 PI-30.4. Supreme Audit Institution (SAI) independence 87. The issue of financial independence of the ASSL in the past, and currently has frequently been discussed by stakeholders and the impact it has on the ASSLs ability to perform its work. The current issue as discussed with the ASSL is not so much the quantum allocated to it, but more so the timing of the release of funds. The PAC emphasized that it has done a lot in the effort to get the fair amount of budget allocations for the ASSL, but as stated the timing of the release of funds has been the issue. 88. The legal framework for the ASSL is the 1991 Constitution of Sierra Leone, the Audit Service Act 2014 and Public Financial Management Act 2016 all of which clearly indicate its independence from the executive. S15 of the Audit Service Act 2014 states that “The Auditor General shall, subject to this Act, act independently in the exercise of his duties under Section 119 of the Constitution of Sierra Leone and shall not be subject to the direction or control of any person or authority.” 89. The appointment of the Auditor General is political as indicated in section 13(1) of the Audit Service Act 2014 “The Audit Service shall have an auditor general who shall be the head of the Audit Service, appointed pursuant to Section 119 of the Constitution of Sierra Leone and on such terms and conditions as may be approved by the President” Score: C. The SAI operates independently from the executive with respect to the procedures for appointment and removal of the Head of the SAI as well as the execution of the SAI’s budget. The SAI has unrestricted and timely access to the majority of the requested records, documentation and information. Impact on the Transparency of Budget and Fiscal Management PI-6. Central Government Operations Outside Financial Reports—Last Completed Fiscal Year 2016 Last Completed Fiscal Year 90. In the recent past and regard to the PEFA 2014, Sierra Leone had about 130 AGAs asked to report to the Treasury and Other Government Accounts Service (TOGAS) in the Accountant General’s Department of the MFED on a quarterly basis. The current situation with regard to this review is that number has significantly been reduced to only 65 reporting entities. 91. Quarterly data are collected for: • Projects administered by project implementation units (PIUs)3 • Subvented agencies4 • Departmental revenues not brought into account5 • Local councils (though this work is still under development) 92. The TOGAS data capture system comprises a spreadsheet and database into which printed statements received from the organizations to be monitored and bank balance details, received from the BSL and commercial banks, are manually entered. This is the raw data on 59 which the Unit’s reconciliation of monetary data6 and fiscal7 data are based. TOGAS reports reconciliations of monetary and fiscal data to the accountant general on a quarterly basis. 93. There is not a regular discipline of the required agencies submitting the schedules and bank statements on time. Ongoing problems remain in obtaining bank statements in respect of those agencies using commercial bank accounts. The proposed STA, together with a 100 percent rollout of the IFMIS and a clear legal mandate would improve reporting. 94. The extent to which data are regularly introduced into the accounts of the AGD and then the fiscal reporting of the Budget Bureau is not evident. PI-6.1. Expenditure outside financial reports 95. The number of subvented agencies (SVAs) required to submit quarterly TOGAS reports has been significantly reduced. In 2017, as at the end of quarter 2, the number of SVAs required is 65. Also in terms of status of submissions, 39 submitted for quarter 2 and 16 have not submitted. This gives a submission rate of 60 percent. The total expenditures by SVAs as indicated by their financial reports are Le 379,306 billion to date. For agencies that do not submit their reports to the TOGAS and are self-financing and do not receive subventions, their total expenditure is Le 263,253 billion.8 These two categories thus make up over 12 percent of total expenditures, considering that not all the information was available to us at the time of review. Score: D. Expenditure outside government financial reports is more than 10 percent of total BCG expenditure. PI-7. Transfers to subnational governments—last completed fiscal year 96. This indicator assesses the extent to which transparent, rule-based systems are applied to budgeting and the actual allocation of conditional and unconditional transfers. Transfers to support subnational government’s expenditure can be made in the form of unconditional grants, where their final use is determined by the subnational governments through their budgets, or through conditional (earmarked) grants to subnational governments to implement selected service delivery and expenditure responsibilities—for example, by function or program, typically in accordance with an agreed-on regulatory or policy standard. The overall level of grants (that is, the vertical allocation) will usually be determined by policy decisions at the central government’s discretion or as part of constitutional negotiation processes, and is not assessed by this indicator. 