Report No. 43282-LR Liberia 2008 Public Expenditure Management and Financial Accountability Review June 2009 Poverty Reduction Economic Management Sector Unit (PREM 4) Afirca Region Document of the World Bank, co-produced with the Government of Liberia, the Africa Development Bank, International Monetary Fund, United Nations Development Programme, Department for International Development, and Swedish National Auditing Office. TABLE OF CONTENTS ACKNOWLEDGEMENTS vi1 PREFACE .......................................................................................................................................................... .......................................................................................................................... IX EXECUTIVE SUMMARY .............................................................................................................................. X ACTION PLAN FOR PFMREFORM ................................................................................................. XXIII 1.A.MACROECONOMICPERFORMANCE ........................................................................................... ............................................................................... 1 B. MA~ROE~~NOMI~PERFORMANCE THELEGACYCIVIL WAR............................................................................................................. AND FISCAL 2 c. MACROECONOMIC OF .............................................................................. 3 2 D. MEDIUM- LONG- DEVEL~PMENTSSINCE2.003 ............................................................. 6 E. TO TERM PROSPECTS THE ECONOMY FOR ............................................................................................... 8 F. LONG-TERM STRUCTURALCHANGES RESOURCE MOBILIZATIONANDFISCAL SPACE................................................................................ 9 .................................................................................................. 9 ................... .................................................................... 10 G. REVENU .................................................................................................................. 12 Grants................................................................................. 15 Government Borrowing and Debt Servicing.......................................................................................... 16 Domestic Debt Servicing.......................................................................................................................... 16 .......................................................................................... 17 H. ......................................................... 18 SUMMAR ATIONS .......................................... 18 2.I . CONCLUSIONS AND RECOMMENDATIONS ....................................................................................... A.PUBLICFINANCIALMANAGEMENT ....................................................................... 20 ................................................................................................................................ 21 SYSTEM 21 B. INTRODUCTION OVERALL ASSESSMENTTHE PUBLIC FINANCIAL MANAGEMENT SYSTEM OF ................................ 21 c. LEGAL ....................................................................................................................... 24 D. FRAMEWORK BUDGET PREPARATION .................................................................................................................... 24 Budget Timeliness, Comprehensiveness, and Policy Linkage . Budget Predictability .................................................................................................................... Shijit toward a Medium-Term Program Budget Framework ... E.Decentralization . . ........................................................................................................................................ 28 BUDGET EXECUTIONANDCONTROL............................................................................................... 28 ................................................................................. 28 Budget Transfers .............................................. Payroll ..................................................................... F. PROCUREMENT ................................................................................................................................ 30 The Legal and Regulatory Framework ........................................................................................... 30 Institutional Framework and Management Capacity............................................................................ 30 Procurement Operations and Market Practices . Private Sector Participation....................................................................................................... ........................................................................................................ 32 ............................................................................. 33 H.Internal Audit ................................................................................................................................ EXTERNAL OVERSIGHT ................................................................................................................... 35 The GeneralAuditing Commission ......................................................................................................... 35 Legislative Scrutiny of the Annual Budget Law ..................................................................................... 36 I.Legislative Scrutiny of External Audit Reports ....................................................................................................................... ................................................ 37 J. DEBT MANAGEMENT DONOR PRACTICES .......................................................................................................................... 37 38 3.K PUBLICEXPENDITUREANALYSIS.............................................................................................. . CONCLUSIONSANDRECO~~~~ENDATIONS ....................................................................................... 39 A. 42 INTRODUCTION ................................................................................................................................ 42 OverallRevenue Trends ........................................................................................................................... 42 Expenditure Trends.... ........................................................................................................... 42 Donor Expenditure ....................................................................................... 44 B. ALLOCATION OF RESOURCES .......................................................................................................... 45 Efficiency .............................................................................................................................. 47 50 4.C .Operational CONCLUSIONAND RECOMMENDATIONS SECTOR EXPENDITUREREVIEWS .............................................................................................. ......................................................................................... 52 ................................................................................................................................................. 53 PrivateExpenditure ......................................................................................................................................... 58 Equity of Health Expenditures................................................................................................................. 61 B .Conclusions and Recommendation EDUCATION SECTOREXPENDIT Introduction................................................................................................................................................ 66 Tertiary andVocational Education.................................................................................................................. 68 Teacher Supply ................................................................................................................................................ 68 TotalPublic Expenditure on Education.................................................................................................. 68 Distribution of Government Expenditure by Level of Education................... The Contribution of ExternalDonorsto Education Financing....................................................................... 71 Detailed Analysis of Education Salary andNon-Salary Expenditur Equity of Spending in Education .... Disparity in Schooling..................................................................................................................................... 76 Disparity inHousehold ......... Disparity inPublic Exp ......... EfJiciency on Spending .................................................................................... 82 Characteristicso f Schools and Outcomes ....................................................................................................... 82 Conclusion and Recommendations ............................................ :......................... 85 LIST OF TABLES: TABLE1.1:KEY ECONOMICINDICATORS. 2003-2008 .................................................................................. 4 TABLE 1.2: FISCALINDICATORS, 2002/03-2007/08 (us$MILLION)............................................................. 5 TABLE1.3: GDPGROWTH ................................. 7 TABLE1.4: RECOMMENDEDREFORMSTO TAXPOLICY .............................................................................. AND INFLATION PROJECTIONS, 2007-2012 (PERCENT) TABLE1.5: PROJECTEDREVENUES,2008/09-2011/12, AND OUTTURN, 2006/07-2007/08 (us$MILLION) 11 ...................................................................................................................................................................... 14 TABLE1.6: REVENUEPROJECTIONS, 2007/08-2011/12 WITH LOWER GROWTH GDP MILLION) ......15 TABLE 1.7:PROJECTEDDOMESTICDEBT SERVICINGCOSTS, 2007/08-2011/12 (us$(US$ MILLIONS) ............17 TABLE1.8: PROJECTEDANNUALDEBTSERVICEAFTERHEAVILY INDEBTEDPOORCOUNTRY(HIPC), MULTILATERAL RELIEFINITIATIVE(MDRI),ANDADDITIONALBILATERALDEBTRELIEF (US$ DEBT MILLION) ....................................................................................................................................................... TABLE1.9: ESTIMATEDCOST BYPILLAR, 2008/2009-2010/201 1(us$MILLION) .................................... 18 19 TABLE2.1:SUMMARYOF 2007PUBLICEXPENDITURE AND FINANCIALACCOUNTABILITY (PEFA) PERFORMANCESCORES ................................................................................................................................ 23 TABLE3.1:VARIATION BETWEENORIGINALBUDGET EXPENDITURE, AND 2006/07-2007/08 (FUNCTIONAL CLASSIFICATION) .......................................................................................................................................... 44 TABLE 3.2: VARIATION BETWEENORIGINALBUDGET EXPENDITURE, AND 2006/07-2007/08 (ECONOMIC CLASSIFICATION) .......................................................................................................................................... 44 EXPENDITURE,2004/05-2007/08 ................................................................................................................. 46 TABLE3.3: FUNCTIONAL CLASSIFICATIONOF PUBLICEXPENDITUREAS A PERCENTAGEOFTOTAL TABLE 3.4: FUNCTIONALCLASSIFICATIONOF PUBLICEXPENDITURE PERCENTAGEOF GDP, 2004/05- IN 2007/08 ......................................................................................................................................................... 46 TABLE 3.5: ECONOMIC CLASSIFICATION OF PUBLICEXPENDITURE, A PERCENTAGEOF TOTAL AS EXPENDITURE,2004/05-2007/08 ................................................................................................................. 48 TABLE 3.6: ECONOMIC CLASSIFICATION OF PUBLICEXPENDITURE, 2004/05-2007/08 ...................................................................................................................................................................... (PERCENT OF GDP) 48 TABLE 4.1: HEALTHEXPENDITURE IMPLEMENTATION RATE, 2004/05-2007/08 ....................................... TABLE4.2: DISTRIBUTION OF PUBLICHEALTHSPENDINGBY ECONOMIC TYPE, 2004/05-2007/08 ..........56 58 iv TABLE4.3: DISTRIBUTION HOUSEHOLDSPENDINGFORHEALTH. 2007 ................................................ OF 59 TABLE4.4: RESOURCESMOBILIZEDFORTHEHEALTHSECTOR. 2006/07 (us$ MILLION) ......................... 59 TABLE4.5: TIME TONEAREST TABLE4.6: STRUCTUREOFHOUSEHOLD'S EXPENDITURE HEALTH,2006/07 ....................................... INFRASTRUCTURE. 63 2007 (MINUTES) ........................................................... ON 64 TABLE4.7: HOUSEHOLDS EXPENDITUREONHEALTH, 2006/07.................................................................. 64 TABLE4.8: NETAND GROSS ........................................................................... 68 TABLE4.9: HOUSEHOLD SPENDINGON EDUCATIONBY LEVEL, 2006/07 ................................................... ENROLLMENT RATES, 2007 71 TABLE 4.10: DISTRIBUTION HOUSEHOLD OF SPENDINGFOREDUCATION,2007 ........................................ 71 TABLE4.1 1:DEVELOPMENT PROJECTED ASSISTANCEFOREDUCATION(US$ MILLION) ......................................................................... PARTNERSANDNONGOVERNMENTAL ORGANIZATIONS' CURRENTAND 72 TABLE4.12: RESOURCESFOREDUCATION,2006/07 (us$ MILLION) .......................................................... 72 TABLE 4.13: UNIT COST ESTIMATESFOREDUCATIONBY LEVEL, 2006/2007 ............................................ 73 TABLE4.14: COMPARISONSOF UNIT COSTAS A PERCENTAGEOF GDPPERCAPITA INSELECTEDSUB- TABLE4.15: PUPILTO TEACHERRATIOINSUB-SAHARAN AFRICA,2004-2006 ........................................ SAHARANAFRICANCOUNTRIES, BY LEVEL ("/.).......................................................................................... 74 76 TABLE4.16: AGE-SPECIFIC ENROLLMENTRATESAND OUT-OF-SCHOOL CHILDREN BY AGE GROUP (PERCENT) ..................................................................................................................................................... 77 TABLE4.17: GROSS ENROLLMENT RATES BY GENDER,LOCALITY, REGION, AND HOUSEHOLD EXPENDITURE QUINTILE, 2007 ..................................................................................................................... 78 TABLE4.18: PERSTUDENTEXPENDITUREONEDUCATION,BYQUWTILE AND LEVEL OFEDUCATION, (US$) ............................................................................................................................................................ 2007 80 LIST OF FIGURES: FIGURE3.1: REVENUEANDEXPENDITUREGROWTH. 2002/03-2007/08 ..................................................... 42 FIGURE3.2: QUARTERLY REVENUEANDEXPENDITUREPATTERNS. 2006/07-2007/08 .............................. 43 2007/08 ......................................................................................................................................................... FIGURE3.3: SHAREOF TOTAL GOVERNMENTEXPENDITURE FUNCTIONAL BY CLASSIFICATION. 2004/05- 46 FIGURE 3.4: SHARE OF TOTAL GOVERNMENTEXPENDITURE ECONOMIC BY CLASSIFICATION 2004/05- 2007/08 ......................................................................................................................................................... 48 FIGURE4.1: EVOLUTION OF PUBLICEXPENDITUREONHEALTH, 2004/05-2008/09 .................................. 55 FIGURE4.2: PUBLICHEALTHEXPENDITURE ?`o OF GDPINSUB-SAHARA AFRICANCOUNTRIES, 2004- 07.................................................................................................................................................................. AS 56 FIGURE 4.3: HEALTHEXPENDITUREBY PROGRAM, 2004/05-2007/08 (PERCENT SHARE) FIGURE4.4: TOTALHEALTHEXPENDITUREAS PERCENTOF GDP, 2004-07 .............................................. ......................... 57 60 FIGURE4.5 :EXTERNAL FIGURE4.6: COMPOSITIONOF HEALTHEXPENDITURE(PERCENTOFTOTAL) ............................................. ASSISTANCEAS A PERCENTAGEOF TOTALHEALTH EXPENDITURE ..................... 61 ..........61 FIGURE4.8: POPULATIONPERTECHNICAL HEALTHPERSONNELRATIO BY REGIONAND CATEGORY .....63 FIGURE4.7: PERCAPITA PUBLICEXPENDITURES CURATIVE SERVICES, 2006/07 (BY REGION) ON 62 FIGURE4.9: GROSS ENROLLMENT RATES INPRIMARYEDUCATION ........................................................... 67 FIGURE4.10: EVOLUTIONOF GOVERNMENT EXPENDITUREON EDUCATION, 2004/05-2008/09 ...............69 FIGURE4.1 1: SHAREOF CURRENT EXPENDITURE LEVEL OFEDUCATION,2004/05-2007/08 (PERCENT) ...................................................................................................................................................................... BY 2007 .............................................................................................................................................................. 75 FIGURE 4.12: RATIO OF AVERAGE TEACHER SALARY TO GDPPER CAPITA IN SUB-SAHARAN AFRICA,. 70 FIGURE 4.13: ENROLLMENT RATESBY AGE AND GENDER.......................................................................... 77 4.14: GROSSENROLLMENT RATES BY COUNTY, 2007 ................................................................................ 79 FIGURE4.15: PERSTUDENTSPENDINGON PRIMARYAND SECONDARYEDUCATIONAS A PERCENTAGEOF PERCAPITA HOUSEHOLD SPENDINGBY QUINTILE, 2007............................................................................ 80 ................................. FIGURE4.17: DISPARITY CONSUMPTIONOF PUBLICEDUCATIONRESOURCES, 2006/07 ....................... FIGURE4.16: CONCENTRATIONCURVEOF ENROLLMENTINPUBLIC SCHOOLS, 2007 81 IN FIGURE4.18: PRIMARYREPETITIONRATESINSUB-SAHARAN AFRICANCOUNTRIES, 2005/06 ..................8284 FIGURE4.19: DROPOUT RATESINPRIMARYAND SECONDARYLEVELS, 2005/06 ........................................ 85 LIST OFANNEXES: ANNEX 1:PUBLIC EXPENDITURE REVIEWS. TABLES.AND FIGURES ........................................................... 90 ANNEX 2: LIBERIA2007 PUBLICFINANCIALMANAGEMENT PERFORMANCEREPORT ............................... 97 ANNEX2.1: PERSONS ANNEX 2.2: PFM-PR CALCULATIONSHEETS-INTERVIEWED .......................................................................................................... 147 .......................................................................................... 148 ANNEX 2.3: PFM-PR- LISTOF MINISTRIES AGENCIES AND ................................................................. 150 V ANNEX 2.4: PFM-PR SUBSIDIESAND TRANSFERS FY 2006/07: SUBSIDIES& TRANSFERS(US$) . ....153 ANNEX 2.5: PFM-PR: BUDGET PREPARATION CALENDARFORFISCAL ~ ~ ~ ~ 2 0 0 7 - 2 0 0 8 ....................... 154 ANNEX2.6: PFM-PR .OPENPROCUREMENTPROCESSFORSMALLPURCHASES....................................... 157 LIST OFANNEX TABLES: TABLEA1.1:FUNCTIONAL CLASSIFICATIONOFPUBLICEXPENDITURE.INPERCENTOF TOTAL EXPENDITURE. 2004/05-2007lO8 ................................................................................................................. 90 TABLE A1.2: GOVERNMENTEXPENDITURE ON HEALTH. 2004/05-2007/08 .............................................. 91 TABLEAl.3: BENEFIT INCIDENCE HEALTH EXPENDITURE EDUCATION. 2000-2007 ................................................... ANALYSIS FORTHE USEOF CAREFACILITIES.LIBERIA2007 92 TABLEA1.4: GOVERNMENT TABLEA1.5: DISTRIBUTION CURRENT EXPENDITURE LEVEL EDUCATIONAND TYPEOF ON 95 OF BY OF EXPENDITUREIN2006/07 ............................................................................................................................. 96 TABLEA2.1: SUMMARY OF 2007 PEFA PFMPERFORMANCESCORES.................................................... 100 TABLEA2.2: NUMBEROF PUBLICFINANCIAL MANAGEMENT(PFM) INSTITUTIONSAND THE VOLUMEOF RESOURCESMANAGEDINEXECUTINGGOVERNMENT PROGRAMS........................................................... 102 TABLEA2.3: ACTUAL EXPENDITUREBYFUNCTIONAL CLASSIFICATION (AS PERCENT OF TOTAL EXPENDITURE) ............................................................................................................................................ 102 TABLEA2.4: ACTUALEXPENDITUREBY ECONOMIC CLASSIFICATION (AS PERCENTOF TOTAL EXPENDITURE) ............................................................................................................................................ 103 LIST OFANNEX FIGURES: FIGUREAl.1: CONCENTRATION CURVESFORUSEOFPUBLIC HEALTH FACILITIES.2007 ......................... 93 FIGUREA1.2: CONCENTRATIONCURVES FORUSEOF PRIVATEHEALTH FACILITIES. 2007 ....................... 94 vi ACKNOWLEDGEMENTS The Public Expenditure and Financial Accountability Review (PEMFAR) for Liberia was initiated in 2007, based on consultations with the Liberian authorities. It is the first comprehensive study o f public financial management systems inLiberia conducted since the civil war (1989-2003). The study was intended to provide the analytical foundation for government and development partner discussions on the phase-out of the Governance and Economic Management Assistance Program (GEMAP), and the broader public financial management reform agenda necessary to lay the foundation for future budget support. The study was undertaken in collaboration with key ministries and agencies and development partners, including the International Monetary Fund (IMF), the African Development Bank (AfDB), the United Nations Development Programme (UNDP), and the Swedish National AuditingOffice (SNAO). The report was prepared based on findings from three missions conducted between October 2007 and January 2008. The aide memoir from the first PEMFAR mission was circulated and discussed in November 2007, at the GEMAP technical team meeting, a forum that brings together government and partners. Subsequent drafts o f the PEMFAR and 2007 Public Expenditure and Financial Accountability (PEFA) assessment were circulated for review in February and May 2008, and comments were received from the Ministero f Finance, Revenue Department, and General Auditing Commission. Development Partners also provided several rounds o f reviews, including the IMF, AfDB, and UNDP. The final version o f the PEMFAR was discussed and validated by stakeholders at the Budget Support Working Group inMarch 2009. The Bank would like to acknowledge the gracious participation of and provision of information from the following key government officials: Hon Antoinette Sayeh, former Minister of Finance; Hon. Augustine Ngahan, Minister o f Finance; Hon. Tarnue Mawolo, Deputy Minister o f Finance; Hon. Elfrieda Tamba, Deputy Minister of Finance; Hon. Roderick Smith, former Deputy Minister o f Finance; John Morlu, Auditor General; Dr. William Allen, Director o f the CSA; Keith Jubah, Chairman o f PPCC; Joseph Neufville, Executive Director o f PPCC; Hon. Walter Gwenigale, Minister o f Health; and Hon. Joseph Korto, Minister o f Education. The Task Team Leader responsible for the preparation o f the product was Emmanuel Fiadzo (AFTP4). The PEMFARwas guided and supervised by Antonella Bassani, Sector Manager (AFTP4). Peer reviewers included Marcel0 Andrade (AFTP4) and Albert0 Agbonyitor (FRM), who provided valuable input and comments. The team benefited from the guidance and support o f the Country Director, Ishac Diwan, and Country Manager, Ohene Owusu Nyanin. Other contributing members o f the Team included Winston Cole, Parminder Brar (AFTFM), Errol Graham, Rebecca Simson (JPO), Anna Kristiina Karjanlahti (PO) and Eric Nelson (AFTP4). Nathalie Lahire (AFTH2), Yi-Kyoung Lee (AFTH2), and Giuseppe Zampaglione (AFTH2). The PEMFAR report was made possible by the contributions o f Bank staf6 partners, and consultants. These included: Executive Summary: Emmanuel Fiadzo (AFTP4) Chapter 1: Martin Brownbridge, Emmanuel Fiadzo, Errol Graham and Rebecca Simson (AFTP4); Chapter 2: Emmanuel Fiadzo (Economist, AFTP4), with input from Winston Percy Cole (Financial Management Specialist, AFTFM), Chukwuma Obidegwu (Lead Economist, AFTP2), Tidiane Toure (Senior Procurement Specialist, AFTPC), (Research Analyst, vii AFTP4), Loxly Epie (AfDB), Mar0 Gabriel (AfDB), Robert Cauneau (consultant), Simplice Zouhon Bi (consultant), and Charles Taylor (consultant); Chapter 3: Simplice Zoun-Bi, Julien Bandiaky (consultant), Chapter 4; John Short (DFID), Julien Bandiaky. The PEFA assessment (annex 2) was conducted by Parminder Brar (Senior Financial Management Specialist, AFTFM), Winston Percy Cole, Emmanuel Fiadzo, Loxly Epie, Lage Olofsson (Swedish National Auditing Office), B o Sandberg (Swedish National Auditing Office), and Rebecca Simson, with support from Amitabh Tripathi (consultant, MOF). The final report also benefitedfrom written inputand advice from: Duncan Last (FAD-IMF), Dominic Sam (Country Director, UNDP) and Malcolm McPherson (consultant, UNDP) and Roger Swagler (consultant) The report was produced in close collaboration with the Government o f Liberia. Substantive input during the data collection, preparation and review process was provided by the Ministries o f Finance, Health, and Education; Civil Service Agency; General Auditing Commission; and legislature. Particular mention must be made to the invaluable guidance and advice provided by the former Minister o f Finance, Hon Antoinette Sayeh and current MinistersofFinance, Hon. Augustine Ngafuan. Esther Bryant (AFMLR),'Glaucia Reis Ferreira, Joan Shopmeyer-Medina (AFTP4) and Lucia Chuo (AFTP3) provided critical administrative support and management services. viii PREFACE Liberia's strong reform performance since the end of the civil war'in 2003 has been widely commended. The 14-year civil war and preceding decade of economic decline and poor governance had a devastating effect on Liberia's institutions, particularly public financial management systems. Since 2006, Liberia has developed its first full PRSP, successfully implemented an IMF-supported staff-monitored program and poverty reduction growth facility, cleared arrears to multilateral institutions, reached the Heavily Indebted Poor Country (HIPC) decision point, and embarked on reform agendas across the public sector. Butmuch still remainsto be doneto strengthengains, modernizesystems and processes, and ensure the sustainability of current efforts. This study was undertakento take stock of the current public financial managementpractices, conditions, and ongoing reform efforts, and provide recommendationsfor a more harmonized and strategic reform agenda. It aims to provide a framework for future reform and donor support, as the transitional Governance and Economic Management Assistance Program (GEMAP) begins its phase-out. It is also hoped that the PEMFAR will provide a basis on which to judge Liberia's readinessfor budget support, and recommendations for moving the government towards increased aid. This study provides a baseline of conditions at the beginning of Liberia's post-conflict recovery, from which Liberia's performance can be judged. Public Expenditure Reviews focused on two strategic sectors: health and education, identified as priority sectors in the PRSP. These expenditure reviews provide initial observations, based on limited data, about the allocative and operational efficiency of public spending. The PEMFAR was finalized in late 2008 and covers developments through the end of the Liberian Government fiscal year 2007/08. The accompanying PEFA assessment (annex 2) was completedin2007, relying on 2006/07 fiscal data. The PEMFAR represents a trulyjoint donor partner and government effort. Ledby the Bank, it includes contributions from the IMF, AfDB, UNDP, and Swedish National Auditing Office, as well as information and input from key line ministries, agencies, and the legislature. ix EXECUTIVESUMMARY 1. The 2008 Public Expenditure Management and Financial Accountability Review (PEMFAR) is the first comprehensive assessment of public expenditure allocations and financial management systems in Liberia by development partners and the Liberian government in the past 20 years. It is a consolidated diagnostic tool to enhance understandingof the country's public expenditure management, assess the fiduciary risks to government and donor resources, and provide recommendations for comprehensive public expenditure reform. This report builds on the findings of the assessments conductedjointly bythe World Bank, the AfricanDevelopmentBank, the International Monetary Fund, United Nations Development Programme, Department for International Development, and the Swedish National Auditing Office, between September 2007 and February 2008 in Liberia. The mission worked closely with the Ministry of Finance, the Departmentof the Budget, the Auditor General, the Public Procurementand Concessions Commission, and other ministries and state-ownedenterprises. 2. The challenges for public expenditure management reform inLiberiaare shaped by the country's history over the past three decades. Liberia has begunto recover from a devastatingcivil war that reducedreal output to less than a third of its peak at the beginning of the 1980s. Production of iron ore, the country's major industry, ground to a halt, the physical infrastructure was destroyed, and public services deteriorated dramatically. Liberia remained in default on its external debt from the mid-1980s until it beganclearing arrears to multilateral creditors in 2007. It is currently one of the most heavily indebted countries, relative to GDP, inthe world. 3. The government led by President Johnson-Sirleaf, inaugurated in January 2006, is committed to comprehensive economic and institutional reform. It has implemented measures to improve fiscal discipline and strengthen financial controls over the use of public funds. Key reforms include a Staff Monitored Program agreed with the IMF in 2006, a subsequent Poverty Reduction and Growth Facility Program, agreedin2008, andthe ongoing Governance and Economic Management Assistance Program. After its successful clearance of arrears with the International Monetary Fund (IMF), World Bank, and African Development Bank (AfDB), Liberia reached the Heavily Indebted Poor Country (HIPC) decision point inMarch2008. 4. Given the institutional collapse over the past decades and the ambitious ongoing reform efforts undertaken by the government to address its weak public financial management (PFM) institutions, this is an opportune time to take stock of the current direction of PFM reforms in Liberia. The PEMFAR provides a comprehensivereview, of past practicesand ongoing reforms and makesrecommendationsfor the future. 5. The PEMFAR assesses Liberia's public expenditure performance along four main dimensions: macroeconomic, institutional, allocations between sectors, and allocations within key sectors. At the macroeconomic level, it describes the external economic constraints and opportunities that will drive growth and revenue mobilization over the medium term and determine the likely fiscal envelope (chapter 1). It next reviews the institutional structures that guide and ensure appropriate resource allocation and assesses the strength and weaknesses ofthe current systems using a standardizedPublic Expenditure and Financial Accountability scoring system (chapter 2). It then assesses how public expenditure has been allocated between sectors and priorities (chapter 3). It lastly reviews the allocation ofresources within two priority sectors, healthand education(chapter 4). X 6. The PEMFAR recommendationsand action plan were developed in consultation with the government of Liberia and reflect national priorities. The original concept note and draft PEMFAR were reviewed extensively by the Ministry o f Finance and the government-ledBudget Support Working Group. Many o fthe recommended actions build on ongoing reform efforts in the context o f the HIPC Initiative and Poverty Reduction and Growth Facility. The Government intends to use the PEMFAR to help guide its PFMreform planning in preparation for a phase-out o f the ongoing Governance and Economic Management Assistance Program (GEMAP). The PEMFAR i s also a response to a call from donor partners for a more comprehensive assessment o f Liberia's PFM system to underpin their assistance strategies, and the report has benefited from their reviews and comments. The PEMFAR will be disseminated broadly to the government and donor partners. The report will be presented and discussed before the Economic Governance Steering Committee, chaired by the President, and the Budget Support Committee. Further workshops will also be held with relevant stakeholders to discuss the findings and recommendations. 7. Note that public expenditure reviews at this stage in Liberia's post-conflict development are significantly constrained by the limited availability and quality o f data, notably about donor resources. In this respect, several PEMFAR recommendations focus on actions neededto strengthen the quality and broaden the coverage o f public expenditures. Key Findings 8. Sustainable economic growth is a prerequisite for creating fiscal space for priority expenditures, because growth in government revenue will primarily be driven by growth in the revenue bases. So, the government should continue to ensure that fiscal policy supports macroeconomic stability by maintaining budget discipline and avoiding domestic financing o f the budget. It should endeavor to persuade donors to gradually channel a larger share o ftheir assistancethrough the government budget to enable itto assume greater responsibility for providing essential public services. Over the medium term, there is scope for strong growth in the budget envelope. But there will still be large shortfalls in budget resources relative to spending requirements, making it essential to improve the allocative and technical efficiency o f expenditures. For both, the government should consolidate and deepen its ongoing public financial management reform efforts to increase absorption capacity, attract more donor resources through the budget, and increase efficiency in public expenditure. 9. Liberia's public financial management system has improved markedly since 2006, and an ambitious reform agenda is under way, which will require strong continued efforts to ensure that reforms take root. Many o f Liberia's financial management structures are transitional, designed to strengthen central controls, curb corruption, and enable centralized planning during reconstruction. Over the medium term, there will be a need to shift toward longer term and more sustainable solutions in some areas and to continue implementing, fine-tuning,and entrenching new systems recently introduced. Both the government and partners should keep inmindthat they are rebuilding an extremely weak and outdated PFM structure, which deteriorated markedly during the civil conflict. Therefore, reform efforts need to focus on "getting the basics right," rather than introducing sophisticated reforms or systems that may absorb a large share o f the government's scarce resources and human capacity. 10. There have been some improvementsin the inter-sectoral allocativeefficiency of expenditures,with a larger share of the budgetallocatedto the key social services. These xi positive shifts inbudget allocations should be maintained. Inaddition, the share of the budget allocated to investments in economic services and infrastructure should increase, if the budget is to support economic growth. Note, however, that government spending represents only a share of public investmentand that significant donor funding devoted to investment is spent outsidethe nationalbudget. 11. Despite recent increases, there is a need to further expand public resources for education and health. Furthermore, the efficiency and equity in public spending on health and education can be improved by shifting more resourcestoward the rural areas and poorest regions, by better integrating donor expenditure in overall budget planning, by shifting a larger share of resources inthe education budget to primary and secondary education, and by encouragingmore girls to remain inschool, thus reducing gender imbalances. MACROECONOMIC PROSPECTSAND FISCAL SPACE 12. Although the Liberian economy grew strongly in the 1960s and 1970s, driven by the production and export of iron ore and rubber, the benefits of growth were distributed inequitably. By 1980, GDP per capita had increased to middle-income country levels, but the economy first slumped and then collapsed over the next two decades. The collapse resulted from deteriorating terms of trade, economic mismanagement and the civil war, which afflicted the country from 1989-1996 and again during 2001-2003. In the first phase of the civil war, the economy contracted by 90 percent in real terms. When peace was restored in 1996, the economy began to recover rapidly, driven by a strong expansion of timber exports. But output fell sharply again after the resumption of fighting in 2001, which ledthe United Nationsto impose sanctions on Liberia's timber and diamond exports. 13. The Liberian economy has the potential for strong growth, but it also faces major risks. Since the signing ofthe ComprehensivePeaceAgreement in2003, which ended the second phase of civil war in Liberia, the economy has startedto recover, with real GDP growth accelerating to about 9.5 percent in 2007. Strong economic growth since 2003/04 enhanced government revenue. Other contributing factors to revenue growth include stronger enforcement in customs administration (including the strengthening of pre-shipment inspection of imports since 2006), strengthening the Large Taxpayer Unit, eliminating the settlement oftax liabilities through noncashpayments, and reducingtax and duty incentives. 14. I n the short to medium term, continuinggrowth of aggregate demand should be sufficient to maintain rapid GDP growth by bringing underusedfactors ofproduction back into use, as long as the economy has sufficient access to foreign exchange to fund the widening trade deficit. Unemployed labor, for example, could be employed in labor-intensive services, construction, and agriculture.' This is a common patterninpost-conflict economies, where growth is driven by a recovery of capacity use and associatedimprovements in factor productivity. However, the current global economic crisis may slow down this recovery process, ifit affects the rate of investmentandtransfers to Liberia. The short-term priority for macroeconomic policy should be to stimulate growth by maintaining budget discipline and ensuring that demand remains buoyant, that jobs are created, and that there is sufficient foreign exchange to meet the economy's need for imports. 'The government has set ahigh priority on labor-intensivepublic works as ashort to mediumterm povertyreducing strategy to provide incometo vulnerable groups and boost aggregatedemand. xii 15. Over the longer term, however, sustainable growth will be possible only with an expansion of the supply side of the economy. This will require investment in public infrastructure, private investment in commercial activities, and investment in human resource development through improving the access to and the quality o f health and education services. The medium-to-longer term effects o f the global economic slowdown are still hard to predict and difficult to measure, but its impact is likely to be felt as a result o f falling demand and prices for Liberia's primary exports, falling remittance inflows, possible falls in foreign direct investment or delays to large investment projects, and a possible fall-off in inflows o f official development assistance. These factors may result in a lower growth rate than previously estimated. But Liberia's growth rate i s expected to remain high in comparative perspective, as production restarts inminingand forestry. 16. The sectors of the economy with the greatest potential to contribute to sustainable growth are timber, mining, agriculture, and services. With the lifting o f sanctions on Liberia's timber exports in 2006, production i s expected to recover. But logging must be sustainable to maintain long-term positive benefits for the country's development. Growth in agriculture will be slower, but because it constitutes about 40 percent o f GDP and employs a high share o f the labor force, its impact on poverty reduction could be significant. Service growth will be boosted by further growth in domestic demand. This includes public services, construction, private consumption and the service inputs required to support the timber and miningindustries, such as transport. 17. Large-scale investments in mining will boost GDP, exports, and government revenue. The government has signed a Mineral Development Agreement with Arcelor Mittal for a US$l.5 billion investment to rehabilitate iron ore production in Nimba County, with production scheduled to start around 2010. The government i s also inthe process oftendering two other iron ore deposits. But a natural resource boom can pose difficult challenges for economic management. Mining itself will create relatively few jobs. There i s also a strong likelihood o f Dutch disease, which would have negative effects on other (more labor- intensive) tradable goods sectors. In many developing countries, including Liberia in the 1960s and 1970s, natural resource booms have not translated into equitable growth and poverty reduction. Instead, they have been associated with inequality, corruption, wasteful public spending, and economic volatility. To avoid these problems, the government should create a stable and business-friendly environment for the private sector, attract private investment into labor-intensive industries, and support the development o f backward linkages from the natural resource industries.It should also ensure that a substantial share o f the rents from resource extraction are captured as public revenue and invested efficiently in human capital development, especially improvements to public education and health services. This inturnwill improve laborproductivity andraiserealwages. 18. The PEMFAR presents an evaluation and quantitative projectionsof fiscal space through fiscal 2011/12. The projections are intended to identify additional budgetary resources to fund priority public expenditures supporting economic growth and poverty reduction. They can also contribute to medium-termfiscal planning and inform debates over fiscal policy. Fiscal space can be generated from four main sources: increases in domestic revenue, increases in donor grants, sustainable public borrowing, and improvements in expenditure efficiency. The PEMFAR's fiscal space projections take the 2007/08 budget outturn as a base and build on it by incorporating macroeconomic projections and the estimated impact of policy changes (such as planned tax policy changes). ... Xlll 19. If economic growth accelerates over the medium term (12.3 percent on average during 2008-12), increased revenue will generate substantial fiscal space by 2011/12. Revenues are projected to grow by US$160 million, or 80 percent in nominal terms and 48 percent in real terms, compared with the 2007/08 outturn. But higher debt servicing will reduce the fiscal space. Domestic debt servicing may require up to US$16 million by 2011/12. External debt servicing requirementsare still subjectto muchuncertainty because of the need to mobilize additional debt relief over and above that given under the HIPC initiative. Butthe government expects that they will be limited to a maximum of 1percent of GDP during the period. There will be almost no scope for domestic or external borrowing in the mediumterm. 20. Revenue growth may be lower than expected if growth prospects are not realized. Given the uncertainty about the possible impact of the global economic slowdown on Liberia's growth, the PEMFAR's fiscal space projections consider an alternative scenario based on slower growth of real GDP. In this scenario, nonmining GDP growth is two percentage points lower than the baseline throughout the projection period, iron ore production is delayed by one year, and forestry extraction is slower than expected. Revenue growth then falls to 72 percent innominal terms and 36 percent inreal terms, comparedwith the 2008 outturn. 21. Core Poverty Reduction Strategy (PRS) spendingneeds are estimated at US$550 million for 2008/09 (comparedwith about US$132 million which is allocated in the budget) and this will fall slightly down to US$526 million by 2010/11. In addition to these core poverty-reducing spending requirements, many other important government administrative costs will have to be financed to maintain the state structure, such as the legislature, Ministry of State, and debt service. The PRS estimatesthat 45 percent of government revenues will be dedicated to non-PRS expenditures in 2008/09, falling to 35 percent of total revenues by 2010/11, or about US$llO million a year. In current prices, total requirements for expenditures on PRS and public administrative costs are estimatedat US$650 million a year over the PRS period (2008/09-2010/11). This amountsto about 70 percent of projectedGDP for 2008/09, falling to 52 percent by 2010/11. The resource shortfall amounts to US$418 million in 2008/09 and falls to US$326 million by 2010/11, assuming that creditors and donors provide additional debt relief or resources to enable the government to avoid any substantial increase in its external debt servicing burden, in line with the government's projections. 22. Despite the projected growth in revenues over the next five years, the requirements for public expenditures are greater than the resources available to fund them. Poverty-reducing expenditure requirements are estimated to be about 300 percent higher than the allocation to these expendituresinthe 2008/09 budget.Attracting more donor aid and improving expenditure efficiency will therefore be crucial if the government is to fbnd its core poverty-reducing spendingrequirements.Donors should begin to channelmore funding through national systems, either through direct budget support or pooled funds. Most donor financing is project based and implemented outside government agencies, which makes coordinating andmanagingpublic developmentprogramsa challenge. THEPUBLICFINANCIALMANAGEMENT SYSTEM 23. A recent assessment of Liberia's PFM system, using standardized Public Expenditure and Financial Accountability (PEFA) indicators, reveals several weaknesses, but shows that it is consistent with those of other post-conflict countries in xiv Africa, such as Sierra Leone. Of the 28 performance indicators and three donor practices indicators, Liberia scored Ds on 15 indicators, Cs on 10, Bs on 2, and an A on 1, while 3 received no score. PEFA indicators pertaining to budget implementation and budget accounting, auditing, and monitoring scored particularly poorly. Liberia performed better on budget credibility and orderliness and on participation in the budget process. GivenLiberia's outdated institutional system for PFM and the severe stress o f the civil war and economic collapse, the scores are relatively strong, reflecting improvements since the end of the civil war. Note, too, that the PEFA scorings look at practices over past years, and thus do not fully capture the strong ongoing reform progress. 24. The legaland regulatoryframework for PFMrequiresrevisionto clarify the roles and responsibilities o f different public agencies, establish internal controls, institutionalize the internal audit function, and formalize reporting requirements. A PFMAct, now before the legislature, is expected to rectify the deficiencies inexisting legislation. 25. Institutional responsibilities for preparing and implementing the budget are being reformed and strengthened. The past two years have seen significant improvements in coordination, thanks to the establishment of the Budget Committee, which includes representatives from the Ministry o f Finance, Bureau of the Budget, Ministry o f State, Ministry of Planning, Civil Service Agency, and General Service Agency. The committee provides oversight o f the budget process and advice to the cabinet on policy proposals and the strategic direction o f the budget. The recurrent and investment budgetshave beenmerged into a single budget, simplifyingthe budgetingprocess. Ministries and agencies are engaged early in the budget process through a budget circular setting out guidelines for budget preparation, sent out by the Bureau o f the Budget. In September 2008, legislation was approved for a merger o f the Bureau o f the Budget with the Ministry o f Finance, which is expected to reduce fragmentation in budget preparation and execution. The legislature's role in the budget review process needs further strengthening. Capacity building and greater communication between legislature and executive are required to ensure that the legislature can adequately oversee the budget process. 26. The national budget document should provide a comprehensiveoverview of the government's fiscal operations. It should include information on prior year budget outturns (presented in the same format as the budget proposal); data on both the debt stock and financial assets, with details for at least the beginning o f the budget year; considerationo f the implications o f new policy initiatives and estimates o f the budgetary impact o f major changes inpolicy or expenditure programs; and information on donor-fundedactivities inLiberia. 27. While the government would like to move toward a medium-term expenditure . framework, it is advised to gradually phase in the elements of such a framework. As a part o f the PRS, a costing exercise was undertaken to guide and ensure policy-based budgeting over the PRS period. The government i s also developing a medium-term fiscal framework, which will help to provide broad parameters for budgetary planning. It i s recommended that this be accompanied by strengthening the macroeconomic analysis capacity within the MoF; producing strategies for key sectors, formulating medium-term objectives, and producing related expenditure plans; linkingbudgeted funds and the results to be produced, defined by measurable indicators; and introducing a system to measure performance. 28. The Liberian government should set up a well functioning system for selecting and managing public investments. A first-level screening o f all project proposals should xv ensure that they meet minimum criteria o f consistency with the government's strategic goals. There shouldbe consideration o f how to finance the costs o f maintaining and operating assets created through public investment projects. And evaluation o f completed projects i s desirable to ensure that learning and feedback from projects create a positive dynamic for improvements over time. 29. The government's personnel databases should be rationalized.There are two databases o f public servants, one operated by the Civil Service Agency and the other by the Bureau of General Accounting in the MoF. Although these databases are regularly reconciled, the lack o f a common classification system makes this cumbersome. Supervisory controls o fthe payroll are weak. Although payroll audits have been conducted inrecent years and ghost workers removed from the payroll, systems for regular maintenance and monitoring o f the payroll need to be strengthened..The planned introduction o f a biometric identification system to verify all civil servants on the payroll and the development o f a Human Resource Management Information System will be milestones, as will the development ofthe IntegratedFinancial Management Information System. 30. A major effort is under way to reform public procurement.A Public Procurement and Concessions Act was enacted in September 2005 which provides for public sector contracts to be subject to meaningful competition. A h b l i c Procurement and Concessions Commission (PPPC) has been established and i s charged with regulating public procurement. Procurement regulations and a manual are currently being prepared with support from a consultancy firm, which will further help to strengthen procurement practices. Procurement unitsand committees have been set up inministriesand agencies, as required under the Act, but these structures lack the human and physical resources to operate effectively. Moreover, there is still some confusion over the roles which different entities should play in the procurement process, especially inthe separation o f functions between procurementunits and procurement committees. As a result, and despite a notable improvement in public procurement practices, they still fall short o f the standards stipulated inthe Act, and incorrect application o f the law'remains common. There is also a recognition that some o f the stipulations inthe Act are too stringent for the Liberian environment. The PPCC has initiated a full review o f the act, which i s expected to lead to amendments to the Public Procurement and ConcessionsAct to be submittedto the legislature inearly 2009. 31. The momentum from the passage of the Public Procurement and Concessions Act should be used. This will require a comprehensive set o f implementingregulations and manuals (currently under draft) and providing procuring entities with detailed instructions and increased capacity on the correct application o f the law. Efforts to strengthen the ability o f private sector operators to meet the requirements o f the Act, including by providing training, reducing the cost o f bidding, and improving the scope and quality o f public procurement reporting, are also needed. 32. A key component of budget execution is the cash management and commitment control system introduced under the GEMAP. This aims to ensure that expenditures remain within the resources envelope, prevent the accumulation o f expenditure arrears, and ensure compliance with the new public procurement legislation. While the current arrangements have re-imposed fiscal discipline and increased transparency and accountability in the budget process, they represent a temporary measure while longer-term financial institutions are re-established. The government may want to consider planning for a more streamlined budget execution process. In the MoF, it i s necessary to shift responsibility for operating the system away from GEMAPtechnical experts to those MoF officials who would xvi normally be in charge o f the payment process. It i s also necessary to rationalize the approval process further, as it involves many steps. 33. The accounting system does not meet internationallyaccepted standards. It is not based on a double-entry method o f book keeping and does not include sufficient reports to permit reporting based on generally accepted accounting principles. It also omits debt and contingent liabilities. As an interim measure, Sunsystems has been introduced, but i s being used primarily for reporting purposes. It will eventually be replaced by an Integrated Financial Management InformationSystem. Implementation should focus on the introduction o f a few core modulesthat can quickly be rolled out without sapping scarce resources. 