97. However, clear criteria for the distribution of grants among subnational governments—for example, formulas for the horizontal allocation of funds— are needed to ensure allocative transparency and medium-term predictability of funds available for planning and budgeting of expenditure programs by subnational governments. Every fiscal transfer from central government to the relevant subnational governments should be taken into consideration. If different formulas or criteria are used for different elements of transfer, the overall assessment may be made on a value-based weighted average. 60 PI-7.1 System for allocating transfers 98. Intergovernment fiscal relations are regulated by The Local Government Act, 2004 (Act No 1 of 2004) (LGA) and the supporting Statutory Instrument which provides the regulations. 99. Every year, local councils receive a transfer for both the discharge of the devolved functions; and toward their administrative costs. The total number of annual grants to local councils each year forms part of the national budget and is published by government notice and in the national newspapers. 100. Each year a seminar is held with all relevant stakeholders (including nonstate actors and numbering over 100 persons) to determine the weightings for the forthcoming budget cycle. Examples of weightings are population size and level of infrastructure. The seminar was held in October 2013 for the FY14 budget. This approach uses allocation criteria and criteria weights to determine the distribution of grants across the 19 local governments. The type and number of criteria has depended on the kind of grants to be transferred and the function to which it is to service. The factors differ from devolved function to devolved function, but each of the formulas incorporates the principles of equity. The formulas are revised annually and updated as necessary. Score: A. The horizontal allocation of all transfers to subnational governments from central government is determined by transparent, rule-based systems. 101. The process by which subnational governments receive information on their annual transfers is managed through the regular budget calendar, which is generally adhered to and provides clear and sufficiently detailed information for subnational governments to allow at least six weeks to complete their budget planning on time. 102. The process by which subnational governments receive information on their annual transfers is managed through the regular budget calendar, which provides clear and sufficiently detailed information for subnational governments to allow at least four weeks to complete their budget planning on time. PI-7.2. Timeliness of information on transfers 103. The LGA states that every local council has to prepare a budget for each financial year three months before the beginning of that year. This does not occur. The BCC which includes a reliable estimate of transfers to local councils has been issued in September in recent years, requesting a budget to be submitted by September 30. Invariably, the budget is not submitted until much later. LGFD recognizes that not enough time is provided and informs the local councils that the budget should arrive before the end of the FY. This is again not necessarily respected as there is no incentive for councils to do so, in particular as transfers are made in a “reliably late” manner. 104. The above process covers the recurrent budget for local councils. However, development grants are not determined until after approval by parliament, which is usually after January 1, although these are not devolved. 61 105. It should also be noted here that while the amounts of the transfers budgeted are usually adhered to, the timing of the transfers has proved “reliably late” over the past few years, to the extent that actual transfers have been made biannually rather than quarterly. Plans are in process to formalize a biannual transfer and thereby try to adhere to budgeted transfer times. Score: C. Substantial delays may be experienced in implementation of the budget procedures. Information on annual transfers to subnational governments is issued before the start of the subnational governments’ fiscal year, which could be after budget plans are decided. PI-9. Public Access to Fiscal Information PI-9.1. Public access to fiscal information 106. This indicator assesses the comprehensiveness of fiscal information available to the public based on specified elements of information to which public access is considered critical (table B.13). Table B.13. Public Access to Fiscal Information Basic Elements Where and When (i) Annual executive budget proposal Yes. Available from the government printers. documentation: A complete set of executive The budget speech is available on the day. budget proposal documents is available to the The estimates are available shortly afterward public within one week of the executive’s in limited numbers (given its size). The submission of them to the legislature. budget speech and summary budget tables are placed on the MFED website shortly after the speech. (ii) Enacted Budget: The annual budget law Yes. As soon as the Act is passed it is approved by the legislature is publicized within two gazzetted immediately. weeks of passage of the law. (iii) In-year budget execution reports: The reports No. The quarterly releases and actual are routinely made available to the public through expenditures are posted on the MFED appropriate means within one month of their website, however not within the alloted time. completion. (iv) Annual budget execution reports The report is Yes. The unaudited annual financial made available to the public within six months of statements are posted on the web and in hard the fiscal year’s end. copy. The financial statements are presented by end March of the follwing year i.e 2015 on 31st March 2016. The audited accounts for FY 2016 are not available until December 2017. (v) Audited External Reports: The statements are Yes. As soon as the Audited statements are made available to the public through appropriate laid before the parliament it becomes a public means within 12 months of completed audit. document and available on the ASSL website. Additional Elements 62 (vi) Prebudget Statements: The broad parameters No. for the executive budget proposal regarding expenditure, planned revenue, and debt is made available to the public at least four months before the start of the fiscal year. (vii) Other external audit reports: All Yes: Performance reports of the ASSL are nonconfidential reports on central government made public. consolidated operations are made available to the public within six months of submission. (viii) Summary of Budget Proposals: A clear, No: The Citizens’ Budget is available but the simple summary of the executive budget proposal or timing of its availability (2weeks) is not met. the enacted budget accessible to the nonbudget At the time of this review it was yet to be experts, often referred to as a “citizens’ budget,” and published for dissemination. where appropriate translated into the most commonly spoken local language, is publicly available within two weeks of the executive budget proposal’s submission to the legislature and within one month of the budget’s approval.. (ix) Macroeconomic forecasts: The forecasts, as No: The economic forecasts while as assessed are available within one week of their assessed in 14.1 is done, however, this is not endorsement. available with one week. Score: D. The government does make make available on time four of the six elements to the public. PI-19. Revenue Administration PI-19.1. Rights and obligations for revenue measures 107. This dimension assesses the extent to which individuals and enterprises have access to information about their rights and obligations, and also to administrative procedures and processes that allow redress, such as a fair and independent body outside of the general legal system (ideally a “tax court”) that is able to consider appeals 108. In this light, the review noted that the NRA has an active website (www.nra.gov.sl), here taxpayers have access to documents that are available for download (such as GST registration Application Form, Income Tax returns) and it provides details of relevant Acts. 109. This review also notes a Public Affairs and Taxpayer Education Department (PATE) which is responsible for stakeholder outreach, including taxpayer education, media relations, publicity, publications and internal communications. The Customs Department holds regular meetings with importers, and the introduction of ASYCUDA has improved transparency and information to importers. The Harmonized System code on customs tariffs is widely available. Also, with regard to administrative procedures and processes that allows redress, section 138 of the Income Tax Act provides for an appeal mechanism. This is a two-stage process whereby objections are first heard by senior managers of the Domestic Tax Department. 63 110. Should resolutions not be made at that stage, an appeal should proceed to a Revenue Appellate Board where the Chair and six Commissioners are appointed by the President. Finally, the process allows a party who is dissatisfied with decision of the Board to appeal to the High Court, within 60 days of the decision. Score: B. The NRA provide taxpayers with access to comprehensive and up-to-date information on the main revenue obligation areas and on rights including, as a minimum, redress processes and procedures. PI-19.2. Revenue risk management 111. This dimension assesses the extent to which a comprehensive, structured and systematic approach is used within the revenue entities for assessing and prioritizing compliance risks. Modern revenue administration relies increasingly on self-assessment and uses risk-based processes to ensure compliance. Resource constraints are likely to dictate that revenue administration processes are focused on identifying payers and transactions with the largest potential risk of noncompliance. An efficient risk management process contributes to minimizing evasion and irregularities in revenue administration as well as lowering the cost of collection for revenue collecting agencies and cost of compliance for payers. The assessors should consider the use of risk management process in registration, filing, payment, and refunds of tax, customs, and social security payments. They should comment on the efficiency of these processes. The assessment should also look into the mitigation measures in place such as audits, investigations, transfer pricing controls, and outreach activities/communication. 