34. The internal audit function needs to be further developed. There is a need for a systematic and disciplined approach to evaluating risk management, control, and governance processes. The PFM Act now being drafted would address this issue with a roadmap o f revised institutional arrangements for internal audit and would secure a border between internal and externalaudit functions. 35. The externaloversight functionis weak but is undergoingsignificantreform and capacity building. The General Auditing Commission is implementingextensive reform and capacity building and produced a first set o f audit reports, submittedto the legislature in late 2008. These efforts will need to be sustained and consolidated to ensure that annual audits are institutionalized. Further training and support to the legislaturewill be required to enable it to perform its oversight role o f reviewingfinancial statements and audit reports. 36. The debt management department of the M o F requires additional capacity and tools. As the HIPC process is unfolding, the government has begun to deal with an unsustainable debt burden (and 20 years of default). However, there i s also a need to institute regular debt-management practices within the MoF and ensure that the debt burden is analyzed during budget preparation. Legislation i s needed to clearly specify the debt- management responsibilities o f the MoF. Debt software tools should be introduced and the debt management staff should be trainedto analyze the debt and producehighquality reports. 37. A complete overview of Liberia's public finances should include informationon the activities of donors. A project database detailing all donor-funded projects needs to be developed. This would require the ministries to declare all direct-donor funding for government-implemented projects and donors to report all the funding they provide in Liberia. Inaddition, the central Project Financial Management Unit inthe MoF could be used for a greater volume o f donor-funded projects, ensuring standardization o f project funds management. PUBLIC EXPENDITURE ANALYSIS 38. Liberia's budgeted expenditures have risen rapidly since 2004/05 as the government's revenue base has recovered, increasingby 194 percent in nominal terms between2004/05 and 2007/08, from 13 percentto 25 percentof GDP.The positive effects o f increased public resources are evident, as the fimctions o f the state are gradually re- established after 14 years o f civil conflict. While government revenue remains insufficient to meet Liberia's reconstruction needs and the country remains heavily dependent on donor resources, the economic and fiscal recovery i s strengthening the government's coordinating and regulatory roles and has contributedto improvementsinbasic social service delivery. xvii 39. The governmentran small cash surpluses in 2005/06, 2006/07, and 2007/08 and expenditure has been unevenly paced across the fiscalyears. The rate o f expenditure has been low inthe beginningo f each fiscal year and rapid toward the end o f the year due to the constraints o f cash-based budgeting, which only allows large capital purchases to be made toward the end o f the year when sufficient cash balances have accumulated; under-projection o f revenues, which has resulted in the passing o f supplemental budgets in the second half of the fiscal years; and capacity constraints within the ministries and agencies to comply with new procurementrequirements and Cash Management Committee controls. So, there may be room to increase spending efficiency through better procurement planning, improved sequencing o f activities, and an eventual transition away from a cash-based budget. As a result, in 2005/06, 2006/07, and 2007/08, there were significant deviations between the original budget estimates and final expenditure, largely due to mid-year budget revisions and the passing o f supplemental budgets. These discrepancies decrease the predictability of the budget and may increase greater discretion inbudgetary allocations as mid-year revisions and supplemental budgets usually undergo less thorough public scrutiny than the annual budgetingexercise. 40. Major shifts have taken place in the allocation of expenditures classified by function .between 2004/05 and 2007/08, reflecting increased attention to the Interim Poverty Reduction Strategy and Poverty Reduction Strategy priorities. The main beneficiaries o f this shift are the social and community services sectors (of which education and health and social welfare constitute the largest sectors) whose share o f the budget rose from 18 percent in 2004/05 to 27 percent in the 2007/08 budget. Debt servicing has also receiveda larger share o fthe budget, rising from 2 to 6.5 percent o ftotal expenditure over the same period. In contrast, administration and security services saw their shares o f the budget fall, from 36 to 29 percent and from 34 to 14 percent, respectively. The share o f economic services declined marginally, from 8 to 7 percent. 41. There have also been positive shifts in the composition of expenditures by economic classification. The share o f wages and salaries in the budget declined from 55 percent in 2004/05 to 36 percent in the 2007/08 budget, while that o f goods and services increased from 21 to 33 percent. The reduction in the share o f the budget allocated to the wage bill was partly made possible by purging ghost workers from the payroll. But transfers and subsidies to public enterprises, commissions, local governments, and other recipients increased as a percentage o f total expenditure from 7.8 percent in 2004/05 to 15.9 percent in 2007/08. This i s due in part to the creation o f new commissions that receive monthly transfers and a rise in direct transfers to ministries and agencies for operating expenses. The budgethas also seen a reductionin investment expenditure, which fell from 14to 10 percent. While this i s partly because donors contribute most o f the public investment expenditure outside o f the national budget, the government's spending on capital investment will have to rise substantially inthe future ifit intends to contribute to rebuilding infrastructure. 42. Liberia's public expenditure patterns cannot be fully assessed without an appreciation of the impact of donor assistance. Total official development assistance was estimated at US$357 million for 2007,2 or more than twice the government's expenditure in the same period. Given the magnitude ofexternal assistance compared with public resources, it will be difficult for the government to improve allocative efficiency and ensure alignment with the PRS, unless improved and disaggregated aid data is collectedand analyzed. Paris Declaration Survey 2007 xviii 43. While Liberia's post-war progress to date is commendable, ensuring that additional resourcesare devotedto Poverty ReductionStrategy policy objectives will be important, as the resource envelope continues to grow in the medium term. As seen in many resource-rich countries, there are risks that increased public resources will fuel greater public consumption at the center, to the benefit o f the highest income groups. Increased budget monitoring and analysis will need to be performed to ensure that the budget i s contributing to stated objectives. Specifically, administrative and personnel expenditure should be aligned with the Civil Service Reform Strategy; resources should shift to investment expenditure, particularly in infrastructure, and to support operation and maintenance o f new physical investments; and greater attention should be given to economic sectors with large employment potential, such as agriculture, which is underfunded. These are some of the top priorities reflected inthe PRS, which focus on economic sectors under Pillar 2: economic revitalization, Pillar 3: governance and rule o f law, and Pillar 4: infrastructure and basic services. In addition, improvements in data availability, reporting, and disseminationwill be crucial for holdings elected officials accountable by monitoring that the use o f public financing is aligned with the PRS. SECTOREXPENDITURE REVIEWS Health Sector 44. Liberia's health status remains poor in regional comparison, and the country is unlikely to reach many Millennium Development Goals due to the deterioration in healthstatus during the 14 years of civil war. Average life expectancy at birth is estimated by WHO (2006) at 42 years, with 44 years for women and 39 years for men. Malaria is the leading cause o f death, accounting for more than 40 percent o f outpatient attendance and up ' to 18 percent o f inpatient deaths at health clinics. Diarrheal diseases are the second leading cause o f morbidity and mortality. However, there have been improvements in health indicators since the end o f the war, with a fall in total fertility rate from an estimated 6.6 birthsper woman in 1986 to 5.2 in2007 and declines inboth infant and childmortality rates. 45. Expenditure allocation in the health sector is driven by the National Health Policy of the Ministry of Health and Social Welfare for 2007-11, which focuses on equitable and affordable health services by providing a Basic Package of Health Service. This package standardises prevention and treatment services throughout the health system to ensure that all individuals have access to the same care. 46. Health expenditure has grown rapidly in the post-war period, and per capita health expenditure is estimated at approximately US$20 for 2006/07, high for Liberia's developmentstatus. Donor funding accounts for an estimated 64 percent o fhealth spending, which is well above the average for low-income countries and reflects Liberia's large reconstructionneeds inthe health sector. 47. In the medium term, the government must demonstrate a strong commitment to the health sector as part of its strategy to improve human capital for Liberia's development. In 2007/08, government expenditure on health accounted for 8.1 percent o f total expenditure, significantly lower than the 2015 Abuja target o f 15 percent o f total expenditure. As Liberiamoves toward implementing a medium term expenditure framework, public expenditure for the health sector should gradually increase. xix 48. There will be a continued need for capital investments in the health sector. Government expenditure i s almost entirely recurrent, while most capital costs are carried by donors. The government is encouraged to develop a capital investment plan as an integral part o f its Public Sector Investment Program in the context o f its medium term expenditure framework, outlining construction and reconstruction priorities, to ensure a more systematic approach to investment and to increase national ownership. 49. A regional analysis of government health expenditure shows that health services are not equitably distributed across regions. The ratio o f persons per physician in the richest region o f the country i s less than half that in the poorest region. Similarly, total expenditure on curative services i s more than three times as highinthe relatively richer South Central region, compared with the poorest South Eastern region. The government needs to ensure more distribution o f health resources to the more remote regions o f the country, both to improve Liberia's health status and to buildpublic trust in state services. 50. Donor commitments, already at a high level, should be maintained in the mediumterm and rise in real terms in line with GDP growth. Liberiawill depend heavily on external financing inthe medium term, until domestic resources are sufficient to maintain constant expenditure. Donors are encouraged to improve the harmonizationand predictability o f aid, for instance through pooled funding mechanisms, such as the recently established Health and Welfare PooledFund, supported by DFID. 51. Sustainability of the government's health program should be secured by introducing alternative sources of domestic funding, such as contributory health insurance and other prepayment or risk pooling mechanisms, rather than relying on out-of-pocket payments. The government should develop a roadmap for introducing such programs as soon as possible. 52. Although health financing will increasingly be channeled through national structures, the government should continue to use NGOs and private providersthrough formal contracting arrangements. This will require strengthened financial management, procurement, and monitoring and evaluation capabilities o f the ministry. The government should also develop a system o f performance- and output-based financing, where financing for nongovernmental organizations or contractors hinges on satisfactory performance. 53. The government has rightly made a basic healthcare package that targets vulnerable populations its first priority. But it should also expand and develop secondary and tertiary care parallel to primary healthcare, and strengthen the referral system. It should also emphasize human resource development, through hrther recruitment and training opportunitiesand improved remuneration. ' Education Sector 54. Enrollment figures suggest that access to education has improved significantly since the end of the civil war, with gross primary enrollment rates increasingfrom 40 percent in 2002 to 86 percent in 2007. The net primary enrollment rate is significantly lower at 37 percent, suggesting that a large number o f over-aged children and youth who missed out on years o f school during the war have returned to school and that many school- age children are not attending school. xx 55. Although government spending on education increased rapidly, from US$7 million in 2004/05 to U S 2 3 million in the 2007/08 budget, it remains low as a proportion of total expenditure at 11.8 percent in 2007/08. In this respect, Liberia i s far from reaching the Education for All Fast Track Initiative benchmark, which targets education expenditure at 20 percent o f total government expenditure. Government distribution o f education expenditure i s also skewed, with Liberia spending an unusually small share o f its budget on primary education and a very high share on tertiary education and administrative and support services. In 2006/07, o f the total education budget, 14.5 percent was allocated to primary education, 11.4 percent to secondary education, 31.6 percent to tertiary education, 7.1 percent to technical/vocational training, and 35.4 percent to administrative and support services. Moreover, private households spent US$27 million on education in 2006/07, which places a considerable burden on poor households, and represents more than twice the government's expenditure on education. This suggests that families place a high priority on educating their children. 56. The rapid expansion in enrollment at the primary level means that many students in the coming years will require places in junior secondary education. The junior secondary education system does not have the space to support the large influx that will be entering the system. Nor can it sustain the numbers when the primary system stabilizes and achieves universal primary completion. The percentage share of expenditure for secondary education has increased from 9.5 percent in 2004/05 to 11.4 percent in 2006/07. To adequately supply the resources for this sub-sector, the allocation will need to be increased. 57. The largest source of education expenditure comes from external aid. These aid flows are very difficult to track, and only tentative data is available. It is estimated that US$40 million was spent by donors on education (with a focus on primary education) in 2007, which accounts for more than half o f total education spending. Better monitoring and targeting o f these resources may improve the efficiency o f education expenditure and outcomes. 58. There are clear gender, regional and income disparities in access to education. Girls, rural children, and children from less wealthy households are less likely to attend school. There are also important regional variations in access to and quality o f schooling. In general, access to primary school is fairly equitable, but disparities increase in secondary and tertiary education. At the primary school level, the parity index for female to male students is 0.96, and for rural to urban students i s 0.89. But these disparities increase at the secondary level to 0.78 and 0.53, respectively. Gross enrollment rates also show highregional variation by county, with a rate o f 29 percent inGrand Bassa, comparedto 129 percent in Grand Cape Mount. More effort should be made to ensure that female and rural children have access to secondary and tertiary education, and resources should be distributed more equitably across the country to reduce regional variations ineducation access. 59. Budget allocations to education should be morestrategic. Within the context o fthe multi-year budget framework, the government should ensure that more public resources go to primary andjunior secondary education, where the social returns are likely to be higher than those in upper secondary and tertiary education. In addition, when allocations are made to investments, the government should indicate the long-term recurrent implications, like those o f the regional training institutes. Some allocations toward basic education services that go beyond primary education should be equity or quality based. As the majority o f external xxi hnding supports primary education, increased capital spending is needed at the secondary education level to provide more schools and learning materials. 60. Greater gains in resource efficiency within the sector could be achieved by lowering the drop-out and grade-repetition rates, especially in grade 1. Interventions should focus on improving overall quality, by improving teachers' qualifications and increasing accessibility to teaching and learning materials. In addition, the government should reduce the pupil to teacher ratio by addressing the very large inequities in teacher distribution. This could be done by providingteacher housing and salary incentives and doing more local recruitment. 61. There is also the need to complete the clean up of the teacher payroll as soon as possible. The disruption and insecurity caused by the war resulted inmany teachers leaving their posts, large numbers of replacement teachers (47 percent in Grand Cape Mount), and concern about the possibility o f ghost teachers. Exercises have been conducted to clean the payroll, but discrepancies remain between school data, human resources data, and Education Management Information System data, suggesting that the payroll i s not completely clean. Cleaning upthe payroll should be completed as soon as possible. 62. Further efficiency gains could come from decentralizing service delivery. An initial step toward decentralization could be enhancing the capacity o f county education offices and improving school management for service delivery. Since the free and compulsory education policy has been implemented, schools no longer benefit from revenues to cover the school operational cost for minor maintenance and small material. School management committees exist but are not fully functional. In many Sub-Saharan African countries, direct flow o f funds to schools to support service delivery has improved accountability and quality. This system may be worth piloting in select counties. One implication would be the need to strengthen local financial management systems to ensure accountability and transparency. 63. To expand resourcesto the sector, the government could promote public-private partnershipand cost recovery at the tertiary level, in keepingwith making expenditure more progressive by spending more at the primary and lower secondary levels. Different models o f public-private partnership and cost recovery systems operating in other countries could be examined for tertiary education, where the unit costs are far higher than those for primary and secondary education. 64. An important element of strategic planning for education is to ensure that its outcomes are aligned to the labor-market needs of the country. The public-private partnership should go beyond investment in school plants or tuition to consultations on curriculum development, life skills training, and other innovative ways o f involving the private sector inmanagingeducation and training services. xxii ACTION PLANFORPFMREFORM 65. The action plan draws on the PEFA results and PEMFAR findings, reflecting the chapter conclusions and recommendations. It buildsheavily on ongoing reform efforts by the government and incorporates and complements the reforms the government i s committed to inthe context ofthe HIPC Initiative and Poverty Reduction and Growth Facility. Itprovides a broader framework for PFM reform, with the aim o f defining a road map as GEMAP phases out. It identifies a number o f quick wins, which are shorter term actions that can be completed within a year and rapidIy contribute to the reform momentum. The medium term actions will require 1-3 years to complete. The action plan is intended as a foundation for further government-partner dialogue about the government's PFMreform agenda. The action plan can serve as a basis from which the government, in collaboration with partners, can develop a narrower, more focused, and strategic PFM reform plan that does not overstretch government capacity. This plan should highlight those activities that are already receiving donor fundingand provide a framework to which other donors can align their support. xxiii Table 1:Action plan Improve the PPC Iegal and in5titutional framework / Develop institutional proc xxiv xxv xxvi xxvii xxviii 1. MACROECONOMICPERFORMANCE 1.1 Liberia has embarked on economic recovery and is rebuilding its institutional capacity following a protracted conflict lasting from 1989-2003. The 14-year civil war had a paralyzingeffect on the country's economic and social service provision, reducingboth real GDP and public expenditure to a third o f their pre-war levels. Today, per-capita GDP i s estimated at US$195,3 and an estimated 63.8 percent o f the population lives below the national poverty line, with 47.9 percent o f the population living in extreme p ~ v e r t yOnly an . ~ estimated 17 percent o f the labor force is employed in formal wage labor. Among paid employees, salaries are low, with the lowest paid civil servants earning approximately US$70 a month. Social indicators are poor, although a recent Core Welfare Indicator Questionnaire (2007) shows some improvements since the end ofthe civil conflict. 1.2 The educational system was severely damaged during the war, resulting in a generationthat grew up with little access to formal education. Today, net primary school enrollment rates remain at 37.3 percent, while the much higher gross enrollment rate (86.3 percent) suggests that older children that missed out on education during the war are returning to school. The youth literacy rate (73 percent) is substantially higher than the overall adult literacy rate (55 percent), and the gender discrepancy in literacy rates is lower among youth than adults. The health system depends heavily on international development partners. Health indicators are poor, with infant and under-five mortality rates o f 72/1,000 and 111/1,000, respectively. Child malnutrition rates are high. An estimated 19 percent o f children under age five are undernourished, 7 percent are wasted, and 39 percent are ~tunted.~ Access to health services i s uneven across the country, with higher access in urban areas. Effective public financial management (PFM) institutions and increased capacity are urgently neededto improve social and economic service delivery to Liberia's needy population. 1.3 During 2003-2005, Liberia was governed by the National Transitional Governmentof Liberia, which mismanaged public finances and engaged inwidespread corruption. This situation prompted the establishment o f the Governance and Economic Management Assistance Program (GEMAP). Set up with the support o fthe donor community in September 2005, GEMAPwas designed to curb the most egregious abuses of public fimds and to ensure fiscal discipline. It also initiated essential PFM reforms in such key areas as procurement. 1.4 Since the election of President Johnson-Sirleaf and the formation of a new government in January 2006, the government has taken actions to improve public financial management. Government revenues have increased markedly and expenditure controls have been strengthened through the establishment o f the cash management committee (CMCo) and the GEMAP interim commitment control system. GEMAP has also helped fill some o f the capacity gaps in the ministries and supported the establishment o f financial and management procedures and systems. An Interim Poverty Reduction Strategy Paper was prepared in April 2007 and a fdl Poverty Reduction Strategy Paper (PRSP), completed inApril 2008, establishes the framework for poverty reduction and determines the budgetary priorities for the medium term. With the completion o f the PRSP and arrears clearance from the World Bank, AfDB, and IMF, Liberia became eligible for debt relief Nominal GDP. IMF, Macroeconomic Framework 2007. Core Welfare Indicator Questionnaire 2007. Liberia Demographic and Health Survey 2007. - 1 - under the Enhanced Heavily Indebted Poor Country (HIPC) initiative, reaching the HIPC decision point inMarch 2008. A. MACROECONOMIC AND FISCAL PERFORMANCE 1.5 This section provides an overview of the macroeconomic and fiscal situation in Liberia, discusses the medium and long term prospects for economic recovery, and presents quantitative projections and an evaluation of fiscal space over the next four years (through fiscal year 2011/12). The projections are intended to identify additional budgetary resources to fund priority public expenditures supporting economic growth and poverty reduction. Fiscal space projections can aid medium-term fiscal planning and inform debates over fiscal policy issues. 1.6 Unfortunately, fiscal space inLiberia dependson variablesthat are very difficult to forecast over the medium term. Problems stem from the volatility and lack o f relevant data (especially on donor aid). Moreover, the impact o f the current global slowdown on Liberia's medium- to long-term growth prospects i s hard to assess. So, the fiscal space projections are tentative. Nevertheless, three important conclusions emerge from the fiscal space analysis: revenues will continue to make the most important contribution to fiscal space; revenues should grow steadily over the medium term; and economic growth i s the most important factor drivingrevenue growth. Despite anticipated strong revenue growth, the government's spending requirements will substantially exceed revenues for core poverty- reducing programs. So, attracting more donor grant aid and ensuring expenditure efficiency will be essential ifthe government is to achieve its fiscal objectives. B. THELEGACYCIVIL WAR OF 1.7 Although the Liberian economy grew strongly in the 1960s and 1970s, driven by the production and export of iron ore and rubber, the benefits of growth were distributed inequitably.By 1980 GDP per capita had increased to middle-income country levels,6 but the economy first slumped and then collapsed over the next two decades (figure 1.1). The collapse resulted from deteriorating terms o f trade, economic mismanagement and the civil war that afflicted the country fiom 1989-1996 and again from 2001-2003. Inthe first phase o f the civil war, the economy contracted by 90 percent in real terms. When peace was restored in 1996, the economy began to recover rapidly, driven by a strong expansion o f timber exports. Output fell sharply again after fighting resumed in 2001, leading the United Nations to impose sanctions on Liberia's timber and diamond exports. Although the economy has resumed its recovery since 2004, the level o f real output in2007 i s still only a third of its peak in 1979. In 1980 GDP per capita reachedUS$890, in 1992 prices (IMF, 2006: 14). - 2 - Figure 1.1: Index ofReal GDP, 1966-2007 (1966 = 100) ......._...... .......... . . ...... ... I 8 O1 . ... .. .... .... . ..........-... .. ............... ...... " "_ ...j I Source: IMF, Article IV Consultation, 2005 (22); IMF, Article IV Consultation, 2006. 1.8 Most of the economy's physical capital-infrastructure and productive capacity-was either destroyed or severely damaged during the civil war. The rest deteriorated from lack o f maintenance. Iron ore production ceased and public provision o f such utilities as electricity and water is extremely limited. The road network is in serious disrepair. 1.9 The government remained in default until 2008 on its external debts from the mid-1980s. Very large fiscal deficits were incurred in the 1980s, but since the late 1990s fiscal deficits on a cash basis were generally small. The government had no access to new foreign loans and there was little scope for selling government debt on a market basis to the very narrow domestic financial sector. The cash fiscal deficits incurred were mostly financed by the Central Bank, although there was little scope for doing this because o f the dollarized economy. Butthe government was frequently in arrears for salaries and goods and services. c. MACROECONOMICDEVELOPMENTS 2003 SINCE 1.10 The second phase of the civil war ended with the Comprehensive Peace Agreement in 2003. The peace agreement brought into power the National Transitional Government o f Liberia, which implemented an economic recovery program called the Results Focused Transitional Framework. After the slump in 2003, output began to recover. Demand rebounded more quickly, buoyed by spending for peace-keeping operations.' Still, public finances remained very fragile-budget revenues fell to just over US$50 million a 'United Nations peace-keeping operations cost approximately US$700 million a year, which is larger than Liberia's GDP. Although much o f this expenditure does not enter Liberia, even if a relatively small share were to enter the country to fund local purchases o f goods and services, the impact on aggregate demand would be substantial. - 3 - year in 2002103 and 2003104. Fiscal discipline was weak and there was evidence o f widespread corruption. 1.11 Following the presidential election in January 2006, the National Transitional Government of Liberia was replaced by the government of President Johnson-Sirleaf. The new government agreed to an ambitious program of reforms, supported by a Staff- Monitored Program and the IMF, beginning in February 2006. Key objectives were recovering the economy, rebuilding public institutions, and restoring credible financial management in the public sector. Performance under the Staff-Monitored Program was satisfactory, and Liberia entered a Poverty Reduction and Growth Facility program in early 2008. In March 2008 Liberia reached the HIPC decision point, due to strongly performing under the Staff-Monitored Program and successfully clearing its arrears with its multilateral creditors. 1.12 Liberia's economy has endured the effects of greater global economic volatility in 2008. GDP and revenue growth in 2008 remained fairly robust at 7.1 percent and 37.1 percent, respectively, while inflation increased to 14 percent in 2008 from 11.7 percent in 2007. Year-on-year inflation spiked at 27 percent in August 2008 as a result o f the global commodity price increase. Given Liberia's high import dependence and large capital accounts deficit, the effects o f the commodity price crisis rapidly passed through to the local economy. The medium- to longer-term effects o fthe global economic slowdown are still hard to predict. Its impact, however, is expected to be substantial and has begun to be felt, with falling demand and prices for Liberia's primary exports, falling remittance inflows, waning interest from foreign investors and delays to large investment projects, and a possible fall-off inofficial development assistance inflows. The growth rate has beenrevised downwards for 2009 and 2010, but may be even lower. Yet Liberia's growth rate is expected to remain fairly high in comparative regional perspective, as some production in the mining and forestry sectors is expected to restart. Table 1.1: Key Economic Indicators, 2003-2008 1 2003 1 2004 I 2005 I 2006 1 2007 2008* Real GDP growth (YO) -3 I.3 2.6 5.3 7.8 9.5 7.1 Nominal GDP (US$ millions) 410 360 530 613 735 870 Inflation 5.0 7.5 7.0 8.9 11.7 14.0 Exports (US$ millions) 109 101 1 I O 158 208 260 Imports (US$ millions) 140 268 294 401 499 760 External official grants (US$ 30 I89 387 303 389 401 millions) *Exchangerate LD/US$ 50.5 54.5 56.5 59.5 63.5 Projection. Source: IMF, Article IC Consultations, 2006,2007, and 2008. 1.13 Strong economic growth between 2003104 and 2007108 enhanced government revenue. Other factors to revenue growth include: more strongly enforcing customs administration (including strengthening pre-shipment inspection o f imports since 2006), strengthening the Large Taxpayer Unit, eliminating the settlement o f tax liabilities through noncash payments,8 and reducing tax and duty incentives. Revenue growth in 2006107 and 2007108 was stronger than initially budgeted. But the effects o f greater economic instability are affecting revenues, and the government may not meet its 2008109 revenue projections. Inthe past some companies had settled their tax obligations by undertaking public works. - 4 - Table 1.2: FiscalIndicators,2002/03-2007/08 (US$ million) Note: Expenditure and balance are on cash basis. Source: IMF, Article IV Consultations, 2006107; Ministryof Finance, Annual Fiscal Report, 2005/06, 2006107, and 2007108. 1.14 Expenditure controls were strengthened by setting up the cash management committee (CMCo) and an interim commitment control system. Under the GEMAP, all local purchase orders and expenditure vouchers from ministries and spending agencies must be submittedto the CMCo, which verifies that each proposed expenditure accords with the spending agency's budget and monthly expenditure allocation and that correct procedures (for procurement and so on) have been followed. The CMCo will not authorize expenditures unless there are sufficient funds in government bank accounts inthe Central Bank of Liberia, thereby ensuring that the government does not run overdrafts with the Central Bank. Since payments cannot be made without CMCo authorization, in principle, this system should prevent any new debt from being incurred (although wages and some other payments are not covered). All government bank accounts have been centralized into two accounts (one denominated in Liberian dollars, the other in U S dollars) in the Central Bank, and checks cannot be drawn on these accounts without the authority o f the committee. Initially, spending agencies struggled to meet the requirements o f the cash management system, especially the procurement requirements, which slowed expenditure. Nevertheless, government spending quadrupled between 2003104 and 2007108 because o f the rapid growth in budgetary resources. Revenue and expenditure growth has outpaced GDP growth, with expenditure increasing significantly inproportion to GDP from 11.6 percent in2003104 to 25.3 percent in 2007108. 1.15 Liberia's public and public-guaranteed external debt is unsustainable. A debt sustainability analysis carried out in 2007 estimated that the nominal value of the debt was - 5 - US$4.8 billion (as o f June 30, 2007).9 This was more than seven times the country's GDP.'' The net present value o f the debt stock was estimated at US$3.3 billion, or approximately 1,700 percent o f exports o f goods and services in 2007." Liberia has begun to resolve its unsustainable debt burden; it cleared arrears to the World Bank, African Development Bank, and IMF in 2007-2008 and subsequently became eligible for debt relief under the Enhanced HIPC initiative. This should reduce the net present value o f Liberia's debt by a common reduction factor o f 91 percent. 1.16 Only limited macroeconomic and fiscal management tools are available to Liberian policymakers.The government is constrained to limit its expenditure to what can be mobilized from domestic revenues. Although the country receives aid o f about 45 percent o f GDP, very little o f this i s channelled through the budget. The domestic banking system is quite shallow-commercial bank deposits amounted to the equivalent o f only 19 percent o f GDP in 2007-and it is unlikely that government debt would be attractive to domestic financial institutions, as the government has a heavy domestic debt burden. Any significant borrowing from the Central Bank would quickly he1growth o f the money supply, given that reserve money comprises only 12 percent o f GDP and that the Central Bank lacks instruments to mop up liquidity. So, the government must avoid domestic borrowing for the budget ifthe Central Bank o fLiberia is to control the growth o f money supply. Exchange rate policy is circumscribed because the Central Bank has very little capacity to intervene in the foreign exchange market. Its capacity to sell foreign exchange is limited because its foreign reserves are small (the Central Bank's net foreign exchange reserves were only US$36 million inDecember 2007). D. MEDIUM-TO LONG- TERMPROSPECTSFORTHE ECONOMY 1.17 The economy has strong potential for growth in the medium to long term, althoughin the short run growth may be constrained by the global economic slowdown. Economic growth commonly recovers in post-conflict countries, even when there i s no significant increase in investment. Growth is driven by a recovery o f capacity use and improvements in factor productivity.12 But there may be delays to this growth acceleration due to lags inlarge-scale investment programs and a fall in demand for exports, precipitated by the economic crisis. Growth is projected to accelerate over the mediumterm (9.2 percent on average during2008-2012). This estimate is basedon only partially reconciled external debt data. With donor support, the authorities contracted an external financial advisor to reconcile commercial debt; the debt sustainable analysis is based on the advisor's initial estimates o f commercial debt stock (estimated at US$1.64 billion). loIDA and IMF2008. This is calculated after the application o ftraditional debt reliefmechanisms. l2Chen, Loayza, and Reynal-Querol2007. - 6 - Table 1.3: GDP Growth and Inflation Projections, 2007-2012 (percent) Source: IMF, Article IV Consultation, 2008. 1.18 The projectionsshown in table 1.3 are subject to uncertainty, as the effects of the global financial slowdown are hard to predict. Thus, it is prudent to include an alternative scenario based on slower growth o f real GDP. Inthis scenario, real GDP growth, excluding the mining sector, is two percentage points lower than the baseline throughout the projection period. The slower growth scenario i s also shown intable 1.3. 1.19 There are several potential sources of growth in the Liberian economy. In the short to medium term, an expansion o f demand should be sufficient to boost output by bringingback factors o fproductionthat are unemployed or underused.Although it is unlikely that there i s much underused real capital, there is widespread underemployment and ~nemployment.'~ This labor could be used in labor-intensive food production, construction, and services as domestic demand expands. Aggregate demand would rise thanks to private consumption fueled by remittances and donor funding. But one constraint on a demand- drivenrecovery could bethe availability of foreign exchange to funda widening trade deficit. Some ofthe increase indemand will spill over into imports and depreciate the exchange rate. 1.20 Eventually, capacity constraints will start to bind on a recovery driven only by domestic demand. Over the longer term, an expansion ofthe economy's supply capacity will be necessary to sustain economic growth, which requires investment to rebuild the economy's productive capital and physical infrastructure. Strengthening human capital, through improvements in health and education, i s also crucial to expanding economic capacity, although the payoff will take some time. Where then should this investment flow? What sectors are most attractive to new investment and what implications will investing in these sectors have on the wider economy? The four sectors with the most potential for growth over the medium to long term are timber, mining, agriculture, and services. 1.21 Timber was at the forefront of the economic recovery in the second half of the 1990s. By 2002, value added in the sector was US$86 million. Most o f the timber was exported. Timber production collapsed to only US$2 million in 2005 because o f the UN imposed sanctions on Liberia's timber exports, but sanctions were lifted in 2006. The Forestry Development Authority projects that timber production and exports could climb to 750,000 cubic meters by 2012, which would bring timber production back to near the level reached in 2002 (809,000 cubic meters). But developments in this industry are subject to l3Preliminary data from ILO, based on the 2007 CWIQ, estimates the labor force at 1.1 million and an unemployment rate o f about 5.5 percent. Among the active labor force, only an estimated 170,000 employees are working inthe formal sector. - 7 - some uncertainty due to the economic crisis and falling investor interest, as well as ongoing discussions about carbon credit schemes and alternative methods o f ensuringrevenue growth while preservinga greater share o f Liberia's forests. A recovery o f timber production would contribute to job creation, but logging must be carried out on a sustainable basis if the industry is to have a long-term future. Growth in agriculture is expected to remain fairly steady at 5-7 percent, but as it is about 40 percent o f GDP and employs a high share o f the labor force, its impact on poverty reduction could be significant. Service sector growth will beboostedby further growth indomestic demand. 1.22 Mining, primarily of iron ore, was the largest sector of the economy before the civil war. Duringthe war, production facilities and rail infrastructure needed to transport the iron ore from the mines to port were destroyed and production ceased. But the long-term prospects for mining are promising, and Liberia has deposits of gold and diamonds. The government has signed a Mineral Development Agreement with Arcelor Mittal for a US$l.5 billion investment over 20 years to rehabilitate an iron ore mine, with its related infrastructure (a railroad and port facilities), inNimba County. This mine is projectedto eventually produce about 18 million metric tons o f iron ore a year.I4 Production i s expected to start in2010 after being postponed due to the economic slowdown. The government has also reached an agreement with China Union for developing a second iron ore mine in Bong County, with projected investment o f US$2.6 billion. The country i s tendering a third iron-ore mine in western Liberia, and several other areas are under exploration. Diamond mining has been restricted to small-scale alluvial mining; large-scale diamond mining may not be commercially viable. In addition, a gold mine at Bea Mountain may be developed over the mediumterm. 1.23 Service sector growth will be boosted by further growth in domestic demand. This includes public services, construction, private consumption, and the services inputs requiredto support the timber and miningindustries(such as transport). 1.24 Growth in extractive industriesand commercial agriculture may be significantly affected by falling global demand for primary products, owing to the global economic slowdown. After the record high prices for commodities in 2007 and early 2008, prices for Liberia's main export commodities have fallen sharply. In addition to the direct effects o f falling export earnings, this may also cause a decreasein investments inthese sectors. E. LONG-TERMSTRUCTURAL CHANGES 1.25 If all of the potential mining investments are eventually undertaken, they will give a huge boost to export earnings, GDP, and government revenue in the long term. The value added from the Arcelor Mittal iron ore mines plus the two other potential iron ore mines could exceed the current level of GDP. Export earnings would increase severalfold. Even if some o f the export earnings from mining are repatriated, there will still be a large increase in domestic demand and a consequent appreciation o f the real exchange rate. As a result, the structure o f the economy outside o f the mining sector could shift toward the production o f non-traded goods-such as construction and services for the domestic economy. l4Maximum production o f 18 million metric tonnes a year from Mittal's iron ore mine is projectedto be achieved after sevenyears o f operation. Valued at today's world iron ore prices, this level o f production would be worth almost U S $ l billion a year. But world iron ore prices have begun falling and are expected to drop further due to the global economic slowdown. - 8 - 1.26 The boom in the mining sector is unlikely to precipitate the "Dutch Disease" in Liberia, at least not in the short to medium term. First, the level o f employment in the miningsector is likely to remain a fairly small proportion o fthe labor market. Consequently, a tightening o f the labor market with the upward pressure on real wages i s unlikely in the short to medium term. Second, although the exchange rate is likely to appreciate, a sharp appreciation o f the exchange rate is not expected because the economy is significantly dollarized (US dollars account for about 90 percent o f the money supply). Indeed, the increased income from the mining boom could spur domestic demand for manufacturing and agricultural tradeables, thereby promoting a greater supply response for exports in the medium to long term. Third, since the Liberian economy is relatively open (trade is almost 100 percent of GDP) and import dependent, a significant proportion o f the foreign exchange from the boom would likely be spent on imports. This would lessen the pressure on money supply and domestically producedgoods. Finally, the government's proactive stanceon fiscal management o f the resources from the potential boom sector, its emphasis on the more equitable distribution of the rents from the mining sector, and its strategy to diversify the economy (including the non-tradable sectors) would help prevent the "Dutch Disease" effect. 1.27 The natural resource booms in Liberia in the 1960s and 1970s did not translate into equitable growth and poverty reduction. To avoid a repeat o f this, the government should provide a stable and business-friendly environment for the private sector. Attracting private investment into the non-tradable sectors and supporting the development of backward links from the natural resource industries will be critical for job creation. It is important to maximize the government rents earned from mining in order to maximize public investment in developing human resources, especially improving public education and health services. Thiswill leadto improvementsinlaborproductivity and help raise real wages. F. RESOURCEMOBLLIZATIONFISCALSPACE AND Framework 1.28 Fiscalspace refers to the potentialfor mobilizingbudgetary resources for public spending `on policy priorities (such as promoting economic growth and reducing poverty) without jeopardizing macroeconomic stability or fiscal sustainability.The term was first used by some Latin American and European governments that argued that higher fiscal deficits were needed for expanding public investment in infrastructure to stimulate g1-0wth.l~has since beenused ina wide range of countries. Projections of fiscal space can It also be compared with estimates o f expenditure requirements to meet policy priorities, evaluate possible budget shortfalls, and guide fiscal policy planning. 1.29 Fiscalspace can be generated from four main sources, sometimes referred to as the fiscal diamond. These are: increases in domestic revenue, increases in donor grants, sustainable public borrowing, and improvements in expenditure efficiency. Additional domestic revenue can be mobilized through tax policy reforms, tax administration reforms, and the natural income elasticity o f the tax system. Evaluation o f borrowing includes domestic borrowing, seignorage, and external borrowing. 1.30 Projections for fiscal space should be realistic and consistent with the government's economic and social development objectives. For example, projections for domestic revenue mobilization should not incorporate tax rate increases, which would undermine the objectives for economic growth by reducing incentives for saving or l5Heller 2005. - 9 - investment or distorting resource allocation. Fiscal space projections can incorporate feasible policy reforms, which create fiscal space. Thus, the projection shows what is possible if good policies are adopted, rather than simply projecting resources based on an unchanged policy environment. Figure 1.2: The fiscal diamond Domestic revenues Donor grants Borrowing Improvementsin expenditureefficiency 1.31 The fiscal space projections for Liberia take the 2007/08 budget outturn as the base. Projections for the four components o f the fiscal diamond over the next four fiscal years (up to 2011/12) are built on this base. Wherever possible, existingmacroeconomic and fiscal projections are incorporated; examples include the IMF's macroeconomic projections and the projections related to external debt in the Debt Sustainability Analysis. Although fiscal space projections are often formulated in terms o f percentage o f GDP, nominal U S dollars are used because GDP itself is projected to grow rapidly over the forecast period; hence much of the expansion o f fiscal space occurring over the medium term would not be apparent if formulated interms o f percentage o f GDP. The nominal U S dollar projections are then converted into real US dollar projections by applying a projection of the government expenditure deflator. Tax and Non-taxRevenues 1.32 Domestic revenues provide the bulk of budget resources in Liberia. They come primarily from taxes on international trade, business and employment income, and to much less extent, domestic sales and excise taxes. Revenues rely too heavily on import taxes, which contributed over 40 percent o f total revenue in2007/08. Income taxes have also risen rapidly and contributed 27 percent o f total revenue in2007108. Consumption taxes contribute less to revenue than would be optimal because o f the narrow base and low rate (7 percent) of the general sales tax. Petroleumtaxes, which comprise customs duties and a sales tax, contribute relatively little to total revenue. The combined tax rate i s low and tax exemptions are ubiquitous. Non-tax revenues have been relatively unimportant in recent years but will become more important when the timber (stumpage fees) and mining (royalties) industries recover. 1.33 The government is implementingmany of the reformsrecommendedby the IMF on tax policy and tax administration.The IMF technical assistance mission in 2007 and 2008 concludedthat although Liberia has established a sound tax system and i s making good reform progress, revenue administration remains weak and the tax system still suffers some - 10- deficiencies. The deficiencies include the small share o f indirect taxes in total revenue and the widespread granting o f tax and duty exemptions.'6 Table 1.4 lists some o f the ongoing and plannedtax policy reforms recommended bythe IMFmissions and the estimated revenue impact o fthese reforms. 1.34 Alongside tax policy reforms, programs to strengthen tax and customs administration contin~e.'~ The domestic tax administration i s being restructured into departments which will focus on the type o f taxpayer. Amendments are being proposed for the Liberian Revenue Code, and procedures are being developed to facilitate registration o f taxpayers, filing o f tax returns, and payment o f taxes. Risk-based systems for verifying taxpayers' compliance with their obligations are being developed, and an appeals process i s beingset up. A stronger tax administration should generatemore revenue over time, but gains will be limited inthe short term. Table 1.4: Recommended Reforms to Tax Policy Note: The reforms have been recommended by IMF technical assistance missions, which also estimated the impact on revenue. Source: Muzondo et al. 2006; Muzondo et al. 2007; Ministry o f Finance, Draft Liberia Revenue Code 2008. 1.35 As noted above, revenues have recovered very strongly since 2003/04. The 256 percent growth in revenue over four years outstripped growth o f nominal GDP. As a result, the revenue to GDP ratio has risen from 13 percent of GDP in 2002/03 to 25 percent in 2007/08. The recovery has been led by trade taxes, which have increased by 235 percent in this period due to the rapid growth o f imports and stronger enforcement by customs (curbing the abuse o f fuel duty exemptions and improving pre-shipment inspections). Income taxes have quadrupled in the last four years (albeit from a very low base), with the recovery o f formal sector activity. In contrast, non-tax revenues have stagnated, mainly because o f the loss o f stumpage fees following the collapse o ftimber exports. l6Kloeden et al. 2007; Muzondo et al. 2006. 17Story et al. 2006; Kloeden et al. 2007. - 11 - 1.36 The tax revenue to GDP ratio o f 25 percent in 2007/08 is highcompared to the 16 percent average for low-income Sub-Saharan economies. This i s surprisinggiven that tax rates are not especially high, exemptions are common, and tax and customs administration are uneven. Liberia's tax effort appearsmuch stronger than that o f other Sub-SaharanAfrican countries with respect to trade taxes. Liberia collected 11.2 percent of GDP in trade taxes in 2007/08, which i s above the Sub-SaharanAfrican average. Liberia's tax effort also appearsto be strong in income taxes, which in 2007/08 were 7 percent o f GDP. One likely explanation for the seemingly strong tax effort is an underestimationo f nominal GDP. G. REVENUEPROJECTIONS 1.37 The methodology for making medium-term revenue projections uses the 2007/08 revenue outturn as a base. Projections are made for each o f the main individual revenue handles, shown in table 1.5. For each revenue handle, the projection incorporates the projected real growth in the revenue base, the projected change in price for ad valorem revenue handles, the estimated impact of tax policy changes where relevant, income elasticity where relevant, and potentialtax administration gains. 