112. The Income Tax Act 2000 has a self-assessment regime as espoused by section 104A. It also classifies its tax payers into large, medium, and small taxpayers. The NRA has between 185 and 200 large taxpayers who contribute over 80 percent to NRA’s revenue. This means that the SMEs only contribute 20 percent of revenue. The NRA has over 3,000 taxpayers in its database and while SMEs make up a large and significant percentage of this figure, the contributions of the SMEs to annual revenue collection is minimal, meaning the country is losing out on this crucial source of revenue. 113. In 2013, the FA2013 introduced the SME Regime. The aim of the regime is to create a separate tax system for the ‘Hard-to-tax Sector’. The FA2013 simplified filing and payment procedures for SMEs, and as a result, it is not a requirement to have audited accounts and withholding tax claims for taxpayers that fall in the category of small/micro class that make a turnover of up to Le 350 million and as low as Le 10 million. Furthermore, the FA2015 has established the Domestic Prepared Scheme which will enable trained accountants and/or auditors to assist small and micro businesses prepare returns for their businesses. It is clear that the SME Regime could have significant implications for not only building a fair tax system but also SME contribution to gross domestic product. As SMEs move toward formalization, the sector is being recognized as a valid component of the economy which will consequently influence government plans for advancing economic infrastructure, and service provision to stimulate economic growth. 64 Score: B. Entities collecting the majority of revenues use a structured and systematic approach for assessing and prioritizing compliance risks for some categories of revenue and, as a minimum, for their large revenue payers. PI-19.3. Revenue audit and investigation 114. This dimension assesses whether sufficient controls are in place to deter evasion and ensure that instances of noncompliance are revealed. Sound audit and fraud investigation systems managed and reported on according to a documented compliance improvement plan must be in place to ensure that once risks have been identified, there is follow-up to minimize revenue leakage. More serious issues of noncompliance involve deliberate attempts at payment evasion and fraud. This may involve collusion with representatives within a revenue administration. The ability of the revenue administration to identify, investigate, successfully prosecute, and impose penalties in major evasion and fraud cases on a regular basis is essential for ensuring that payers comply with their obligations. This dimension assesses use of audits and fraud investigations managed and reported on according to a documented compliance improvement plan. Dimension 19.2 assesses the extent to which a comprehensive, structured and systematic approach is used within the revenue entities for assessing and prioritizing risks. 115. The Audit Unit of the NRA has been strengthened through an audit capacity building program, and has adopted integrated and effective risk-based audit systems. Since 2013 there has been a separation of policy and collection in NRA. Audit is carried out by Policy Division. There are clear audit plans made each year (completed to about 70 percent in 2013 with companies not audited prioritized for the subsequent year). There are 19 criteria used for risk assessment. Full financial audits are carried out. 116. The recently installed Domestic Taxpayer Information System software has improved the planning of audits which now includes GST audits. 117. Customs declarations are computerized. Sierra Leone has installed ASYCUDA and is fully operational at the Customs Headquarters covering trade using Queen Elizabeth Port which covers over 90 percent of trade. It has been rolled out to Lungi airport and there are plans to extend the service to the border posts with Liberia and Guinea. ASYCUDA includes a risk assessment module which determines the status of an import (green = automatic import; yellow = document check; red = full inspection, and blue = postclearance audit). The Postclearance Audit Unit was set up in 2011 to identify imports for postclearance audit based on a comprehensive risk assessed annual audit plan. In 2013 the unit recorded 146 audits (the target was 144). 118. To combat fraud, NRA instituted a Revenue and Intelligence Investigation Unit in 2012 with assistance from the US Treasury. It is now operational for both Customs and Domestic Taxes, and is linked to both the Police and the Anticorruption Commission. It also conducts operations within NRA. Score: A. The NRA undertakes audits and fraud investigations managed and reported on according to a documented compliance improvement plan, and complete all planned audits and investigations. 