1.38 The real growth of GDP (excluding mining") is used as a proxy for growth in the revenue base for all income taxes, taxes on immovable property, domestic taxes on goods and services, entrepreneur income, transfers from enterprises and public institutions, fees and charges, and incidental sales o f goods and services (see table 1.3).19 1.39 The price change for all income taxes and for the entrepreneur and property income of government is proxied by the change in the GDPdeflator. The price change for domestic taxes on goods and services, tax on immovable property, administrative services charges, and incidental sales o f goods and services is proxied by the consumer price inflation. N o price change is included into projections for fines, stumpage fees, and relatedcharges. 1.40 For international trade taxes, the growth in the base is projected in nominal terms and proxied by the nominal growth of imports or exports. Nominal growth in the taxable import base i s proxied by nominal growth in GDP, excluding the mining sector. Export duties levied on timber exports are taken from the financial forecast o f the Forestry Development Authority. 1.41 An income elasticity of 2 percent a year is incorporated into the projections for all income taxes. These taxes should be naturally income elastic and have displayed very strong buoyancy over the last four years. A gain o f 1 percent a year is applied to all tax handles from 2008/09 onward to reflect the revenue gains from the planned tax and customs administration reforms. Maritime revenue, which has been relatively stable over the last five years, i s projected at a constant US$12 million per year, the average o fthe last five years. 1.42 Projections for non-tax revenue from timber, which comprises stumpage fees and related charges, are taken from the financial forecast of the Forestry Development Authority. 1.43 The following tax policy reforms are incorporated into the revenue projections: Miningis excluded because investmentswill be long term. Beyond that, corporate income taxes will be low at first because deductions and mininginputsare not taxable. 19Real growth in GDP for the fiscal year is calculated as the average of the two real growth rates for the relevant calendar years shown intable 1.3. - 12- 0 The top marginal rate o f income tax is reduced from 35 percent to 30 percent in 2009. The reduction i s not projected to have any impact on labor supply or pre- tax corporate profits. 0 The threshold for payment o f personal income tax i s raised in 2008/09, at a cost o f US$0.2 million inlost revenue. 0 The general sales tax rate is increased from 7 percent to 8 percent in 2008/09, to 9 percent in 2009/10, and to 10 percent in 2010/11. Demand for consumer goods (the tax base) is reduced by the tax-induced price change. It is assumed that all o f the tax increase i s passed on to consumers and that the average price elasticity o f demand for consumer goods i s -1. 0 Excise duties on alcohol are increased to 35-50 percent in 2008/09, which i s projectedto raise US$0.4 million. 0 All specific rate taxes and fees (such as customs duty and sales tax on petroleum products) are increased in line with consumer price inflation from 2008/09 onward, with the exception o f court fines and Forestry Development Authority stumpage, land rental, and other fees. 1.44 The outturns for 2007/08 include the budgeted extraordinary revenue payments of US$15 million from Arcelor Mittal and US$5.6 million of license fees from Global Systems for Mobile Communication(GSM) companies, both of which are not repeated in future years. The other component o f extraordinary revenue-payments o f overdue and outstanding taxes-is projected at 2 percent of the previous year's tax revenue in eachyear o f the forecast period. 1.45 A US$3 million paymentto the Special County DevelopmentFund has also been made by Arcelor Mittal in each year beginningin 2007/08. The resumption o f large scale iron ore mining operations will generate substantial royalty and income tax revenues for the government in the long term, but most o f these revenues will fall outside of the timeframe covered here." Income tax is not projected to become payable until several years after production begins, for the reasons noted above, and so i s not included in the forecast period covered here. 1.46 Revenues are projected to increase sharply between 2007/08 and 2011/12. But only after 2011/12 will taxes paid by the miningsector start to increase rapidly and boost the revenue to GDP ratio. Table 1.5 shows the revenue projections over the four year period. Total revenue is projected to rise in nominal terms by 80 percent over the four years, from US$201 million in 2007/08 to US$361 million in2011/12. Revenuejumped by 37 percent in 2007/08, driven by the collection o f US$20 million in one-off extraordinary revenues. It is expected that revenues will remain buoyant in 2008/09 as a result o f continued signing fees paid by mining and forestry concessionaires. As a share o f GDP, revenue jumps from 21.8 percent in2006/07 to 25.1 percent in 2007/08, and to 25.9 percent in2009/10. It falls to 24.1 percent by 2011/12. The sharp increase in revenue to GDP ratio is a result o f strong tax growth and the large extraordinary revenue from the mining and forestry sectors in 2007/08, 2008/09, and 2009/10. Taxes on international trade continue to contribute the largest share of 2o The amount of revenue the government could earn from the Mittal Steel iron ore mine is sensitive to the world price o f iron ore. In2007 the price reached historically high levels, and had it remained at these levels over the long term, the government could have generated revenues o f around $200 million a year by the seventh year of the project (about 2014). Prices have begun to decline, and ifworld prices continue to decline in line with the projections inthe WE0 forecast, government revenues will be much lower. By year seven they will amount to about $20 million and will only reach a peak o f around $50 million after more than 20 years ofthe project (Daniel et al, 2007: 60-61). - 1 3 - revenue and stay relatively constant at 40-45 percent o f total revenue. The major change in distributingrevenue is the large increase inthe share o f non-tax revenue-from 5 percent o f total revenue in 2006/07 to 20 percent in 2011/124riven mainly by royalties, county development fimd revenues paid by Arcelor Mittal, and the recovery o f stumpage fees and relatedcharges from timber production. Table 1.5: ProjectedRevenues, 2008/09-2011/12, and Outturn, 2006/07-2007/08 (US%million) Source: Authors' projections. 1.47 The growth of revenue over the medium term is driven primarily by inflation and growth in GDP, imports, and exports.Revenues in2011/12 are higher than in2007/08 inabsolute terms byUS$160million, ofwhich US$95 million is directly attributableto non- miningGDPgrowth and the indirect effect ofGDP growth onthe real growth o f imports. The recovery o f the timber industry contributes US$46 million through growth in stumpage fees and related charges and customs duties on timber exports. Miningrelated revenues contribute US$8.5 million. Tax administration reforms generate about US$5 million and the income elasticity o f income taxes contributes US$4 million. The net impact o f tax policy changes i s almost neutral: the gains o f about US$6 million fi-om increased general sales tax rates are offset by losses due to the reduction inthe top marginal income tax rate. - 14 - 1.48 The revenues in table 1.5 are shown in nominal terms. After being deflated to derive a projection o f growth inreal terms, total revenues are projectedto grow by 48 percent between 2007/08 and 2011/12, comparedwith 80 percent growth innominal revenues. Table 1.6: Revenue Projections, 2007/08-2011/12with Lower GDPGrowth (US$ million) Non-tax revenue Source: Authors' projections. 1.49 Using a lower growth estimate and assuming delays to the development of the mining and forestry sector yields substantiallylower revenue estimates. By 2011/12 the slower growth and delayed developments in forestry and mining reduce total revenue by US$44 million, to US$317 million compared with US$361 million in the higher growth scenario. Since the revenue projections are sensitive to the growth rate o f real GDP and projections o f the latter are inevitably clouded by much uncertainty, an alternative scenario that incorporates the lower GDP growth projections i s shown in table 1.3. In addition to lower growth, this second revenue scenario also assumes delays to extracting iron ore by one year and slower concessioning o f forestry resources. Import growth rates are also lowered in linewith the lower real GDP growth, holdingthe import to GDP ratios constant. Inflationand the GDP deflator are unchanged from those usedinthe revenue projections intable 1.5. The impact o f the lower real GDP growth projections and slowdown in mining and forestry are summarized intable 1.6. Grants 1.50 Liberia has received rants from donors averaging more than US$300 million a year in the last three years?'But only a small share o fthese resources has beenchannelled through the budget. The government received US$1.5 million in budget support in 2006/07 and US$5.7 million in 2007/08. Some project aid i s channelled through the government's systems, with the projects implemented by line ministries, but i s not recorded inthe budget estimates or in fiscal reports, making it difficult to estimate the magnitude o f this. Unfortunately, while a system is being developed for recording donor aid commitments and disbursements, this system is not yet fully operational. 1.51 A large share of the aid received by Liberia has been spent outside the government budget because much of it is emergency humanitarian aid and because donors lack confidence in the capacity of public institutionsto effectively manage aid. But over the medium to long term, the objective of donor aid is likely to shift away from purely emergency humanitarian aid and toward more conventional developmental assistance, such as infrastructure projects, which are normally implementedby governments. This should encourage donors to channel a larger share o f their aid through the government budget, especially if the capacity o f the public service to absorb aid is strengthened. Furthermore, 21This figure does not include any of the approximately $700 million a year which is spent on peacekeeping operations. - 15 - multi-donor groups have been established to coordinate aid programs in key sectors, including health, education, and infrastructure. I.52 There are many difficulties obtaining forward-looking projectionsof aid from donors in developing countries. Future donor aid budgets are subject to uncertainty, particularly because many developed countries are facing fiscal constraints, which may affect bilateral aid levels. And it i s difficult to identify a uniform definition o f aid applicable to all donors. Since only partial projections from some donors are available, any quantitative projections o f grant aid would be essentially speculative and are omitted from this chapter. Projections o f donor aid flows to Liberia should improve with the establishment o f the aid management unit within the Ministry o f Finance, which has recently begun to collect aid data. In the medium term, it should be possible for the government to channel more aid through the budget, but in the short-term much o f the donor assistance will continue to take the form ofproject grants. GovernmentBorrowingand Debt Servicing 1.53 In principle, fiscal space can be created by government borrowing if it is consistent with debt sustainability and macroeconomic stability. But the scope for new borrowing in Liberia is heavily circumscribed by the government's unsustainable external and domestic debt burdens and by the lack o f macroeconomic space for domestic borrowing. The government hasundertaken no new domestic borrowing since 2002/03 inorder to protect macroeconomic stability. The government has little creditworthiness to issue domestic debt to the commercial banks or the nonbank private sector because it already has outstanding domestic debt and arrears estimated at over US$300 million in 2007. So, any domestic financing o fthe budget would have to rely mainly on credit from the Central Bank o f Liberia. But because reserve money is very limited-less than 12 percent o f GDP at the end o f 2007-and the Central Bank lacks instruments to sterilize injections of reserve money, any government borrowing from the Central Bank wouldjeopardize the Central Bank's control o f the money supply. So, it is prudent for the government to avoid any domestic financing o fthe budget over the medium term. DomesticDebt Servicing 1.54 The government will need to find resources for paying off its verified domestic debts in a phased manner over the medium to long term. An exercise to verify domestic claims on govemrnent was conducted in 2006. Of more than US$900 million o f assessed claims, US$304 million were assessed as valid (of which US$261 million comprised claims by the Central Bank and US$24 million comprised salary arrears), US$317 million were assessed as contestable (mainly because of inadequate documentationto prove validity), and US$292 million were rejected. 1.55 The government's domestic debt strategy is to repay the valid claims and any of the contested and rejected claims that are proved valid. Except for the debts owed to the Central Bank, claims will be discounted, with the rate o f discount varying according to the size o f the claim (larger claims will be subject to a larger discount). The valid debt to the Central Bank and other financial institutions will be restructured into securities with a face value o f US$278 million, a maturity o f 30 years, a grace period on principal repayments o f 10 years, and an interest rate that increasesevery five years for the first 15 years and is then held constant. So, in the first five years the servicing o f debt to financial institutions will comprise only the low interest rates, amounting to US$2.4 million a year, but debt servicing - 16- costs will rise sharply after 10 years when principal repayments become due, to around US$19 million. The interest costs o f servicing the debt to financial institutions during 2007/08-2011/12 are shown intable 1.7. Table 1.7: ProjectedDomestic Debt ServicingCosts, 2007/0&2011/12 (US$ millions) Sozrrce:Republic o f Liberia 2007 (42), IMF. 1.56 The government plans to establish a Domestic Debt Trust Fund, into which appropriations from the budgetwill be made, to provide funds for clearingarrears and servicing debt. It has been proposed that the government should allocate 5 percent o f total revenues to the trust fund. The projected payments into the fund shown in table 1.7, for 2008/09-2011/12, are derived by multiplying the projected revenues in table 1.5 by 5 percent. Whether these payments are sufficient to cover debt payments inthe medium term or whether surpluses are accumulated that could be used to fund debt service payments will depend on how large a share of contestable and rejectedclaims are proved valid. External Finance 1.57 Even with HIPC debt relief provided by all creditors, the government will face some increase in external debt servicing costs since the country was not servicing its debt prior to enteringHIPC and only made token paymentsof US1.2 million a year?2 Without further exceptional support for debt service payments, the expected debt service after HIPC would rise to an average o f about US$80 million a year, as shown intable 1.8. As this i s clearly unaffordable, the high level o f debt service payments due during the interim period requires further exceptional deferral o f payments and that creditors and donors provide additional debt relief. Most o fthe bilateral creditors have agreedto provide relief beyond that required under the HIPC initiative. The government anticipates that external debt service will therefore be limited to a maximum o f 1 percent o f GDP during the interim period, which amounts to between US$7 million and US$12 million a year. Once the completion point i s reached, the provision o f HIPC and Multilateral Debt Relief Initiative (MDRI) will reduce the debt servicing burden to levels that can be met without placing an excessive burden on the budget. 22 This debt service was being paid on IMF, IDA, and AfDB loans andpaid into a suspense account. - 17- Table 1.8: Projected Annual Debt Service After Heavily IndebtedPoor Country (HIPC), Multilateral Debt ReliefInitiative (MDRI), and Additional Bilateral Debt Relief (US$ million) 2008/09 2009/10 2010111 ~2011112 Total debt service after HIPC 91.2 73.4 70.2 25.4 1 10.3 and MDRf debt relief i--- Multilateral 2.8 5.7 7.1 I_-._-- - 5.3 2.3 _I- ', '+^.- Official bilateral 56.3 --.-45.5 50.2 . -. 13.8- ' .. . 3.7- - Commercial 35.1 i7.4 17.4 6.0 . 3.8 Total debt service after HtPC, 92.6 68.4 63.6 30.6 8.9 MDRI, and additional bilateral I debt relief I Note: Projections assume the decision point is reached in early 2008 and completion point in late 2010. Source: IDA and IMF2008, table A4. Improving Expenditure Efficiency 1.58 Achieving expenditure efficiency improvements will require public financial management (PFM) reforms recommended in this report, which require a medium- to long-term perspective. It is not clear that significant efficiency gains can be generated inthe four-year timeframe o f these projections. That fact, along with data limitations, means that efforts to quantify potential gains are not very useful. H. SUMMARY OF FISCALSPACEPROJECTIONS AND THEIR IMPLICATIONS 1.59 If economic growth accelerates over the medium term, increased revenue mobilization will generate substantial fiscal space by 2011/12, with revenues growing by 80 percent (or US$160 million) in nominal terms compared to the 2007/08 outturn. Higher debt servicing, however, will reduce the amount o f fiscal space available. Domestic debt servicing may require up to US$16 million by 2011/12. External debt servicing requirements are still subject to much uncertainty because o f the need to mobilize additional debt relief over and above that given under the HIPC initiative. But the government expects that they will be limited to a maximum o f 1 percent o f GDP during the Poverty Reduction Strategy period. There will be almost no scope for mobilizing budget resources from domestic or external borrowing over the mediumterm. 1.60 Revenue growth may be lower than expected if growth prospects are not realized and if there are further delays to the extraction of iron ore and forestry products. Since the impact of the global economic slowdown on Liberia's growth is uncertain, revenue projections are tentative. If growth is two percentage points lower than projected, iron ore production i s delayed by one year, and forestry extraction i s slower than expected, then revenue estimates fall to US$317 million by 2011/12. This represents a growth in revenue o f 58 percent innominal terms and 28 percent inreal terms over the 2007/08 outturn. 1.61 No projectionswere made of the fiscal space that could be generated from donor grants or improved spending efficiency. But the fiscal space projections can be used to estimate the resource shortfall between the projected budget resources for primary expenditures and the estimates o f public expenditure requirements. Ifthe requirements are to be fully funded, the resource shortfall would then have to be filled through grants, additional debt relief, andexpenditure efficiency improvements. - 18- 1.62 The full Poverty Reduction Strategy, completed in April 2008, includes a costing chapter, which lays out the country's anticipated funding needs for the fiscal years 2008/09-2010/11. It includes anticipated expenditure for all core poverty-related sectors in line with the Poverty Reduction Strategy pillars, including security, economic revitalization, governance and rule o f law, and infrastructure and basic services. It is anticipated that much o fthis expenditure will be expended outside the government budget, through different project and pooled fundingarrangements. Table 1.9: EstimatedCost by Pillar,2008/2009-2010/2011 (US$ million) Source: Expenditure estimates: PRSP 2008; Revenue estimates: authors' projections. 1.63 The core Poverty Reduction Strategy (PRS) spending needs are estimated at US$550 million in 2008/09 (compared with the US$132 million allocated in the budget). This will fall slightly to US$526 million by 2010/11 (table 1.9). 1.64 In addition to these core poverty reducing spending requirements, many other government administrative costs will have to be financed in order to maintain the state structure, such as financing o f the legislature, Ministry o f State, and debt service. The PRS estimates that 45 percent of government revenues will be dedicated to non-PRS expenditures in 2008/09, falling to 35 percent of total revenues by 2010/1I,for a total o f approximately US$l10 million a year. 1.65 In current prices, total requirements for expenditures on PRS and public administrative costs are estimated at US$650 million a year over the PRS period (2008/09-2010/11). This is about 70 percent of projected GDP for 2008/09, falling to 52 percent o f GDP by 2010/11. The resource shortfall is US$418 million in2008/09 and falls to US$326 million by 2010/1I-assuming that creditors and donors provide additional debt relief or resources to enable the government to avoid any substantial increase in its external debt servicing burden in line with the government's projections. This shortfall could be funded largely by donor grants. Initial estimates for the PRS suggest that donor commitments are already substantial for the PRS period. The resource shortfall could also be funded, at least partly, by improving expenditure efficiency. Given that the noncore expenditures comprise 62 percent o f primary expenditures in the 2007/08 budget, there appears some - 19- scope for making allocative efficiency gains by shifting budgetary resources from the noncore to the core spending requirements over the medium term. I. CONCLUSIONSANDRECOMMENDATIONS 1.66 Achieving and sustaining rapid economic growth should be the policy priority for the government and its donor partnersover the medium term. It should be possible in the short to medium term to accelerate GDP growth simply by ensuring that aggregate demand i s sufficiently buoyant, given that there i s widespread unemployment, and i s supported by adequate inflows o f foreign exchange to meet the growing import requirements. There i s a strong case for maintaining or even increasing donor aid to support economic activity. Rapid economic growth will generate more income and employment and i s crucial for boosting government revenues. 1.67 The government should continue to maintain macroeconomic stability. It should maintain prudent fiscal policies, avoid deficit financing, and implement an appropriate debt- management strategy. It should also continue pursuing long-term growth objectives over immediate gains inrevenue, to ensure that pro-poor growth is sustained. 1.68 Over the long term, sustainable economic growth will depend on attracting private investment, rebuilding the country's infrastructure, and supporting human capital development. A favorable investment climate can be developed by streamlining business procedures, implementing planned tax reforms and tax administration reforms, and improving the management and infrastructure o f Liberia's ports. 1.69 Despite the projected growth in revenue over the next five years, there will be large resource shortfalls because the requirements for public expenditures are greater than the resources available to fund them. Poverty-reducingexpenditure requirements are estimated at about 300 percent higher than the allocation to these expenditures inthe 2008/09 budget. Attracting more donor aid and improving expenditure efficiency will therefore be crucial ifthe government is to fund its core poverty-reducing spending requirements. Donors should begin to channel increased funding through national systems, either through direct budget support or by contributing to pooled funds. Most donor financing is currently project based and implemented outside o f government agencies, which makes coordination and management o f public development programs challenging. - 20 - 2. PUBLICFINANCIAL MANAGEMENT SYSTEM A. INTRODUCTION 2.1 Liberia's public financial management system has improved markedly since 2006 through an ambitious reform agenda agreed between the government and the international partners. But many o f Liberia's financial management structures are transitional, designed to strengthen central controls, curb corruption, and enable centralized planning during reconstruction. So, there is a need to shift toward longer-term and more sustainable solutions in some areas and to continue to implement, fine-tune, and entrench new systems. Note that the authorities are rebuildingan extremely weak and outdated public financial management system, which deteriorated markedly duringthe civil conflict. Reform efforts thus need to focus on getting the basics right, rather than introducing sophisticated reforms or systems that may absorb a large share o f the government's scarce resources and human capacity. 2.2 This chapter provides an analysis of Liberia's public financial management system. It starts with a summary o f the findings o f a 2007 Public Expenditure and Financial Accountability (PEFA) analysis. This is the first systematic baseline assessment of the performance o f the public financial management (PFM) system inLiberia inthe post-war era, which identifies its key strengths and weaknesses. The chapter then provides a detailed analysis o f recent PFM reform progress and identifies remaining areas in need o f strengthening through a review of eight key dimensions that promote good PFM outcomes: the legal framework, budget preparation, budget execution and control, procurement, accounting and auditing, externaloversight, debt management, and donor practices. B. OVERALLASSESSMENT OF THE PUBLICFINANCIAL MANAGEMENT SYSTEM 2.3 A comprehensive Public Expenditure and Financial Accountability Assessment was carried out in late 2007 by representatives from the African DevelopmentBank, Department for International Development, International Monetary Fund, United NationsDevelopment Programme,Swedish NationalAuditing Office, and World Bank, in collaboration with Liberian authorities. The PEFA assessment was reviewed by the PEFA Secretariat. It covers 2004/05-2006/07 and builds on the analytics undertaken by various development partners over the last five years. The assessment benefited from input from key civil servants, ministers, legislators, and PFM advisors working in the Ministry o f Finance. 2.4 The assessment used the Public Expenditure and Financial Accountability performancemeasurementframework.This framework, usedby about 80 countries, tracks country PFM performance over time. It comprises 28 high level indicators that assess the functioning of the PFM system components and 3 indicators that assess donor performance. Each PEFA indicator i s given a rating from A (highest) to D (lowest). Note that the PEFA scorings look at practices over past years, and thus do not fully capture ongoing reform progress. The scores aim to provide a baseline that will allow for comparison with performance in future years (see annex 2 for a more detailed presentation o f the PEFA framework). 2.5 Table 2.1 presentsthe ratingsagainst each of the 31PEFA indicatorsfor Liberia. In general, the scores are consistent with other post-conflict countries in Africa, such as Sierra Leone. There i s 1 A score, two B scores, 10 C scores, and 15 D scores, with three - 2 1 - PEFA indicators receiving no score. Considering Liberia's outdated institutional system for PFM and the severe stress to which it has been subjected because o f the civil war and economic collapse, the scores are relatively strong and reflect improvements since the end o f the civil war. 2.6 The processes for budget execution, reporting, and auditing scored especially poorly, along with the practices of donors. Budget outturn, comprehensiveness, and transparency received slightly better scores on average, particularly for revenue and expenditure performance, budget classification, and access to fiscal information. The indicators that were almost all consistently poor, with scores o f D/D+, reflected remaining weaknesses in tax collection, payroll controls, quality and timeliness o f annual financial statement, and internal and external audit. Good scores were given for budget classification, availability o f information on resources received by service delivery units, and expenditure and revenue outturn compared with original approved budgets. Two PEFA indicators about the financial information provided by donors and the proportion o f aid managed by using national procedures bothreceived scores o f D+. 2.7 The poor PEFA scores suggest substantial scope for improvingthe PFM system, provided that sufficient resources and sustained efforts are dedicated to this objective and that there is strong politicalsupport for PFM reforms.Many important reforms are already in progress and starting to show results, but the PEFA assessment focuses on the previous fiscal year (2006107) and so does not necessarily capture ongoing reform efforts. Andthe PEFA framework is a standardized assessment tool, which does not fully account for the transitional systems that were established in post-conflict Liberia due to eroding institutional capacity (such as the Cash Management Committee, cash-based budgeting, and the large scale of donor project financing). Annex 2 provides a description o f the PEFA indicators and the reasons for the scores. - 22 - Table 2.1: Summary of2007 Public Expenditureand FinancialAccountability (PEFA) Performance Scores m P I - 3 T ~ g z trevenueoutturn comoaredwith original aoorovedbudget e 1R IPI-4 I Stockandmonitoringofexoenditurepaymentarrears - . . . . .-... . B. Key cress-cutting issues: Comprebensivenessand transparency I I II PI-19 Competition, value for money, and controls inprocurement I C D+ PI-20 IIEffectiveness II c+ . I of internalcontrols for non-salaryexpenditure I c+ I PI-27 , I I L . Legislativescrutiny of the annual budget law I c+ I c+ -23 - C. LEGAL FRAMEWORK 2.8 The legal and regulatory framework for PFM in Liberia is contained in the ExecutiveLaw of 1972 and the Revenue Code of Liberia Act of 2000; amendments have been made to them over the past years. The Public Procurement and Concessions Commission was established through the Public Procurement and Concessions Act of September 2005. An act ofthe national legislature establishedthe GeneralAuditing Office as an autonomousagency within the executive branchofthe government. The 1986Constitution of the Republic of Liberia made provision for the legislature to create the General Auditing Commission. The commission is placedunder the legislative branch, giving it independence from the executive. 2.9 The existing PFM legal and regulatory framework needs to be developed comprehensively. In accordance with section 2204 of the Revenue Code of Liberia Act 2000, the MinistryofFinance has developedand the cabinet has approveda set of 54 interim financial rules that cover mainly expenditure controls. But these rules are not sufficiently comprehensive as they do not cover the whole PFM cycle. They also do not provide clarifications of reporting requirements or of roles and responsibilities for managing public funds. Organizational structures need to be specified with clear roles and responsibilities as well. Other elements to be developed within a PFM fiamework include establishing internal controls, institutionalizing the internal audit function, formalizing reporting requirements, and establishing a code ofconduct for public officials. 2.10 Assisted by the International Monetary Fund, the World Bank, and other donors, the government has developed a new public financial management act, with enabling regulationsand operationalmanuals.This new act addresses the discussed issues by covering the whole PFMcycle-budget planning, procurement, accounting, reporting, and external oversight. The act has been submitted to the parliament for approval in January, 2009. The new framework will feature: a Organizational structures with clear roles and responsibilities for all those involved, endingthe fragmentationof responsibility for preparingand implementing the budget. a Internal controls and an institutionalized internal audit function. a Formalizedreporting requirements. a Coverage of the entire PFM cycle, including budget planning, procurement, accounting, reporting, andexternaloversight. D. BUDGET PREPARATION 2.1 1 The institutional responsibilities for preparing and implementing the government budget have been fragmented, with no single agency having overall responsibility for control and coordination. The Ministry of Finance (MoF) has the responsibility for revenues, treasury functions, and, since late 2005, the implementation of the cash management system. Until August 2008, the Bureau of the Budget (BOB),which was independent from the MoF, had the responsibility for preparing the full budget. Legislation was recently passed for merging the bureau with the MoF, and the two are planning the merger, which will simplify and improve coordination of the budgeting process and budget execution. There have also been marked improvements in budget preparation coordination in past fiscal years, with the establishment of the Budget Committee, which - 24 - includes representatives from MoF, BOB,Ministry of State, Ministry of Planning, Civil Service Agency, and General Service Agency. The Budget Committee provides oversight of the budget process and advice to the Cabinet on policy proposals and the strategic direction of the budget. The recurrent and investment budgets have been merged into a single budget, which has helped to simplify the budgeting process. Ministries and agencies are engaged early in the budget process through a budget circular setting out guidelines for budget preparation, sent out by the BOB.The legislature's role in the budget review process, however, needs strengthening. Capacity building efforts and greater communication between the legislature and executive are required to ensure that the legislature can perform its oversightrole andreview andpass the budget ina timely manner. Budget Timeliness, Comprehensiveness,and PolicyLinkage 2.12 Although the timeliness of enacting the budget has improved modestly, late passage continues to pose challenges to the government. Budgets in recent years were approved betweentwo and three months into the beginning of the new fiscal year. The legal framework governing the budget calendar can be improved. The Revenue Code does not stipulate a deadline for submission of the budget estimates to the legislature and there is no provision for the MoF to operate on a specialwarrant inthe event of a delay inapproving the budget. These omissionsare expectedto be rectified with the new PFMAct, now undergoing legislative review. 2.13 While the comprehensiveness of the budget document improved in the current fiscal year, it does not provide a complete overview of the government's fiscal operations. The budget circular contains policy priorities, fiscal policy objectives, recommendations for budget preparation, and lists of key activities. The budget documents for 2007/08 and 2008/09 contain more detail than in previous years, including preliminary revenue forecasts and the first claims on these revenues. In addition, the recurrent and developmentbudgetswere merged into a single budget starting in2006/07. 2.14 The government has limited extra-budgetary expenditures to an estimated 1-2 percent, but some critical fiscal information is still missing.23First, the budget lacks data on the debt stock and financial assets, both of which should include details at least for the beginning of the current year. Second, the budget does not contain information on the prior year's budget outturn, presented inthe same format as the budget proposal. Third, there is no information on the implications of new policy initiatives or estimates ofthe budgetary impact of major changes in policy or expenditure programs. Fourth, the large volume of aid resourcesare mainly channeledoutside of the national budget and not adequately reportedin the budget or in a separate aid database. Total aid flows are estimated at 100-200 percent of government resources. They have a large impact on macro and fiscal conditions and pose a challenge to coordinating and managing public services. Finally, the budget classification system does not comply with international standards; the items of expenditures are currently being revised with support from the World Bank and the United States Agency for International Development. 2.15 The link between the budget and the government's policy objectives has been strengthened through a shift toward a rudimentary program budgetingapproach. This i s in line with the Interim Poverty Reduction Strategy for the 2007/08 fiscal year, which 23 Section 2206 o f the Revenue Code stipulates the f o m and details to be contained inthe budget document as determined by the President. - 25 - defined the government's spending priorities inareas such as health, education, public works, and agriculture. The completed full Poverty Reduction Strategy (PRS) (April 2008), which included a costing chapter, laid the basis for the 2008/09 budget. Although the 2007/08 and 2008/09 budget documents have improved markedly compared to earlier years, the absence of a program segment in the chart o f accounts limits directly linking expenditure reports to PRS objectives. BudgetPredictability 2.16 Despite recent improvements, the government's ability to produce a macroeconomic framework needs to be strengthened. Its ability to project revenues and expenditures remains limited and relies heavily on estimates from the International Monetary Fund (IMF). The MoF is developing a medium-term fiscal framework, which will address some o f these shortcomings by settingthe broader parameters for macroplanning. But further capacity building i s necessary to strengthen the capacity o f the macrofiscal unit o f the MoF and the revenue department to develop independent projections and regularly reassess and revise projections. 2.17 Revenue projectionshave improved recently and the 2007/08 budget for the first time included some description of the macroeconomic framework. The process o f estimating tax revenue is well structured. Sector ministries and agencies submit revenue estimates to the MoF where various tax- and duties-collecting units-such as the Bureau o f Customs and Excise, the Large Taxpayer Unit, the Medium Taxpayer Unit, and the Small Taxpayer Unit-submit formal revenue estimates. The MoF uses a mix o f tax ratio scenarios and trend analysis to forecast and fine tune the assumptions and estimates. The process considers fiscal year historical performance, including pre-shipment inspection statistics, policy measures, and other applicable macroeconomic forecasts (such as GDP growth estimates). 2.18 Expendituresneed to be more carefully projectedduring budgeting. The estimate of recurrent expenditures for each ministry is usually determined by the previous year's budget, with some indexing for inflation and wages. This process does not allow room to reassess the government's recurrent expenditure priorities, as determined by the Poverty Reduction Strategy (PRS). Capital expenditure financed by the budget i s largely directed toward the acquisitions o f equipment and vehicles. But the most important part o f the investment expenditure is not incorporated into the budget because it is externally funded by donors. 2.19 Given the cash budgeting system, the wage bill projection is consistent with the budget ceilings, and no arrears accumulate. The budget circular recommends that estimates o f payroll expenditures be calculated on the basis o f the number o f employees as o f December 31 o f the previous fiscal year. The line ministriesprepare their wage bill estimates for the coming year for submission to the Department o f the Budget. The payroll and professional service fees from the preceding fiscal period are used as a baseline to determine personnel costs for the next fiscal year.24The cost o f consultancies, professional services, and vacant positions that are expected to be filled are factored into the personnel cost estimates. So, the wage bill estimates are based on the database and on recruitment plans. Remaining 24The Department for International Development is funding a Civil Service Capacity Buildingproject through consultancy services from Adam Smith International. Manpower hearings, and position controls are planned. - 26 - wage arrears come from the transitional period, and there has been no recent accumulation o f arrears. Including staffing levels ineach unit should help identify skill gaps inthe units. 2.20 There are some limits to a proper assessment of the wage bill. There are two databases o f civil servants, one at the Civil Service Agency and one in the Electronic Data Processing Unit in the Ministry o f Finance. While these databases are relatively consistent, they require regular reconciliation exercises to ensure consistency. Adding and removing names from the civil servant database remains slow and unstructured. A major problem i s that allowances are not based on clear guidelines. Lacking a consistent classification system has even led to employees beingpaid at wrong grades. The government is working to address some o f these problems with a Human Resource Information Management System, which will be linked to the IntegratedFinancial ManagementInformation System. Shift toward a Medium-Term ProgramBudgetFramework 2.21 The government has indicated its desire to shift to a multi-year budget to help guide investment and development spending. At present, a one-year budget is prepared by the Ministry o f Finance, and all appropriations lapse at the end o f the fiscal year, except for such sensitive appropriations as the County Development Fund (which is transferred into an escrow account if unallocated in the current fiscal year). As part o f the Poverty Reduction Strategy, the government undertook a three-year costing exercise, yielding costing estimates by fiscal year and sector required to meet the strategy objectives. It also estimated the budgetary resources needed to achieve the objectives and indoing so provided a first step to a multi-year budget. The costing exercise identified the multi-year financing gap that will affect the government's ability to rebuild the economic and social infrastructure needed for economic revitalization and poverty reduction. Another step toward multi-year planning i s developing a medium-tern fiscal framework, outlining changes to the broader macro conditions. 2.22 The Liberian government is encouraged to phase in elements of a multi-year budget framework over time, as capacity to manage the process is built. A multi-year program budget approach assumes certain prior conditions, notably a stable macrofiscal framework; technical capacity to define programs, establish performance indicators, and assess performance; and clear procedures and respect for prevailing regulations. It is recommended that the government ensure that: The macroeconomic analysis capacity within the MoF i s strengthened to improve the predictability o fthe resource envelope and expenditure priorities. 0 Strategies for key sectors (especially those with long-term planning and building cycles) are produced, medium-term objectives formulated, and related expenditure plansproduced. 0 There i s a strong link between budgeted finds and the results as defined by measurable indicators. 0 There is a system o f informationon performance. Preliminary steps incritical sectors could be taken inpilot ministries by identifying programs and key performance indicators and reporting them inthe budget document. 2.23 The Liberian government should set up a well-functioning system for selecting and managing public investments as domestic and external resources increase. This system should ensure that donors increasingly channel resources through the national budget. - 27 - An institutional arrangement is needed to ensure a first-level screening o f all project proposals to ensure that they meet minimum criteria o f consistency with the government's strategic goals. This will requireformal and well-publicized guidance on the technical aspects o f project appraisal, consistent with the technical capacity o f ministries and departments. This guidance would define cost-benefit and cost-effectiveness criteria-with larger projects requiring more rigorous tests o f financial and economic feasibility and sustainability. Upstream investment in training for project-evaluation techniques is important for an effective public investment system. Costs are generally incurred to maintain and operate assets created through public investment projects, and there should be consideration o f how these costs should be funded. Evaluating completed projects i s also desirable to ensure that learning and feedback from projects will create a positive dynamic for improvement. Decentralization 2.24 The government is discussing the need to decentralize political power, decisionmaking, and government authority, but its approach to decentralization is cautious and gradual. During the PRS period, the government intends to undertake extensive consultations and begin developing a roadmap and policies for decentralization. Financial decentralization to ministries, agencies, and county administrations will be considered, but not inthe immediate future, as capacity constraints, logistical challenges, and constitutional restrictions need to be addressed first. The decentralization o f the procurement function, as specified under the PPC Act, helped increase ownership over the budget, while central controls are maintained through the payment process and Cash Management Committee approval requirements. In the short term, continued focus on financial management capacity buildingat the ministryand agency and county level will be pivotal, to gradually strengthen capacity for an increasingly decentralized financial management system. E. BUDGET EXECUTION AND CONTROL CashManagement 2.25 To ensure prudent cash management, ministries and agencies prepare monthly cash plans. The DoB releases allotments based on the cash plans, and the budget is executed based on the processes described in the "Recording, Accounting and Reporting" section below. The MoF does not consolidate its cash balances on a daily basis. Balances held on ministries' and agencies' bank accounts in the Central Bank and commercial banks are not included in the cash availability statement prepared for decisionmaking by the Cash Management Committee (CMCo). Cash availability i s based on balances from the bank statement rather than from the cash book. BudgetTransfers 2.26 A Budget Transfer Act was passed in early 2008, amending section 2212 of the Revenue Code, to allow transfers of up to 20 percent between agencies. This represents an important improvement, although the limits remain high by international standards (normally between 5 and 10 percent). Payroll 2.27 The ElectronicData ProcessingUnit processesthe government payroll using the GovernmentAccounting Payroll System for about 42,000 employees and retirees (civil -28 - servants, teachers, police, military, and pensioners). Previously, civil servants from departments or districts were paid in bulk, but the government shifted to issuing individual checks to all civil servants, to clean up the payroll and ensure that payment reaches each civil servant. Some remuneration to employees, such as allowances, are paid outside the system, and there have been instances when checks were generated for names that are not in the system. The MoF is taking steps to bring the payments closer to the civil servants, by undertaking payment exercises through paymasters that distribute checks directly to employees inthe county capitals. Cashiers from Central Bank travel with the MoF pay team to cashthe checks, and logistical support i s obtained from UnitedNations Mission inLiberia, at a cost. This means that civil servants only have to travel to their county capital to receive payment, instead o f to Monrovia. Initially this strategy resulted in 2,000 unclaimed checks a month on average. However, the large number o f checks issued is causing difficulties for the MoF inreconciling the payroll bank account. 2.28 The Civil Service Agency is planning to implement a biometric identification systemto verify all civil servants who should be on the payroll. It also plans to rationalize the classification system to ensure consistency in basic pay for the job families and grades. This will be an important milestone before implementing the Human Resource Management Information System. 2.29 Supervisory controls over the payroll are weak. Controls exist for changes in personnel records and payroll, but the full integrity o f the payroll cannot be ensured because changes are not adequately checked before the payroll i s run. Audit trail reports, to show before and after details for changes and to track users, do not exist. A structured payroll audit to identify control weaknesses i s not undertaken. But the county payment system has ensured that fewer ghost workers collect payment, as evidenced by the significant number o f unclaimed checks each month. Revenue CollectionProcedures 2.30 Tax administration has undergone significant reforms in the past year, in line with IMF recommendations.The previous system o f bureaus and the centralized collection unit for aged debt have been dissolved, and three new tax units (large, medium, and small taxpayer) now have their own collection and enforcement responsibilities. This has helped improve the accuracy o f information on balances and adjustments resulting from payments, new assessments, and accruing interest and penalties. Records from the small taxpayer units, however, still may not be entirely reliable. The customs administration i s being strengthened by establishing a customs one-stop-shop at the Freeport in Monrovia, although the premises are still awaiting full connectivity. 2.3 1 Transparency of taxpayer obligations is improving. Regulations for tax administration and procedure are contained in sections 50-74 o f the Revenue Code, and notices for tax collection are displayed inconspicuous places at MoF. Executive Orders 2 and 8 o f the National Transitional Government o f Liberia provide that all tax and non-tax revenues should be paid directly to the Central Bank. The MoF has developed an in-house Bank Pay Slip and Flag Receipt System to handle the recording and accounting for revenue collections. The Bank Pay Slip i s linked online to the pre-shipment office and the Central Bank. MoF can therefore project taxes due from goods arriving through the ports. By flagging receipts recorded inthe Bank Pay Slip against lodgement in Central Bank o f Liberia bank accounts, MoF can track assessments and collections. Procedure notices are also displayed inthe MoF. - 2 9 - 2.32 There is room to improve effectiveness in collecting tax payments, although reconciliation efforts have been strengthened. Most accounts are reconciled daily- including tax assessments, collections, arrears records, and receipts-by the Treasury and the Central Bank. Also, taxes assessed are reconciled against collections and payments through regular desk and issues-oriented audits. There are also semi-annual and annual comprehensive audits to verify that payments are consistent with taxes assessed. Taxpayer Delinquency Investigations are also conducted regularly to ensure tax compliance. Vendors are required to have a Tax Payer's Identification Number. The number should be used as a unique vendor code inthe accounting system at MoF and linked to the Ministryo f Commerce business registration system. It should also be a requirement for banks to open business accounts. Similarly, Social Security Numbers should be used to track payment o f employment tax. 2.33 Despite this progress, the amounts collected do not correspond to the recovery potential of the departments in charge of revenue collection. The government still faces major problems, such as false declarations, under declarations, low compliance, and leakages. Although audits have been introduced, revenue recovery could be improved through a strengthened audit fimction. Fake receipts are also a serious threat to the receipt system. Tighter internal controls are needed. With an improved accounting system, receipts shouldbe system generated and linked to a cash book used as the main reference for bank reconciliation. The government is tryingto obtain receipts with security-related features. F. PROCUREMENT The Legal and RegulatoryFramework 2.34 The Public Procurement and Concessions (PPC) Act, which came into effect in January 2006, is Liberia's first significant step toward subjecting public sector contracts to meaningful competition. Overall, the Act is comprehensive and provides a framework for a sound procurement system. But some aspects o f the Act require amendment, including revising the thresholds (which are too low), changing the rules governing exploration licensing procedures, and clarifying the roles and responsibilities o f the Public Procurement and Concessions Commission (PPCC). A committee has been established to review the Act and, with support from an external consultancy firm, is drafting an amendment to submit to the legislature in early 2009. In addition, the PPCC, with support from a consultancy firm, i s developing detailed implementing regulations and manuals, with the aim ofproviding procuring entities with detailed instructionsonthe correct application o f the law. Institutional Framework and ManagementCapacity 2.35 Procurement units and procurement committees have been set up, and the procurement staff are eager to comply with the Procurement Act. Butthese procurement structures lacked the resources-including staff, basic computer systems, access to transport, and manuals and instructions-to operate effectively. Procurement directors in some entities lacked the technical capacity and authority to resist interference from other divisions in the procurement process. 2.36 There is still some confusion over the roles and responsibilities within the Liberian public procurementsystem. Some procuring entities do not recognize the needfor a strict separation o f fhctions between the procurement unit, which is the executing body, - 30 - and the procurement committee, which provides supervision. Evaluating proposals i s sometimes done by members o f the procurement committee. And in some ministries, procurement responsibilities are spread out over different units, with shopping being conducted by one unit and national competitive bidding by another, coordinated only at the level o f the minister. The PPCC, whose primary function i s to regulate the system and prevent departures from the procurement law, i s routinely solicited to provide no-objections on procurement decisions, a function that i s normally the responsibility o f procurement committees and compromises the PPCC's neutrality as a regulator. ProcurementOperationsand Market Practices 2.37 While Liberia's procurement law is basically sound, public procurement falls short of the high standard the law sets. For instance, not much has changed in actual shopping practices. Staffs o f the procuring entities still go to three selected suppliers for written quotations on the spot instead o f soliciting them through a writtenrequest. Shopping is still carried out for contracts far above the threshold o f US$2,000 stipulated inthe schedule (although it is generally recognizedas being too low). 2.38 National competitive bidding practices also need improving to conform to the Act. The advertisement contains a call for bids instead o f inviting bidders to purchase the bidding documents at the cost o f reproduction. So, bidders are deprived o f important information that they should consider when preparing a bid. Consequently, important provisions-such as the delivery time, penalties for late delivery, and requirement for spare parts availability-are left to the contracting stage or are not addressed at all. The criteria for evaluation are also not specified. 2.39 Despitethese shortcomings, there hasbeen notableimprovementin procurement practice in the past year. Procurement staff and committee members now have some grasp of the key requirements of the Act. The break in procurement performance came largely as a result o f PPCC workshops given to the heads o f procuring entities. This suggests that further targeted training would be highly beneficial. A more systematic training program for procurement officials is being developed with support from a consulting firm. Most o f the interimprocedures ofthe last fiscal year that were contrary to the provisions o f the Act have been discontinued. There are strong indications o f a change in attitude toward procurement and a change o f habits inresponseto the Act. 2.40 The establishment of the Cash Management Committee in the Ministry of Financehashelpedidentifyand overturn many breachesof the procurementlaw, as the committee returnsvouchers that are not in compliance with the PPC Act. The Ministry o f Finance has developed and issued a checklist with Cash Management Committee requirements to help ministries and agencies comply with procurement and payment procedures. Private Sector Participation 2.4 1 Private sector representatives have responded positively to the Public Procurement and Concessions Act. They believe it will promote transparency, accountability, and equal opportunities for businesses in Liberia. But they noted that challenges remain. These include a lack o f training in preparing biddingdocuments, the high cost o f bidding documents and bid security, and poor access to credits from banks. - 3 1 - G. ACCOUNTING ANDAUDITING Accounting 2.42 Section2204 of the Revenue Code requiresthe government of Liberia to keep its books on a double entry and accrual basis of accounting. But because o f the absence o f core skills and competencies, this provision is inappropriate at this stage in Liberia's development process and should only be considered after sufficient capacity has been established. Inthe interim,the system is operating on a cashbasis o f ac~ounting.~~ 2.43 The government, on the advice of its donors, has set up a system to account for its revenues, expenditures, assets, and liabilities involving the Bureau of the Budget, Bureau of General Accounting, the Cash Management Committee, the Office of Comptroller General, the Central Bank of Liberia, and ministries, departments, and agencies. The system has increased transparency and accountability, enabled initial budget policy targets to be met, and ensured that payment arrears are not accumulated. 2.44 Each ministry and agency is responsible for submitting a monthly cash plan to the Department of the Budget (BOB), and the department prepares a monthly cash allotment based on the cash needs. Copies o f the cash allotment from the DoB are sent to the ministry or agency and the Bureau of General Accounting (BGA) for cross checking. Following receipt o f its cash allotment from the Department o f the Budget, the ministry or agency prepares a local purchase order. The purchase order-plus a purchase voucher, invoice, and copy o f the BOBcash allotment-is then sent to the Bureau of General Accounting. After verifying compliance with procurement procedures, the BGA puts its seal on the local purchase order and returns the documents to the ministry or agency. This seal o f verification allows the ministry or agency to proceed with procurement. On receipt o f the goods and services, the ministry or agency forwards the sealed local purchase order, invoice, and other vouchers to the BGA, which then checks for conformity with code and other formalities. It also performs a physical audit when the goods and services are valued at US$5,000 or more. A BGA report including all the supporting documentation is forwarded to the Cash Management Committee (CMCo) secretariat. The CMCo secretariat prepares a list o f vouchers for payments to be approved by the CMCo, which includes the Ministry o f Finance, the Ministry o f Planning, the Ministry o f State, and an externally recruited financial controller financed by the donor community. The CMCo also receives from the Comptroller General's office details o f the government o f Liberia cash balances with the Central Bank. After reviewing the list o f vouchers for payment and consideringthe available cash balances, the CMCo prepares a list of approved payments. One copy of that list is forwarded to the Comptroller General for payments and distribution o f the checks through paymasters, and another copy i s forwarded to the Central Bank for verification. 2.45 The accounting system is incomplete. The system does not cover debt and contingent liabilities. As mentioned, transactions for investment expenditures financed by donors are not inthe budget and not recorded inthe accounting system. 2.46 At the Bureau of General Accounting the processing of transactions is principally manual, although some procedures have been computerized using 25There i s only one Chartered Public Sector Accountant in the Bureau o f General Accounting. There is an ongoing World Bank capacity building project within the Ministry o f Finance for about 60 students that are undertaking an MBA program. The first batch is expected to graduate in early 2009 and will be posted within MoF and other ministries and agencies. - 32 - standalone spreadsheets. This naturally presents inconveniences for information dissemination. The government uses a single-entry method o f accounting, and the accounting i s on a cash basis. So, transactions and events are only recognized once cash has been received or disbursed. 2.47 There is a need to introduce a funds accounting system into the general accounting system so the government can track and report the activity of individual funds. The funds (General Fund, Special Revenue Fund, Capital Project Fund, and so on) should be sparingly used and carefully controlled by creating special appropriation accounts that provide a record o ftransactions financed by means o f special resources. 2.48 There is a need for reform of the cash management system in the mediumterm. While the current arrangements have increased transparency and accountability in procurement, they represent a temporary measure designed as a part o f the Governance and Economic Management Assistance Program to re-impose fiscal discipline while longer-term financial institutions are re-established. The government may want to plan for a more streamlined budgetexecution process with fewer steps inthe payment process. Information Systems 2.49 The information systems currently in place are fairly rudimentary. Different units within the Ministry of Finance (MoF) maintain separate spreadsheets, making reconciliation o f fiscal data challenging and weakening the credibility of the government's financial reports. Financial reporting i s expected to improve when the Integrated Financial Management Information System is implemented and the BOBmerges fully with the MoF. 2.50 To improve the accounting system, the Ministry of Finance introduced the Liberia Expenditure Control and Accounting Program (developed in-house) and then transitioned to SunSystems, an interim data-capturing and reporting system. SunSystems does not provide a full accounting package and focuses primarily on facilitating financial reporting. It will be replaced by an Integrated Financial Management Information System (IFMIS). Implementation will begin with the core modules-budget execution and commitment control, and a general ledger that includes bank reconciliation and reporting. 2.51 SunSystems, and its predecessor, built a platform for implementing IFMIS by exposing key officialsto a computerizedenvironment. Use o fthese software programs has also improved the quality o f reporting. But extensive change management and capacity buildingwill have to be done to ensure that all stakeholders understand the objectives of the IFMIS implementation. The Deputy Minister for Expenditure has been designated as the IFMIS Champion and has established a number o f IFMIS implementation committees. In addition, a change management advisor has been hired to help facilitate the change process. Drawn from key institutions, the teams will cover applications users, application management and support, and information and communications technology support to form the Government Implementation Team. The team will be required to work closely with the IFMIS vendor to ensure transfer of skills and knowledge required to maintain the system independently in the long term. A remuneration policy to retain these core skills should be developed. A new chart o f accounts will needto be developed. -33 - FinancialReporting 2.52 The government does not produce financialstatements but has begun producing annualand quarterlyfiscal reportsthat detail revenues, allotments, and expenditures of the various ministries and agencies. Financial statements have not been produced since before the war, and the current accounting procedures limit the government's ability to fulfill reporting requirements. Current legislation requires financial statements to be based on accrual accounts, which is not realistic under present cash budgeting procedures. This has been addressed in the new Public Financial Management Act, currently with the legislature. Given the government's inability to produce full financial statements, its annual fiscal reports-produced for 2005/06, 2006/07, and 2007/08-and quarterly outturn reports mark an important step toward increased transparency and accountability. The reports have been produced ina timely manner and are published innewspapers and on the MoF website. 2.53 Section 21.2 (c) of the ExecutiveLaw 1972 requires the Minister of Finance "to report the financial activitiesand financial position of the government to the President and to the legislature." The new Public Financial Management Act stipulates that quarterly reports are to be submittedwithin 45 days o f the end o f the quarter, and a final account o f the national budget is to be submittedwithin three months o f the end o f the fiscal year. Existing legislation does not stipulate any deadlines for reporting. 2.54 Special funds, such as the County DevelopmentFundand departmentalaccounts transactions, are not disaggregated by economic or functional classification in fiscal reports. The County Development Fund was introduced in 2007/08 mainly to fund community-driven development projects. According to the guidelinesfor the management o f the funds, transfers should be made to the County Development Fund bank accounts from which project activities are executed. The nature o f these expenditures is not recorded in the fiscal reports by either economic or functional classifications-they are recorded solely as transfers. The MoF needs to develop procedures to capture these details when the counties start submittingreturns to the Bureau o f General Accounting. The chart o f accounts will also needthe requiredsegments for these entries to be passed. InternalAudit 2.55 The internal audit function in Liberia is weak. The role o f internal audit in management control responsibilities should be better defined. Currently, internal audit functions combine the role o f financial controller and financial police. The audit carries out both mandatory pre-controls before decisions are taken and special investigations ordered by ministersafter the fact. 2.56 The 1972 Act assigned responsibilityfor internal audit to the Auditor General. This arrangement was enacted while the Auditor General Office was within the executive branch, but since the General Auditing Commission is now under the legislative branch, this arrangement i s less appropriate. Internal audit i s conventionally required to function as an aid to management and to have the resources, capacity, and independenceto report directly to the chief executive o f the entity. The Public Financial Management Act before the legislature i s expected to provide a roadmap o f revised institutional arrangements for internal audit and secure a border between external and internal audit functions. 2.57 At present, the majority of ministries do not have functioning internal audit units. These units exist within the ministries o f Finance, State, Justice, General Service - 34 - Agency, Health and Social Welfare, and Lands, Mines, and Energy. The ministries of Public Works, Internal Affairs, and Education are settingup or revitalizing their internal audit units. None of these, however, have a clear mandateor reporting structure. The Ministry ofFinance internal audit unit has an operational manual that was issued in May 2004, but not widely used. The General Auditing Commission has drafted a memorandum of understandingwith the Institute of Internal Auditors26for the institute to partner indeveloping the internal audit fimction for the government. The memorandum of understanding proposes as a first step settingup ajoint working group to assess the needs, timing, and overall approachto building thispartnership.The goals are: Adopting and recognizing the Instituteof InternalAuditors' International Professional PracticesFramework as the authoritative guidance for internal auditors, thereby providing staff with globally-recognized standards by which to practice, and ensuring the independenceandobjectivity ofthe function. Helping establish an Institute ofInternal Auditors inthe country that canprovide sustainability for training and building capacity inboth the public and private sectors. Embracing professionalcompetencethrough a hands-ontraining, certification, and continuing educationprogram. 2.58 While it is useful that the General Auditing Commission is providing some guidance for establishing an independent internal audit function, the Ministry of Finance should take the responsibility for internal auditing under the executive. An internal audit strategy was recently completedby the Ministry of Finance, with input from the GeneralAuditing Commission. This strategy helps specify the roles and mandates of internal auditors in relation to internal control systems. It is further buttressedinthe proposedPublic Financial ManagementAct. Internal audit regulationswill need to be developedto implement the strategy. H. EXTERNAL OVERSIGHT The GeneralAuditingCommission 2.59 Over the last three years, the government has taken important steps to establish an externalaudit function.Prior to 2005, the Auditor General was required to report to the President. This called into question the independence of the Auditor General and compromised his ability to monitor the executive branch. And because the Auditor General reportedto the President, the legislaturehadno formal means of verifLing the executive's use of budgetary resources. In 2005, the arrangements were changed. The General Auditing Commission, headed by the Auditor General, was created as an independent, autonomous body of the government by amending section 53.2 of the Executive Law 1972. The General Auditing Commissionnow reportsto the legislature. 2.60 To strengthen the external audit function, capacity must be improved and the legislative framework refined. Legislative review is in progress, and the General Auditing 26The Institute o f Internal Auditors represents the global profession of internal auditing, which has over 135,000 members in 166 countries. Its members agree to practice internal auditing in accordance with The Institute o f InternalAuditors' Code of Ethics and the International Standards for the Professional Practice o f Internal Auditing. - 35 - Commission is working to strengthen the capacity of its national auditors through training programsandtechnical assistance. 2.61 A five-year strategic plan for the General Auditing Commission is being developed. The plan will provide a focus for further building the capacity of the institution and form the basis for a convergenceof support by donors andthe government. 2.62 The International Organization of Supreme Audit Institutionscode of ethics and auditing standards and the African Organization of English-speakingSupreme Audit Institutions audit manual and Quality Control Manual have been customized to local circumstances. These manualsand codes needtime to be fully implemented. 2.63 The General Auditing Commission is introducing a more modern auditing technique. The first materiality and risk assessment (risk index) on a national level has been implemented to focus on key government institutions. Through the assistance of a human resource and a legal consultant, the General Auditing Commission has developed a series of human resources guidelines and forms, including a code of conduct, staff confidentiality forms, asset declaration, and conflict of interest forms. 2.64 The Auditor General is addressingthe challengeof low human capacity. All staff from the GeneralAuditing Commissionwere laid off and required to reapply for theirjobs. A recruitment exercise was undertaken, with about 1,400 applicants. Nearly 1,000 applicants took a test in Monrovia and Gbarnga(Bong County) on generalauditing and accounting; one third were successful. The commission interviewed 220 people and selected 92 to begin a training program in late 2007. To conduct external audits in 2008, the General Auditing Commission has contractedexperiencedauditors from neighboring countriesto lead teams of Liberian auditors. By pairing Liberian auditors with more experienced auditors from other countries, the GeneralAuditing Commissionhopes to facilitate knowledgetransfer. 2.65 The GeneralAuditing Commission has recently completed a first set of external audits, includingan audit of the Ministry of Finance. Further audit reports are currently under preparation. Annual fiscal outturn reports were prepared for 2005/06, 2006/07, and 2007/08. The General Auditing Commission made use of the 2006/07 outturn report to conduct a more comprehensive and reliable transaction-based audit for 2006/07. In the medium term, it will be necessary to introduce an Annual Attestation Audit to match the PFM development in Liberia. These efforts will need to be sustained and consolidatedto institutionalize annualaudits. LegislativeScrutinyof the Annual Budget Law 2.66 Legislative scrutiny of annual budget law and financial statements is important to ensure transparency and accountability. The Ways and Means Committee in the legislature is responsible for scrutinizing the annual budget when it comes,beforethe House of Representatives. However, the Ways and Means Committee of the House and the Senate lacks technical in-housecapacity to analyzethe budget and guide a meaningful debate. When the 2007/08 budget was submitted by the executive, the legislature asked the General Auditing Commissionto review and provide technical comments. 2.67 Timely and complete annual reports of the financial status and affairs of all government agencies are required to be submitted to the Presidentand the legislature pursuantto Section 2205 Chapter 22 of the Revenue Code of Liberia (Act of 2000). The - 36 - executive is also expected to submit the mid-year review o f the budget to the finance committee o f the legislature. At present, annual and quarterly fiscal reports are made publicly available on a timely basis, although the quality o f these reports needs improvement, and the legislature's capacity to meaningfidly scrutinize the reports is limited. 2.68 An act limitingthe sue of in-year budget transfers by the executive to 20 percent between ministriesand agencies has recently been passed. But the budget document does not include information on spending against the budget in previous years. So, when assessing the budget, the legislature cannot confirm that expenditure inpast years remained within the permitted limits. LegislativeScrutinyof ExternalAudit Reports 2.69 Timely consideration of audit reports from supreme audit institutions and follow-up on previous recommendations are important to hold the executive accountable. The executive is expected to report to the legislature in the form o f an annual financial statement accompanied by an audit opinion from the General Auditing Commission. The Ministry o f Finance is not producing financial statements but has published annual and quarterly fiscal reports, which are available to the legislature. The General Auditing Commission recently submitted a first set o f audit reports to the legislature, based on the MinistryofFinance fiscal reports. I. DEBTMANAGEMENT 2.70 Liberia's debt management capacity has been severely eroded by the decade and a half of civil conflict.The Liberian government has neither the technical personnel nor the systems to adequately record, analyze, and report on domestic and external debt. Consequently, Liberia's debt data records and reporting are incomplete, and information on commercial debt stock is based partly on creditors' claims. As part of its debt resolution strategy, the government-with support from the World Bank and the Swiss Agency for Development and Cooperation-hired financial and legal advisors to help establish a preliminary inventory o f claims from external commercial creditors on Liberia. According to these estimates, Liberia's stock o f external commercial claims at end-June 2007 was about US$1.575 billion. 2.71 Since December 2007, Liberia has cleared substantial arrears to multilateral creditors (including the IMF, the World Bank, and the African Development Bank), which paved the way for reaching the Heavily Indebted Poor Countries Initiative decision point in March 2008.27At that time, Liberia qualified for US$2.8 billion o f debt relief innet present value terms.28Subsequently, inApril 2008, Liberiareached an agreement with its Paris Club creditors under Cologne Terms that cleared Liberia's arrears toward these 27Arrears to the World Bank were cleared in December 2007 through a bridge loan that was repaid with the proceeds of a grant financed with an exceptional IDA allocation. Arrears to the African Development Bank Group were cleared in December 2007 through an operation under the post-conflict countries framework. Arrears to the IMF were cleared in March2008 through a bridge loan that was repaid with new financing from the Poverty Reduction and Growth Facility and Extended Fund Facility. 28Such assistance is equivalent to a common reduction factor o f 90.5 percent, the second largest common reduction factor thus far under the HIPC Initiative. See "Liberia. Enhanced Heavily Indebted Poor Countries (HIPC) Debt Initiative: President's Memorandum and Recommendation and Decision Point Document," February 27,2008, IDAR2008-0038. - 37 - creditors. In addition, on an exceptional basis, the Paris Club creditors granted Liberia a deferral o f all payments falling due untilend-December 2010. 2.72 The government has begun implementing its domestic debt resolution strategy, finalized in January 2007. The strategy addresses the approximately US$300 million o f valid domestic claims, including payments to the Central Bank o f Liberia and salary arrears to civil servants. Given the extremely limited fiscal space in the face o f the significant development needs (including for investment in infrastructure and delivery o f key social services), it i s expected that repayment o f the domestic debt will be stretched out over 30 years. 2.73 In June 2008 the government formed a comprehensive debt management strategy. The strategy is focused on three key objectives: reducing and rationalizing Liberia's external debt by end-2009; rebuilding the institutional capacity to monitor debt levels, analyze debt sustainability and risks, and service debt in a timely manner; and building the foundation for Liberia to begin contracting limited amounts o f highly concessional targeted finance to support its development programs and over the long-term re-entering international financial markets. The government has established a debt management unit within the Ministry of Finance and is training personnel and acquiring a modern debt-management system. The debt management responsibilities o f the MoF should be clarified and software tools for debt management, such as the Commonwealth Secretariat Debt Recording and Management System, should be introduced. J. DONOR PRACTICES 2.74 Information on donor funded projects, as stated earlier, is not included in the budget. Indeed, there is no database o f all aid flows into the country. The absence o f a project database impairs efficient allocation o f limited resources, since programs that may already have been funded by direct donor support could be duplicated by the government or by another donor. It is crucial to have a database o fall projects to be usedby the government and donors to improve coordination o f their interventions. The Ministry o f Finance has established an Aid Management Unit to improve reporting o f aid flows and to track alignment with the Poverty Reduction Strategy. The Aid Management Unit, in collaboration with the Liberian Reconstructionand Development Committee and the Ministry o f Planning and Economic Affairs, has begun to develop a database for aid management. As part o f the budget preparation process, ministries and agencies should provide information on donor fundedprojects executed through them. 2.75 While aid-tracking mechanismsare instituted,the MoF should request donors to provide information on funds for government-executed projects committed by sector for the next three years, as well as the amount disbursed to date. The MoF should work with the executing agencies to ensure that financial returns are submitted and reconciled against the project bank accounts. 2.76 The central Project FinancialManagement Unit in the MoF could be used for a greater volume of donor-fundedprojects.This will ensure standardization o f management o f all project funds and aid the eventual transition from direct donor support to donor funding going through the budget, as the Paris Declaration recommends. There have been some improvements inthis direction. The majority o f World Bank and African Development Bank funds are now managed bythe Project FinancialManagement Unit, and the education pooled fundbeing establishedmay use the Unitfor financial management. - 38 - K. CONCLUSIONSAND RECOMMENDATIONS 2.77 The public financial management system in Liberia has improved markedly since 2006, but it is in need of further reform and strengthening. The legal and regulatory framework for public financial management should be revised to clarify the roles and responsibilities o f different public agencies, establish internal controls, institutionalize the internal audit function, and formalize reporting requirements. A public financial management act before the legislature is expected to rectify the deficiencies inexisting legislation. 2.78 A budget document is needed that provides a comprehensive overview of the government's fiscal operations. This would include: 0 Information on the prior year's budget outturn (presented in the same format as the budget proposal). 0 Data on boththe debt stock and financial assets, with details for at least the beginning o fthe budget year. e Consideration o f the implications o f new policy initiatives and estimates o f the budgetary impact o f major changes inpolicy or expenditure programs. 2.79 While the government would like to move toward a medium-term expenditure framework, it is advised to gradually phase in elements of such a framework. As a part o f the Poverty Reduction Strategy, a costing exercise was undertaken. This important first step toward medium-term budgeting will help guide and ensure policy-based budgeting over the course o f the Poverty Reduction Strategy period. The government is also developing a medium-term fiscal framework, which will help provide broad parameters for budgetary planning. The government should also ensure that: 0 The macroeconomic analysis capacity withinthe MoF i s strengthened. 0 Strategies for key sectors are produced, medium-term objectives are formulated, and related expenditure plans are produced. 0 There is a strong link between budgeted funds and the results as defined by measurable indicators. 0 A system is introducedto measure performance. 2.80 The Liberian government should set up a system for selecting and managing public investments.An institutional arrangement is needed for a first-level screening o f all project proposals to ensure that they meet minimum criteria o f consistency with the government's strategic goals. Upstream investment in training for project-evaluation techniques i s important for a public investment system. There should be consideration o f the fundingneededto maintain and operate assets created through public investment projects. An ex-post evaluation o f completed projects is also desirable to ensure that there i s learning and feedback from projects that will create a positive dynamic for improvement over time. 2.81 The personnel databases in the Civil Service Agency and the Electronic Data Processing in the Bureau of General Accounting should be rationalized. This will produce more accurate personnel expenditure projections and promote more consistent classifications. A biometric identification system to verify all civil servants on the payroll and - 39 - a Human Resource Management Information System will also be important for developing the IntegratedFinancialManagementInformation System. 2.82 Tax administrationneedsto be improved.Amounts collected should correspondto actual recovery potential of the departments in charge of revenue collection. Collecting tax payments should be more effective, and fake receipts should be addressed by tighter internal controls. 2.83 The momentum from passing the Public Procurement and Concessions Act should be used. This will require a comprehensive set of implementing regulations and manuals. Procuring entities should be provided with detailed instructions and increased capacity on the correct application of the law. Efforts to strengthen the ability of private sector operators to meet the requirements of the Act are also needed, including by providing training and reducing the cost of bidding. Efforts are also neededto improve the scope and quality ofpublic procurementreporting. 2.84 The cash managementsystem should be reformed. While the current arrangements have re-imposedfiscal discipline and increasedtransparencyand accountability inthe budget process, they represent a temporary measure designed as a part of the Governance and Economic Management Assistance Program while longer-term financial institutions are re- established. The government may want to consider planning ahead for a more streamlined budget executionprocess inthe future. 2.85 The benefits brought about by the introduction of the Liberia Expenditure Control and Accounting Program and Sunsystems should be expanded by the timely introduction of an Integrated Financial Management Information System. Implementation should focus on a simple and basic systemthat will not sap scarce resources, by beginningwith implementing a few core modules. 2.86 The internal audit function needs to be developed further. There is a need for a systematic and disciplined approachto evaluating risk management, control, and governance processes. The Public Financial ManagementAct being drafted would provide a roadmap of revised institutional arrangements for internal audit and secure a border between internal and external audit functions. 2.87 The externaloversight functionis weak but is undergoingsignificant reform and capacity building. Extensive reform and capacity building efforts are under way at the General Auditing Commission, and audit reports are expected in the current fiscal year. These efforts should be sustained and consolidated to ensure that annual audits are institutionalized. Training and support to the legislature is requiredto enable it to perform its oversightrole by reviewing financial statements and audit reports. 2.88 The debt management departmentof the Ministry of Financerequiresadditional capacity and tools. As the Heavily Indebted Poor Countries process is unfolding, the government has begun to deal with an unsustainabledebt burden (and 20 years of default). But there is also a need to build .capacity within the MoF to institute regular debt- managementpractices and ensure that the debt burden is analyzedduring budget preparation. Legislation is needed to specify the debt-management responsibilities of the MoF. Debt software tools are neededincombination with capacity building of the debt managementstaff to analyzethe debt and producehigh quality reports. - 40 - 2.89 A complete overview of Liberia's public finances should include information on donors' activities. A project database, detailing all donor-funded projects, should be developed. This would require the ministries to declare all direct-donor funding for government-implementedprojects and donors to report all funding they provide in Liberia. The central Project Financial Management Unit in the MoF could be used for a greater volume of donor-funded projects and thus ensure standardization of all project hnds management. -41 - 3. PUBLIC EXPENDITUREANALYSIS A. INTRODUCTION 3.1 Inthis chapter we examine the reliability of the budget as an instrument for the government's policy objectives, by analyzing expenditure outturns in relation to the original budget and how strategic and efficient resource allocation is. We also discuss donor resources and their contribution to overall public service delivery. We assess operational efficiency in aggregate terms across the government budget, before reviewing operational arrangements from a sectoral perspective in Chapter 4. We also consider the economic composition of expenditure, with an emphasis on the balance between investment and recurrent expenditure. Overall RevenueTrends 3.2 Revenues have recovered very strongly since 2002103, growing by 263 percent in nominal terms over the past five years from US$57 million in 2002103 to US$207 million in 2007/08. This represents an average annual real revenue growth rate o f 21 percent and translates into a significant increase in the ratio o f revenue to GDP. Revenues were 12.7 percent o f GDP in 2002/03 as a result o f the low level o f revenue, the very low inflows o f grant aid to the budget,29 and the government's inability to borrow because o f its unsustainable external debt burden and the very shallow domestic financial market. Revenues have since grown to about 25 percent o f GDP in 2007/08 as a result o f measures taken to improve tax administration by expanding pre-shipment inspection o f imports, abolishing noncash payment of taxes, and enforcing revenue centralization. The recent recovery o f revenues has taken place even though Liberia has yet to begin collecting revenue from the forestry sector-due to the United Nations ban on timber exports, which was lifted first in 2006. Figure 3.1: Revenueand Expenditure Growth, 2002103-2007108 200- .,.+ --25.0 -- 5.0 Source: Ministry o f Finance, Annual Fiscal Reports, 2004105-2007lO8; IMF, Article IV Consultations, 2008. ExpenditureTrends 3.3 Expenditure in Liberia has risen rapidly since 2004105, as the government's revenue base has recovered. Expenditure grew by an average o f 23 percent a year over the 29Most aid is disbursed outside o f the budget. - 42 - past five years, from US$69 million in 2002/03 to US$203 million in 2007/08. Liberia i s operating under a cash budget policy, where borrowing to finance the budget deficit is not permitted. The government ran small surpluses in 2005/06 (0.6 percent o f GDP), 2006/07 (2.0 percent o f GDP), and 2007/08 (2.4 percent o f GDP). 3.4 Expenditure has not been evenly paced over the course of the budgetyear, which suggests that improved planning could increase expenditure efficiency. The rate o f expenditure has been low in the beginning o f each fiscal year and rapid toward the end. In 2006/07, expenditure in the last quarter was almost four times higher than expenditure inthe first quarter (figure 3.2). This i s a result o f many factors, including late passage o fthe budget, which prevented any spending in July and August; the constraints o f cash-based budgeting, which allow for large capital purchases to be made only toward the end o f the year when sufficient cash balances have accumulated; under-projection o f revenues, which has resulted inpassing supplemental budgets inthe secondhalfofthe fiscal year; andcapacity constraints within the ministriesand agencies to comply with new procurement requirements and Cash Management Committee controls. When these controls were first introduced it was not unusual for 50 percent o f vouchers to be returned to the ministries due to errors or lack o f supporting documentation. So, there i s room to increase expenditure efficiency through improved procurement planning, improved sequencing o f activities, and an eventual transition away from a cash-based budget. Figure3.2: Quarterly Revenue and ExpenditurePatterns,2006107-2007/08 FY 07/08 QI QII Qlll QN _~ ual Revenue Ependiture 10 Rewnue projection W Actual rewnue Expenditurei Source: Ministry o f Finance, Annual Fiscal Reports, 2006107-2007108. 3.5 Budget performance in the past two fiscal years shows some discrepancies betyeen the original budget and expenditure. Expenditure in the general administrative services category exceeded the original budgets by 13 percent in 2007/08 and 33 percent in 2006/07. Its share o f total expenditure was consistently underestimated in budget estimates. Conversely, transfers to other institutions and general claims were significantly lower than budgeted. The large budget variances are partially explained by the passage of supplemental budgets in 2006/07 and 2007/08, which influenced allocation patterns (the revised budget estimates are included intable 3.1). But budget variance also results from transfers between budget lines, after review and approval by the Budget Committee. In2008 an Act was passed to limit to 20 percent in-year budget transfers between ministries and agencies without legislative approval. Prior to this law, there were no legal restrictions on in-year budget transfers. - 43 - Table 3.1: Variation between Original Budgetand Expenditure,2006/07-2007/08 (functionalclassification) Source: Ministry ofFinance, Annual Fiscal Reports, 2006107-2007108; National Budget 2006107-2007108. 3.6 Variation analysis by economic classification shows modest variations between budgeted and actual expenditure in 2007/08. Expenditure on personnel and goods and services exceededthe budgeted amounts by 7 and 11percent, respectively, while debt service payments were significantly lower than planned. In. 2006/07 there were significant discrepancies between the budget and final reported outturn, with expenditure on personnel falling short o f the budget by 27 percent, while transfers and subsidies exceeded the budget by 19 percent. Butthis may be a result of poor reporting. The annual fiscal outturnreport for 2006/07 includes a column titled "other," which covers US$34 million o f expenditures not yet reconciled at the time o freport completion. Table 3.2: Variation between Original Budget and Expenditure,2006/07-2007/08 (economic classification) Source: Ministryo f Finance, Annual Fiscal Reports, 2006107-2007108; National Budget 2006107- 2007108. Donor Expenditure 3.7 An analysis of public expenditure patternsin Liberia is incompletewithout some understanding of donor-financed expenditure patterns. Given Liberia's low implementation and financial management capacity, donors have been hesitant to provide budget support. Most financing has been project-based and executed by nongovernmental organizations, private contractors, and special Project Implementation Units within line - 44 - ministries. Availability o f data on aid flows is poor, and only estimates o f total donor- financed expenditure are available. 3.8 Total official developmentassistancewas estimated at US$357 million for 2007;' or more than twice the government's expenditure in the same period. Liberia's official development assistance to GDP ratio is one o f the highest in the world. This excludes assistance fi-om the U.S. Department o f Defense o f almost US$60 million annually year and the United Nations Mission in Liberia's peacekeeping operations. In addition to official development assistance, Liberia receives significant support for the social sectors from private foundations, including the Soros Foundation, the William J. Clinton Foundation, and the Bill & Melinda Gates Foundation. The largest donor is the United States government, which provided almost US$lOO million in 2007. The combined contribution o f the United Nations agencies amounted to almost US$90 million, and the World Bank's contribution o f about US$40 million represented the third largest contribution. Given the ratio o f external assistance to public resources, it will be difficult for the government to improve allocative efficiency and ensure alignment with the Poverty Reduction Strategy unless improved and disaggregated aid data i s collected and analyzed. B. ALLOCATION OFRESOURCES 3.9 Government expenditures can be classified under six functional classification headings. These headings are: General administrative services: public administration, management o f the economy, administrationo f internal affairs and external relations, and national legislature. Security services: justice and law enforcement, safety and national security expenditures. Social and communily services: government services such as education, health, youth affairs, gender, and public works. Economic services: spending to support activities such as agriculture, forestry, commerce, mines, industry, telecommunications, and energy. Institutions and entities: transfers to commissions, agencies, and state-owned enterprises. Other general claims: debt payments and transfers to other organs o f government. 31 30 Paris Declaration Survey 2007. 31 This category includes the County Development Fundallocations. - 45 - Figure 3.3: Share of Total Government Expenditure by Functional Classification, 2004105-2007108 I Relative Share of Categoriesof Expenditure 2004-07 100% 80% Debt serwces 60% Others entities $aEconomicseruces m Social and Community sewces 40% Secunty serwces 20% 0% 2004/05 2005/06 2006/07 2007/08 Source: Estimates based on Ministry o f Finance Annual Fiscal Reports. Table 3.3: Functional Classification of Public Expenditure as a Percentage of Total Expenditure, 2004105-2007108 Source: Estimates based on Ministry o f Finance Annual Fiscal Reports. Table 3.4: Functional Classification of Public Expenditure in Percentage of GDP, 2004105-2007108 Source:Estimates based on Ministry o f Finance Fiscal Reports. 3.10 Spending on general administrative services has made up at least one-third of government expenditures for 2004/05-2007/08, but fell modestly in 2007/08. The category represented 6 percent o f GDP in 2004/05, 3.2 percent in 2005/06, 7.1 percent in - 46 - 2006/07, and 7.7 in2007/08. Certain items within this category have risen sharply, including expenditure on the legislature, which rose from 5.5 percent o f total expenditure in2005/06 to 7.7 percent in2007/08. 3.11 Security services expenditures have steadily decreased, from 5.6 percentof GDP in2004/05 to 3.7 percentin 2007/08.As a proportion o ftotal government expenditures, they declined from 34 percent in 2004/05 to 14 percent in 2007/08. But donor contributions mask the true volume o f security expenditure. Defense expenditure was drastically reduced to 14 percent o f total expenditure in 2007/08, but most o f the costs for armed forces are financed by the United Statesgovernment and not channeled through the national budget. 3.12 Socialand communityservices, mainly education and health, have beenthe main beneficiaries of a remarkable shift in government spending, increasing as a share of total expenditure from 18.1 percent in 2004/05 to 27.4 percent in 2007/08. The increase reflects the government's emphasis on human resources development and represents an alignment with Poverty Reduction Strategy objectives. In addition, a large proportion o f donor aid is channeled to these sectors, which i s not reflected inthe national budget. 3.13 Despite the emphasis that the Poverty Reduction Strategy and the interim strategy have put on growth and job creation for poverty reduction, economic services receivedonly a modestshare of government spendingbetween200447. Spending onthis sector fluctuated between 4.4 and 7.8 percent of total expenditures. Expenditurespeaked at 1.8 percent o f GDP in 2007/08. Expenditure on agriculture, which is listed as a Poverty Reduction Strategy priority, rose from 0.9 percent o f total expenditure in 2004/05, to 2.6 percent in 2006/07, before falling to 1.8 percent in 2007/08. Meanwhile, the expenditure on institutions, entities, and other general claims increased as a percentage of total expenditure from 2.1 percent in 2004/05 to 15.9 percent in 2007/08 as a result o f increased debt service payments and the inclusion o f County Development Funds (transferred to county authorities) under this rubric. (For more detaileddata see annex 1, table A.l.) OperationalEfficiency 3.14, Expenditures can be classified as either recurrent or capital expenditure. Recurrent expenditure includes wages and salaries, other goods and services (including non-wage operation and maintenance), interest payments, and subsidies and other current transfers. Capital expenditure covers payment for the purchase or production o f new or existing durable goods.32 3.15 Overall, government expenditure has become progressively more efficient in the post-war period, with a decreasing share of expenditure on wages and salaries and an increasing share of expenditure on goods and services. The rapid rise in revenue enabled the government to increase productive expenditure without having to reduce expenditure on wages, salaries, and other administrative costs. 32Pradhan 1996; World Bank 2005. - 47 - Figure 3.4: Share of Total Government Expenditure by Economic Classification 2004105-2007/08 100% 90% 80% 70% II 60% E Debt sewes I 50% j o Subsidies and Transfers I 40% Goods and serwces Wages and salanes 30% 20% 10% 0% 2004/2005 2005/2006 2006/2007 2007/2008 Source: Estimates based on Ministryo fFinance Annual Fiscal Report. Table 3.5: Economic Classification of Public Expenditure, as a Percentage of Total Expenditure, 2004/05-2007/08 Source: Estimates based on Ministryo f Finance Annual Fiscal Reports. Table 3.6: Economic Classification of Public Expenditure, 2004/05-2007/08 (percent of GDP) 1 Source: Estimates based on Ministryo f Finance Annual Fiscal Reports. 3.16 The centra1government's expenditureson wages and salaries declined from 55.4 percentof total spending in 2004/05 to 31.8 percent in 2006/07 and rose to 35.7 percent in 2007/08. The change was facilitated by increased revenues. It also reflected improved - 48 - government control over the wage bill and gains from efforts to remove "ghostsy7from the payroll. A rising share of wages and salaries in 2007/08 reflects the salary increases granted to civil servants inresponseto the extremely low wage structure. 3.17 The number of employees has been reducedby almost 30 percent,from 49,856 to 34,995,33 between January 2006 and September 2007. Monthly pay for the lowest paid civil servants has risen fiom a base o f US$15 in 2005/06 to US$30 in 2006/07, US$55 in 2007/08, and US$70 in 2008/09. While this falls short o f a living wage and i s clearly insufficient to attract qualified technical staff,34 it has signaled the direction o f the government's civil service reform strategy. As the government is committed to salary increases and deconcentration o f wages, personnel expenditure is likely to remain constant, or increase modestly, as a proportion o f total expenditure in the medium term. The Civil Service Reform Strategy (2008) expresses the government's intention to raise wages in line with available resources, thereby maintaining fiscal balance. 3.18 Expenditure on goods and services increasedas a proportion of total expenditure from 21 percent in 2004/05 to 37.5 percentin 2006/07 and fell to 32.7 percentin 2007/08. This expenditure category declined from 3.5 percent o f GDP in 2004/05 to 2.9 percent in 2005/06 before increasing significantly to 7.2 percent in 2006/07 and 8.6 in 2007/08. Expenditure on goods and services is driven by the need to make available a minimum o f supplies (office goods, computers, furniture, and so on) to enable public administrators to provide basic services. Despite this significant increase, though, spending on goods and services i s still insufficient to meet the needs o f the line ministries, and supply shortages are common. 3.19 Transfers and subsidies to public enterprises, commissions, local governments, and other recipients increasedas a percentage of total expenditure from 7.8 percent in 2004/05 to 15.9 percent in 2007/08. This represents an increase from 1.3 percent o f GDP in 2004/05 to 4.1 percent in 2007/08. This is in part due to new commissions that receive monthly transfers and a rise in direct transfers to ministries and agencies for operating expenses. Transfers and subsidies were highat US$32 million in2007/08, as the government finances a large number o f permanent and transitional commissions with specific functions (Truthand Reconciliation Commission, Governance Commission, and so on) and subsidizes some agencies and state-owned enterprises that are not yet generating revenue. The efficiency of this expenditure i s hard to assess because transfers and subsidies can cover any type o f expenditure (salaries, goods and services, or capital expenditure). 3.20 Interest on public debt increasedas Liberia began clearing domestic arrears. As a proportion o f total spending, this expenditure category has increased fiom 1.8 percent in 2004/05 to 5.5 percent in 2007/08. This is due to arrears payments to civil servants and the government's decision to progressively normalize financial relations with the private sector and external official creditors. 3.21 Capital expenditure was 2.3 percent of GDP in 2004/05 and was only slightly higher at 2.7 percent in 2007/08. Butthere has beena reallocation o fresources toward more growth-supporting spending, compared to previous years when capital expenditure was 33Number ofpersonnel on Government of Liberia payroll (Liberia Civil Service Agency 2007). 34Low-skilled workers in the nongovernmental organization sector are paid about US$200 a month (World Bank, ESES 2007). - 49 - mostly security expenditure for war and transition. Additionally, much o f the nation's public investment is donor-financed outside o fthe national budget. c. CONCLUSIONAND RECOMMENDATIONS 3.22 Budgeted expenditures have risen rapidly since 2004/05 as the government's revenue base has recovered. The positive effects o f increased public resources are evident across sectors, as the functions of the state are gradually re-established after 14 years o f civil conflict. While government revenue remains insufficient to meet Liberia's reconstruction needs and the country remains heavily dependent on donor resources, the economic and fiscal recovery is strengthening the government's coordinating and regulatory roles and improving basic social services. 3.23 The government ran small cash surpluses in 2005/06,2006/07, and 2007/08, and expenditure has been unevenly paced. Expenditure has been low inthe beginningo f each fiscal year and rapid toward the end o f the year for a number of reasons, including the constraints o f cash-based budgeting, which allow for large capital purchasesto be made only toward the end o f the year when sufficient cash balances have accumulated; under-projection o f revenues, which has resulted inthe passing o f supplemental budgets inthe second half o f the fiscal year; and capacity constraints, particularly in previous years, within the ministries and agencies to comply with new procurement requirements and Cash Management Committee controls. Thus, there may be room to increase the expenditure efficiency with better procurement planning, improved sequencing o f activities, and an eventual transition away from a cash-based budget. In 2005106, 2006107, and 2007/08 the original budget estimates deviated significantly from final expenditure, largely due to mid-year budget revisions and the passing o f supplemental budgets. These discrepancies decrease the predictability o f the budget and may increase greater discretion in budgetary allocations, as mid-year revisions and supplemental budgets usually undergo less thorough public scrutiny than the annual budgetingexercise. 3.24 Major shifts have taken place in allocating expenditures classified by function between 2004/05 and 2007/08, reflecting increased attention to Interim Poverty Reduction Strategy and Poverty ReductionStrategy priorities. The main beneficiaries o f this shift are social and community services (of which education, health, and social welfare are the largest sectors), whose share o f the budget rose from 18 percent in 2004105 to 27 percent in 2007108. Debt servicing has also received a larger share o f the budget, rising from 1.8 to 6.5 percent o f total expenditure over the same period. In contrast, administrative and security services saw their shares o f the budget fall, from 36 percent to 29 percent and from 34 percent to 14 percent, respectively. The share o f economic services fell from 36 percent to 29 percent. 