65 PI-19.4. Revenue arrears monitoring 119. This dimension assesses the extent of proper management of arrears within the revenue entities by focusing on the level and age of revenue arrears. Revenue administrations need to have a critical focus on the management of arrears to ensure that debts owed to the government are managed actively and that appropriate processes are adopted focusing on expediting the payment of collectable debt. This will ensure that revenue administrations maximize the collection of arrears before they become uncollectable. In order for the arrears management process to be considered comprehensive, it should allow for capturing information on revenue arrears and facilitate collection of those arrears in the year they occur. 120. The bulk of arrears has been held by PEs, whether for corporation tax or GST. There are no arrears on payments on imported items. The collection of arrears has improved over the review period, in particular as a result of some high visibility cases as the NRA has sought to shut down PE operations until arrears are paid. It seems that slowly but surely the tradition of parastatals not paying their full tax burden is at least being addressed from the side of the NRA. In 2010 an offsetting arrangement9 between government institutions and parastatals was agreed which paved the way somewhat for all institutions to adhere to their full liabilities whether tax or utilities invoices. 121. The NRA has also increased its capacity for arrears collection, in particular after the merger of the GST and the Income Tax Departments to form the Domestic Tax Department in 2011. Within this, restructuring the Large Taxpayers Office was reconstituted and capacitated. There is a system to collect arrears the day after they are identified, as liabilities and penalties are communicated to taxpayers, followed up by reminders and telephone calls and eventually the imposition of sanctions, such as the withdrawal of a tax clearance certificate and “naming and shaming.” Score: A. The stock of revenue arrears at the end of the last completed fiscal year is below 10 percent of the total revenue collection for the year, and the revenue arrears older than 12 months are less than 25 percent of total revenue arrears for the year. PI-24. Procurement 122. In 2006, the National Public Procurement Authority (NPPA) was established by the Public Procurement Act of 2004 revised in February 2016. Since inception there has been a marked improvement in both central and local government activities over-sighted by the Procurement Authority. The NPPA was established, together with a regulatory and monitoring body the Independent Procurement Review Panel (IPRP), serving as complaints handling body. We note that this body is now not functioning since 2013. Institutional arrangements were put in place for procurement within ministries and other government institutions including local councils. PI-24.1. Procurement monitoring 123. This indicator analyses how the efficiency and effectiveness of the procurement system is being monitored. It assesses the extent to which prudent monitoring and reporting 66 systems are in place for ensuring value for money and fiduciary integrity. This dimension looks at the availability and analysis of data for goods, services and works contracts. 124. Section 14 of the Public Procurement Act of 2016 sets out the key functions of the NPPA. One of these functions is for the NPPA to prepare an annual procurement report. Annual monitoring and assessment of procurement activities were undertaken by the NPPA over the past years. This assessment covers key spending MDAs and public sector entities, with the aim to assess compliance in procurement methods. 125. The NPPA has completed the 2015 assessment of procurement activities of 135 procuring entities out of a target of 145 which include MDAs, local councils, and other public sectors entities. This represents 93 percent of total entities targeted. 126. The data provide information on procurement activities planned for 2015, procurement activities executed and those unplanned for. It also provides the percentage of procurement planed for as against those executed. 127. In terms of numbers, 4,439 procurements were executed against a total of 3,801 planned procurements resulting to 638 procurements unplanned during 2015. Contracts to the value of Le1.481 trillion were executed against planned of Le 1.372 trillion resulting to Le 109 billion executed but not planned. Total value of procurement executed that was planned for amounted to 93 percent. The summary of the analysis is shown in table B.14. Score: B. Databases or records are maintained for contracts including data on what has been procured, value of procurement and who has been awarded contracts. The data are accurate and complete for most procurement methods for goods, services and works. Table B.14. Analysis of 2015 Procurement Data Percentage Total Executed Planned Not Planned Planned Number 4,439 3,801 638 86 Contract 1,481,342,767,114 1,371,673,518,528 109,669,248,585 