3.25 There have also been positive shifts in the composition of expenditures by economic classification. The share of wages and salaries in the budget declined from 55 percent in 2004/05 to 36 percent in the 2007/08 budget, while that o f goods and services increased from 21 to 33 percent. The reduction in the share o f the budget allocated to the wage bill was partly made possible by purging ghost workers from the payroll. The number o f employees declined by nearly 30 percent since 2006. As detailed in the Civil Service Reform Strategy (2008), the government intends on a gradual pay and pension reform, which will include raising salaries and decompressing the salary scale. While the pay reform will be based on availability o f public resources, it is likely to exceed the gains made from cleaning the payroll. Thus the budgetary personnel category is unlikelyto fall further. The budget has - 50 - also seen a reduction in investment expenditure, which fell from 14 percent to 10 percent. While this is partly due to donors contributing most o f the public investment expenditure outside o f the national budget, the government's spending on capital investment will have to rise substantially ifit intends to contribute to rebuilding infrastructure. 3.26 Liberia's public expenditure patterns cannot be fully assessed without considering donor assistance. Total official development assistance was estimated at US$357 million for 2007,35 or more than twice the government's expenditure in the same period. This excludes assistance from the US. Department o f Defense o f almost US$60 million annually year and the United Nations Mission in Liberia peacekeeping operations. The government should begin collecting disaggregated data on donor flows to improve allocative efficiency and ensure alignment with the Poverty Reduction Strategy. 3.27 While Liberia's post-war progress to date is commendable, ensuring that additional resourcesare devotedto Poverty Reduction Strategy policy objectiveswill be important, as the resource envelope continues to grow in the medium term. As seen in many resource-rich countries, there are risks that increased public resources will fuel greater public consumption at the center, to the benefit o f the highest income groups. Increased budget monitoring and analysis will need to be performed to ensure that the budget is contributing to stated objectives. Specifically, administrative and personnel expenditure should be aligned with the Civil Service Reform Strategy; resources should shift to investment expenditure, particularly in infrastructure, and to support operation and maintenance o f new physical investments; and greater attention should be given to economic sectorswith large employment potential, such as agriculture, which i s underfunded. These are some o f the top priorities reflected in the Poverty Reduction Strategy, which focus on economic sectors under Pillar 2: economic revitalization, Pillar 3: governance and rule of law, and Pillar 4: infrastructure and basic services. In addition, improvements in data availability, reporting, and dissemination will be crucial for holdings elected officials accountable by monitoring that the use o f public financing i s aligned with the Poverty Reduction Strategy. 35Paris Declaration Survey 2007. - 5 1 - 4. SECTOREXPENDITUREREVIEWS 4.1 This chapter presentsan analysis ofthe levels,trends, efficiency,andeffectiveness of public expendituresinthe healthand educationsectors inLiberia.The chapter examines the status o f policies and strategies in each sector, service delivery performance, allocative and operational efficiency o f expenditures, and geographical and gender disparities. A. HEALTH SECTOR EXPENDITURES Introduction 4.2 This section examines the distribution and management of health expenditures in Liberia.It assesses the adequacy of aggregate fundingfor the health sector inLiberia and trends over recent years; analyzes the distribution o f health expenditures and alignment with health sector priorities, inorder to identify scope for efficiency gains; and examines the equity o f health expenditures and provides recommendations for improving their distribution to help promote broad-based poverty reduction. 4.3 While Liberia's healthcare system shares many features with those of other least developed countries, the country's post-conflict context poses additional challenges. A public expenditure review during this phase o f Liberia's post-war reconstruction i s to inform the public policy debate as Liberia reforms institutions and public financial-management systems. It also sheds light on the important role o f donor funding for basic services and thereby highlights the importance o f managing and coordinating external resources efficiently to maximize sustainable health outcomes duringthis critical phase. 4.4 Public expenditure reviews at this stage in Liberia's post-conflict development process are significantly constrainedby the limited availability and quality of data in the health sector. Results from the Liberia Demographic and Health Survey36and Core Welfare Indicator Q~estionnaire~~provide new and improved data on health status and access to health services. But they must be treated as preliminary, as they are still beingdiscussed and reconciled against earlier data. Similarly, sector expenditure data i s poor. Analysis is difficult becauseo f the large volume o f external assistance, most o f which is channeled outside o f the budget and is poorly recorded. Expenditure analysis includes 2007/08 fiscal data but i s focused on 2006/07 as the primary fiscal year o f analysis, as detailed data for 2007/08 was not yet available at the time o f completion o fthis study. Objectivesof the Health System 4.5 The health sector is included under the Infrastructureand Basic Services pillar of the Poverty Reduction Strategy (and previously under the interim strategy). The strategy strongly commits to the provision o f health services. A National Health Policy for 2007-2011 is beingimplemented, which focuses on expanding access to a basic package o fhealthcare services through investments in infrastructures, human resources, and decentralized management. The 36LiberiaDemographic and Health Survey 2006/07. 37CWIQ 2007. - 52 - government aims to ensure access to a basic package o f health services in 70 percent o f all clinics by 2010. The basic package standardizes prevention and treatment services throughout the health system to ensure that all individuals receive the same package o f care. 4.6 The basicpackage of healthservices is a minimumpackage to be made availableas an integratedwhole. Rather than an assortment of vertical and parallel programs, it focuses on six national health priorities: maternal and newborn care, child health, reproductive and adolescent health, communicable disease control, mental health, and emergency care. For the duration o f the National Health Policy period (2007-2011), the government will suspend user fees at public healthcare facilities to ensure universal access to the package. The government also aims to decentralize the management o f healthcare provision to the county level. Health Service Delivery 4.7 The central government's contributionto the health sector is split among several institutions: the Ministry of Health and Social Welfare (MOHSW), the John F. Kennedy Medical Center, the Institute o f Biomedical Research, and the Phebe Hospital. The Phebe Hospital has only recently received its own allocation, though it had received transfers from the MOHWS in the past. All institutions receive monthly transfers directly from the Ministry o f Finance, and liquidate the funds at the end o f each month. The MOHSW provides policy and strategy oversight to these institutions, but not financial oversight. 4.8 MOHSW's departments include: health services; planning, research and development; social welfare; and administration.Within these departments, many divisions, as well as hospitals and health centers, receive monthly transfers from the MOHSW, which they liquidate by presenting receipts o f expenditure. This liquidation process has been problematic, so disbursement o fthese transfers i s often stalled. 4.9 The government is only one among many public healthcare providers in Liberia. The majority o f health expenditure is channeled through nongovernmental and other organizations, which provide direct support to government clinics and hospitals. Government oversight o f these nongovernmental organizations and related programs i s weak, and no consistent financial reporting to the MOHSW takes place-although coordination efforts are improving. 4.10 MOHSW is developingan Office of FinancialManagement(with support from the Departmentfor InternationalDevelopment),which is addressingaccountingand financial control. It is also rebuildingthe resource mobilization capacity inthe External Aid Unit within the Department o f Planning, Research, and Development. A pooled fund for health has been established, to be managed by the Office o f Financial Management, and it is hoped that donors will increasingly channel their support to the health sector through the pooled fund. Through these measures, the MOHSW aims to better oversee and coordinate health financing and thereby better allocate resourcestoward the basic package o f health services. Health Status 4.11 Liberia's population is estimated at 3.5 million, with an annual growth rate of 2.1 percent.Average life expectancy at birthis estimated by WHO (2006) at 42 years44 years for - 53 - women and 39 years for men. Liberia's health status remains poor in regional comparison, and the country is unlikely to reach many o f the Millennium Development Goals due to the deteriorationinhealth status during the 14 years of civil war. But there have beenimprovements inhealth status since the end of the war. Total fertility rates have dropped significantly from an estimated 6.6 births per woman in 1986 to 5.2 in 2007. Yet considerable differences remain between fertility rates inrural areas (6.2 births) andurban areas (3.8 births).The use o f modern family planningmethods among women remains low at 11.4 percent.38 4.12 Based on data from the 2007 Liberia Demographic and Health Survey, the infant mortality rate is estimatedat 72/1,000 live births, comparedto 117/1,000 in 1999/2000. This i s below the Sub-Saharan Africa average o f 102/1,000 live births. The under-five mortality rate decreasedfrom 194to 111per 1,000 live births, which compares favorably with the Sub-Saharan Africa average o f 171/1,000 live births, althoughthe methodology usedto estimate the mortality rate was different in 1999/2000 and 2007. These figures should be interpreted with caution, however, because o f the highlevel of displacement or heaping of births in2000. Child nutrition is poor. an estimated 19 percent o f children under-five years are undernourished, 7 percent are wasted, and 39 percent are stunted.39 4.13 Malaria, acute respiratory infections, diarrhea, tuberculosis, sexually transmitted diseases,worms, skin diseases, malnutrition, and anemia are the most commoncauses of ill health. Malaria accounts for more than40 percent o f outpatient attendance and up to 18 percent o f inpatient deaths. Diarrheal diseases are the second leading cause o f morbidity and mortality. HIV prevalence estimates vary widely, but the latest estimates from the 2007 Liberia Demographic and Health Survey place the prevalence rate at 1.5 percent-significantly lower than the previous estimate o f 5.2 percent, which was based on sentinel surveillance among pregnant women. Given that prevalence rates are significantly higher in some neighboring countries (CBte d'Ivoire's is 7 percent), Liberia faces the risk o f an increase inHIV/AIDS. 4.14 Access to water from improved water sources is 51 percent;' a significant improvement from 2005 estimates of 24 percent.Nationwide, 39 percent o f households have access to sanitation but significant rural and urban disparities exist-with sanitation available to 60 percent o f urban residents but only 30 percent of rural residents.41 4.15 There are 354 functioning health facilities in Liberia, including 18 hospitals, 50 health centers, and about 286 health clinics>2 The government estimates that about 500-550 health facilities will be needed for all communities to be within a 10 km radius of a health facility. On average, it takes almost three hours for members o f rural households to reach the closest health facility.43 According to the 2006 Rapid Assessment o f the Health Situation in Liberia, there were about 4,000 full-time and 1,000 part-time staff, including 122 physicians, 668 nurses, 297 midwives, andmore than 3,000 other trained and untrained healthprofessionals. This 38 Liberia Demographic and Health Survey 2007. 39 Liberia Demographic and Health Survey 2007. 40 CWIQ 2007. 41 CWIQ 2007. 42 RAR 2006. 43 CWIQ 2007. - 54 - i s less than one-third o f the total health workforce o f 14,000 recommended by WHO for a country with Liberia's p ~ p u l a t i o n . ~ ~ Health Expenditure 4.16 This section examines the total amount of expenditure on health from the government, international donors, and private contributions and compares this to key economic and fiscal indicators. With a focus on government expenditure, it examines the breakdownofhealth expenditureby program and over time. 4.17 One of the most striking features of health expenditure in Liberia is the rapid expansion of the health sector in the post-war period. Access to healthcare fell markedly duringthe Liberian civil war. Public expenditure on social services reached anextreme low-point toward the end o f the war in 2002/03, as government resources dwindled and aid flows contracted. Since 2004/05 public expenditure, especially health expenditure, has increased rapidly. Government Health Expenditure 4.18 Inline with the increase ingovernment revenues, spending for health since 2004/05 increased from US$4.2 million (5.5 percent of total spending) to US16.7 million in 2007/2008 (8.1 percent of total spending). Spending i s projected to increase to US$21.4 million in 2008/09 but will drop as a proportion of total spending to 7.2 percent. There is a similarly positive picture inper capita public health spending, with a steadily rising value from US$1.3 in 2004/05 to US$3.4 in2006/07 and US$4.8 in2007/08. Figure 4.1: Evolution of Public Expenditure on Health, 2004/05-2008/09 'ercentage of total Expenditure $US expenditure millions (nominal) 10.0 - 25.0 9 0 -" 8 0 -. 20.0 7.0 -. 6 0 -. 15.0 -+-Share of total expenditure 5.0 -. 4 0 - 10.0 --eHealth 3.0-. expenditure 2.0 -- 5.0 I0 -. 0 0 -i 0.0 :ource: Ministry o f Finance, Annual Fiscal Reports, 2005lO6-2007/08. 4.19 Inrelationto GDP, government expenditure on health more than doubled from 0.9 percent to 2 percent between 2004/05 and 2007/08. But this remains fairly low compared with other post-war countries, such as Mozambique (3 percent in 2004) and Rwanda (4 percent in 44 CWIQ 2007. - 55 - 2004) (figure 4.2). Comparedwith Sub-SaharanAfrica, Liberia's government health expenditure in proportion to GDP is around the average. However, it is worth noting that compared with many other developing countries, Liberia receives little budget support. Given the limited amount of donor project financing captured by the budget, little external assistancefor the health sector it is capturedinthe budget, thus understating the public per capita health expenditure. Figure 4.2: Public Health Expenditure as % of GDP in Sub-Sahara African Countries, 2004-07 10 -' a -. 6 -. 4 2 0 Source: World Bank, HNP data from 2004 for Africa compared against Ministry of Finance 2006107 expenditure for Liberia. 4.20 In addition to an increase in resources dedicated to health expenditure, the budget implementation rate has improved over the past three years. The budget implementation rate increased from 66 percent in 2004/05 to 98 percent in 2007/08. Poor implementation rates in previous years are a reflection of high levels of in-year budget transfers (especially during the National Transitional Government of Liberia period) and low absorptive capacity, especially in 2005/06 whennewpublic procurement regulations came into effect. Table 4.1: HealthExpenditure Implementation Rate, 2004/05-2007/08 Fiscalyear Budget Expenditure Expenditure (US%million) (US%million) (percent of budget) Source: Ministry o f Finance, Annual Fiscal Reports, 2004105-2007108. 4.21 Public health expenditure is almost entirely current expenditure, with only 5-10 percent of total sector expenditure dedicated to capital spending. Capital spending doubled from a low base of US$0.5 millionin2004/05 to US$l.2 million in2006/07 and thenfell back to - 5 6 - US$0.7 million in 2007/08. Low government capital health expenditure i s justified by the fact that a large proportion o fthe external assistance i s dedicated to important investmentexpenditure (with the majority of reconstruction costs for health clinics and hospitals externally funded). Government resources are usedto pay salaries to ensure that the administrativehealth structure is supported and strengthened. (For more detailed data, see Annex 1,table A.2.) 4.22 The estimated breakdown among the MOHSW's different programs is shown in figure 4.3. Expenditure over the past three years shows positive trends: curative services combined with transfers to the John F. Kennedy Medical Center and Phebe Hospital (which provide curative services) has remained the largest expenditure category and has increased as a proportion o f expenditure since 2005/06. Administrative costs have decreased proportionately, indicating that the increase in health resources over the past years has been devoted to service provision rather than an increase inthe administrative structure. Figure 4.3: Health Expenditure by Program, 2004/05-2007/08 (percent share) 100% 80% 60% 40% 20% 0% 2004/2005 2005/2006 2006/2007 2007/2008 Source:Estimates based on Ministry o f Finance, Annual Fiscal Reports and National Budgets, 2004l05- 2007/08. 4.23 Analyzing expenditure by economic classificationshows that in 2006/07 about 29.8 percent of the current health budgetwas spent on wages and salaries,while 37 percentwas spent on goods and services. The wage bill has remained relatively constant over time, while expenditure on goods and services almost doubled in proportion to total expenditure between 2004/05 and 2006/07. But the economic classification system masks costs in the transfer category, making it hard to draw conclusions from this data. Inmost years, more than 50 percent o f health expenditure was classified as transfers. Transfers include payments to the three non- MOHWS agencies, and these transfers do not capture how the funds are allocated (between wages, goods and services, and so on). Transfers also include payments to MOHWS clinics throughout the country for operation and maintenance costs. - 57 - Table 4.2: Distributionof Public Health Spending by Economic Type, 2004/05-2007/08 Source: Ministry o f Finance, Annual Fiscal Reports, 2004105-2007108; National Budget, 2004105- 2007/08. External Resources 4.24 As noted, donor assistance constitutes a large proportion of Liberian health expenditure. Many o f these resources are disbursed directly to nongovernmental organizations and contractors rather than through the MOHSW. It is important to fully account for external assistance in addition to government resources. Otherwise, there is danger o f duplicating spending and wasting scarce financial and physical resources. To be fully effective, policy and planning in the health sector require a view o f the expenditure flows both in total and by components. 4.25 The total external contribution to the health sector is hard to quantify. Multilateral and bilateral donors committed US$89 million to the health sector through the results-focused transitional framework (January 2004-March 2006), or roughly US$40 million a year. The MOHSW estimates that external assistance to the health sector was US$45 million in 2006/07 (although these estimates exclude some o f the private aid flows contributed by nongovernmental and faith-based organizations). Future commitments for 2007/08-2010/11 suggest that US$90- 120 million a year inexternal aid will be provided to the Liberian health sector. Private Expenditure 4.26 Despite increasing public funding for the health sector, private health expenditures remain significant. Data from the 2007 CWIQ indicate that in more than half o f all cases, households seeking healthcare in the month prior to administering the questionnaires reported paying for the services. More than 40 percent o f those who did not seek healthcare during their last illness reported cost as the main barrier to access. Total private expenditure on healthcare in 2006/07 i s estimated at US$15.5 million (US$4 per capita), 51 percent higher than government health expenditure during the same period. Health expenditure represents about 2 percent of average household expenditure. Private health expenditure i s primarily for drugs (40 percent), medicaltreatment (24 percent), and consultations (21 percent) (table 4.3). - 58 - Table 4.3: Distribution of Household Spending for Health, 2007 Source: Estimates based on 2007 CWIQ survey. Reconstructing the Resources Mobilizedfor Health 4.27 Table 4.4 presents a reconstruction o f the health resources and expenditure available for the 2006/07 fiscal year. Total expenditure on health is estimated at US$70.8 million, o f which US$10.3 million was fiom government sources, US$45 million from donor sources, and US$15.5 million from private sources (households). This means that government funding represents only 14.5 percent o f total spendingon health, while contributions from donors account for 63.6 percent and contributions from households account for 21.9 percent. Table 4.4: Resources Mobilized for the Health Sector, 2006/07 (US$ million) Source: Ministry o f Finance, Fiscal Report 2006/07; IMF, Article IV Consultation, 2008; CWIQ 2007. - 59 - 4.28 Liberia's total health expenditure is 11.5 percent of GDP, which is high compared with other low-income African c~untries.~' Liberia's per capita spending on health, US$20, i s similar to significantly more developed countries, such as Kenya, Malawi, Uganda, and the Gambia. Spending i s significantly higher than in many neighboring countries at a similar developmental level, such as Sierra Leone `(US$7) and Guinea-Bissau (US$9). This partly reflects Liberia's large health investment needs, as much o f the health infrastructure was destroyed during the war. Thus, Liberia's per capita expenditure may mask low expenditure on curative services. Itcan also reflect the country's likely underestimated GDP figures. Figure 4.4: Total Health Expenditure as Percent of GDP, 2004-07 14 $3 I 12 10 I--- - s a 8 I 6 4 4 I 4 2 0 Source: HNP data from 2004 compared against Ministry o f Finance 2006/07 expenditure for Liberia. 4.29 In proportion to total health expenditures, external resources for health are higher inLiberia thaninany other African country (figure 4.5). Donor contributions are expected to remain high inthe short term, but may decrease in the mediumto long term. Additionally, it i s expected that funds will increasingly be channeled through the government's budget as opposed to directly to nongovernmental organizations. This will put pressure on the MOHSW to increase its financial management capacities and improve service delivery or increase its capacity to directly contract nongovernmental organizations to continue service provision. Itmay also create potential for gains inexpenditure efficiency, as greater economies o f scale can be used, resources can be pooled, and management structures merged, since there may be less need for many small implementing agencies, eachwith their ownmanagement and administrative structures. 45World Bank; HNP. - 60 - Figure 4.5: External Assistance as a Percentage of Total Health Expenditure 60 50 al40 S $ 30 P 20 10 0 Source: HNP data from 2004 compared against Ministry of Finance 2006107 expenditure for Liberia. Figure 4.6: Composition of Health Expenditure (percent of total) Compositionof Health Expenditure(% of total) 100% 90% 80% 70% 60% 50% 40% Liberia (2006/07estimate) Averagefor African countries (NHA2004) I Source: HNP data from 2004 compared against Ministryo f Finance 2006/07 expenditure for Liberia. Equity of HealthExpenditures 4.30 The goal of the government is to develop an efficient and affordable health system for all Liberians. Equity of healthcare delivery is analyzed in terms of regional disparities and income disparities in access to healthcare. The analysis i s divided into three main sections; the first two examine inequalities in public spending on health, access to health facilities, and - 61 - availability and quality o f healthpersonnel across differentregions; and inequalities inhousehold expenditure on health. The third section provides a benefit incidence analysis o f public spending. Equity of GovernmentHealth Expenditure 4.31 One important area of expenditure analysis is the regional distribution of public spendinginhealth.Liberia's financial information system does not allow for a direct analysis o f regionalization, as most expenditures are centralized. The only level o f care that can be compared across regions is curative services. The average per capita spending at the curative care level varied from US$0.5 inNorth Central to US$3.1 in South Central in2006/07. 4.32 Per capita allocation of public expenditures is highly regressive, with a proportionally greater share of expenditure on curative care in relatively richer regions. Ranking the regions by descending poverty incidence shows that South Eastern A (76.7 percent) is the poorest region, followed by North Western (76.3 percent), North Central (68.1 percent), South Eastern B (67.2 percent), and finally South Central (58.9 percent). As shown infigure 4.7, South Central, which has the lowest poverty incidence, has the highest per capita expenditure on curative services. Figure 4.7: Per Capita Public Expenditureson Curative Services, 2006/07 (by region) South Central South EasternB NorthCentral NorthWestern South EasternA Source: Calculations based onNational Budget and Ministry of Finance, Annual FiscalReports, 2006107. 4.33 Access is one of the main reasons for not seeking care, especially in rural areas. In table 4.5, access i s measured by the distance from the nearest health facility. The data give the average time it takes to reach various types o f facilities. At the national level, health clinics are on average two hours walking distance from the household. In rural areas it takes almost three hours on average to reach the nearest clinic or hospital, compared with an average walking distance o f 30 minutesinurban areas. - 62 - Table 4.5: Time to Nearest Infrastructure, 2007 (minutes) Source: Tsimpo and Wodon 2007. 4.34 Another way to measure equity of health expenditure is to review disparities between regions in the availability and quality of health personnel. Liberia has been unable to sustain networks of physicians and qualified midwives in the poorest regions of the country. The richest region of South Central, with 38 percent of the population, has 53 percent of the country's physicians, anda ratio of 7,23 1inhabitants per physician. Incontrast, inNorthCentral, the population to physician ratio reaches 15,140. The distribution of midwives and nurses among regions is likewise unequal (figure 4.8). Figure 4.8: Population per Technical Health Personnel Ratio by Region and Category SouthCentral South Eastern B North Central NorthWestern South EasternA Liberia Note: Includes fulltime and parttime employees. Source: MOHSW 2006. HouseholdExpenditure on Health 4.35 Table 4.6 provides data on private healthcare spending by households. The largest expenditure interms of the share of total spending for health is for drugs (39.2 percent of total spending). This is followed by spending for medical treatment (injections, bandages, and so on), at 25.8 percent, and spending for consultations, at 22.3 percent. As a share of total expenditure, - 63 - table 4.7 shows that health spendinghas a higher cost for the poor. But in absolute value, more affluent householdstendto spend significantly more on average. 4.36 On a per capita basis, households in the top decile o f the population (ranked according to consumption per equivalent adult) spend almost eight times as m u c h as households in the bottom decile. The total private spending for health is estimated at US$15.5 million. Table 4.6: Structure of Household's Expenditureon Health, 2006107 Source: Tsimpo and Wodon 2007 Table 4.7: Household Expenditure on Health, 2006/07 Source: Tsimpo and Wodon 2007. 4.37 An analysis of the benefit incidence of public spending for health shows that for most public facilities, public spending for health seems to be allocated in roughly similar proportions t o the various household groups ranked by expenditure decile (as shown in Annex 1, figure A.l). For private service providers, traditional healers tend to be used more - 64 - frequently by the poor, and other types o f providers are used more frequently by the better off. Public spending for health does not appear to be regressive probably because in Liberia a large share o f health services are provided by private institutions, which tend to be used more frequently by better off households. (For more detailed data please see Annex 1, table A.3). ConclusionsandRecommendations 4.38 Health expenditure has grown rapidly in the post-war period. Per capita health expenditure i s estimated at US$20 for 2006/07, which is high for Liberia's development status. Donor funding accounts for an estimated 64 percent o f total health expenditure, which is well above the average for low-income countries and reflects Liberia's large reconstruction needs in the health sector after 14 years o f civil war. 4.39 I n the mediumterm, the government should demonstrate a strong commitment to the health sector as a part of its strategy to improve human capital for Liberia's development. In 2007/08 government expenditure on health was 8.1 percent o f total expenditure, which is significantly lower than the 2015 Abuja target o f 15 percent. As Liberia moves toward implementing a medium-term expenditure framework, this increased commitment to healthmust be reflectedinthe government's medium-termplanning. 4.40 There is a continued need for capital investmentsin the health sector. Government expenditure is almost entirely recurrent expenditure, while most capital costs are carried by donors. The government should develop a capital investment plan as part o f its Public Sector Investment Program in the context o f its medium-term expenditure framework, outlining construction and reconstruction priorities, to ensure a more systematic approach to investment expenditure andto increase national ownership. 4.41 Health services are not equitably distributed across regions.The ratio o f persons per physicians in the richest region o f the country is less than half that in the poorest. Total expenditure on curative services is more than three times as high in the relatively richer South Central region, compared to the poorest South Eastern region. The government needs to increase distribution o f health resources to the more remote regions o f the country, both to improve Liberia's health status and to buildpublic trust instate services. 4.42 Donor commitments,which are already at a high level, should be maintainedinthe mediumterm and rise inrealterms inlinewith GDP growth. Liberia will continue to depend heavily on external financing in the medium term until domestic resources are sufficient to maintain constant expenditure levels. Donors are also encouraged to improve harmonizationand predictability o f aid-for instance, through pooled funding mechanisms, such as the recently established Health and Welfare Pooled Fund, supported by the Department for International Development. 4.43 The government's health program should be secured by introducing alternative sources of domestic funding, such as contributory healthinsuranceand other prepayment mechanisms, rather than relying so heavily on out-of-pocket payments.The government is encouraged to develop such programs as soon as possible. - 65 - 4.44 Although health financing will increasingly be channeled through national structures, the government should continue to use the services of nongovernmental organizations and private providers, through formal contracting arrangements. This will require strengthened financial management, procurement, and monitoring and evaluation o f the ministry. With support from development partners, the government is'developing a system where continued financing for organizations or contractors is contingent on satisfactory performance. 4.45 The government has rightly made a basic healthcare package, which targets vulnerable populations, its first priority. But there is a need to continue expanding and developing secondary andtertiary care andto strengthen the referral system. Continued emphasis should be placed on human resource development, boththrough further recruitment and training opportunities, and improved remuneration. B. EDUCATION SECTOR EXPENDITURES Introduction 4.46 This section analyzes educationexpenditure patternsagainstperformance.Itfocuses on an analysis o f trends ineducation expenditure, distribution of resources within the sector, and equity o f access to educational resources. 4.47 Liberia has increased access to education since the end of the war, as evidenced by rises in both gross and net enrollment rates. Both government expenditure and donor support for education have increased rapidly. While the performance o f the education sector i s improving, continued growth inresources will be necessary to sustain an increase in access and quality o f services. Efficiency should also be increased by allocating more resources to primary education. Issues o f equity require a focus on programs that encourage girls to stay in school past the primary level and a more equitable regional distribution o f educational resources. While this chapter focuses on government expenditure, a full picture o f educational expenditure in Liberia cannot be achieved without an analysis o f donor expenditure. Unfortunately, data on donor expenditure is poor, and even aggregate figures are limited. Improved data on donor aid flows are necessary to fully analyze the efficiency o f public expenditure and ensure optimal resource allocation. 4.48 The education system in Liberia is composed of primary, secondary, and tertiary levels. Pre-primary education covers three years, followed by six years o f primary education (grades 1 to 6). Secondary education consists of three years o f junior secondary high school, followed by three years of senior secondary high school. Vocational and technical education and training are available in some secondary high schools, as well as at some tertiary educational institutions. The tertiary levels encompass junior colleges (baccalaureate and graduate levels), university colleges and universities (Cuttington University College and the University o f Liberia), and polytechnics. Partnership is a strong aspect o f education and training services in Liberia, with numerous private sector, community-based, faith-based, and concession-sponsored organizations providing education and training services alongside government educational institutions. - 66 - 4.49 The results from the 2005/06 school census show that there are a total of 3,855 schools with over a quarter concentrated in Montserrado County. Just over half o f these schools are public. About 17 percent are private for-profit schools, three-fourths o f which are located in Montserrado. Overall, only slightly more than half o f all students go to government schools, with private for-profit schools serving more than a third o f the students, and religious and community schools serving the remainder. 4.50 The Ministry of Education is responsible for the overall administration and management of education and training. The Minister o f Education is assisted by three deputy ministers responsible for administration, instruction, and planning and development. According to the 2002 Education Law, the Ministry o f Education has a strong decentralized management structure with education offices at county and district levels. Inpractice, however, because o f the strong residual effect o fthe years o f conflict, decisions are mostly made at the national level and instructions passed on to the counties and districts. There are 15 county education offices, each headed by a county education officer with a support staff, and 86 district education offices each headedby a district education officer. Primary and Secondary SchoolEnrollment 4.51 The gross enrollment rate at the primary level increased from 72.7 percent in 2000 to 86.3 percent in 2007. The lack o f proper infrastructure, teachers, and security in the country and the high cost o f education (linked to user fees) for families led to a sharp decline in enrollment in 2002, especially for girls.46To stem this decline, the 2002 Education Law made primary education free and compulsory, butresources were lackingto fully implement this law. Figure 4.9: Gross Enrollment Rates in Primary Education .... ....... . .... ~ I I 20 .............................................................. // --"I 0 2000 2002 2007 Source: CWIQ 2007; Ministryo f Education2005. 4.52 A return to peace and renewed efforts by the government since 2006 to promote school enrollment have increased gross primary school enrollment by almost 50 percent between 2002 and 2007. Many children and youth with limited exposure to formal education have returned to school, which also explains the discrepency between net and gross enrollment rates. A school census,from 2005/06 suggested that only 15 percent o f students inthe first year 46NationalPolicy for Girls Education, Ministry o f Education, 2005. - 67 - o fprimary school were o f the right age (6-7 years old), while halfthe students were 11-20 years old. 4.53 The gross enrollment rate at the secondary level has also increased steadily, from an estimated 32 percent in2000 to 51.3 percent in 2007. The big discrepancy between 2007 net and gross secondary school enrollment (15.2 percent versus 51.3 percent) suggests that many secondary school children are above the official secondary school age. Table 4.8: Net and Gross Enrollment Rates, 2007 Source: CWIQ 2007. Tertiary and VocationalEducation 4.54 In2007,1,400 students graduated from the University ofLiberia. The unitcost at the University is US$242 (2007/08), twice Liberia's GDP per capita. This figure is comparable to most other Sub-Saharan African countries. Incontrast, for middle- to high-income countries, the average is about 30 percent o f GDP per capita. 4.55 Beforethe war, Liberia had a number of public vocationaVtechnica1 institutions, but few have been resuscitated and they receive little government support. Much vocational education today i s provided by nongovernmental organizations and church groups. There are many testing and certification standards and weak links between vocational training and private sector skill requirements. TeacherSupply 4.56 There are about 28,000 teachers in Liberia, and about half (46 percent) work in public schools. Over 60 percent o f all public school teachers are unqualified. At present, teachers are only being recruited to replace those who leave. But the rapid expansion o f primary enrollment and the plight o f volunteer teachers-many o f whom have been teaching for years without pay-will soon require the recruitment o fteachers into the public sector. Total Public Expenditure on Education 4.57 Government expenditure on education accounted for between 1.6 percent and 3.2 percent of GDP between 2004/05 and 2007/08. This compares unfavorably with other post-war countries such as Burundi (5.1 percent in 2005), Rwanda (3.8 percent in 2005), Mozambique (3.7 percent in 2004), and Sierra Leone (3.8 percent in 2005). The figures for other countries in Sub-Saharan Africa range from 0.6 percent o f GDP (Equatorial Guinea in 2002-03) to 13.8 percent (Lesotho in2005). - 68 - 4.58 Education expenditure has increased sharply, from US$7.3 million in 2004/05 to US23.3 million in2007/08. Educationexpenditure is expected to increase to US$28.0 million in2008/09. This represents anaverage increaseinreal spending for education of 26 percent per year. This mirrors the increase intotal government expenditure (whichincreasedfrom US$76.3 millionin2004/05 to US$203.4 millionin2007/08), meaningthat the share of total expenditure allocated to education has remained relatively steady at around 10-12 percent. This is still low compared to other African countries and far below the Education For All Fast Track Initiative benchmark of 20 percent. Figure 4.10: Evolution of GovernmentExpenditureon Education,2004/05-2008/09 Percentage of Education Expenditure2QQ4iQ5-2QQ7iQ8 total expenditure /.. 1 3 0 I Expenditure US$ millions (nominal) I " I -------- / c25 1 2 I II 0 4 : - L o 2004/2005 2005/2006 2006/2007 2007/2008 2008/09 Source: Ministry of Finance, Annual Fiscal Reports, 2004105-2007lO8. Distribution of GovernmentExpenditure by Level of Education 4.59 About 14.5 percent of all government resources for education went to primary education (including pre-primary level) in 2006/07 (figure 4.11). The priority given to this sub-sector decreased between 2004/05 and 2005/06-with a reduction in resources from 17.1 percent to 7.5 percent, but with an increase to 14.5 percent in 2006/07. Preliminary estimates based on the budgetary allocation for 2007/08 suggest that the share to primary education will have increasedto 29.5 percent. 4.60 The share for secondary education increased from 9.5 percent in 2004/05 to 11.4 percent in 2006/07 and to 19.1 percent in 2007/08. The share for the tertiary level increased from 24.6 percent in2004/05 to 31.6 percent in2006/07. Estimates for 2007/08 indicate that the share of spending on tertiary education decreased to 27.1 percent. Finally, the allocation to technical and vocational education decreased substantially from 7.6 percent in 2004/05 to 4.9 percent in2007/08. - 69 - 4.61 The final category for expenditure is administrative and support services. Expenditure in this category decreased between 2004/05 and 2006/07 in relative terms, comprising around -- percent o f the total expenditure on education in 2006/07. Allocations for 2007/08 show a further reduction to 19.4 percent. Figure 4.11: Share of current Expenditure by Level of Education, 2004/05-2007/08 (percent) 100% 90% 80% Administrativeandsupport services 70% 0 TechnicaVVocationalTraining 0 Higher education 60% Secondaryeducation 50% Ri Pnrnaryeducation 40% 30% 20% 10% 0% 200412005 200512006 200612007 200712008 Source: Ministry o f Finance, Annual Fiscal Reports, 2004105-2007lO8; National Budget, 2004105-2007108. Household Contributions to Education Financing 4.62 Parents who send their children to private school pay significantly higher school fees. Those with children in government and government-assisted schools report spending on fees, even though tuition fees were officially abolished in all primary schools in 2002. In this section, data from the 2007 Core Welfare Indicators Questionnaire (CWIQ) survey are used to explore the magnitude o fhousehold spending on education. 4.63 In 2006/07 total private spending for education was estimated at US$27 million, significantly higher than the total public expenditure on education of US$12.2 million in the same year. Table 4.9 provides data on private education spending by household, for those households that have at least one child enrolled in school. The total amounts that households spend on primary and secondary education are larger than the amount the government spends. For higher education, the contribution o f households i s about the same as that o f the government, constituting 51.8 percent o f total spending. - 70 - Table 4.9: Household Spending on Education by Level, 2006/07 Source: Estimatesbased on the 2007 CWIQ survey. 4.64 The largest expenditure for education is for school fees at the primary school level. School fees for secondary education also absorb a large part o f the private education budget o f families, as do school uniforms (table 4.10). Table 4.10: Distributionof Household Spending for Education, 2007 Source: Estimatesbased on 2007 CWIQ survey. The Contributionof External Donors toEducation Financing 4.65 I t is difficult to estimate the contribution of external donors to education financing in Liberia. External aid to Liberia is estimated to be higher than total government expenditure, making it hard to estimate the total resource envelope for education. Much donor support i s channeled outside the budget and directly to implementing agencies. A survey conducted to capture all the disbursements for education estimated that donors' disbursements to the sector in 2006/07 were US$38 million, more than three times the amount o f government spending (table 4.1 1). -71 - Table 4.11: Development Partners and Nongovernmental Organizations' Current and Projected Assistance for Education (US$ million) Source: Education Donors survey 2008. 4.66 Table 4.12 presents an estimate of the resources and expenditure for 2006/07. Total expenditure on education i s estimated at about US$77.2 million, o f which US$12.2 million came from government sources, US$38 million came from donor sources, and US$27 million came from private sources (households). This means that government funding represents only 15.8 percent o f total spending on education, with contributions from donors accounting for about 49.2 percent and contributions from households for 35 percent. Table 4.12: Resources for Education, 2006/07 (US$ million) Source: Ministry of Finance, Annual Fiscal Reports, 2004/05-2007/08; Education donor survey 2008; CWIQ 2007. - 72 - DetailedAnalysis of EducationExpenditure in2006/07 Salary and Non-Salary Expenditure 4.67 Salary expenditures made up 23.4 percent of total public spendingon education in 2006/07. But this masks the true proportion o f personnel costs, as a significant proportion o f transfers and subsidies, which accounts for 49 percent o f the education budget, are spent on personnel. 4.68 The major components of non-salary expenditure are transfersand subsidies (49%) and goods and services (23%). Capital expenditure accounts for only 5 percent of the education budget. Estimates of the Unit Cost 4.69 The cost per student (unit cost) from government resources is shown by levels in table 4.13. For primary education, the cost was about US$6.9 per student in2007, an equivalent o f 4 percent o f GDP per capita. Note that unit costs gradually increase with higher levels o f schooling. But since a large proportion o f education expenditure i s from external sources and not accounted for inthe budget, it is hard to estimate the total unit cost. The differences inunit cost by level of education may be less pronouncedifexternal sources were included. Table 4.13: Unit Cost Estimates for Educationby Level,2006/2007 a. Number o f students inthe University o f Liberia, at around 15,289, is assumed to represent 80 percent of total number of students in public schools inthe higher education level. Source: CWIQ 2007. - 73 - Table 4.14: Comparisons of Unit Cost as a Percentage of GDP Per Capita in Selected Sub- Saharan African Countries, by Level (%) Primary education Secondary education Higher education All levels Burkina Faso 31.5 Burundi 73.3 Lesotho 1145.7 Lesotho 34 Lesotho 25.1 Lesotho 50.8 Eritrea 1082.5 Botswana 33 Kenya 24.2 LIBERIA 49.3 Botswana 438.4 Senegal 32.8 Senegal 21.0 Botswana 40.2 Rwanda 404.5 Burkina Faso 32.1 Cape Verde 20.3 Senegal 40.1 Burundi 348.8 Burundi 31.3 Burundi 19.1 Ghana 34.4 Chad 333.9 Kenya 28.8 Ethiopia 17.7 Uganda 34 Kenya 284.5 Ghana 24.8 Seychelles 16.7 Chad 28 Senegal 270.2 Cape Verde 24.1 Botswana 15.7 Cape Verde 26.4 Congo 245.7 Seychelles 22 South Africa 14.3 Kenya 24.1 Ghana 209.4 Eritrea 20.7 Ghana 12.8 Benin 21.2 Benin 200.9 Benin 17.8 Mauritius 11.8 Mauritius 19.7 Burkina Faso 92.6 South Africa 17.6 Benin 11.5 Burkina Faso 19.6 Uganda 88.9 Rwanda 17.2 Uganda 11.3 Congo 18.2 Madagascar 75 Mauritius 17 Eritrea 11.1 - Seychelles 17.7 Zambia 68.2 Uganda 15.1 Madagascar 8.4 South Africa 17.6 LIBERIA 63 Congo 14.6 Chad 6.8 Eritrea 15.2 Cape Verde 74.5 LIBERIA 12.7 Cameroon 6.1 Zambia 8.2 Cameroon 66.6 Chad 12.4 Zambia 5.4 Ethiopia 7.9 South Africa 50.1 Zambia 11.3 LIBERIA 4.0 Cameroon 2.6 Mauritius 36.9 Cameroon 6.6 Equatorial Congo 3.4 Guinea 2.4 Equatorial Guinea 0.9 Source: UNESCO Institute for Statistics Global Education Digest. (http://stats.uis .unesco.orglReportFolders1reportfolders.aspx). 4.70 The average salary of teachers is the first factor in the unit cost of education. Figure 4.12 compares average teacher salaries at the primary level in relation to GDP per capita in Liberia with other countries. The Fast Track Initiative benchmark for teacher salaries inprimary schools-based on the structural parameters in countries demonstrating the best performance in the goal of universal completion of primary school-is 3.5 times the GDP per capita. Liberia is at the benchmark level, with primary teacher salaries at 3.5 times the GDP per capita. But due to Liberia's low GDP per capita, inabsolute terms Liberian primary and secondary teacher salaries are low, at about TJS$55 a month, insufficient to attract secondary school graduates to pursue teaching degrees. - 74 - Figure 4.12: Ratio of Average Teacher Salary to GDP per Capita inSub-Saharan Africa, 2007 m -Ei n c Source: UNESCO Institute for Statistics Global Education Digest htlp://stats.uis.unesco.org/ReportFolders/repo~folders.aspx); 4.71 The second important factor inthe unit cost is the pupil-teacher ratio at the school level. InLiberia, the pupil-teacher ratio is estimated at 28 inprimary school and 13 insecondary. A comparison with other Sub-Saharan African countries and with the Fast Track Initiative benchmark indicates that the pupil-teacher ratio in Liberia's primary schools is on the low side and less than the benchmark o f 40. For secondary education, the pupil-teacher ratio in Liberia is also much lower than that o fthe comparator countries. 4.72 The comparisons of education unit costs above suggest that Liberia's public education expenditure per student is extremely low in absolute terms and low in relationship to per capita GDP. But significant donor aid for education is not accounted for in the government budget and i s not reflected in the unit cost estimates. Given that average classroom size i s relatively small inproportion to other African countries, Liberia's low unit cost reflects low teacher salaries and insufficient expenditure on educational resources. This has negative effects on the quality o f education and on the ability o f students to absorb teaching. - 75 - Table 4.15: Pupilto Teacher Ratio in Sub-Saharan Africa, 2004-2006 Source: UNESCO Institute for Statistics Global EducationDigest http://stats.uis.unesco.orgiReportFolders/reportfolders.aspx). Equityof Spending inEducation Disparity in Schooling 4.73 Disparities exist in schooling. Girls, rural children, and children inpoorer households all have a lower likelihood o f participating informal education at primary and secondary levels. The reasons for these disparities are not always the same across groups. Insome cases, disparities in enrollment rates result from disparities in initial access to schooling. In other cases, they result from higher drop-out rates among children from vulnerable groups. 4.74 Age specific enrollment rates show a clear differential by gender (figure 4.13). According to the 2005 school census data, a gender divide does not emerge until age 15, with girls less likely than boys to be in school above age 15. There are two possible reasons for this divide. First, it is possible that females initially have equal access to education and then start to drop out faster when they reach age 15 and beyond. We observe respective dropout rates o f 4.8 percent and 5.4 percent for males and females respectively at the senior secondary level, which corresponds to the age group 15-17 years. A second possibility i s that there were gender disparities when the current over 15 cohort first entered school, and the subsequent cohorts did - 76 - not have a gender disparity in initial access. Ifthis is the case, then the current pattern of gender disparity will disappear over time. Table 4.16: Age-Specific Enrollment Rates and out-of-school Children byAge Group (percent) ! Maie Female Total Age 011t-of- Out-of- Out-of- (years) Enrolled school Enrolled school Enrolled school 6-11 37 63 40 GO 38 62 12-14 75 25 78 22 77 33 15-17 81 19 75 25 78 27- Source: Ministry of Education 2006; CWIQ 2007. Figure 4.13: Enrollment Rates by Age and Gender 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 Ages Source: Ministry o f Education 2006; CWIQ 2007. 4.75 Gross enrollment rates, a measure of education coverage, show that the more favored subgroups are male, urban children, and children from the highest consumption quintile. For each o f the groups, table 4.17 presents the parity index-the ratio of the gross enrollment rate o fthe lagging subgroup to that o fthe more favored subgroup. A parity index o f 1 indicates perfectparity (that is, the two groups have the same gross enrollment rates). The gender gross enrollment rate parity index o f 0.96 at the primary level as a whole implies that primary coverage for females is 96 percent as high as it i s for males. It indicates high parity o f the two groups at the primary level. At the secondary level the gender disparity is high with a parity index of 0.78 o f that o fmales. - 77 - Table 4.17: Gross Enrollment Rates by Gender, Locality, Region, and Household Expenditure Quintile, 2007 Source: CWIQ 2007. 4.76 Table 4.17 shows that the gap in each set of comparisons is smallest at the primary education level. The gap between the lagging group (female) and the favored group (male) broadens rapidly as the level o f education rises. The gap between the rural area and the urban area i s also higher at the secondary level. Furthermore, there i s a clear gradient in educational coverage by household expenditure quintile, with children from poorer households less likely to be enrolled inschool inthe primary and, evenmore, secondary levels. 4.77 At the county level, primary gross enrollment rates are high for Grand Cape Mount, Lofa, Grand Kru, Bomi, Margibi, and River Gee and very low for Grand Bassa (value of 28.8). The parity index between the highest (Grand Cape Mount) and the lowest (Grand Bassa) i s 0.22 for primary education. It i s even lower at the secondary level with parity index o f 0.08 between Grand Bassa and Grand Kru(figure 4.14). - 78 - Figure 4.14: Gross Enrollment Rates by County, 2007 bgions Counties m RivergeeCounty d = :; Maryland County W Grand Kru County 4 Shoe County :0E3 Rver Cess County a 4 GrandGedehCounty MontserradoCounty [7;Secondary d E MargibiCounty Primary a 6 Grand Bassa County C E GbarpoluCounty 3 Grand Cape Mount County .6 Bomi County c 7 NimbaCounty r Z q s Lofa County u Bong County 0.0 20.0 40.0 60.0 80.0 '00.0 PO.0 MO.0 GER inpercent Source: CWIQ 2007. 