93 value awarded (Le) Original 1,481,342,767,114 n.a. n.a. n.a. contract value Actual cost 1,459,552,981,844 n.a. n.a. 98.5 (Le) Difference n.a. n.a. Le 21,789,785,270 n.a. (Le) 67 Source: Ministry of Finance and Economic Development Budget Bureau. PI-24.2. Use of competitive procurement methods 128. The second schedule of the National Public Procurement Act 2016 sets out procurement thresholds that should go through open competition. Further in sections 41and 46, the Act also provides guidelines as to when a restricted bidding or sole sourcing is to be undertaken. The minimum thresholds for open competition for goods, services and works are given below: • Goods Le 60,000,000 circa ($12,376) • Services Le 60,000,000 circa ($12,376) • Works Le 150,000,000 circa ($30,941) 129. Information relating to total contracts awarded showing those below and above the thresholds and those awarded through open competition is shown in table B.15. Table B.15. Value of Contracts Awarded Using Open Method Value of contracts that are above open competitive threshold (Planned): Category Amount (Le) Value of Goods above the Threshold 829,729,765,057 Value of Works above the threshold 517,678,320,660 Value of Services above the threshold 70,907,935,035 Total 1,418,316,020,752 Value of contracts executed using open competition(Actual) Category Amount (Le) Value ICB Goods 592,465,456,392 Value NCB Goods 102,651,389,983 Value ICB Works 467,083,357,729 Value NCB Works 48,808,786,580 Value ICB Services 52,660,570,597 Value NCB Services 9,769,869,950 Total 1,273,439,431,231 Percentage of actual contract executed 90 percent using open method as against planned Source: Ministry of Finance and Economic Development Budget Buerau. 130. The percentage of contracts (value terms) that are above the open competitive threshold executed through the open competition is 90 percent. Data on the justification for the nonuse of open competition have not been provided by NPPA in the past. They, NPPA officials said that they do have records noting that there were justifications given and kept on file. However, in the past the officials stressed that the records were available but at the time of request of the information they had not put them together and also the consultants were time bound concluded their fieldwork before the information could be given to them. 68 131. Examples were given as to the case of the Ebola emergency where the use of methods not open was used as this was a genuine case and was justified. However, they ensured that the correct documentation was used. Score: A. PI-24.3. Public access to complete, reliable and timely procurement information 132. The dimension assesses the level of public access to complete, reliable and timely procurement information including; legal and regulatory framework for procurement, government procurement plans, bidding opportunities, contract awards (purpose, contractor and value), data on resolution of procurement complaints and annual procurement statistics are made available to the public through appropriate means. 133. The only procurement information available is bidding opportunities, procurement plans, contract awards, and the legal and regulatory framework. These were available during the time of this assessment on the NPPA website at www.publicprocurement.gov.sl. Information on data on resolution of procurement complaints and annual procurement statistics were not published. 134. At least four of the key procurement information elements are complete and reliable for government units representing most procurement operations and are made available to the public in a timely manner (table B.16). Table B.16. Availability of Key Procurement Information Document Availability Legal and regulatory framework for Yes procurement Government procurement plans Yes Bidding opportunities Yes Contract awards (purpose, contractor and value) Yes Data on resolution of procurement complaints No Annual procurement statistics No Score: B. PI-24.4. Existence of an independent administrative procurement complaints system 135. The IPRP was established by the Procurement Act 2004 and is still recognized in the revised Act of 2016 Section 20, serving as complaints handling body. This provides a solid legal and regulatory framework for the appeals process. In the first instance appeals are made to the Head of the Procuring Entity and in the second instance referred to the NPPA and finally to the IPRP. 136. The lack of adequate resources available to the IPRP should be addressed as this possibly affects the actual and perceived independence of the panel. For example, the IPRP has no premises, no website, and no budget with which it can obtain the specialist advice it may require to conduct a thorough hearing. Consequently, the IPRP relies on the NPPA to 69 provide specialist procurement advice; the NPPA website is used to post the IPRP appeal judgments; the lack of premises compromises the integrity of appeals files which are retained by panel members. Membership of the IPRP is rotational, and the term of the panel has expired and not yet replaced. Membership of the IPRP was dissolved in 2013 and not replaced as at time of this assessment. 137. Section 65(2) of the Act requires an administrative fee and compliant fee to be levied in order for an application to be considered by the IPRP. According to the regulations, this fee is Le 2,000,000 (Le 2 million). The reason for this, as put forward by NPPA, is to forestall unnecessary cases and complaints being brought to the appeals process. This fee is however refunded if the complainant’s appeal is upheld by the panel. Score: D. The performance is less than is required for a C PI-8. Performance Information for Service Delivery 138. This dimension measures the extent to which information is available on the level of resources actually received by service delivery units of at least two large ministries (such as schools and primary health clinics) and the sources of those funds. The information captured by ministries on resources should support the comparison of service performance with the actual resources received. The reasons for selecting the ministries for this dimension should be explained in the report narrative. PI-8.3. Resources received by service delivery units 139. The availability of information (financial and nonfinancial) which can be used to demonstrate the resources (in kind and cash) that were actually received by front-line service delivery units compared with the overall resources known to have been made available to the sector(s) is vital to improving transparency and accountability. 140. The indicator is focused on tracking the flow of relevant information, and in doing so following the money, through all levels of government to determine whether the public financial management systems effectively support front -line service delivery by ensuring all intended resources are received in a timely manner. The extent to which existing government information systems are adequate to capture this information is considered together with any relevant alternative information sources such as inspections, audits, ad hoc assessments and Public Expenditure Tracking Surveys (PETS) that can add value. 141. The PEFA focus is on primary schools and primary health care units although the same tracking principles are relevant across all services. Overall, while PETS are a useful diagnostic tool which can aid the identification of problems – they are not a substitute for the ongoing improvements to continuous monitoring systems which capture all resources. 142. In Sierra Leone, PETS are the subject of paragraph 164D of the Financial Management Regulations 2007, which sets out the authority and procedures for implementing recommendations though a Steering Committee. In 2001, the Economic Policy Research Unit of the MFED established the PETS Task Team. 70 143. In Sierra Leone, the last PETS survey was carroed out in 2011, but none his been carried out ever since. Score: D. Though the data collection system exists, it is not complied into annual reports and therefore cannot be assessed as a regular feature of a monitoring system. 1 Source Budget Bureau ‘’Guidelines and Strategy to Restore Fiscal Discipline and Strengthen Expenditure Control’’ Note, Dated January 2013 (received from Ministry of Finance and Economic Development Budget Bureau February 2014). 2 Majority refers to 50 percent or more (by value). 3 Donor funded projects which go through the Ministry of Finance and Economic Development should be captured by the TOGAS analyses. 4 Autonomous government agencies reporting their accounts to the Accountant General’s Department. 5 Those not collected by the National Revenue Authority. 6 Independent bank balances received from BSL and some commercial banks and copies of bank statements provided by some project implementation units’s and subvented agencies. 7 The statements of income and expenditure received on a quarterly basis from project implementation units’s and subvented agencies. 8 These include the RMFA, SLMA, EPA, NATCOM, GUMA, and SLRTC. 9 Some parastatals were owed payments by government budget units, for example, for telecoms and utilities. As a result, they did not pay their full tax burden in lieu of owed monies. This was addressed in 2010 to some extent by an offsetting arrangement. Appendix C. Broadening Access to Fiscal and Budget Data 1. Governments are increasingly sharing financial activities with citizens through mobile devices (tablets/smart phones). Some of the advanced applications provide online payment and search options as well. Interactive data visualization options, graphical user interfaces, feedback mechanisms, advanced search and reporting options, and innovative tools (searchable interactive maps of public information) have broadened the access to fiscal and budget data through mobile applications.xxv Singapore 2. The Ministry of Finance (http://www.mof.gov.sg) home page is easy to navigate and provides extensive information about the budget process. Budget archives present the details of policies, public spending, and results since 1996. Individuals, civil society, and businesses can post their views and suggestions through a public consultation section. Mobile applications can be downloaded from the website to learn