4.78 The majority of new primary school entrants are not of the official age of six years. The gross intake rate i s the ratio o f new entrants in primary grade 1, regardless o f age, to those who are at the official age o f entry for primary grade one (6 years). The gross intake rate for primary school grade 1 i s about equal among boys (80.5%) and girls (77.1%). The rate tends to overestimate the share o f children in the age cohort entering primary grade 1 because o f the inclusion o f multiple age cohorts in the numerator. The net intake rate is similar to the gross intake rate but only includes new entrants who are at the official age for primary grade 1. The net intake rate is 9 percent for male and 8.8 percent for female. The big differences between the gross and net intake rates for both sexes confirms that the majority o f new primary school entrants are not o fthe official age. Disparity in Household Expenditure on Education 4.79 The unit cost of household spending on primary and secondary education is higher for the richest quintile than for the poorest quintile (table 4.18). On average, households spend more than the government per primary and secondary student, but less than the government on higher education. But poorer households spend more in relation to their overall household expenditure. Figure 4.15 shows the differences from the poorest to the richest - 79 - households in per student expenditure on primary and secondary education as a share of per capita household expenditure. Table 4.18: Per Student Expenditure on education, by Quintile and Level o f Education, 2007 (US$) Source: CWIQ 2007. Figure 4.15: Per Student Spending on Primary and Secondary Education as a Percentage o f Per Capita Household Spending by Quintile, 2007 1 Percentage IS Q1 (Poorest) Q2 0 Q3 oQ4 Q5 (Richest) I Primary education Secondary education Source: Estimates basedon CWIQ 2007. Disparity in Public Expenditure on Education 4.80 A benefit incidence analysis conducted with the 2007 Core Welfare Indicators Questionnaire (CWIQ) data shows that public spending on education is pro-poor, with the exception o f the post-secondary level, but gender and regional disparities are severe. Assuming the unit costs of enrollment are similar for all students attending a given cycle, the estimates o f the number o f students enrolled provide us with shares of total spending per cycle that are allocated to the various deciles. As shown in figure 4.16, at the primary, junior secondary, and senior secondary levels, public spending for education i s allocated more to the poorest deciles than to the other deciles. At the post-secondary level, most of the public spending goes to students from the wealthier segments o f the population. That public spending for education appears to be pro-poor i s relatedto the fact that inLiberia a large share of students are enrolled inprivate as opposed to public schools, with the poor more likely to use public schools than those from better-off households. - 80 - Figure 4.16: Concentration curve o f enrollment in public schools, 2007 0 1 2 3 4 5 6 7 8 9 10 -Equity -- Primary-- Secondary 1- - -Secondary2- -Post secondary -Overall PublicschoolI Source: Tsimpo and Wodon 2007. 4.81 The CWIQ survey shows that public education expenditure distribution is gender biased. By combining information on the proportion o f population groups accessing different levels o f education (from the 2007 CWIQ survey) and the total expenditure on education by level, it i s possible to determine how equitably public resources are distributed across groups. Table 4.19 shows the percentage of education resources benefitting different sub-groups, including male and female, those inurban and rural localities, and by region. For instance, males benefit from almost 70 percent o f public education expenditure, while females benefit from the remaining 30 percent (largely due to the low female enrollment rates in and high proportion o f public spending on tertiary education). Also, the locality and regional disparities are severe, with the majority o f the 5 to 25 year olds who live inrural areas benefiting from only 38 percent o f public education expenditure. Also, the South Central region benefits disproportionately compared with its share o fthe age cohort. - 81 - Figure 4.17: Disparity in Consumption o f Public Education Resources, 2006/07 Source: Estimates based on CWIQ 2007 and Ministry o f Finance, Annual Fiscal Report, 2006107. 4.82 The disparity by gender i s expected t o improve over time as the new cohort o f girls 12 years o f age and younger advance to higher levels of education. Ensuring successful transitions to secondary schooling for this group o f girls i s vital for closing this gap. Only 27.4 percent o f students at the tertiary level are female. Also, efforts to reduce disparities at the locality and regional levels will be neededto close the existing gap inaccess to education. Efficiency on Spending inthe Education Sector Characteristicsof Schoolsand Outcomes 4.83 There was an average o f 28 pupils f o r every teacher at the primary level in Liberia in2005/06 and 14 students per teacher at the junior secondary level. These figures are low compared to other Sub-Saharan African countries. There are variations across the 15 counties; for example, the student-teacher ratio in Lofa is 105. More evenly distributed teacher deployment i s needed inLiberia. 4.84 In the school year 2005/06, there were around 17,442 teachers in primary school. Female teachers make up around 12.9 percent o f the primary school teaching force. Compared with the average o f 38 percent for 20 other countries inAfrica, Liberia has a low share of female teachers in primary schools. There is little regional variation in the share o f female primary school teachers. About 22.7 percent o f female primary school teachers are volunteers, and most are financed by community contributions. 4.85 There are very large inequities in teacher distribution, with fewer teachers, and even fewer qualified teachers, inthe rural counties. At the primary level, Montserrado County - 82 - has a pupil to teacher ratio o f 30, and 68 pupils per qualified teacher. By contrast Lofa County has a pupilto teacher ratio o f 105 and 282 pupils per qualified teacher. Efforts to attract teachers to remote rural areas are needed, recruiting teachers locally and providing them with housing and salary incentives. 4.18: Primary Pupilto Teacher Ratio in Sub-Saharan Countries, 2005/06 Source: UNESCO Institute for Statistics Global Education Digest, httv://stats.uis.unesco.orp/ReuorlFolders/re~o~folders.aspx);Estimates based on 2005106 school census 4.86 About 60.4 percent of primary school teachers (or about 10,526) are unqualified. Unqualified teachers are defined in Liberia as teachers without teaching certificates from recognized Teacher Training Institutions. Again there are significant differences across regions. The share o f unqualified teachers is 81 percent in Sinoe County. Similar to the primary level, there are relatively high percentages o f unqualified teachers in the junior and senior secondary schools, 44.8 percent, and 49.7 percent respectively. 4.87 Completion rates measure the proportion of children who attain schooling of a given grade. InLiberia the important milestones are completing grades six, nine, and twelve because they correspond to completing primary, junior secondary, and senior secondary school, respectively. The gross completion ratio refers to the number o f students, regardless o f age, completing the final year o f each level of education, divided by the population o f the official completion age o f the level. The 2005/06 school census data yields a gross completion ratio for primary school o f 91.9 percent, junior secondary school o f 42.8 percent, and senior secondary school o f 21.4 percent. But this indicator overstates actual completion performance since the gross completion ratio includes a large number o f over-age students. There i s disparity in the completion rates at all levels o f education by county (annex table 3). While in Lofa, Grand Kru, and Montserrado the primary school gross completion ratio is over 100 percent, in Grand Cape Mount it i s below 30 percent. L o w completion rates are also observed at the junior secondary and senior secondary levels. - 83 - 4.88 Grade repetition rates and dropout rates are two important indicators for measuring student flow efficiency. Liberia has average grade repetition rates o f 6.9 percent in primary education. We observe a highlevel of repetition ingrade 1. The repetition rates decrease withthe grades. The secondary level is characterized by a fairly low repetitionrate. Compared to other Sub-Saharan African countries, Liberia is on the lower side with an average of 5 percent, while in Burkina Faso the rate i s 25 percent (figure 4.18). Still, this level i s relatively high, especially inthe lower grades, and reducing the repetition rate should be a policy concern o f the government. Figure 4.19: Primary RepetitionRates in Sub-Saharan African Countries, 2005/06 35 30 25 20 15 10 5 0 Source: Estimates based on 2005106 school census. 4.89 Dropout rates are high at grade 1 in primary education with a rate of 7.6 percent. Further investigations are needed to identify the causes of highdropout rates. Dropout rates fall as children progress to higher grades and are lowest inthe final year of senior secondary school, at 3.8 percent. - 84 - Figure 4.20: Dropout Rates inPrimary and Secondary Levels, 2005/06 Grade1 Grade2 Grade 3 Grade4 Grade5 Grade6 Grade7 Grade8 Grade9 Grader)Grade 11 Grade P Grades Source: Estimates based on 2005106 school census. Conclusion and Recommendations 4.90 Enrollment figures suggest that access t o education has improved significantly since the end o f the civil war in2002/03, with gross enrolment rates increasing f r o m 72.7 percent in2000 to 86 percent in2007. Butlarge discrepancy remains inthe gross enrollment rate of 86 percent and net enrolment rate of 37 percent, suggesting that many over-aged children and youth who missed out onyears o f school duringthe war have returned to school. 4.91 In comparison to total government expenditure and in proportion to GDP, government expenditure on education in Liberia remains low, even compared with other post-conflict countries. In this respect, Liberia is far from reaching the EFA Fast Track Initiative benchmark. Government distribution o f education expenditure is also skewed, with Liberia spending an unusually small share o f its budget on primary education and a very high share on tertiary education. In 2006/07, out o f the total education budget, 14.5 percent was allocated to primary education, 11.4 percent to secondary education, 31.6 percent to tertiary education, 7.1 percent to technical/vocational training, and 35.4 percent to administrative and support services. Moreover, private households spent US$27 million in 2006/07-more than twice the government's expenditure on education-which places a considerable burden on poor households. This suggests that families place a highpriority on educating their children. 4.92 The rapid expansion in enrollments at the primary level means that many students in the comingyears will require places injunior secondary education. Currently, the junior secondary education system has the places neither to support the large influx coming through the system nor to sustain the numbers that should be coming through when the primary system stabilizes and achieves universal primary completion. The percentage share o f expenditure for secondary education increased from 9.5 percent in 2004/05 to 11.4 percent in 2006/07. To adequately supply the resources needed for this sub-sector, allocationwill need to be increased. 4.93 The largest source o f education expenditure comes f r o m external aid. These aid flows are very difficult to track, and only tentative data is available. It is estimated that US$38 million was spent by donors on education (with a focus on primary education) in2006/07, which - 85 - accounts for more than half o f total education expenditure. Better monitoring and targeting o f these resources may helpto improve the efficiency o f educational expenditure and outcomes. 4.94 Disparities in education exist. Female children, rural children, and children from less wealthy households all have lower likelihood o f attending school. There are also important regional variations in access and quality o f schooling. In general, access to primary school is fairly equitable, but disparities increase in secondary and tertiary education. At the primary school level, the parity index for female to male students i s 0.96, and for rural to urban students is 0.89. However, these disparities increase at secondary level to 0.78 and 0.53 respectively. Gross enrollment rates also show high regional variation by county, with a rate o f 29 percent in Grand Bassa, compared with 129 percent in Grand Cape Mount. More effort should be made to ensure that female and rural children have access to secondary and tertiary education. Resources could be distributedmore equitably across the country to reduce regional variations in education access. 4.95 Because of the limited fiscal space, the government's strategy for education in the short-to-medium term should focus on: 0 Improving the efficiency ofthe allocation and expenditure. 0 Buildinga partnership with the private sector to increase resources. Ensuringthat education and training are better alignedwith labor market needs. 4.96 Budget allocations to the education sector should be based on a sector strategy. In the multi-year budget framework, the government should ensure that relatively more public resources are channeled to primary andjunior Secondary education, where the social returns are likely to be higher relative to those inupper secondary and tertiary education. Inaddition, when allocations are made to investments, they should indicate the long-term recurrent implications, like those o f the regional training institutes. Some allocations toward basic education services (those that go beyondprimary education) should be equity- or quality-based. As the majority o f external funding is for primary education, increased capital spending is needed at the secondary education level to provide more schools and learning materials. 4.97 Greater resource efficiency could be achieved by lowering the dropout rate and grade repetition rates, especially in grade 1. Interventions should focus on improving quality, by improving teachers' qualifications and increasing accessibility to teaching and learning materials. The government is advised to reduce the pupil to teacher ratio by addressing large inequities in teacher distribution. This could be done by increasing local recruitment and providing teacher housing and salary incentives. 4.98 There is also the need to clean up the teacher payroll. The disruption and insecurity caused by the war resulted in many teachers leaving their posts, large numbers o f replacement teachers (47 percent in Grand Cape Mount), and concern about the possibility o f ghost teachers. Recent exercises have been conducted to clean the payroll, but discrepancies remain between school data, human resources data and Education Management Information System data, suggesting that the payroll i s not yet completely clean. . - 8 6 - 4.99 Further efficiency gains could come from decentralizingservice delivery. An initial step could be enhancing county education offices' capacity and improving school management for service delivery. Since the free and compulsory education policy has been implemented, schools no longer benefit from revenues to cover operational costs for minor maintenance and small material. School management committees do exist but are not fully functional. In many Sub-Saharan African countries, a system o f direct flow o f funds to schools for service delivery has improved the accountability and quality of service delivery. This system may be worth piloting in select counties. One additional benefit o f this direct fund flow would be the need to strengthen local financial management systems, which is a prerequisite for accountability and transparency o f public funds. 4.100 In order to expand resources, the government should explore promoting public- private partnershipsand cost recovery at the tertiary level. This is inkeeping with making expenditure more progressive by spending more at the primary and lower secondary levels. Different models o f public-private partnership and cost recovery systems in other countries should be examined, particularly for tertiary education, where the unit costs are far higher than those for primary and secondary education. 4.101 An important element of the strategic planningfor education is to ensure that its outcomes are better aligned to the labor market needs of the country. The public-private partnership should therefore go beyond investment in school plants or tuition. It should include consultations on curriculum development, life skills training, or other innovative ways o f involving the private sector inmanaging the delivery o f educational and training services. - 87 - BIBLIOGRAPHY Backiny-Yetna, Prospkre, Quentin Wodon, Rose Mungai, and Clarence Tsimpo. 2007. "Poverty in Liberia: Level, Profile and Determinants." Preliminary Draft. World Bank. Chen, Siyan, Norman V. Loayza, and Marta Reynal-Querol. 2007. "The Aftermath o f Civil War." WPS4190, World Bank. Government of Liberia. 2007. "Preliminary Multi-Year Estimates o f Revenue and Core Poverty-Related Expenditures 2006/07-2009/10." Liberia Partners Forum, Washington, DC. Heller, Peters. 2005. "Understanding Fiscal Space." IMF Discussion Paper 05/4. International Monetary Fund, Washington, DC. International Development Association and International Monetary Fund. 2008. "Liberia: Enhanced Heavily Indebted Poor Countries (HIPC) Initiative Preliminary Document." International Monetary Fund.2003. "Selected Issues and StatisticalAnnex." IMFCountry Report N o 03/275. International Monetary Fund. 2005. "Selected Issues and Statistical Annex." IMF Country ReportN o 05/167. International Monetary Fund. 2006. "Liberia: First Review o f Performance under Staff- Monitored Program-Staff Report." IMF Country ReportNo 06/412. International Monetary Fund. 2007. "Liberia: Third Review o f Performance under Staff- Monitored Program-Staff Report." IMF Country Report N o 07/356. International Monetary Fund. 2008. "Liberia: Fourth Review o f Performance under Staff- Monitored Program-Staff Report." IMF Country Report No 08/108. Muzondo, Timothy R., Richard Bird, Gloria Reid and Zuhtu Yucelik. 2006. "Liberia. Reform of Tax Policies andRevenue Administration." FAD, IMF. Muzondo, Timothy R., and Michael Welling. 2007. Liberia: Tax Reform Program and the ECOWAS Common ExternalTariff." FAD, IMF. Schiller, Christian, Davina Jacobs, Suhas Joshi and Sanjeev Kumar. 2006. "Liberia: Public Financial Management Reforms." FAD, IMF. Story, Thomas, Frank Bosch, William Crandall and David Hesketh. 2006. "Liberia: Modernizing Tax and Customs Administration." FAD, IMF. Kloeden, David, BrianJones, Andrew Masters, Kevin Woodley and Aftab Haider. 2007. "Liberia: Status and Next Steps inRevenue Administration Modernization." FAD, IMF. Daniel, Philip, Paulo Medas and Emil M. Sunley. 2007. Liberia: Reforming the Fiscal Terms for Miningand Petroleum." FAD, IMF. Republic ofLiberia. 2008 Poverty Reduction Strategy. Monrovia. - 88 - Republic ofLiberia, Ministry ofEducation. 2006. 2005/2006School Census.Monrovia. Republic o f Liberia, Ministry o f Education. 2005. National Policy for Girls Education. Monrovia. Republic o f Liberia, Ministry o f Finance. 2006. Preliminary Report on 2005-2006 Fiscal Outturn, Monrovia. Republic of Liberia, Ministry o f Finance. 2006. Report on 2006-2007 Fiscal Outturn, Monrovia. Republic of Liberia: Ministry o f Finance. 2007. "Strategy for the Resolution of Domestic Debt andArrears." Monrovia. Republic of Liberia, Ministry of Health and Social Welfare. 2006. Liberia Rapid Assessment of the Health Situation in Liberia. Monrovia. Tsimpo, Clarence and Quentin Wodon. 2007. "Education in Liberia: Basic Diagnostic usingthe 2007 CWIQ Survey." Preliminary Draft. WorldBank. Tsimpo, Clarence and Quentin Wodon. 2007. "Health in Liberia: Basic Diagnostic using the 2007 CWIQ Survey." Preliminary Draft. World Bank. World Health Organization, National HealthAccount Series. 2007. http://www.who .int/nha/en/. UNESCO Institute for Statistics Global Education Digest. http://stats.uis.unesco.orrr/ReportFolders/reportfolders.aspx - 89 - Annex 1: Public Expenditure Reviews, Tables, and Figures Table Al.1: Functional Classification o f Public Expenditure, in Percent of Total Expenditure, 2004105-2007lO8 - 90 - Table A1.2: Government Expenditure on Health, 2004/05-2007/08 Source: Ministry of Finance. - 91 - I N a I Figure Al.1: Concentration Curves for use of Public Health Facilities, 2007 100 90 80 70 60 50 40 30 20 0 1 2 3 4 5 6 7 8 9 10 -Equity -- - -Government hospital -- - - Government health center Government health clinic-- Other public facility Total public healthA Source:Authors' estimates based on 2007 CWIQ survey. - 93 - Figure A1.2: Concentration Curves for use of Private Health Facilities, 2007 10000 90 00 80 00 70 00 60 00 50 00 30.00 20.00 0 1 2 3 _ 4 5 6 7 8 9 10 Private hospital/clinic - - Pharmacy %:e doctoddenht _-_ 4-Mobileclinic/blackbaggeddrugpeddle- Other private facility +Traditional healer Total private health Source:Authors' estimates based on 2007 CWIQ survey. - 94 - Table A1.4: Government Expenditure on Education, 2000-2007 Source: Ministry of Finance. - 95 - s b 0 0 w .I E 8 L 1 c, .I W G 3a 43 0 ca Q) W e cd I \D m I 43 0 Annex 2: Liberia 2007 Public Financial Management Performance Report BACKGROUND 1. This i s the first systematic baseline assessment o f the performance o f the public financial management system in Liberia. It builds on the analytics undertaken over the last five years by various development partners. It is based on the public financial management Performance Measurement Framework-a monitoring framework used by around 80 countries. This framework tracks country public financial management (PFM) Performance over time. It has been developed by the Public Expenditure and Financial Accountability (PEFA) partners, which include the World Bank, IMF, European Commission, Norway, Department for International Development, and Switzerland. The PEFA indicator set comprises 28 high-level indicators that assess the various components o f the public financial management system and 3 that assess donor performance. The information provided by the indicators can be used for developing a PFM reform program and for monitoring progress over time. Note, however, that the PEFA indicatorsmeasure the outcome o f the system rather than specific actions associated with PFMreforms. 2. Each indicator contains one or more dimensionsto assess the key elements o fthe PFM process. Two methods of scoring are used. Method 1 (Ml) i s used for all single-dimensional indicators and for multi-dimensional indicators where poor performance on one dimension i s likely to undermine the impact o f good performance on other dimensions o f the same indicator. A plus sign i s givenwhere any of the other dimensions are scoring higher. 3. Method 2 (M2) is based on averaging the scores o f individual dimensions of an indicator. It i s prescribed for selected multi-dimensional indicators, where a low score on one dimension does not necessarily undermine the impact o f a highscore on another dimension o f the same indicator. Though the dimensions all fall within the same area o f the PFM system, progress on individual dimensions can be made independent of the others and without having to follow any sequence. A conversion table is provided for 2, 3, and 4 dimensional indicators to arrive at the overall score. In both scoring methodologies, the `D' score i s considered the residual score, to be applied ifthe requirements for any higher score are not met. 4. The overall assignment has beencarried out by the Public Expenditureand Financial Accountability Review (PEMFAR) team, with participation from the African Development Bank, DFID, IMF, UNDP, the Swedish National Auditing Office (SNAO), and the World Bank. The report was prepared based on findings from three missions. The first was conducted in October 2007 by Chukwuma Obidegwu, Tidiane Toure, Winston Cole, Emmanuel Fiadzo, Eric Nelson, and Rebecca Simson from the World Bank; Loxly Epie and Mar0 Gabriel from the African Development Bank; and consultants Robert Cauneau, Simplice Zouhon Bi and Charles Taylor. A second follow-up mission was conducted in November 2007 by Parminder Brar, Winston Cole, Allister Moon, Emmanuel Fiadzo, and Rebecca Simson from the World Bank. A mission by Lage Olofsson and B o Sandberg from the SNAO inJanuary 2008 provided further input on the audit function. The missionsworked closely with the Ministry of Finance (MoF), the Bureau o f the Budget (BOB),General Auditing Commission, Public Procurement and Concessions Committee (PPCC), andvarious Ministries and State-Owned Enterprises. The report was reviewed by the Ministry o f Finance, - 97 - IMF, UNDP, and DFID inFebruary to April 2008 and revised based on comments and input received. The report was also reviewedby the PEFA Secretariat inlate 2008. 5. The report was prepared around late 2007 and early 2008, but the PEFA scoring provides a snapshot o f performance in 2007. Some later information has been provided for clarity, but the ratings do not take into account developments in2008. INTRODUCTION 6. Liberia's public financial management systems have improved since 2006 through an ambitious reform agenda agreed to by the government and internationalpartners. Key reforms include the development o f financial rules, changes made to the budget calendar, and a more inclusive budget formulation and implementation process. The allotments issued by the Bureau o f the Budget (BOB)have been linkedto cash and procurement plans submitted by the line ministries and agencies. The implementation o f commitment control has ensured that no new arrears have been incurred. Improvements in the business process and the work environment in the MoF have decreased the processing time o f payment requests with qualitatively better checks. Payment o f civil servant salary has improved-salaries have been paid consistently on a monthly basis. MoF has begun to pay employees in the counties by using Central Bank mobile payment centers, United Nations Mission in Liberia, and by placing pay masters inthe counties. There have also been efforts to computerize some o f the business processes o f the PFM cycle-the automation of the EDP payroll, the check printing for the CMCo approved payments and the implementation o f the Liberian Expenditure Control and Accounting Program for cash plans, monitoring o f allotment, tracking vouchers, and reporting on expenditure. These have laid the grounds for implementation of a more sophisticated Integrated Financial Management Information System (IFMIS). Significantly, there is substantial government leadership and ownership o f this reformprocess. 7. There i s also considerable support for capacity-building efforts through a combination o f donor support and indigenous ownership, making the whole process more sustainable in the long run. A two-year Masters inManagement program-a collaborative initiative o f the MoF, Civil Service Agency, Liberia Institute o f Public Administration and the University o f Liberia-is a significant step forward. The first batch has 29 trainees who on completion o f the program (December 2008) will be placed in finance divisions o f key ministries and agencies. These students and other MoF staff will later participate inthe Accountancy Bodies o f West Africa (ABWA) technician examination school funded by a Low Income Countries Under Stress grant. A second batch o f 30 trainees joined the program in February 2008. In addition, the MoF has initiated regular on-the-job training and workshops to build capacity and take the reform process to the various stake holders inthe PFM cycle. 8. The following report provides a detailed assessment o f Liberia's Public Financial Management system, using a standardized set o f 28 PEFA indicators plus 3 PEFA indicators pertaining to donor practices. Since this i s the first time the final set o f performance indicators (PI) have been used in Liberia and shared with all stakeholders, the report contains some additional details on the scoring methodology. Note, however, that the PEFA scorings look at practices over past years and thus cannot fully capture the strong, ongoing reform progress. The scores aim to provide a baseline for comparison with performance in future years. The - 98 - PEFA framework is a standardized assessment tool for PFM institutions and thus does not fully capture the strength of special, transitional PFM arrangements established in post- conflict settings such as Liberia. 9. Liberia's PEFA indicators are mixed, with 13 o f 28 indicators scored as D/D+, and 10 indicators scored as C (two PEFA indicators did not receive a score). The indicators relatedto donor performance were also scored poorly with two scores o f D and one indicator that could not be scored. PEFA indicators pertaining to budget implementation and budget accounting, auditing, and monitoring scored particularly poorly. However, given Liberia's outdated institutional system for PFM and the severe stress to which it has been subjected due to civil war and economic collapse, the scores are relatively strong and reflect strong performance since the end o fthe civil war. - 99 - Table A2.1: Summary of 2007 PEFA PFM Performance Scores A. PFM-OUTTURNS: Credibilityof the budget Scores II1PI-1 Aggregateexpenditure outturncomparedwith original approvedbudget B PI-2 Composition of expenditure outturncomparedwith original approvedbudget D PI-3 IIIAggregaterevenueoutturn comparedwith original approvedbudget III A III I PI-8 I TransDarencvofintereovernmentalfiscalrelations I Noscore I II PI-9 II Oversightof aggregatefiscal risk from other public sector entities II D PI-10 Public access to key fiscal information C II C. BUDGETCYCLE II PI-14 Effectivenessof measuresfor taxpayer registration andtax assessment C PI-15 IIEffectiveness incollection o ftax payments II D+ II III PI-25 PI-23 Availability of information on resourcesreceivedby service delivery units D PI-24 IIQuality andtimeliness of in-year budget reports II C II I Quality andtimeliness ofannualfinancial statements I D I PI-27 Legislative scrutiny ofthe annualbudgetlaw C+ PI-28 Legislative scrutinv of externalaudit reDorts I ' I No score D.DONORPRACTICES I II D-1 I Predictabilityofdirect budget support I Noscore I Financial information provided by donors for budgeting and reporting on project and D-2 Drogfamaid D I D-3 IIProportion of aidthat i s managedby use ofnational procedures lI D I - 100- Assessment of the PFMsystems, processes, and institutions Description of the legal and institutional framework for PFM 1. The main legal and regulatory framework for PFM in Liberia is contained in The Executive Law 1972 and The Revenue Code o f Liberia Act o f 2000; amendments have been made to them over the past years. The Public Procurement and Concessions Commission(PPCC) was established though the Public Procurement and Concessions Act o f September 2005. 2. The General Auditing Office (GAO) was established by an act o f the national legislature as an autonomous agency withinthe executive branch o f the government. The 1956 act was later amended by an executive law o f 1972 to enforce its mandate and responsibilities. The National Transitional Legislative Assembly in June 2005 amended chapter 53 o f the Executive Law o f 1972 to change the name o f the GAO to General Auditing Commission and to place it under the legislative branch o fthe government. 3. Draft legislations are also with the legislature to move the Bureau o f the Budget from the Office o f the President to the Ministry o f Finance. This will improve the collaboration o f the two offices and ensure a more strategic and integrated approach to budget formulation and monitoring. 4. The legislature is proposingto establish a Legislative Budget andFinance Office (LBFO) to analyze budget documents and fiscal reports from the executive. This proposal should be reviewedwith plans to develop a PFM Act. 5. A `Budget Transfer Act' was passed in early 2008, amending section 2212 o f the Revenue Code, to allow transfers o f maximum 20 percent between agencies. This represents an important improvement, as there hadbeen no formal limits to transfers. 6. Based on Section 2204 (b) o f the Revenue Code, "To prescribe for all government agencies, the forms, systems, and procedure for administrative appropriations and fund accounting ....."; the Ministry o f Finance framed a set o f interim financial rules (February 2006) that prescribe procedures for public spending and form the basis o f approval for all expenditure by the Cash Management Committee. These rules are applicable to all ministries, agencies, and public corporations receiving public funding. Based on the experience with implementing these rules over the last 20 months and across two fiscal years the, Ministry o f Finance has revised these sets o f rules. These revised rules have been approved by the Cabinet (December 2007) and have been issued to ministries and agencies. The Ministry o f Finance also introduced the interim commitment control system (July 2006) to ensure that the government did not accumulate any fresh arrears. Inaddition, the government has introduced through an executive ordinance a policy on foreign travel (April 2006) and domestic travel (January 2007) to regulate the travel entitlements o f government employees. There has been an effort to strengthen the institutional framework and make it more accountable and transparent. 7. The PFM legal and regulatory framework needs to be developed in a comprehensive manner to identify roles and responsibilities o f the involved institutions, establish internal controls, and institutionalize the internal audit function; formalize reporting requirements; and establish a code of conduct for public officials. The government i s currently working to develop - 101 - a comprehensive PFM Act, which aims to address some o f these shortcomings, with technical support from the IMF. 8. The government's institutional framework is summarized in the table below, which shows the number o f PFM institutions and the volume of resources managed in executing government programs. Table A2.2: Number ofPublic Financial Management (PFM)Institutions and the Volume of ResourcesManaged in ExecutingGovernment Programs PFMInstitutions Institutions Number of Entities %oftotal public expenditures(2007/08 budgeted) Central government* 46 93.1% State-owned enterprises 16 3.9% Autonomous government 10 3.0% agencies** Sub-national governments - - *Includes ministries, departments, and agencies ** Includes all commissions/agencies receiving direct transfers. OVERVIEW OF EXPENDITURE PATTERNS 9. Expenditure in Liberia has risen rapidly since 2004/05, as the government's revenue based has recovered. Expenditure grew by an average o f 23 percent a year over the past five years, from US$81 million to an estimated US$199 million between 2005/06 and 2007/08. The tables below detail expenditure by functional and economic classification. Due to incomplete reconciliation o f expenditure, the fiscal reports from 2005/06 and 2006/07 include expenditure labeled as "other." Table A2.3: Actual Expenditure by Functional Classification (as percent of total expenditure) - 102 - Table A2.4: Actual Expenditure by Economic Classification (as percent oftotal expenditure) A. PFMOUTTURNS:CREDIBILITY OFTHEBUDGET PI-1: Aggregate expenditure outturn compared with original approved budget 10. The budget i s a planning instrumentthat determines how policy objectives are translated into actual achievements through public service delivery. The effectiveness o f the budget as a planning tool i s reflected in the performance o f outturn against the original budget. This first section of the report analyzes the credibility o f the budget based on expenditure, revenue, and arrears. 11. Comparing the original budget total to aggregate outturn (excluding debt payments) shows that expenditure was fairly close to the budgeted amount in FY06/07. In FY05/06 however, actual expenditure was 14.5 percent lower than the budget estimate. This was largely due to revenues falling short o f projections; only US$86 million was collected in revenue, compared with original estimates o f US$98 million. The FY05/06 budget was also influenced by the elections in 2005 and change in government in early 2006, which disrupted both revenue collection and expenditure patterns. InFebruary 2006, the National Transitional Government o f Liberia (NTGL) was replaced by the democratically-elected President Johnson-Sirleaf administration. While the NTGL period was marked by widespread misuse o f public funds, the new government put stringent mechanisms inplace to control public expenditure and developed a recast budget for February to June 2006, passed first in April 2006. These new expenditure controls and delays inpassing the recast budgetresulted inslower spending than targeted during the second half of the fiscal year. 12. Data for 2004/05 has been included, but the fiscal report for that year is o f low quality, and data i s poorly disaggregated. Actual expenditure exceeded budgetedexpenditure inFY06/07 becausetotal revenue collection was higher thanprojected. Budgeted expenditure US$ million Actual expenditure YO (excluding debt service) US$ million Difference variation FY06107 119.3 123.1 3.8 3.2% FY05106 86.1 73.6 -12.4 -14.5% FY0410547 69.2 75.1 5.9 8.5% 47Data on debt service expenditure for 2004/05 is unreliable, as fiscal reporting was poor. - 103 - B rn iginal (i) difference The B between actual primary Approved Actual expenditure and the Variation originally budgeted primary expenditure 2006107 119.3 123.1 3.2% 2005l06 2004105 69.2 8.5% Inonly one ofthe last three years (2005106), has the deviationexceeded the budget by more than 10%. PI-2: Compositionof expenditure outturncomparedwith original approved budget 13. A comparison o fbudgeted and actual expenditure shows significant variance for 2006/07 and 2005/06, while 2004/05 has been excluded due to lack o f disaggregated data. Variance in expenditure composition exceeded overall variance in primary expenditure by 19 percent in 2006/07 and 12 percent in2005/06. In2005/06, the budget was recast inearly 2006 after the new administration took office, which resulted in considerable discrepancies between the original budget-prepared by the NTGL-and final expenditure. In2006107 a supplemental budget was passed in 2007 as a result o f strong revenue performance, which altered the distribution o f resources between the spending entities. Until 2008, there were no limits to the volume o f budgetary transfers that could be authorized without legislative approval. The National Budget Committee approved all budgetary transfers. An act was passed in early 2008 limiting interagency budgetary transfers to 20 percent. D nal (i)Extenttowhich D variance inprimary expenditure composition Variation exceeded overall deviation in primary 2005106 12% expenditure 2004105 Not available. The variance inthe compositiono fprimary expenditure exceeded the deviation inoverall primary expenditure by more than 10 percent intwo out of the last three years. Data was not available for 2004105. PI-3: Aggregate revenue outturn comparedwith original approved budget 14. The government's biggest source o f revenue is taxes on international trade, followed by taxes on income and profits and domestic taxes on goods and services. Liberia has for several decades operated a flag o f convenience for international vessels; it generates approximately 10 percent o f its revenue from vessel registration fees. Liberia's revenue has increased rapidly since - 104 - 2006, owing largely to an increase in international trade as the economy rebounded and GDP grew, as well as strengthened tax administration, which increased custom and excise revenue. Actual Revenue US$ Estimated Revenue US$ m m Difference % Variation FY06/07 129.917 148.34 18.423 12% FY0.5106 97.894 85.575 -12.319 -14% FY04/0548 80.000 82.687 2.687 3yo 16. This variance in predicted and actual revenue collection is partly a reflection o f the difficulty o f predicting GDP growth and revenue growth during the immediate post-war period. Revenue collection was lower than expected in FY05/06, owing to uncertainty surrounding the 2005 elections-which disrupted revenue collection-as well as the lack of confidence in the NTGL. Since the present administration took office in early 2006, tax administration has been strengthened significantly, abuse of duty waivers has been reduced, and confidence in the Liberian economy has grown. This had lead to higher rates o f investment and importation and a growth in the tax base. `The underestimated revenue in 2006/07 is also a function o f the cash- based budgeting system and the government's commitment not to borrow funds, which encourages greater caution and modest revenue predictions, as the government cannot borrow to smooth spendingincase o f a revenue shortfall. PI-3 Aggregate revenue A budget. (i)Thedifference A between actual revenue Approved Actual collection compared Collection Variation with original revenue 2006107 129.9 148.3 12% estimates inthe budget 2005106 -14% 2004105 82.7 Actual domestic revenue collection fell below 97 percent of budgeted domestic revenue estimates in only one of the past three years. Inthe other two years, collection exceeded the revenue estimates. PI-4: Stock and monitoring of expenditure payment arrears 17. The Liberian government has faced a complex debt situation. The country had remained ina state of default for over 20 years and faces anenormous debt burden. Itbegan accumulating arrears to the IMF in 1986. The country's debt service payments virtually ceased over the course ~ 48 Data on debt service expenditure for 2004/05 is unreliable, as fiscal reporting was poor. - 105 - o f this period. The nominal value o f its external debt was estimated at US$4.8 billion as o f end- June 2007, o f which 96 percent was in arrears. Multilateral, bilateral, and commercial creditors each hold approximately a third o fthe debt stock. Inaddition, the government has begunto make payments against its sizable domestic debt stock. NOMINAL DEBT I S T O C K US$ million External 4,798 Multilateral 1,615 Bilateral 1,543 Commercial 1.640 Domestic 914 18. Data on the stock o f arrears has been generated through a debt reconciliation process, completed in April 2007. Nearly all records were destroyed during the civil conflict and had to be reconstructed based on creditor records, where the integrity o f the creditor's records could be trusted. The government used technical assistance to track and verify commercial debt. Liberia's domestic debt was reviewed in 2006-07, resulting in the development o f a domestic debt strategy. The nominal value o f domestic claims was estimated at US$914 million in 2006. All claims were subsequently reviewed and claims o f US$304 million deemed valid. Valid claims will be honored after applying different discount rates (between 0 and 88 percent) dependingon the size ofthe debt. The government has set upa special trust fundfor domestic arrears clearance and repayment o f the Central Bank debt. The government i s working to institute systematic debt monitoringpractices. 19. For the purposes o f PEFA indicator 4, only interest arrears are considered expenditure arrears, and unpaid amortization o f loan principal i s not included inthe calculation. Since much o f Liberia's interest arrears have become capitalized-particularly arrears from multilateral creditors-the total arrears stock to be treated under this indicator i s far lower than Liberia's total arrears stock. Disaggregated data showing interest arrears and principal arrears for commercial and domestic debt could not be determined for the purpose of this review. However, interest arrears on multilateral debt alone was $847 million as o f June 2007, which represents over 600% o f expenditure (2006/07)-this far exceeds the threshold for a "D" score on this indicator dimension. 20. Liberia has begunto resolve its unsustainable debt situation, successfully clearing arrears to the World Bank and AfDB in December 2007 and to the IMF in March 2008. Upon clearing arrears to its multilateral creditors and successfully entering an IMF Poverty Reduction and Growth Facility program, Liberia reached HIPC decision point inMarch2008-with multilateral and bilateral creditors agreeing to debt relief with a common reduction factor o f 91 percent. It i s expected that Liberia will qualify for further debt relief under MDRI once HIPC completion point is reached. Most bilateral creditors have indicated an intention to forgive 100 percent o f Liberia's debt. Negotiations with commercial creditors i s ongoing and proceeding well. Domestic debt payments have begunto be made in accordance with the government's domestic debt strategy, including repayments to the Central Bank o f Liberia. - 106 - 21. The Government has accumulated no new external or domestic debt since 2002/03. Since 2006, it has runa cash budget policy and implemented strict commitment controls, which do not allow any further deficit financing. Liberia will not be in a position to contract new external loans, other than Poverty Reduction and Growth Facility credits, until it has reached HIPC completion point. I arrears (i) stockof The The total stock o f interest arrears is in excess o f 10 percent of D expenditure arrears as total expenditure. percentage of total expenditure for the Due to lack o f data disaggregating interest arrears from corresponding fiscal year amortization arrears, the ratio o f arrears to expenditure cannot be determined. However, interest arrears on multilateral debt alone stood at $847 million inJune 2007, and thus well exceeds 10 percent of total expenditure. (ii)Availability o f data Data on debt stock is available throughadhoc exercises, including a C for monitoring the stock major debt reconciliationexercise completedinApril 2007. o f expenditure payment arrears B KEY CROSS-CUTTING ISSUES: COMPREHENSIVENESSAND TRANSPARENCY PI-5: Classificationofthe budget 22. The budget is presented by administrative classification and broken down by object (economic) classification. Revenues are classified ina format consistent with GFS classifications o f revenue at a main head of classification level and at a sub-heading level. For expenditure, the current object classification, which is compatible with GFS 1986, was significantly rationalized for the 2007/08 budget.49Inparticular, inconsistent usage o f detailed object codes between line ministries was reduced, and a standard classification was issued as part o f the 2007/08 budget documentation. For the 2008/09 budget, the authorities are preparing a revised object classificationcompatible with GFS 2001. 23. Prior to the 2007/08 budget, the BOBwas using excel spreadsheets to develop the budget, which resulted in problems with duplicate codes. When the Liberian Expenditure Control and Accounting Program database was introduced inJune 2007/08, several pages o f duplicate codes were identified and cleared up. Each budget line now has a unique code. The 2007/08 budget now has a list o f all approved codes for objects o f expenditure. 24. The 2007/08 budget also introduced the notion o fprograms, although inpractice most o f these were simply administrative sub-divisions of the various organs o f the state and the line ministries. The budget document includes descriptions, measurable objectives, and associated activities for each program. This new budgeting approach is inits first year andwill be improved 49The Bureau o f Budget is receiving support from USAID inthe area o f budget classifications as well as broader reforms to the budget preparation process. - 107 - over the coming years, duringwhich time the definition o f what constitutes a program is likely to evolve. 25. Functional classification is not currently used in Liberia, although the budget uses the notion o f sectors. The organs o f the state and the line ministries are grouped in four sectors," with two additional sections for subsidies to institutions and claims on government. The authorities are considering upgrading to a COFOG based functional classification. 26. There i s no formal coding o f poverty-reducing spending items. However, where programs are considered to be contributing to poverty reduction, additional text has been provided in the 2007/08 budget to indicate the link to policy priorities as defined in the interim poverty reduction strategy paper. These links will be upgraded in 2008/09 budget to the new poverty reduction strategy paper, although it i s unlikely that tracking o f such expenditures at the transaction level will be possible inthe immediate future. 27. The distinction between subsidies and transfers i s not always clear as shown in annex 3. Transfers are made to institutions such as Monrovia City Corporation, Forestry Development Authority, and JFK Hospital, butthese institutions are listed under sectors and not inthe separate category for transfers. Similar institutions that receive transfers and spend from operational bank accounts and submit returns to MoF are classified as subsidies/transfers. The classificationneeds to be consistent to allow meaningful analysis. 28. Other classifications are being considered for budget and accounting, in particular fund and location identifiers. The pressure from the legislature to identify spending by county is encouraging the authorities to consider the use o f a geographic location code. The use o f fund code will also be important once the budget beginsto reflect aid flows. 29. Liberia has yet to develop an adequate chart o f accounts. Transactions are booked against the detailed budget classification. The development o f a full chart o f accounts remains a priority inLiberia, particularly inthe context of the proposed design and implementation of an IFMIS, which i s expected to start inthe short term5'. 30. The new chart o f accounts should aim to cater for the following elements, but a detailed analysis should be done taking into account the reporting requirements for the various users o f the financial statements and the budget controlj2 level before a decision is made for the number o f segments and digits. e Fundsource (CRF, ownrevenue, County Development Fund, donor, and so on) e Administrative (ministry/department/agency and budget unit) a Function a Progrardproject General administrative services, security services, social and community services, and economic services. 51The proposed IFMIS conceptual design document has been prepared with the World Bank, and bid documents are being finalized at this time. The development o f a chart o f accounts will be required prior to the implementation o f the IFMIS. 52Some donor projects are controlled at the component or sub-component level while others go down to the activity level thereby requiringmore digits for a project segment. - 108 - 0 Geographic location 0 Economic 31. The new chart of accounts should be directly linked to the IMF GFS 2001 classification to ensure that GFS reports can be produceddirectly from the accounting system without need for a bridging table. The GFS codes need not be in the main chart o f accounts for coding transactions but as side tables duringthe setup o fthe system. C (i)Theclassification The government budget is classified by organizational units with C systemused for some ministries and agencies (for example, the Ministry o f formulation, execution, Health) including programs as part o fthe organization codes. The and reporting o f the other available segmentis economic classification. The revenue central government's object codes are in line with GFS revenue coding structure. budget The iPRS i s not directly linked inthe classification system; this makesit difficult to strategically formulate and allocate resources to meet the objectives of the iPRS and produce reports for monitoring. The 2007/08 budget however has narratives for the various programs in the ministry and agency, thus some indirect linkcanbe made. Budget formulation and execution is based on administrative and economic classification that can produce consistent documentation. Source: 2007108 Annual Budget documents, informationprovided by Bureau of Budget, and Ministryof Finance. PI-6: Comprehensiveness of information included in budget documentation 32. The information inthe budget documents has been evolving and improving, but further improvements are needed to meet the minimumrequirements o f PI-6. Initial steps taken in the FY 2007/08 budget documents provide far greater informational content then inprevious years. In particular, the document has begun to make the essential linkages between iPRS and budgetary allocations, albeit at a fairly aggregated level. The president's budget speech discusses the macro-economic and fiscal framework background for the budget. The addition o f summary tables and graphs i s also positive. Furthermore, the addition o f information on personnel numbers has helped focus attention on both the numbers o f civil servants and their distribution within the government. The efforts made to provide descriptions, objectives, and activities o f programs are important for establishing a program-budgeting approach in Liberia. Unlike the budget for 2006-07, which assumed a static exchange rate o f L$58:US$1, the 2007/08 budget has linked the conversion rate to the market exchan e rate as o f the last day o f the previous month, as published by Central Bank o f Liberia!3 Notwithstanding these improvements, weaknesses remain and the budget does not contain all details required under Liberian law, as 53 An A c t to Approve the Budget for the Flscal Penod begnningJuly 1,2007 and endingJune 30,2007 Provldingfor the Expenditure of the Governmentof the Republic of Liberia. - 109- stipulated in section 2206 of the Revenue Code. The following outlines the situation regarding the informationbenchmarkshighlighted inPI-6: 0 Macro-economic assumptions (not met): The FY 2007/08 budget document does not include any table outlining the macro-economic assumptions used in formulating the Liberian draft budget. While the "Message to the National Legislature" document details the expectation that GDP growth will average 10 percent over the period 2007-2012, no basis i s offered for this assumption. The document also reiterates that the government is committed to a stable exchange rate to the USD o f approximately 60 to the USD. Although there is some discussion on historical inflation levels, there is no forecast o f inflation for the coming period included inthe budget document. Fiscal deficit (met): Liberia i s currently operating under a cash budget policy where borrowing to finance the budget deficit are not permitted. The government has committed to spendonly as revenues permit; a statement that "there shall be no deficit financing" i s included in the preface to the budget. The budget document includes a table which compares total revenue projection against total appropriations and indicates that surplus revenue from 2006/07 will be used to cover this deficit infull. Thus the fiscal balance is projected to be zero. 0 Deficit financing (met): The budget document indicates that the difference between revenue and expenditure will be financed from 2006/07 surplus revenue. The government intends to runa fully balanced budget and thus the fiscal balance i s zero. 0 Debt stock (not met): Liberia has completed an exercise to verify the existing stock o f domestic debt owed by the government including domestic arrears. The FY2007/08 "Message to the National Legislature" contains a statement on existing domestic debt and an estimate of the outstanding unverified domestic debt (US$9 14.0 million) and verified domestic debt (US$304 million). N o breakdown o f this stock i s included. At the time o f the passage o f the budget, Liberia was negotiating with its international creditors on a proposal for debt relief on its external debts, through the HIPC initiative. Some details o f these outstanding stocks o f external debt are included inthe budget document inthe "Message to the National Legislature," but it does not provide a complete picture. 