more about the details of budget plans and execution performance. A budget quiz and download pages provide additional information in an easy-to-understand format. New Zealand 3. The New Zealand Treasury website provides access to extensive fiscal and budget information through mobile applications and other interactive tools (Twitter, Facebook, YouTube, and so on). A new budget application (NZ Budget) was launched in December 2012 to present 2012 and 2013 budget data on mobile devices (iPad, iPhone, Android), together with budget-related videos, budget speech, executive summary, and key facts for taxpayers. A tablet version provides additional information on the Budget Economic and Fiscal Update and the Half-Year update. Russian Federation 4. The Ministry of Finance is developing e-Budget as an integrated FMIS solution to improve service delivery and transparency, following the rollout of the Federal Treasury Automation System in 2012. As a part of ongoing reforms, a mobile application (Public Services) was developed for smart phones/tablets (on Android, iOS, Windows Phone) for checking tax obligations, applying for or renewing driver’s licenses, paying traffic fines, and other services. The Ministry of Finance portal also provides information on reaching retirement, migration, obtaining grants and social assistance, and acquiring real estate. Germany 5. The Ministry of Finance/Bundesministerium der Finanzen provides access to federal budget information through websites, social media, and mobile applications. The interactive website presenting the Federal Budget/ Bundeshaushalt (http://www.bundeshaushalt-info.de) revenues and expenditures (according to budget sections, functional category, and the 72 spending groups of all government institutions) demonstrates good practice. Results are displayed both graphically and in tabulated form, and can be downloaded as open budget data (PDF, XLS). 73 Appendix D. List of Persons Met Government of Sierra Leone Mr. Momodu Kargbo, Minister of Finance and Economic Development Mr. Brima Bangura, Deputy Minister of Finance and Economic Development Mr. Chernor Maju Bah, Deputy Speaker of Parliament and Chairman, Parliament Accounts Committee Mr. Edmund Koroma, Financial Secretary Mr. Ansu Tucker, Senior Deputy Financial Secretary Mr. Prince Cole, Senior Deputy Financial Secretary Ms. Princess Johnson, Acting Director PFM Reforms Unit Ms. Claudia Johnson, Finance Officer PFM Reforms Unit Mr. Alpha Sesay, Director, Integrated Project Administration Unit Ms. Nancy Kamara, Head of Finance, Integrated Project Administration Unit Mr. Alex Kaija, Finance Officer, Integrated Project Administration Unit Mr. Lamin Vandy, Project Director, PFMICP Mr. Richard Williams, Accountant General (AG) Mr. Lawrence Caulker, Deputy AG Mr. Raymond Coker, Assistant AG Mr. Tamba Momoh, Deputy Auditor General, Auditing Service Sierra Leone Mr. Simeon Jonjo, Director, Information and Communication Technology (ICT) Mr. Max Bailor, Systems Manager Mr. Mathew Dingie, Director, Budget Bureau Mr. Kandeh Sesay, Director, Internal Audit Department Mr. SNK Lansana, Deputy Director, Internal Audit Department Mr. Abdul Rahman Koroma, Assistant Director Operations, Internal Audit Department Mr. Abdulai Samu, Assistant Director Admin, Internal Audit Department Mr. Sahr Jusu, Director, Debt Management Unit, Mr. Alimamy Bangura, Director, Economic, Policy, and Research Unit (EPRU) Mr. Alfred Coker, Head of Monitoring and Evaluation, NPPA Mr. Joseph B. Dauda, Head of Finance Department, NPPA Mr. Mohamed Musa, Head of Capacity Building Department, NPPA Mr. Alie B. Kargbo, Head of Research Unit, NPPA Mr. Brima F. Barmeh, Senior M&E Officer, NPPA Mr. Idrissa Kanu, Director, Revenue and Tax Policy Mr. Tamba Momoh, Deputy Auditor General Mr. Selvin Bell, Deputy Auditor General Mr. Festina Macualey, Head of Unit, Accountant General Department Mr. Adams Kargbo, Director, Local General Finance Department Mr. Adams Tommy, Acting Deputy Director, Local General Finance Department, Mr. Salieu Kamara, Economist, Local General Finance Department, Bo City Council Mr. Victor Kalie Kamara, Chief Administrator, Bo City Council Mr. Foday Daboh, Finance Officer, Bo City Council Mr. Augustine Amara, Chief Administrator, Bo City Council Mr. Kemoh Sartie, Finance Officer, Freetown City Council 74 Mr. Mohamed Madina Bah, Finance Officer, Freetown City Council Mr. Sidi Bah, Director, Nonstate Actor Mr. Abdul Rahman Bayoh, Director, HRMO Mr. Jalloh, Human Resources Director, HRMO Mr. Nyakeh Nyamah, Director IT, HRMO Mr. Peter Sam-Kpakra, Director, Multilateral Project Department Development Partners Ms. Sarah Somoudi, DFID Mr. Daniel Grotino, Program Officer, EC World Bank Mr. Sheikh Sesay, Acting Country Manager, Sierra Leone Roberto O. Panzardi, Task Team Leader, IPFMRP Mr. Victor Boakye-Bonsu, Task Team Leader, PFMICP Mr. Sydney Godwin, Financial Management Specialist, Sierra Leone Civil Society Mr. Abdul Rahman, District Coordinator West Mr. Lamin, Transparency International Mr. Feli, Democracy and Development Associates Private Sector Mr. Ahmed Nano, Executive Secretary, Sierra Leone Chamber of Agribusiness Development