0 Financial assets (not met): There is no information on financial assets in the current budget document. 0 Prior year's budget outturn (not met): The prior year's budget outturn i s not included inthe budget document for FY2007/08. 0 Current year's budget (met): The original and revised estimates of the current year are included in the 2007/08 budget document and presented in the same format as the budget proposal. - 110- 0 Summarized budget data (not met): Summarized data for both revenue and expenditure are included in the budget documentation. Estimates o f this year's revenues and expenditures are included but no figures for the previous year's outturn. Budget implications of new policy initiatives (not met): The FY2007/OS budget i s a major improvement over previous years, as efforts have been made to identie measurable program objectives based on the current program definitions. In addition, the major policy objectives defined in the poverty reduction strategy paper for each administrative unit are included inthe budget documentation. There i s limited discussion, however, o f new policy initiatives or the budgetary impacts o f any revenuepolicy changes or changes to expenditure programs. 33. Inaddition to the detailed informationon revenuesand expenditures, and inorder to be considered complete, the annual budget documentation should- include information on the following elements: Elements Available 1. Macro-economic assumptions, including at least estimates of aggregate growth, Not met inflationand exchange rate. 2. Fiscal deficit, defined according to GFS or other internationally recognized Met standard. 3. Deficit financing, describing anticipated composition. Met 4. Debt stock, including details at least for the beginning ofthe current year. Not met 5. Financial assets, including details at least for the beginning o f the current year. Not met 6. Prior year's budget outturn, presentedinthe same format as the budget proposal. Not met 7. Current year's budget (either the revised budget or the estimated outturn), Met presentedinthe same format as the budget proposal. 8. Summarized budget data for both revenue and expenditure according to the main Not met heads o f the classifications used (ref. PI-5), including data for the current and previous year. 9. Explanation o f budget implications o f new policy initiatives, with estimates o f Not met the budgetary impact o f all major revenue policy changes and/or some major changesto expenditure programs. - 111- PI-6 Comprehensivenessof informationincludedinbudget documentation (i) oflistedinformationinthe Share The recent budget documentationfulfils 3 ofthe 9 budget documentation most recently informationbenchmarks as summarizedinthe table issued by the central government (in above. order to count inthe assessment, the full specification ofthe information benchmark must be met) PI-7: Extent of unreported government operations 34. All revenues, includingnon-tax revenues collectedby line ministries and agencies, are by law to be credited to the main treasury account. In general, revenues are credited to the main treasury account. However, amounts credited to line ministry accounts through transfers in one budgetyear, but not consumedthat year, are carried over for use insubsequent years, without re- appropriation. They are reflected inexpenditure reports as spent at the time o f transfer to the line ministry. These are estimated to be about 1-2 percent of total expenditure in total across all ministries (excluding state-owned enterprises and other autonomous agencies, as the size o f these balances could not be e~timated).~~Revenues collected by Liberia's embassies abroad are also not systematically credited to the main treasury account or fully reflected in the expenditure reports, althoughthese revenues are small. 35. Some expenditure is being undertaken outside o f the normal procedures. Monthly cash allocations are made to line ministry accounts for small, operational cash-based transactions, which are expensed at the time o f transfer. The details o f these expenditures are therefore not adequately captured in the expenditure reports. There have also been cases o f more significant spending for urgent operations using a force account, which i s inaccordance with the PPC Act, but circumvents normal procurement and payment procedure^.^^ Although reports on this spending are expected to be sent to the Ministry o f Finance, these are not integrated into the expenditure reports. 36. Donor funds are not reflectedinthe budget or the expenditure reports. Although the level o f official aid flows into Liberia is estimated to be up to twice Liberia's domestic revenue resources, other than small levels o f budget support, none o f these resources are currently reflected inthe budget documents, quarterly or annual expenditure reports. However, at present, all o f Liberia's donor funding i s grant funding, as the government i s not in a position to borrow, thusno loanfinancedoperations are going unreported. 37. Liberian Reconstruction and Development Committee with assistance from UNDP has set up a databaseto track donor-fundedactivities, but it does not yet contain comprehensive data. BOBalso designed forms for the 2007/08 budget preparation for ministries and agencies to 54This estimate is based on balances on line ministry accounts at month end, which for August 2007 was about US$3 million, or 1.5 percent o f budget (only data at end-August was made available at the time o f the assessment). 55One particular recent instance o f this is a construction o f a temporary by-pass to avoid congestion arising out o f the donor-funded rehabilitation o f a key road out o f Monrovia. - 112- complete to record all donor-funded programs implemented through the government for inclusion in the budget, but the response was low. The government is committed to working closely with donor partners to improve donor aid reporting, as clearly articulated inthe poverty reduction strategy. PI I I D+ I un vernment (i) ofunreported Level The actualdetailed spendingby ministriesand agenciesfor cash B extra-budgetary transferredto departmentalbankaccountsare not capturedinthe expenditure overallfiscal reports. The level of unreportedextra-budgetary expenditure(other thandonor-fundedprojects) is estimatedat 1-5 percentoftotal expenditure. (ii) Incomeiexpenditure Informationon donor financed projects included in fiscal reports D information on donor- i s seriously deficient. funded projects Source: Revenue Code of Liberia Act 2000,2007/08 Annual Budget documents, discussions with Bureau o f Budget, Ministry of Finance, Central Bank o f Liberia, Ministryof Public Works, and Liberia Reconstruction and Development Commission. PI-8:Transparency ofIntergovernmental FiscalRelations 38. There are no subnational governments inLiberia, as Liberia i s a unitary state by law. The counties are a deconcentrated part o f central government, county superintendents are appointed by the president, and county authorities do not have any independentrevenue collection rights, nor do they currently directly manage any spending beyond small community development funds. 39. A County Development Fundwas started by the Legislature in 2006/07 to fund locally- identified projects at the county level. The County Development Fund was established in 2006/07, but so far spending has been slow. It was established with a US$2.5 million annual budget, to be disbursed for projects identified by local committees at the county level, with spending outside o fthe normal commitment control procedures. Concerns over financial controls and weak capacity at the county level have led to delayed implementation o f the scheme, pendingthe preparation o f guidelines and procedures, including reporting, to be followed inthe execution o f these funds. As a result the entire unspent amount was transferred to an escrow account opened at Central Bank with the approval o f the Cash Management Committee. Funds are divided equally between the 15 counties. 40. The National BudgetAct 2007/08 provides for a more detailed procedure on the County Development Fund that has an appropriation of US$3 million divided equally between the 15 counties. The Act stipulates the disbursement procedures and provides for the establishment in each county o f a Project Management Committee that shall be responsible for identifying and implementingcounty development projects. The Ministry o f Finance in consultation with the Ministry o f Internal Affairs and County Administrators has formulated expenditure and reporting guidelines for the County Development Fund. All the counties are opening bank accounts and setting up procurement committees for the implementation o f the development projects. Given - 113 - the limited banking network, it i s going to be a challenge implementing activities inthe proposed decentralized framework. PI-$Transparency of No score Intergovernmental FiscaiRelations I I (i) Transparency and There i s no subnational government inLiberia. objectivity inthe horizontal allocation amongst subnational govemments- unconditional and conditional transfers (both budgetedand actual allocation) ~~~ (ii)Timeliness and There is no subnational Government inLiberia. reliable informationto S N governmentson their allocations (iii) of Extent There i s no sub-national government inLiberia. consolidation of fiscal datafor general u government I ;ource: Ministry o f Finance, County Development Act o f2006, draft Guidelines and Procedures for County Development Fund. PI-9: Oversight of aggregate fiscal riskfrom other public sector entities 41. Financial information from autonomous government agencies i s provided during budget preparation, but not routinely monitored. Autonomous government agencies are expected to provide information to the Bureau o f Budget when they are making their annual budget request. The quality o f this financial information varies considerably, given the general PFM capacity constraints. Some o f these entities are receiving significant funds from external donor sources, but these are often not adequately reported. There is no unit inthe MoF dedicated to monitoring the activities o f autonomous government agencies. At the line ministry level, the capacity to monitor AGAs that are under their purview i s generally poor. 42. The fiscal risks o f commercial public sector entities are still only partly identified. The annual 2007/08 budget document identifies-in the section on Institutions and Entities- subsidies to a number o f state-owned public sector entities. There is no information inthe budget document indicating the reasons for the subsidies, nor i s there any information on contingent liabilities that may arise from these entities. Under the GEMAP arrangement, co-signatory arrangements were introduced into the National Ports Authority, Roberts International Airport, Liberia Petroleum Refining Corporation, Forestry Development Agency, and Bureau o f Maritime Affairs. Inaddition, a Chief Administrator was assigned to the Central Bank o f Liberia and with co-signatory powers and a PFM Advisor exercised co-signatory powers in the MoF. Thus, there is significant oversight o f these state-owned enterprisesthrough the GEMAP reporting mechanisms. There also exists a Bureau of State Enterprises, which is responsible for oversight and regulation with regards tostate-owned enterprises. However, this bureau i s weak and ineffectual. - 114 - (i) ofcentral Extent Quarterly and annual monitoring of key revenue-generating state- D government monitoring owned enterprises takes place inthe context of GEMAP. However, no ofAGAsREs government-led systematic annual monitoring of autonomous government agencies and state-owned enterprises currently takes place. (ii)Extentofcentral There is no subnational government inLiberia. No score government monitoring of SN governments' fiscal position PI-10: Public access to key fiscal information 43. Public access to key fiscal information is improving, although much remains to be done. The availability o f the annual budget documents56and the publication o f annual and quarterly fiscal reports on the MoF ~ e b s i t e ,as~ ~well as the publication o f contract awards o f over US$25,000 in the PPCC's quarterly bulletin, are all positive steps. The 2006/07 annual fiscal outturn report was publishedin the newspapers and spurred significant public debate about the allocation and use o f funds. However, because the quarterly reports are not posted within one month o f completion o f the quarter, Liberia does not meet the benchmark for this performance indicator. 44. The Auditor General provided comments on the 2007/08 budget document when it was submitted to the legislature. This attracted significant interest from civil society, as evidenced by the large number o f radio discussions and newspaper articles with detailed budgetanalysis. 45. Liberia does not yet produce annual financial statements, as the accounting function is not yet fully restored and only operates on a single entry basis. However, the government has prepared and published, within 4 months o f the end o f the fiscal year, a detailed annual fiscal outturn report for 2006/07,58 which provides aggregate information on revenue and expenditure outturns with detailed explanations. Nevertheless, without an accounting base, these reports cannot be considered substitutes for financial statements. When this report was published inlocal newspapers and on the MoF website, it provided an opportunity for the public to scrutinize and hold the executive accountable. Some ministries and agencies and expected beneficiaries o f allocations questioned the accuracy of the report. This led to the MoF holding a workshop to discuss andexplainthe report. 46. External audit reports are not available, as the General Auditing Commission has yet to complete its first set o f audits. Following the adoption o f the Audit Act o f 2005, which gave autonomy to the General Audit Office, and the appointment o f an Auditor General, the office i s 56http://www.mofliberia.org/nationalbudget.htm http://www.mofliberia.org/expenditurereports.htm 58http://www.mofliberia.org/fiscalreports.htm - 115- beginningto establish itself. It has begunperforming audits o f the 2006/07 fiscal year for select ministries,andexpects to complete its first report shortly. 47. The PPCC began publishing notices o f all contracts awarded above US$25,000 in its quarterly bulletin. However, the bulletin was only published once (Fall 2007) - no subsequent issues were produced. The PPCC website does not contain up-to-date information about contracts awarded. 48. Information on the resources for primary service units i s not publicly available. The demand for such information, however, i s growing, especially from members o f the legislature who want to know what resources are being used within their counties and districts. A proper chart o f accounts will track resources to cost centers and locations (geo-codes) through the accounting system. 49. While the MoF i s setting up the appropriate structures and systems to improve the PFM system, it has made significant effort to ensure public access to key fiscal information. 50. Elements o f informationto which public access i s essential include: o sectors (such as elementary schools or - 116- (i) oftheabovelistedinformation Share The government makes available to the public 1 c inthe budget documentation most ofthe 6 listedtypes of infonnation. recently issued by the central government (inorder to count inthe assessment, the full specification ofthe information I I I I Lbenchmark must be met) Source: Ministry of Finance website (www.mofliberia.org), Public Procurement and Concessions I Commission quarterly bulletin and website (www.ppcc.gov.1r). c(1). POLICY-BASED BUDGETING PI-11: Orderlinessand participationinthe annual budgetprocess 51. The fiscal year for the government o f Liberia i s July 1to June 30. 52. The budget calendar i s driven by the deadline for submission o f the budget to the Legislature, set in each year's Annual Budget Law for the next year. For the last two years, this has been set for May 20, allowing Parliament 40 days to debate the budget before their Constitutionally-defined obligation to pass the budget before the start o f the fiscal year. BOBis responsible for preparing an annual budget calendar to meet this deadline, to be approved by the National Budget Committee.59So far, the budget calendar6' is determined each year, subject to specific emerging issues (such as the poverty reduction strategy process), and it has been too soon to establish a more stable timetable o f budgetary events. However, the budget guidelines and the associated issues policy guidance from the President are an expected outcome o f the mid-year Cabinet review, which usually takes place in early January. The BOBfollows up the issuance o f the budget circular with workshops for line ministries. Capacity constraints in line ministries are still hindering their ability to submit their budgets on time. Budget hearings are heldby the Budget Committee, withBOBproviding technical support. 53. The Cabinet is actively involved inpreparingthe issues policy guidance and the budget circular, which includes ceilings. These two documents provide the policy and technical framework for the preparation o f the annual budget. The circular includes indicative ceilings set for each sector and each ministry, as well as specific guidelines on how to cost various input elements, such as wages, goods and services for the budget submissions. However, given the one-year focus o f the budget, the significant demands on very constrained resources, and the lack of adequate capacity inline ministries, line ministries often do not respect the ceilings set in the circular when making their budget submissions. The burden then lies on the Budget Committee, which has to challenge excessive line ministry submissions duringthe budget hearings. 54. The legislature has passed the budget with a 1-2 month delay in the last 3 years. The dates o f submission to and adoption by the legislature for the last 3 budgets are as follows: 59The establishment o f a Budget Committee chaired by the Minister o f Finance has improved coordination of the budget preparation process. The Budget Committee brings together the Ministry of Finance, the Bureau of Budget, the Ministryo f Planning, the Ministry of State, the Civil Service Agency, and the General Services Agency. 6oAnnex 3 provides a report on implementation of the 2006/07 budget calendar. - 117- 0 FY2007/08 submitted 17thMay, 2007 and adopted 17thJuly, 2007. 0 FY2006/07 submitted 29th June, 2006 and adopted 23rd August, 2006. 0 FY2005/06 submitted after start offiscal year and adopted 19thAugust, 2005. PI-ll Orderlinessand B (i)Existenceof,and The BOBissues abudget calendarto ministriesand agencies, together B adherence to, a fixed with the budgetguidelines.According to the calendar, ministriesand budget calendar agencies are given four weeks for receipt ofthe guidelinesto submissionof first estimates. However, ministries and agencies continue to experience some difficulties incompleting their estimates on time, although there are have beenmarked improvementsinthe timeliness of submissions andpassageofthe budget over the years. (ii)Guidanceonthe The Cabinet is active indrafting the policy and sectoral ceilings A preparation of budget through the Budget Committee [ministerial as well]. Their submissions collective decisions are contained inthe budget circular that is then distributedto the ministries and agencies. The budget circular is simple but clear. There i s a Budget Committeewhich i s a sub-group of central cabinet agencies [Finance, Budget, Mino f State for Financial Affairs, Planning, Civil Service Agency and General Services Agency]. They offer the first advice on the proposed policy. Their suggestions then go to the fuller Cabinet, including the President. Then the budget circular is finalized. (iii) budget timely The financial year for the governmentis 1stJuly to 30th June. C approval by the legislature 2007108 submittedon 17th, May, 2007 andpassedby the Legislature on 17th, July, 2007. 2006/07 submittedon 29th, June, 2006 andpassedby the Legislature on 23rd, August, 2006. 2005/06 submittedafter start of fiscal year and passed by the Legislature on 19th, August, 2005. Inthe lastthree years, the budgethasbeenapprovedwithintwo months of the start ofthe new fiscal year. PI-12: Multi-year perspectivein fiscal planning,expenditurepolicy, and budgeting 55. Liberia i s not currently usinga multi-year budget, although the government has expressed an interest ina medium-term expenditure framework. Except for certain sensitive appropriations such as the County Development Fund (which i s transferred into an escrow account if unalloted inthe current fiscal year), all appropriations lapse at the end of the fiscal year. This makes it difficult to plan and initiate multi-year development programs since fundingmay be discontinued at the end o f the fiscal year. The government has begun to initiate a program approach in the budget, although this i s built around administrative units within each ministry rather than outcomes. As the program approach develops, this will create pressures to develop a medium- term horizon. However, experiences with medium-term economic frameworks in other developing countries has been mixed, and there i s a risk that it may sap scarce government - 118- resources. Another critical blocking factor in the development o f realistic medium-term fiscal projections is the poor quality o f data on macro-economic projections.61 56. The linkages between policies and budget allocations i s gradually being strengthened. In the current fiscal year, the interimpoverty reduction strategy paper was used to guide allocation decisions. Based on the interim poverty reduction strategy paper, it was decided that resources should be concentrated in key pro-poor areas such as health, education, public works, and agriculture, However, weak capacities in line ministries have so far limited the preparation o f sector strategy documents and more detailed budgetplanning at the sectoral level. There i s also a need to improve reporting o f donor-funded programs and incorporate them into the sector strategy and budget planning processesto enable a budgetingprioritization consistent with policy goals. Moreover, the absence of a program or function segment inthe chart o f accounts makes direct linkage o f expenditure reports to interimpoverty reduction strategy. 57. In the preparation process for the 2008/09 budget--cunently under review by the legislature-allocation decisions were guided by the full poverty reduction strategy, which was beingfinalized at the time o f budget preparation. As part o f the poverty reduction strategy paper, the government undertook a three-year costing exercise, yielding costing estimates by fiscal year and sector required to meet the poverty reduction strategy objectives. It also estimates the anticipated budgetary resources to be devoted to achieving the poverty reduction strategy objectives and indoing so provides a first step toward a multi-year budget. These estimates were developed based on dialogue and workshops with line ministryrepresentatives, but are subject to some uncertainty and will need further scrutiny in each budget year. Nonetheless, the costing exercise also illuminates the multi-year financing gap that will affect the government's ability to rebuild the economic and social infrastructure needed to underpin economic revitalization and combat poverty. It thereby illuminates the support required from donors. The government plans to institute poverty reduction strategy coding inthe chart o f accounts and an improved donor aid flow tracking system, to enable monitoring o fpoverty reductionstrategy expenditure. 58. A preliminary debt sustainability analysis has been prepared for Liberia, based on partially reconciled external debt data. This work has beenprepared by the lMF, incollaboration with the authorities and the World Bank, as part of the decision point HIPC document, and also on low-income country debt sustainable analysis, domestic and external, for the Poverty Reductionand Growth Facility document. Work on debt stock has been completed, along with a debt strategy to clear these over time. However, there are still large numbers of contestable claims, which are being reviewed. 59. Liberia has consolidated its previously separate recurrent and development budgets into a single coherent budget. The budget is now prepared and documented under a unified budget cycle led by the MoF and the Bureau o f Budget. It is important to note, however, that a significant part o f investment expenditure is donor funded and therefore is off-budget (see indicator PI-7). 61The Bureau o f Statistics will not be able to produce GDP estimates untilat least 2009, possibly intime for the preparationo f the 2009/10 budget, according to Ministry o f Planning and Economic Affairs. -119- (i) multi-year fiscal Government prepares a single year budget only. Allocations are D forecasts and functional done by administrative units for broad sectors since there i s no allocations functional classification o fthe budget. (ii) andfrequency scope A preliminary debt sustainable analysis for external debthasbeen C o f debt sustainability preparedby the IMFandWorld Bank, incollaborationwiththe analysis government, inpreparation for the HIPC process. (iii)existence o f costed Pillar strategies were developed and costed as a part o f the C sector strategies poverty reduction strategy process. However, detailed individual costed sector strategies are not available, and most ministries and agencies lack the capacity to plan and cost the activities in line with fiscal forecasts. (iv) linkages between Investments are largely donor funded and the projects are D investment budgets and executed by Project Implementation Units within the government forward expenditure or NGOs and other civil society organizations, without an estimates effective information sharing arrangement during budget preparation to ensure comprehensiveness and consistency with sector strategies. The budget therefore does not take account o f recurrent cost implications o f donor-funded projects for inclusion in forward estimates. (11) PREDICTABILITY AND CONTROL INBUDGETEXECbTION 60. Inpost-conflict Liberia, revenue administrationis beingrebuilt incapacity, infrastructure, systems, and procedures from a severely degraded base. While substantial reforms-which have helped to boost revenue significantly-have been achieved to date, continued efforts are required to achieve an efficient revenue administration and increase revenue collection. 61. The department o f revenue has been reformed significantly over the past two years. Tax administration has shifted from a tax type organization-with bureaus for direct taxation, internal revenue, and functional activities such as audit and collections-to an organizational structure based on tax payer size. The MoF currently has two revenue administrationbureaus, the Bureau o f Customs and Excise and the Bureau o f Internal Revenue. The Bureau o f Internal Revenue in turn, contains the large, medium, and small tax payer units-and each unit is responsible for tax collection and auditing o f its respective tax payer group. Procedures are being streamlined to support wider-scale automation, through the introduction o f an Integrated Tax Administration System in 2008/09. And capacity-building exercises are ongoing to improve the human resource capacity. PI-13: Transparency oftaxpayer obligations and liabilities 62. The tax administration and policy legal setting is provided by the Liberia Revenue Code (LRC) o f 2000, which covers the various tax regimes for goods, services, income, and profits. - 120- The code has been widely publicized and i s available on the MoF website. However, the code lacks specificity in certain areas-such as fiscal regimes for extractive industries-which has allowed the government some discretionary powers when setting tax rates. After the LRC's enactment in 2001, the national transitional government and its predecessor issued many regulations and administrative circulars, including some changes to tax bases and rates that should have been legislated. Additionally, tax concessions were granted in breach o f the law. Accordingly, in early 2006 the new government suspended all LRC regulations and circulars pending their review, revision, or repeal. This left ministries and agencies with the responsibility o f interpreting the act, which increased the risk of inconsistencies and discretion. Presently, the LRC i s beingreviewed with support from the IMF.A revisedcode i s expected to be submittedto the legislature shortly, containing more detailed tax provisions, particularly for the extractive industries and other concessionaries. 63. The assessment and determination o f customs values and duties i s relatively predictable and reliable given the use o f pre-shipment inspectionservices. 64. The MoF has made a significant effort to increase taxpayer services and education programs. This effort has included publishing information about the tax system and ongoing reform inthe newspapers, partaking inradio talk shows, and producing other materials. Each tax division contains a registration return service, where tax payers can seek guidance on regulations and procedures. However, a tax or customs administrationwebsite does not yet exist. 65. The LRC provides for the establishment o f an independent Board of Tax Appeals; the MoF has recommended the establishment o f the board and is awaiting the President's decision to appoint members. A department of appeals was established within the MoF in October 2007 to review and assess complaints, but it i s not yet fully operational. The appeals procedure has not been formalized, and very few appeals have been received or processed. Tax payers regularly petitionthe ministry directly for an assessmentto be reconsidered. (i) andcomprehensivenessoftax Clarity The LiberiaRevenue Code has been circulated C liabilities widely and is available on the MoF website. However, regulations and circulars were suspended in2006, which creates room for uncertainty and discretion around implementation o fthe law. A review of the LRC is ongoing, which will lead to further amendments and greater specificity inthe LRC. PSI services provide reasonably reliable tax estimates for importers. (ii)Taxpayeraccesstoinformationontax Taxpayer education and services programs exist, liabilities and administrative procedures although web-based information is not yet available. i c (iii) Existence and functioning o f a tax A division for appeals has been established within C appeals mechanism the MoF, but is not yet fully operational. The LRC also mandates the creation o f an independent Board o f Tax Appeals, which has yet to be established. PI-14: Effectiveness of measures for taxpayer registration and tax assessment - 121- 66. A Taxpayer Identification Number system under the MoF administration was introduced in2001; the original 17-digit numbers were replaced by a simpler 9-digit identifier in2007. The identification number database i s computerized, although it i s badly structured and requires more entries (including duplicates) than the manual records used by tax administrators. Efforts are under way to establish linked registers o f taxpayers among the MoF, Customs, and Ministry o f Commerce, but these plans have not yet been implemented. At present, the Taxpayer Identification Number register i s not linked to the registers o f other agencies. The number i s useful for tracking assessmentto actual payment and banking. 67. The taxpayer audit system has been reformed with the decentralization o f the audit function to the individual taxpayer units (large, medium, small). The units conduct desk audits, issue-oriented audits, and annual comprehensive audits. Audits are based on a risk assessment, butaudit staffs need training. I c- (i)Controls intaxpayer registration Taxpayer registration is mandated, although design, C system management, and utilization o f TIN information needs improvement. No link currently exists with Ministryo fCommerce. However, customs payments do require use o f the TIN. (ii) Effectiveness o f penalties for non- The regime for PSI and penalties for under C compliance with registrationand declaration have been reformed, and are applied declaration obligations fairly consistently. Enforcement o f penalties for other tax breaches remain weak, butare gradually improving. The revenuecode i s under revision, andthe penaltiesfor non-compliance havebeenstrenethened. (iii) Planning and monitoring o f tax audit The tax audit function has been reformed and is C and fraud investigationprograms now included inthe functions o f each individual taxpayer unit.An annual audit plan i s developed based on risk assessment, although the audit manual has not been finalized and guidelines for assessment not yet determined. PI-15: Effectiveness in collection of tax payment 68. Responsibility for the collection o f taxes has been reformed in line with IMF recommendations. The previous system o f bureaus and a centralized collection unit for aged debt have been dissolved and the three tax units (large, medium and small) now have their own collection and enforcement responsibilities. This has helpedimprove the accuracy o f information on balances and adjustments resulting from payments, new assessments, and accruing interest and penalties. Records from the small taxpayer units, however, may still not be entirely reliable. The Ministry's listing o f tax arrears and collection show increased collection rates in recent years, but the ratio o f overdue taxes to collection rate remains low, at 55 percent in 2006/07. Overall, however, the total value o f tax arrears are low. 69. Executive Orders 2 and 8 o f the National Transitional Government o f Liberia provide that all tax and non-tax revenues should be paid directly to the Central Bank o f Liberia. The MoF has developed an in-house Bank Pay Slip (BPS) and Flag Receipt Systemto handle the recording - 122 - and accounting for revenue collections. The BPS is linkedonline to the pre-shipment office. The Central Bank and MoF can therefore project taxes due from goods arriving through the ports. By flagging receipts recorded in the BPS against lodgment in Central Bank o f Liberia bank accounts, MoF is able to track assessments and collections. Procedure notices are also well displayed at the MoF. 70. Despite significant improvements in the customs administration management, there are still large leakages, especially from the customs at the Monrovia Freeport, which handles the majority of imports/exports. Giventhat customs revenue accounted for 47 percent o f government revenues inFY06/07, revenue performance could increase markedly through improved customs administration. There have been reports o f fraudulent customs documentation, particularly false declarations, fake receipts, and abuse o f duty free waivers. The MoF, with support from the IMF, has developed a new customs modernization strategy to be implemented over three years, which proposes to outsource customs administration and streamline and automate the administration process. The government has recently instituted measures to strengthen destination inspections, inresponse to concerns about false declarations. 71. Revenue collections receipted by the Central Bank are well controlled, accounted for, and reconciled. The MoF reconciles its accounts daily based on data from the Central Bank on deposits, and a monthly comprehensive reconciliation is conducted in preparation for the production o f the monthly revenue reports. Procedures exist for the payment information to flow through the systemfor accurate reconciliation o f deposits. ctiveness in collection of tax I n+ (i)Collectionratioforgrosstaxarrears, Although arrears are not relatively large in annual being percentageo ftax arrears at the revenue terms, the debt collection ratio is low. At beginningo f a fiscal year, which was the aggregate level over the past three fiscal years collected during that fiscal year (2004/05,2005/06 and 2006/07), approximately 50% of all overdue taxes were collected. (ii)Effectiveness o ftransfer of tax All tax payments are made directly to Treasury A collections to the Treasury by the revenue accounts at the Central Bank o fLiberia. administration (iii) Frequency o f complete accounts Tax deposits to the Central Bank Treasury accounts D reconciliation between tax assessments, are reconciled daily, and comprehensively collections, arrears records and receipts reconciledmonthly. However, weaknesses exist in by the Treasury reconciling assessmentto actual payment due to lack of capacity by the tax audit staffs to undertake annual reconciliationswithin three months o f the end o fthe fiscal year. PI-16:Predictabilityof the availability of funds for commitment of expenditures 72. The Budget Committee comprising the MoF, Bureau of the Budget, General Service Agency, Civil Service Agency, Ministry o f Planning and the Ministry o f State has been resuscitated. Cash plans are reviewed and, once approved, allotments are released by BOB. Allotments are now based on collective and rational decisions based on cash plans. Trainings have been held on cash planning, and ministries and agencies are now submittingdetailed cash plans based on their annual work plans. - 123 - 73. Each ministry and agency has a focal analyst in BOBresponsible for all its budgetary matters and visits are made on site to provide hands-on training incompleting budget forms. 74. Predictability in the availability o f funds for commitment o f expenditures has improved through the simplification o f budget release forms by BOB;releases are now done within 2 days o f receipt as opposed to up to 2 weeks in the past. Consolidated data on allotments and expenditure i s now also more accessible, due to the introduction o f the Liberian Expenditure Control and Accounting Program-an automation system developed in-house that is being used inthe BOBto capture the annual appropriation and print the budget documents. The program is only partially implemented within the BOB,and is currently not linked to the payment process through the BGA and CMCo. However, the Liberian Expenditure Control and Accounting Program has resulted ina major improvement incoding through the removal o f around 22 pages o f duplicate codes. Cash plans are also entered into the program to ensure that the annual appropriationi s not exceeded. 75. The timeliness o fthe payment process for suppliers has also improved significantly inthe past year, but involves a large number o f steps and therefore risks hold-ups. This i s due to the transaction first having to be processed through the commitment control system, then comply with the procurement procedures and then be resubmitted for payment to MoF. However, the time taken to process payments i s gradually improving through trainings heldfor comptrollers in voucher preparation, financial rules and coding. The business processes have also been streamlined over the past two years. MoF at a recent town hall meeting listed the following accomplishments o fthe expenditure department inM o F . ~ ~ Before January 2006 After January 2007 1. Purchase orders 14 days to process 7 days to process 2. Vouchers 17 days to process 7 days to process 3. Physical audit 7 days to verify 2 days for verification 4. Checks 12 days to process 5 days to process 5. Payroll Payments were not current Civil Servants paid o n or before the 30" o f each month 6. Decentralization N o payments made and no Payments made in 14 Counties paymasters located inleeward with Paymasters stationed in counties leeward Counties 7. Computer systemiIT N o operational IT system in IT systems with 60 computers Exp.Dept. inExp.Dept. 8. Employees Exp. Dept. had up to 385 Staff reduced to 226 employees employees 9. Physical infrastructure N o clear goals and objectives 75% o f the Exp. Dept. and updated facilities renovation completed with goals and obiectives 76. Despite these accomplishments, Line Ministries and spendingunits still voiced claims o f serious delays in processing vouchers. Institutions that are supposed to be receiving statutory transfers also suffer from similar delays. 62Ministry o f Finance Interactive Town Hall meeting "TAKING BUDGET REPORTING TO THE PEOPLE", Monrovia City Hall, November 14,2007. - 124 - bility inthe funds for f (i) towhichcashflows Extent A cashflow forecast is preparedfor the fiscal year, andmaybe are forecast and monitored. reviewedinternallybytheBOBandMoF, ifrequired, for the purpose o fallotment control, withchangesapprovedby theNBC. (ii)Reliability and horizon o f The Bureau of Budget provides ministries and agencies with C periodic in-year information to monthly or quarterly (certain independent agencies) allotments, spending agencies on ceilings based on their submitted annual cash flow plans, andthese for expenditure allotments establish an upper limit for expenditure commitment by ministries and agencies. Due to the cash-basedbudgeting system, expenditure commitments cannot be made in excess of the monthly allotment, but ministries and agencies can plan andsequenceprocurements based on their approved cash flow plans for the entire fiscal year. (iii)Frequency and Inrecentyears, significant in-year adjustmentstobudgetallocations B transparency o f adjustments to havebeendone in conjunctionwiththe passageofa supplemental budget allocations which are budget, conductedonce ayear, which has undergonecabinetreview decided above the level o f andpassagebythe Legislature. management o f spending agencies Smaller in-year adjustmentsto the budgetbetweenministriesand agencies are reviewed and approvedbythe inter-ministerialBudget Committee. PI-17: Recordingand management of cash balances, debt and guarantees 77. After the war, the government inherited an enormous debt overhang and extremely poor and incomplete debt records. Since then, an extensive debt reconciliation exercise has been undertaken, and records reconstructed based on creditor records, where the integrity o f the creditor's records could be trusted. Externaltechnical advisors helped to verify commercial debt. This debt reconciliation data have laid the basis for Liberia's entry into HIPC. In addition, a domestic debt verification exercise was undertaken in2006-07, and the government developed a strategy for resolution o f domestic debt and arrears. 78. Although the debt management unit at the MoF has received some initial capacity building and significant external support with the debt verification exercises, systematic and regular debt-monitoringpractices and reconciliationshave not been instituted. 79. In principal, the cash budgeting system eliminates new domestic borrowing by the government. The government has accumulated no new external debt since 2002/03. Since 2006, ithas runa cash budgetpolicy and implemented strict commitment controls, which do not allow further deficit financing. Liberia will not be in a position to contract new external loans, other than Poverty Reduction and Growth Facility credits, until it has reached heavily indebted poor country completion point. However, as Liberia regains its creditworthiness, the institutional framework for contracting new loans will require strengthening. The Minister o f Finance is authorized to make decisions on government borrowing and can issue loan guarantees on behalf of the government, incollaboration with the Ministry o f Justice. However, no formal limits have - 125 - been set to constrain the borrowing by statement o f expenditures or the levels o f guarantees for such debts. 80. The government maintains its accounts at the Central Bank o f Liberia. Balances o f accounts held at Central Bank are sent to MoF daily but are not consolidated and monitored for proper cash management. The mission was unable to obtain evidence o f reporting and monitoring o f balances held at commercial banks or o f balances o f ministry and agency accounts at the Central Bank. 81. The government operates in two currencies: United States Dollar (US$) and Liberian Dollar (LD$). Current presidential orders stipulate that all tax and non-tax revenues are to be paid directly into the government US$ or LD$ accounts in the Central Bank. While most o f the cash resources o f the government are concentrated in two general accounts, other government accounts exist in both the Central Bank and in commercial banks. These accounts are primarily for donor-funded programs, which are not processed through the national budget, and for operational funds, which are approved by the CMCo as direct transfers to ministries and agencies. The MoF also has a payroll account inthe Central Bank from where salaries are paid. 82. The MoF does not have internal access to the balances of the ministries and agencies, although it has requested the Central Bank to provide them. The amounts involved are substantial (US$ 1.7 million at 22 August 2007), and the existing system undermines the discipline o f a single source for all expenditures from the budget. The existence o f a separate payroll account in the Central Bank also underminesthe principle of a single treasury account, although there arejustifiable reasons for its continued existence inthe short to mediumterm. The separate payroll account was established due to the difficulty in reconciling payroll checks against bank statements. 83. To increase transparency and accountability, contain expenditures within the quarterly cash limits, enable initial budget policy targets to be met, and ensure that arrears are not built up by ministries and agencies, the government has set up a system of accounting for its revenues, expenditures, assets, and liabilities. This accounting system involves the BOB,BGA, ministries, departments and agencies, the CMCo, the Office o f Comptroller General, and the Central Bank o f Liberia. While the current arrangements have increased transparency and accountability inthe payment process, they represent a temporary measure designed as a part o f GEMAP and aiming to re-impose fiscal discipline while longer-term financial institutions are re-established. There i s a need for the government o f Liberia to plan ahead and design a more streamlined procurement and payment process that gives greater autonomy to the line ministries once the GEMAP period comes to an end. - 126 - (i) ofdebtdata Quality A debt reconciliation exercise was completed inNovember 2007, C recording and reporting which forms the basis for a debt sustainability analysis (underlyingthe IUPC document). 100percent ofLiberia's multilateral debt and 99 percent of its bilateral debt was reconciled, and initial estimates o f commercial debt stocks have been prepared. Debt data records were destroyed or lost during the war, thus the government had to rely on creditor's records in order to determine current debt stock. Regular debt management practices, such as quarterly or annual reconciliations, havenot yet been instituted, although the government's debt managementunit is beingstrengthened. ii)Extentofconsolidation The government has three main treasury bank accounts inCentral C o fthe government's cash Bank.Cash balances are calculated on aweekly basis, butthe balances system does not include consolidation o f bank balances. Off- budgetrevenue collected by spendingagencies i s not tracked, but represents a limited amount compared with total budget revenue. (iii)Systems for The cashbudgeting system eliminates inprincipalnew domestic B contracting loans and borrowingbythe government. The governmenthas also stated, as a issuance o f guarantees matter ofpolicy, that no loansor guarantees are allowed at present (this will bere-visitedwhenLiberiareachesHIPCcompletion). The MinisterofFinance is authorizedto make decisions on government borrowing and can issue loanguaranteeson behalfofthe government, L incollaboration withtheMinistryofJustice. Fomiallimits to constrain the borrowingby state-owned enterprises or the levels ofguaranteesfor such debts are not currently legislated. PI-18: Effectiveness of payroll controls 84. The payroll i s currently run on the Government Accounting Payroll System (GAPS), which i s processed inLiberian dollars. GAPS handles payroll for the majority o f all government employees. Given that wages and salaries constitute 35 percent o f the 2007/08 budget estimate, GAPS accounts for a significant proportion o f total expenditure. However, allowances are not processed through the payroll, but paid out through the CMCo process and issued in US$, and therefore not subject to the same level o f control as the payroll. Inaddition, certain ministries and agencies from the security sector are not currently using GAPS and instead pay their employees through the regular CMCo process in US$. These include the Armed Forces o f Liberia, the Liberia National Police, the Special Security Service, and the National Security Agency. In addition, state owned enterprises and commissions or agencies which receive direct transfers also use the CMCo process instead o f GAPS to pay their employees. They submit a single payroll voucher, and a check for the net amount i s issued to the institution concerned, which inturn pays its employees. 85. A Personnel Action Notice is used to place employees on the payroll (only those paid through GAPS). The notice i s initiated by the organization in which the employee works and submitted to the Civil Service Agency (CSA) for endorsement. It is submitted to the BOBto check for funds availability inthe case o f additions to the payroll. The notice is then sent back to the CSA, which forwards the approved Personnel Action Notice to BGA for authorization o f - 127- input into the GAPS. A dummy payroll is printed and verified by a payroll auditor in BGA before checks are printed. Actions on PAN sometimes take upto 3 months to process. Vetting the payroll has continuedwith the MoF requesting all ministries and agencies to provide a certified list o f all employees, against which payments are made. This exercise has been completed for all the ministries and agencies except the Ministry o f Education, which has the largest payroll, with approximately 17,000 employees. 86. The grading structure is complex and inconsistent. There are more than 350 different classification codes in the payroll system and 98 different gross salaries ranging from L$294 - 8,870 for classification. This makes it difficult to apply titles, grades, and salary levels consistently across the civil service and provides scope for greater discretion when determining salary and grade level. The Civil Service Agency (CSA) will need to review the classification structure for the whole Civil Service to ensure consistency. There is no direct linkage between personnel records heldby the CSA andthe EDPpayroll, althoughthe two databases are regularly reconciled. A key requirement for an efficient payroll system i s a predefined classificatiodgrading structure in the payroll database that determines the rate o f pay o f all employees. This should form a part o f the civil service reform agenda that the government i s currently developing. 87. Over the past year the MoF has begun to decentralize the payment o f salaries to the civil service. The salary payments are now made every month, each employee i s issued an individual check, and the disbursements are made in the county capitals for civil servants based outside o f Monrovia. Each ministry and agency and county has dedicated paymasters who coordinate with UnitedNations Mission inLiberia and Central Bank to transport cash for these payments. While this process is resource-intensive, it has ensured that civil servants can receive payment within reasonable distance from their work place and has enabled the government to verify the existence o f each civil servant and remove ghosts from the payroll (on average 2,000 checkdmonth went unclaimed in 2007). In 2008, on a pilot basis, the MoF hopes to begin paying salary and allowances o f senior civil servants through direct debits. This initiative would depend on the employees opening their bank accounts with commercial banks. However, extending this initiative to the entire civil service may not be possible in the short term, due to the lack o f bankingfacilities inmany counties and the hightransaction costs for the banks o f administering a large number o f bank accounts for low-wage earners with little propensity for saving. 88. The CSA i s also planning to implement a biometrics system. The process o f capturing employees in the biometrics system has to be carefully planned to ensure that only genuine employees get into the system. The opportunity should also be taken to update HR details o f employees and establish a proper payroll record-keeping system at the MoF, CSA, and line ministries. - 128 - PI-18 EffeHiveness of n+ payroll contrafs (i) ofintegration Degree A database ofpersonnel recordsis maintained at the CSA, although the C and reconciliation quality ofthe records i s low. The CSA personnelrecords are not between personnel directly linkedto the EDP payroll. although the two records are records and payroll data reconciled at least every six months. (ii)Timeliness of changes Delays intaking action on PersonnelAction Notice (PAN) can take D to personnel records and morethan three months. the payroll (iii)Internal controls of Controls through the PAN exist for changes inpersonnelrecords and C changes to personnel payroll, but supervisorycontrols are weak to ensure the full integrity of records and the payroll the payroll because changeseffectedare not adequately checkcd before the payroll is run.Audit trail reportsdo not exist to show before and after details for changes effectedor to track the user. (iv) Existenceofpayroll No structuredpayroll auditshavebeen undertakeninthe past three D audits to identify control years. weaknesses and/or ghost workers PI-19: Competition, value for money, and controls in procurement 89. The Public Procurement and Concessions (PPC) Act dated September 2005, which came into effect in January 2006, is Liberia's first significant step toward subjecting public sector contracts to meaningful competition. Overall, the PPC Act is comprehensive and provides a framework for a sound procurement system. However, some aspects should be improved. A comprehensive set o f implementing regulations and manuals have not yet been produced, but the Public Procurement and Concessions Commission (PPCC) is working on these documents with a consulting company. 90. There still appears to be some confusion about the roles and responsibilities o f the key players o f the Liberia public procurement system. Some procuring entities do not recognize the need for a strict separation of functions betweenthe Procurement Unit, which is the executing body, and the Procurement Committee, which provides supervision. Proposals are sometimes evaluated by members o f the Procurement Committee. In some ministries, procurement responsibilities are spread out over different units-with shopping conducted by one unit and national competitive biddingby another, coordinated only at the level o f the minister. The PPCC, whose primary function i s to regulate the system and prevent departures from the procurement law, i s routinely solicited to provide no-objections on procurement decisions, a function that is normally the responsibility o fprocurement committees. 91. While Liberia's procurement law is basically sound, the practice o f public procurement falls short of the standard which the law sets. For instance, not much has changed in shopping practices. Staffs o f the procuring entities still go to three selected suppliers to obtain written quotations on the spot instead o f soliciting them through a written request. Shopping i s still carried out for contracts far above the threshold o f US$2,000 that is stipulated inthe Schedule (although it is generally recognized as beingtoo low). - 129- 92. National competitive bidding practices also need improving to conform to the Act. The advertisement contains a call for bids instead o f inviting bidders to purchase the bidding documents at the cost o f reproduction; thus, bidders are deprived o f important information that they should consider when preparing a bid. Consequently important provisions-such as the delivery time, penalties for late delivery, requirement for spare parts availability-are often left to the contracting stage or not addressed at all. The criteria for evaluationare also not specified. 93. The Cash Management committee in the MoF has helped identify and overturn many breaches o f the procurement law. The MoF has issued a checklist with CMCo requirements, to help ministries and agencies to comply with procurement and payment procedures. 94. The MoF has observed the following common deficiencies in procurement when processing local purchase orders or vouchers: Invoices are not consistent interms o f itemdquantitiesprocured (shopping). Delivery notes are not consistent with invoices or local purchase orders and are not signed. Procurement committee meetings minutes are not attached and/or not signed. Single sourcing is notjustified. Specifications are not drawn properly resulting in entities resorting to quoting "end user requirement" asjustification for preferring a particular bid. Advertisements are not prepared with lots inmind: ministries issue one advertisement for anything from stationery, fuel, computers to vehicles. The period of advertisement and closing of bid submission and opening are not spaced out as per the provisions of the Act. There is misunderstanding about the term "responsive bids." There is a lack o f clarity about payment terms (linked to delivery), advance payment (linked to a bank guarantee), andmode ofpayment. 95. Despite these shortcomings, procurement practice has improved markedly inthe past six months. Procurement staff and committee members now have some grasp o f the key requirements o f the Act. The break in procurement performance came largely as a result o f the PPCC workshop given to the heads o f procuring entities inMay 2007. This suggests that further targeted training would be highly beneficial. Most o f the interimprocedures inplace during the last fiscal year that were contrary to the provisions o f the Act have been discontinued. There are strong indications o f a change inattitude toward procurement and a change o f habits inresponse to the Act. 96. Inaccordance withthe PPC Act, a Complaints and Appeals ReviewPanelwas appointed in early 2008 to investigate complaints about procurement irregularities. The commissioners have been appointed and received some initial training, but have not yet begun to review - 130- complaints. Some complaints have been received by the PPCC, but have not been passed on to the Complaints and Appeals Review Panel ina timely and consistent manner. Pi-19 Iue ntr (i)Useofopen Although procurementopportunities are advertisedindaily newspapers competition for award o f for large purchases, direct transfers are sometimes usedto bypass some contracts that exceedthe procurementcontrols exercised by CMCo. The PPCChas also found nationally established that procurements usingthe shoppingmethod have at times been monetary threshold for approvedeven though they exceededthe threshold for small purchases, small purchases although reliabledata is not available. (ii)Justification for use of Justification for use of less competitive methodsi s weak. C less competitive procurement methods (iii) Existence and PPCAct, 2005, Part VI11COMPLAINTS AND REVIEW C operation o f a PROCESS-sections 125-129 stipulatesthe process to address procurement complaints procurementcomplaints by the complaints panel. mechanism A Complaints andAppeals Review Panelhasbeenestablished and commissionerstrained. However, the panelhas not yet begunto review and address complaints. PI-20: Effectiveness of internal controls for non-salary expenditure 97. The Cash Management Committee is responsible for approving all payments for the Government o f Liberia. The Committee, headed by the Minister o f Finance, has representatives from the Ministry o f State, Bureau o f Budget, and Ministry o f Planning. The committee meets biweekly to approve payments for government. As one o f the GEMAP requirements, a donor- funded external advisor acts as a member o f the CMCo and has co-signatory authority on the Committee. 98. Each ministry and agency is responsible for submitting to the BOBa monthly cash plan, which is the basis o f monthly cash allotments issued to ministries and agencies by the BOB. These cash plans are reviewed and approved by the BOB,which issues allotments to the ministries and agencies and BGA, which i s responsible for processing payment requests. 99. Following receipt o f its cash allotment from the BOB,ministries and agencies prepare a local purchase order. The order and the supporting documents-invoices and copy o f the BOB cash allotment-are then sent to the BGA. After checking for appropriation, allotment, conformity o f object code, and verifying compliance with procurement procedures and financial rules, the BGA approves the purchase order by putting a special seal on it and returns the documents to the ministries and agencies. This seal o f verification allows the ministries and agencies to proceed with the procurement. On receipt o fthe goods and services, the ministry and agency forwards the voucher along with the original sealed local purchase order, invoices, and delivery note to BGA. The BGA then checks for conformity with the order and processes the voucher for payment. - 131- 100. A Physical Audit Unit (PAU) inthe MoF's Department of Expenditure is responsible for certifying receipt o f supplies and works before payments are processed. The P A U works independently o f the Internal Audit Department in the MoF and sometimes extends their work beyond physical verification to compliance with laws and other internal controls. The PAU auditdcertifies every transaction relating to goods and services above the threshold o f US$5,000 before the payment is approved by the CMCo. 101. The vouchers certified by the PAU and approved by the BGA are forwarded to the Cash Management Committee (CMCo) secretariat. The CMCo secretariat prepares the list o f vouchers for payments, which i s reviewedand approved by the CMCo. The CMCo also receives from the office o f the Controller General of Accounts details o f the government cash balances with the Central Bank, prepared on the basis o f Central Bank statements. However, cash position i s based on bank statement balances at Central Bank rather than the cash book balances held by the Controller General o f Accounts. 102. After reviewing the list o f vouchers for payment prepared by the CMCo secretariat, and after considering the available cash balances, the CMCo approves the payments. The approved signed list for the CMCo meeting is forwarded to Central Bank. A copy is sent to the Office of the Controller General o f Accounts for printingo f checks and disbursement. 103. Although non-salary expenditures are subject to the commitment control procedures outlined above, approximately 16 percent o f the 2007/08 budget i s approved by the CMCo as direct transfers to state-owned enterprises and commissions (6.9 percent), for ministries and agencies for operational expenses, and occasionally to ministries and agencies for execution o f projects under the Force Account method, which is allowed under the PPC Act (9.1 percent). Reports o f these expenditures are required before the next transfer can be authorized, but controls inplace at the ministries and agencies to ensure that operational funds or other transfers are expended appropriately are weak. - 132 - (i)Effectiveness of Line ministries prepare monthly cash plansbased on the annual B expenditure commitment approvedh-evisedbudget, the BOBonly issues monthly allotments controls based on these cash plans. Local purchase orders are authorized by the MoF on availability o f allotments and all payment vouchers go through a strict commitment control process with final approval for expenditure issued by the CMCo based on cash availability. However, once transfers to line ministries for smaller operational expenses from the consolidated account are approved through the commitment control process, control procedures at the line ministrv level are weak. (ii)Comprehensiveness, There is no comprehensive PFMAct and regulation inplace but C relevance, and the MoF has issued financial rules, domestic and foreign travel understanding o f other ordinances and commitment control procedures for preparation internal control rules or and processing of vouchers. These rules however are not procedures comprchensive innature and cover only part o fthe whole budget execution and reporting process. Duplicate procedures exist in recordingtransactions in manual books and spreadsheets.Controls over the use o ftransfers for oDerationa1exDenses are weak. (iii)Degreeofcompliance Compliance with rules is high, as all payments are carefully B with rules for processing reviewed by the CMCo andthe external GEMAP advisor. and recording transactions. Occasionally, simplified or emergency procedures have been used without adequatejustification (such as for the Jellah Road Project, in2007). The quality ofreportingon monthly transfers o f operational funds and subsidies to ministries and agencies and independent commissions and corporation is not fully comprehensive, and compliance with rules for processingthese transactions cannot be properly assessed. PI-21: Effectiveness of internal audit 104. The internal audit function inLiberia isweak and needs to be developed. Comprehension o f the role o f internal audit in relation to management control responsibilities is not well developed. The existing internal audit functions fulfill a role that combines the role o f financial controller (by carrying out mandatory precontrol before decisions are taken) and the role o f financial police (by carrying out special investigations ordered by ministers). 105. Internal control systems are designed to ensure that: e Obligations and costs comply with applicable law. e Assets are safeguarded against waste, loss, unauthorized use, andmisappropriation. e Revenues and expenditures, liabilities, and recovery o f undue paid amounts applicable to budget spender operations are recorded and accounted for properly so that accounts and reliable financial and statistical reports may be prepared and the accountability o fthe assetsmaintained. - 133 - 0 Programs are efficiently and effectively carried out to meet objectives in accordance withapplicable law andmanagement policy. 0 All management risks have been detected and addressed by an appropriate and proportionate control. 106. Under the 1972 Act, responsibility for internal audit i s assigned to the Auditor General. This arrangement was enacted when the Auditor General's Office was placed under the Executive branch. Since the General Auditing Commission i s now under the Legislative branch, this arrangement is less appropriate, because internal audit is conventionally requiredto function as an aid to management and to have the resources, capacity, and independence to report directly to the chief executive o f the entity. The PFM Act that i s beingdrafted and the proposed new law on the General Auditing Commission could help to address this issue and provide a roadmap for the revision o f the institutional arrangements for internal audit, inorder to secure a clear division between external and internal audit. 107. At present, most of the government ministries do not have internal audit units. The exceptions include the ministries o f Finance, State, Justice, General Service Agency, Health and Social Welfare, and Lands, Mines, and Energy.The ministries o f Public Works, Internal Affairs, and Education are setting up or revitalizing their internal audit units. None o f these, however, have a clear mandate or reporting structure. The MoF internal audit unit has an operational manualthat was issued inMay 2004 but not widely used. The General Auditing Commission has drafted a memorandum o f understanding with the Institute o f Internal Auditors63for the institute to partner in developing the internal audit function for the government. The memorandum o f understanding proposes as a first step setting up a joint working group to assess the needs, timing, and overall approach to buildingthispartnership. 108. While the General Auditing Commission is providing some useful guidance for the establishment o f an independent internal audit function, the MoF, however, should take full responsibility for the internal audit under the executive. An internal audit strategy i s under draft by the MoF, with input from the GAC. This strategy should specify the roles and mandates o f internal auditors in relationship to internal control systems. Any legal amendments should be assessed duringthe ongoing process o f developing the comprehensive PFMlaw. 63The Institute o f Internal Auditors represents the global profession of internal auditing, which has over 135,000 members in 166 countries. Its members agree to practice internal auditing in accordance with the Institute o f Internal Auditors' Code o f Ethics andthe InternationalStandards for the Professional Practice o fInternal Auditing. - 134 - ess of t D+ (i)Coverageandquality The internal audit function inLiberia i s weak and needs to be D o fthe internal audit developed. There is no evidence that the work of the internal audit function units meets InternationalStandards for the ProfessionalPractice o fInternal Auditing (ISPPIA), nor that the basic role of internal audit is fully recognized intheory or practice. The PhysicalAudit Unit at the MoF has mainly concentrated in verifying the receipts o f goods and services beforepayment. (ii)Frequencyand Reports are not regularly produced or submitted to the MoF o f D distribution o f reports General Auditing Commission. (iii) of Extent There are few reported instances o f formal internal audit C management responseto recommendations made to the chief executive by the internal internal audit findings auditing units, and little evidence o f recommendations being acted upon. There have been some positive developments within the MoF, and there are indications that, at least in one instance, reports produced by the MoF internal auditing unit resulted in dismissal and court prosecutions. C (111)ACCOUNTING,RECORDING,AND REPORTING PI-22: Timelinessand regularityofaccountsreconciliation 109. The responsibility for the accounting o f government expenditures and revenues rests with the MoF. The MoF has an expenditure department headed by the Deputy Minister o f Expenditure and a revenue department headed by the Deputy Minister for Revenue. Within the department for expenditure lies the Bureau o f General Accounting (BGA), which i s responsible for processing the payment o f all goods and services (including allowances and payroll), as well as the Controller General of Accounts (CGA), which has as its principal function the reconciliation o f the Liberian government bank accounts with the Central Bank and issuing o f checks. The General Service Agency, a separate government entity, i s responsible for the recording andtracking o f assets. 110. The credibility o f reports and financial statements depends on an effective and timely bankreconciliationprocess to ensure the accuracy and authenticity o ftransactions going through the cash book and bank statements. Bank reconciliationreports should be independently checked and countersigned by a senior officer and unreconcileditems inthe cash book or bank statements should be monitored and cleared monthly. 111. The Government of Liberia maintains three bank accounts: US Dollar General Account, Liberian Dollar General Account, and Liberia Dollar Payroll Accounts. All accounts are with the Central Bank o f Liberia. Ministries and agencies also maintain bank accounts at the Central Bank for operational purposes. Accounts for donor-funded projects are (as per the donor requirements) at commercial banks. 112. Central Bank of Liberia sends daily cash balances to the MoF, which i s usedas the main guide by the CMCo to ascertain the government's cash position before approvingpayments. The cash availability statement for these meetings is prepared by the CGA's office and i s independently verified by the CMCo Secretariat. Liberia operates a single entry system, so - 135 - payments recorded by the BGA are not contra-entered into the cash book maintained by the CGA. This makes reconciliation o f bank statements cumbersome, and routine bank reconciliationstatements are not produced. 113. Government employees and pensioners (about 45,000) are issued individual monthly checks for salaries and allowances. Because o f the volume o f checks and the lack o f infrastructure, Central Bank adds the Liberian Dollars payroll check inbatches o f 50 or more and posts the bulkamount inthe statement, renderingbank reconciliation impossible. 114. Monthly transfers to ministries and agencies for operational expenditures and transfers to independent commissions and corporations are not reconciled monthly or at the end o f the fiscal year. These transfers, which account for 16 percent o f total expenditure (2007/08), are approved and paid through the regular CMCo review process. Ministries and agencies are required to provide financial reports on the use o f funds before its next transfer i s made. However, lack o f reconciliationo fthese transfers compromises the control system inplace. n ere is no effective bank reconciliation in place inthe absence of D reconciliations cash books to reconcile against. Transactions inbank statements are ticked against CMCo listings. (ii)Regularity o f Monthly cash advances for operating cost are paid into bank D reconciliation and accounts of ministries and agencies that are not reconciled clearance o f suspense monthly and are not cleared at the end of the fiscal year. Use of accounts and advances suspense accounts is uncommon. PI-23: Availability of information on resources received by service delivery units 115. Data on resources made available to primary service delivery units i s not available, and no tracking exercises have been completed inthe past three years. There are an estimated 4,146 primary schools inLiberia64and 354 health f a ~ i l i t i e sMost . ~ ~ resources for these service delivery units-including school materials, drugs, medical supplies and equipment-are procured/processed centrally by the relevant ministry (Ministry o f Education and Ministry o f Health and Social Welfare) and distributedto the units. While individual units may have systems for tracking items received, these records are not systematized, and not centrally collected and consolidated. Small funds for operations and maintenance are transferred to certain schools and clinics, but liquidation of these payments has been challenging. Moreover, no Public Expenditure Tracking Surveys (PETS) have been conducted since the end o f the war to give an indication o f the availability o f resources at service delivery units. The tracking process i s complicated by the fact that a large proportion o f resources at schools and health centers are provided by NGOs or other organizations, and central tracking and recording o fthis support i s minimal. 64MinistryofEducation, Education census 2007l08. 65Republic o f Liberia, Poverty Reduction Strategy, 2008. - 136 - I I D ---1 (i)Collection and processing o f information to No comprehensive data collection on D demonstrate the resources that were actually received resources to service delivery units in a (incash and kind) by the most common front-line major sector has taken place within the service delivery units (focus on primary schools and last three years. primary health clinics) inrelation to the overall resources made available to the sector(s), irrespective o fwhich level o fgovernment is responsible for the operation and funding of those units PI-24: Quality and timeliness of in-year budget reports 116. The government produces quarterly fiscal reports that detail revenues or allotments and expenditures o fthe various ministries and agencies. Butthese reports are not based on reconciled accounting data. Moreover, The fiscal reports could be strengthened by comparing expenditure with the approved or revised budget, to allow decisionmakers to effectively review and monitor budgetexecution. The reports also fail to highlight early warning signals and emerging slippages inthe implementation ofthe budget. These reports do mark an important step toward increased transparency, and the reports are publishedin newspapers and posted on the MoF website. The timelines o f the reports have improved over time, as illustrated by the list below, detailing when the following reports were posted on the website: 1. Quarter IReport 2006-07 -November 27,2006 2. Quarter I1Report 2006-07 - February 2,2007 3. Mid-Year Review 2006-07 - March 2,2007. 4. Fiscal Report 2006-07 - October 21,2007 117. A limited number o f reports are produced by various departments during the implementation o f the budget. The Bureau o f General Accounting (BGA) produces a report showing appropriations, allotments, un-allotted balances, and expenditures on a commitment basis (from vouchers recorded in the accounts payable ledger). This report i s produced monthly and distributed to the President, the National Assembly, Minister o f Finance, and Bureau o f the Budget. The BGA also produces a weekly payroll expenditure report-a weekly list indicating the vouchers pending, budget allotments against each request, including their status (complete or incomplete), and the amounts paid-and a monthly outstanding payable report. Inaddition to the above, the Cash Management Committee secretariat produces a monthly report o f CMCo approved expenditures. 118. The CGA prepares a daily cash-position report o f the government's main US$ and LD$ accounts at the Central Bank o f Liberia. This report show opening balances, deposits, credits, checks issued, outstanding items, and bank balances. A weekly cash availability report is also prepared. This report is distributed to the Minister o f Finance and the Deputy Minister o f expenditure and feeds into the cash management committee process. The CGA also prepares a monthly summary o f expenditures on a cash basis, showing objects o f expenditures and year to date cumulative totals. Finally, the CGA also produces a monthly expenditure report for ministries, by major budget codes with explanations o f purpose. Together, these reports allow comparison against allotments for main administrative headings. - 137 - 119. The government's fiscal reports could be strengthened by comparing expenditure with the approved or revised budget, to allow decisionmakers to effectively review and monitor budget execution. Furthermore, the reports fail to highlight early warning signals and emerging slippages inthe implementation o fthe budget. reports are useful and provide an important measure of PI-25: Qualityand timeliness of annualfinancialstatements 120. The government does not currently produce financial statements, but it has begun producing annual fiscal reports which detail revenues/allotments and expenditures o f the various ministries and agencies. This marks a significant improvement from earlier reporting produced by the MoF in 2003-2005, which contained only a one-page spreadsheet of expenditure. Financial statements have not been produced since before the war, and the current accounting procedures will not support the production o f cash-based financial statements supported by bank reconciliations. Current legislationrequires financial statements to be based on accrual accounts, which i s not realistic under present cash-budgeting procedures. This can be addressed inthe new PFM law, currently under draft. Given the government's inability to produce full financial statements, its annual fiscal reports, produced for 2005/06 and 2006/07, are an important step toward increased transparency and accountability. The reports are published in newspapers and posted on the MoF website. 121. Section 21.2 (c) of the Executive Law 1972 requires the Minister o f Finance "to report the financial activities and financial position of the Government to the President and to the Legislature". N o deadline i s set for the submission of the financial statements, which i s normally three months after the year end inmost countries. Inaddition, accounting and reporting standards for financial statements are not currently specified in the law, but it i s hoped that this will be rectified by the newPFM law. - 138 - c (i) Completeness o f the The responsibility for the preparation o f government C financial statements accountslfinancial statements lies with the MoF. Government o f Liberia financial statements were not prepared for the fiscal year o f 200415, but fiscal reports were prepared for 200516 and 200617. A review ofthe draft o fthe 2006/7 report leadthe mission to conclude that they were not complete financial statements. They contain information on revenue and expenditure by the various agencies and balances unspent for the fiscal year, but they lack a balance sheet o f financial assets and liabilities and contain no accounting policy note. (ii) Timeliness o f The reports for 2005/06 and 2006107 are not complete financial No score submission o f the statements, and thus the timeliness o f submission cannot be financial statements assessed. (However, it i s noteworthy that the government's fiscal reports were submitted in a timely manner, within 6 months o f the end o f the fiscal year. 2006/07 Fiscal Outturn Report submitted Oct 18,2007 2005/06FiscalOutturnReport submitted Sept 5,2006) (iii) Accounting standards The government does not have standards for accounting and No score used reporting and financial statements are not produced; only annual fiscal outturn reports are produced. C (IV) EXTERNALSCRUTINYAND AUDIT PI-26: Scope, nature and follow-up of external audit 122. Over the last three years the government has taken steps to establish an independent external audit function. Prior to 2005, the Auditor General was required to report to the President. This called into question the independence o f the Auditor General and compromised his ability to monitor the executive branch o f the government, as the President i s the head o f the executive. Moreover, because the Auditor General reported to the President, the legislature had no formal means o f verifying the executive's use o f budgetary resources. In 2005, these arrangements were changed. The General Auditing Commission headed by the Auditor General (AG) was created by amending section 53.2 of the Executive Law 1972 as an independent autonomous body of the government, with a duty to report to the legislature. Furthermore, although the President was mandated to appoint the Auditor General, the appointment is by and with the consent o f the Senate. Moreover, the powers and duties of the Auditor General were better defined. 123. During the NTGL period, the Auditor General's recommendations did affect accountability in specific areas. The special audit report on misuse of foreign travel allowances and advances in 2005 was noteworthy as it provided evidence o f widespread corruption and laxity inthe use o fpublic funds. 124. To strengthen the external audit function, there is a need to improve the legislative framework, increase the General Auditing Commission's financial independence and strengthen capacity. Action i s being taken on all three fronts. An act to repeal and replace chapter 53 o f the Executive Law o f 1972 to grant financial independence to the General Auditing Commissionhas - 139 - been submittedto the Legislature, but has yet to be approved. Financial independence is being improved through MoF providing General Auditing Commission with quarterly releases, with replenishmentson the basis of accountabilities provided. There is, however, a need for a new and separate act on the General Auditing Commission fully in line with international recognized auditing standards (Lima and Mexico declarations) and best practice in order to strengthen the external audit. 125. The new Auditor General i s addressing the personnel problem. A recruitment exercise has been undertaken in which about 1,400 applicants were registered. In Monrovia, Gbarnga, Bong County, 998 applicants took a test on general auditing and accounting; 341 were successful (Audit 256 and Administration 85). A first round of 220 applicants was interviewedby a nine- personpanel. O fthese, 92 were selectedand startedtraining on 19November 2007. 126. The International Organization of Supreme Audit Institutions code of ethics and auditing standards and the African Organization of English-speaking Supreme Audit Institutions audit manual and Quality Control Manual have been customized to local circumstances. These manualsand codes needtime to be fully implemented. 127. The General Auditing Commission is presently introducing a more modern auditing technique. A first materiality and risk assessment (risk index) on a national level has been carried through to strategically focus onkey government institutions. 128. Usingthis risk-based approachto assess 84 ministries and agencies, the GAC ranked the top 10 "High Priority Auditees" will audit the first five starting on 26 November 2007, with an audit entrancemeeting withthe MoF. 129. With a human resources and legal consultant, the General Auditing Commission has developed a series of HR guidelines and forms, including a code of conduct, staff confidential forms, asset declaration, and conflict of interest forms. 130. A five-year strategic planis inan early stage of development. This will provide a focus for buildingthe capacity of the institution and form the basis for donors and the government to converge their support. 131. Annual fiscal outturn reports were prepared for 2005/06 and 2006/07. The General Auditing Commission intends to use the 2006/07 outturn report to conduct a more comprehensiveand reliable transaction-basedaudit for 2006/07. 132. In a somewhat longer perspective, it would be necessary to introduce an Annual Attestation Audit to matchthe PFMdevelopment inLiberia. - 140 - the last 25 years, although there have been reviews o f activities that fell short o f internationalstandards. (ii) Timeliness o f No audit reports have been submitted to the legislature. submission o f audit reports to legislature (iii) Evidence o f followup A limitedkelect number o f recommendations given inthe Auditor D on audit General's comments to the draft National Budget were followed recommendations up by the legislature. This is however not an audit report. PI-27: Legislativescrutiny of the annual budgetlaw 133. Legislative scrutiny o f the annual budget law and financial statements i s very important to ensure transparency and accountability. The Ways and Means Committee inthe legislature is responsible for scrutinizing the annual budget when it is laid before the House o f Representatives. 134. The Ways and Means Committees o f the House and the Senate lacks technical in-house capacity to properly analyze and scrutinize the budget to guide meaningfd debate. They have drawn on support from the General Auditing Commission to review and provide technical comments on the budget. There are plans to increase capacity within the Ways and Means Committees, and in the legislature more generally, to perform budget analysis and improve the budgetreview function. 135. According to Section 2205 Chapter 22 o f the Revenue Code o f Liberia (Act o f 2000), timely and complete annual reports o fthe financial status and affairs o f all government agencies are requiredto be submitted to the President and the legislature. The executive i s also expected to submit the mid-year review o f the budget to the finance committee o f the legislature. However, other than the monthly reports produced by the BGA, the MoF is not regularly submitting comprehensive in-year budget outturn reports to the legislature and ministries and agencies. 136. Section 2212 o f the Revenue Code o f Liberia Act o f 2000 spells out the procedures concerning in-year budget amendments. Inthe code, rules exist for in-year budget amendments by executives, but they do not stipulate any limit to the amount that can be transferred. These procedures require that a request be made by the concerned ministry or bureau to the Director of the Budget. The Director o f Budget reviews it and submits the request to the President for approval. A budget law now in draft form proposes strict limits on the extent and nature o f amendments. 137. The budget document does not include actual spending against budget for the current year to enable the legislature to confirm that the budget was executed based on the Appropriation. There i s a proposal to amend section 2212 o f the Revenue Code-"The Budget - 141 - Transfer Act" to allow transfer o f 20 percent66 between agencies. This has not yet been approved. 138. The BOBhas prepared policy guidance andrequest submission instructions on budgetary transfers to ensure transparency and fiscal discipline while allowing the executive the flexibility to request changes in the approved budget. The guidingprinciple is that transfers must be done only after serious consideration and must be fully documented and submitted for all requisite approvals. Instructions for completing inter- and intra-ministerial forms for budgettransfers have also been designed. 139. The legislature is proposing to set up a Legislative Budget Finance Office (LBFO). This should be reviewed in light of the need to develop a comprehensive legal and regulatory framework that will ensure that roles and responsibilities for the management o f public funds are clarified in the areas o f macro-economic forecasting, planning, budget preparation, budget execution, recording and accounting, internal auditing, external auditing, and legislative oversight. c+ C C respected Committee lacks the requiredtechnical capacity and procedures to effectively scrutinize the budget (iii)Adequacy o ftime for The budget year for the government i s 1 July to 30 June . The B the legislature to provide 2007/08 National Budget o f Liberia was submitted to the a response to budget Legislature on May 20, 2007. The budget task effect in July 2007, proposals-both the which allowed the legislature more than a month to review the detailed estimates and, budget proposals. (However, the budgets for 2005/6 and 2006/7 where applicable, for were not submitted to the legislature until August 2005 and 2006 proposals on macro-fiscal respectively, which left little time for the legislature's scrutiny.) aggregates earlier in the budget preparation cycle (time allowed inpractice for all stages combined) (iv) Rules for in-year Although section 2212 o f the Revenue Code of Liberia Act of C amendments to the 2000 spells out the procedures concerning in-year budget budget without approval amendments, the lack of limits allows extensive administrative by the legislature. reallocation as well as expansion of total expenditure. PI-28: Legislative scrutiny of external audit reports 140. Timely consideration of audit reports from supreme audit institutions and followup on recommendations are important to hold the executive accountable. The executive i s expected to 6630 percent is high comparedwith international standards which is normally between5-10 percent - 142 - report to the Legislature inthe form o f annual financial statements with an audit opinion from the General Auditing Commission. Since external audit reports have not beenproduced since before the civil war, this indicator is not applicable. No Score (i)Timelinessof No external audit reports have beenproduced to date. No Score examination of audit reports by the legislature (for reports received within the last three Yeas) (ii) ofhearingson Extent No external audit reports have been produced to date. No Score key findings undertaken by the legislature (iii)Issuance of No external audit reports have been produced to date. No Score recommended actions by the legislature and implementation by the executive D.DONORPRACTICES D-1:Predictability of Direct Budget Support 141. The government receives little direct budget support from donors. The vast majority o f donor funding goes through projects and programs outside the national budget process. In the past three years, no budget support has been recorded in the initial revenue forecasts in the budget; rather it has been included in supplemental budgets prepared and passed in the second half of the fiscal years. In2006/07, the Chinese government provided US$1.5 million inbudget support, included ina supplemental. It was not included inthe original budget revenue estimates as it was not anticipated. For the 2007/08 budget, no direct budget support was projected in the revenue estimates. Later, agreements were reached with the World Bank and France for small amounts o f budget support, which were included in a supplemental budget passed in 2008, and funds were disbursed. - 143 - D-I Pr of Direct Port ,, (i)Annualdeviationof No budget support was included in annual revenue forecasts (in No Score actual budget support the national budget), as none was anticipated by the government. from the forecast Midway through 2006107 a single tranche budget support provided by the donor operation was agreed to with the Chinese government. agencies at least six weeks prior to the government submitting its budget proposals to the legislature ~~~ (ii) timelinessof In-year Donor disbursements have been made as one-off payments rather No Score donor disbursements than quarterly disbursements. Disbursements were not scheduled at the beginning o f the fiscal year, and therefore their timeliness cannot be assessed. D-2:Financial information provided by donors for budget, reportingon projects, and program aid 142. The poor quality o f donor reporting on funding and disbursements poses a major challenge to the government's sectoral planning and macrofiscal assumptions. Current donor flows are estimated at US$300-US$500 million for 2007, which is one to two times as much as the entire government's budget. A large number o f both development and humanitarian donors are providing support to Liberia, including multilateral organizations, bilateral donors, and private foundations. The majority o f donor funding flows through NGOs, contractors, or UN agencies - this i s largely a legacy o f the humanitarian nature o f aid, as well as a result o f concerns from donors that the government's capacity to execute development programs i s insufficient. Some donors have begun implementing projects through special implementation units embedded within ministries, usingdonor procedures and controls. 143. Donor activities are coordinated through the Liberia Reconstruction and Development Committee, chaired by the President, which meets monthly and brings together key cabinet ministers and donor partners. A Liberian Reconstructionand Development Committee secretariat i s housed within the office o f the President. The secretariat has begun to establish an aid database, but data are currently incomplete. Donors do not currently provide complete budget estimates or regularly report disbursements. Liberia participated inthe Paris Declaration Survey for 2007, and based on the responses from six o f the largest donors, ODA was estimated at US$360 million for 2007 (this excludes support to the security sector, debt relief, and funding from private donors). Several other ad hoc exercises have been carried out to determine the magnitude o f donor support to Liberia, by the government, UNDP, and the IMF, but no clear and consistent criteria have been applied for comparison or verification. The government i s committed to improving donor reporting and data compilation as part o f its efforts to track the poverty reduction strategy implementation. It i s encouraging donors to fully participate with requests for information. While funding levels are high, the government's challenge will be to ensure that donor funding is aligned with the government expenditure priorities. - 144 - man use anal res (i)Overall proportion of Other than small quantities of direct budget support, all aid funds D aid funds to central are managed using donor procedures. governmentthat are managed through national procedures Source: Budget documents, interviews with BOBand Liberian Reconstruction and Development Committee. - 146 - 144. Tracking andmonitoring aid flows is complicatedby a number of factors, including: The many levels of contracting and sub-contracting. For instance, the World Bank may execute a grant through UNDP, which inturn disburses funds to anNGO. Consistent criteria need to be applied to ensure that funding is not countedtwice. Distinctions between pledges, commitments, and disbursements. Donors tend to report funding as disbursedwhen it has been made available to the implementing agency and not whenthe actual funds havebeen spent. Annual budget processes in bilateral donor countries, which prevent bilateral donors from committing or pledging funds for more than ayear inadvance. Significant volumes of private funding, particularly through large foundations-such as the Soros Foundation, Clinton Foundation, McCall-McBainFoundation-which may be difficult to track ifthe foundations lack people inLiberia. n (i)Completeness and Most donors do not provide budget estimates for disbursement o f D timeliness of budget project aid three months in advance o f the government's fiscal project support (ii)Frequencyand Few donors provide regular reports to MoF on disbursements. D coverage of reporting by (Although grants implemented through the project financial donors on actual donor management unit do provide regular disbursement reports to flows for project support MoF.) D-3: Proportion of aid that is managed by use of nationalprocedures 145. Most donor funding is implementedoutside of national structures, often through NGOs or contractors. Some donors, primarily the World Bank, AfDB, and DFID, execute projects directly through the government, using special implementation arrangements-such as the Special Implementation Unit, established at the Ministry of Public Works with support from the World Bank and the Project Financial Management Unit at the MoF-and utilizing the donors' procurement and financial management procedures. Only small quantities of budget support, provided by the World Bank and the Government of France, have gone through government procedures. 146. Many donors are reluctant to use government procedures due to concerns about weaknesses in the government's systems. Procurement systems remain weak, accounting systems are largely manual, and external audits are only being conductedfor the first time inthe current fiscal year. - 145 - Annex 2.1: Persons Interviewed Antoinette Sayeh Minister Ministryo f Finance Tarnue Mawolo DeputyMinister, Ministryo fFinance Administration Roderick Smith Deputy Minister, Expenditure Ministry o f Finance Elfrieda Tamba Deputy Minister, Revenue Ministryo fFinance Elizabeth Tubman Ministry o f Finance Aletha Browne Assistant Minister, ExpenditureMinistry o fFinance Sam Russ Director. Debt Management Ministrv o f Finance Y J Jerry Tamba Taylor Assistant Ministryo fFinance Amitabh Tripathi Advisor Ministryof Finance Karikolraj Kasirajan Advisor Ministryo fFinance Daniel Honig Assistant to the Minister Ministry o f Finance Concessions Commission William Allen Director Civil Service Agency Representative Ways and Means ~Emmanuel J. Nuquay Committee, Legislature Jeremy Tunnacliffe Charge d'Affaires European Commission Kamil Kamaluddeen Economic Advisor UNDP Douglas Carey Economic Advisor Embassy o f the U S A - 147 - Annex 2.2: PFM-PR Calculationsheets - 1. This annex contains calculation sheets for PFM Performance Indicators PI-1 and PI- 2. 2. The FY06/07 expenditure report presented some o f the expenditure data differently from the original budget. A sum ofUS$5.24 million, allocated as lump sums to two budget lines inthe original budget for payment o f pre-NTGL salary arrears and other domestic debt holders, was reported against the budget lines o f individual ministriesand agencies inthe fiscal report. These funds were used to settle payment arrears, but were reported against the budget lines of the individual agencies that the claimants had worked for. In order to correct for this report discrepancy, the FY06/07variance analysis excludes the salary arrears payments. I I I I - 148 - - 149 - Annex 2.3: PFM-PR List of Ministries and Agencies - List ofministries andagencies and autonomousgovernmentbodiedstate-ownedenterprises. -GeneralAdministrativeServicesSector 1. National Legislature 2. Ministry o f State for Presidential Affairs 3. Office o fthe Vice President 4. Bureau o fthe Budget 5. Ministry o f Finance 6. Ministryo f InternalAffairs I 7. Ministry o f Planning and Economic Affairs 8. Civil Service Agency 9. General Services Agency 10. Ministry o f Information, Culture and Tourism 11. General Auditing Commission 12. Ministry o f ForeignAffairs 13. Liberia Institute o f Public Administration 14. National Elections Commission 15. Liberia Institute o f Statistics and Geo-Information Services I 16. Judiciary 17. Ministryo f Justice 18. Ministry o f Defense 19. National Security Agency 20. Special Security Service I 21. Ministry o fNational Security 22. National Bureau o f Investigation I -Social & CommunitvServicesSector 23. Ministry o f Education - 150 - 24. Ministry of Health and Social Welfare 25. John F. Kennedy Medical Center 26. Ministry o f Labor 27. Ministry of Youth and Sports 28. Ministry of Public Works 29. National FoodAssistance Agency 30. Center for National Documents and Records Agency 3 I.Agriculture and Industrial Training Bureau 32. Ministry of Rural Development 33. Ministry o f Gender and Development 34. Liberia Institute for Bio-medical Research 35. Phebe Hospital Economic ServicesSector 36. Ministry of Agriculture 37. Ministry of Lands, Mines, and Energy 38. Ministry of Commerce and Industry 39. Ministry of Posts and Telecommunications 40. Cooperative Development Agency 4 1. Ministry of Transport ~ 42. Liberia Domestic Airport Agency 43. Bureau of State Enterprises 44. EnvironmentalProtection Agency 45. Monrovia City Corporation 46. Forestry Development Authority Institutions, Entities & Other General Claims 2. Subsidy to Mano River Union 3. Subsidy to Liberia Copyright Office -151 - 4. Subsidy to Liberia Water and Sewer Corporation 5. Subsidyto Liberia Produce Marketing Corporation 6. Subsidy to National Housing Authority 7. Subsidyto Liberia IntellectualProperty System 8. Subsidy to National Oil Company o f Liberia 9. Subsidy to Liberia Industrial Property System 10. Subsidy to Liberia Electricity Corporation 11. Subsidyto Monrovia Transit Authority 12. Subsidy to Liberia Telecommunications Corporation 13. Subsidy to Bureau o f Maritime Affairs 14. Subsidy to Liberia Agency for Community Empowerment 15. Subsidyto Liberia Industrial Freezone Authority 16. Subsidy to Liberia RubberDevelopment Authority Transfers 1. Transfer to National Investment Commission 2. Transfer to Liberia Repatriation, Rehabilitation?and Reintegration Commission 3. Transfer to Human Rights Commission 4. Transfer to Governance Reform Commission 5. Transfer to Truth and Reconciliation Commission 6. Transfer to Public Procurement and Concession Commission 7. Transfer to National Commission on Disarmament, Demobilization? Rehabilitation?and Reintegration 8. Transfer to Liberia Reconstruction and Development Commission I 9. Transfer to National Commission on Disabilities AutonomousEducational Institutions - 152 - mcnt Authorin Transferto National Investment Commission 652,422 647,290 647,290 5,132 Transfer to Liberia Repatriation, Rehabilitation,andReintegration Commission 510,000 506,250 506,250 3,750 Transfer to HumanRights Commission 229,867 229,792 229,792 75 Transfer to GovernanceReform Commission 235.000 231,108 231,108 3J92 Transfer to Truth and ReconciliationCommission 1,642,075 1,642,075 1,642,075 0 Transfer to Public Procurement and ConcessionCommission 200,000 200,000 200,000 0 Transfer to NCDDRR 848,701 848,701 848,701 0 Transferto Liberia ReconstructionandDevelopment Commission 178,685 177,860 177,860 825 Transfer to NationalCommission ITOTAL I 9.408.480 I 9.385.593 I 9.385.593 I 22.887 I - 153 - Annex 2.5: PFM-PR:Budget Preparation Calendar for Fiscal Year 2007-2008 Nr Projected Tasks/activities Description date - Develop budget preparation Preparation o f formats to guide ministries and 1 15- budget agencies in generating and BOB,MoF, and 25.Nov. formulation presenting programmatic M P E A needed time to formats informationincluding understandthe objectives, outputdtargets, inputs, and costs, and estimates o f fees, charges, and donor financing - Hold two training workshops, 27. NOV- Training on one for stafflanalysts of BOB, 2 1.Dec. content and MoF, and MPEA, and another BOB,MoF, and utilization o f for staff o f line ministries and M P E A budget agencies includingFinancial preparation Comptrollers and Directors o f formats PlanningIPrograms, on the content and utilization o f budget formulation formats Conduct a cash planning 3 11- 3'd Cash workshop to assist ministries MoF/CMCO- 12.Dec. Planning and agencies inreviewing and Secretariat Workshop usingthe currently cash plans; include in staffs/analysis o f BOBand MoF/CMCo- Secretariat 4 Preparation and Prepare mid-year review conduct o f mid- process and assist ministries financial data 1. Dec.- year review and agencies in carrying not assembled 6.Jan. related activities; prepare documentation for Cabinet - mid-vear review sessionhetreat 5 Prepare revenue forecasts in Revenue consultation with ministries forecasts and agencies Cabinet with respect to policy initiatives and tax administration measures consultations with impact on revenue levels. - 6 Cabinet mid- Guide Cabinet and assist year review President inundertaking mid- 8- 13.Jan. sessiodretreat year review Prepare Draft Budget 7 Prepare budget Guidelines and subject them to guidelines debate by Cabinet at Mid-Year Review SessiodRetreat - 154 - Projected Tasks/ Description Responsible Date activities parties 15 Jan Issuance o f President issues policy guidance policy guidance and budget Issue budget guidelines guidelines Dissemination o f budget guidelines and training o f line Jan. 18 19 - Budget ministries and agencies staff Workshop for including Financial Comptrollers line ministries and Directors o f and agencies Planning/Programs various ministries and agencies on preparation o f budget submissions. Analysts and Sector Heads visit ~ Provide Jan. 22 - technical LineMinistries/Agencies to Feb 22 assistance to further provide guidance on the line ministries use o f forms and with submission and agencies procedures. Receive agency BOBand MoF receive and review Feb. 22 budget FY2007108 expenditure and non- submissions tax revenue estimates from line ministries and agencies. Initial quantitative review made. Consultations with responsible Feb. 23 - authorities inagencies to ensure Mar. 15 Review and compliance and completeness in analyze terms o f budget guidelines. submissions Qualitative analysis o f submissions conducted to inform the Budget Director's recommendation to the Budget Committee Provide insight into the programs and plans o f ministries and Mar. 19- Conduct agencies with a view to assessing 30 Budget the actual needs. Agencies are Hearings challenged with potential trade- offs they must consider. Analysts and Sector Heads collate Compile notes the recommendations for agency Apr. 2 - 5 from budget budget revisions as made by hearings Budget Committee duringthe hearings. - 155 - Nr. Projected TasWactivitie Description Responsible Date s Parties - Revenue estimates revised to make it consistent with Update revenue emerging realities: policy forecasts shifts, new trends in collection, changes inthe macroeconomic - forecasts, new growth factors. Final draft sent All the necessary adjustments 20 08-May to President are incorporated into the budget. Final draft ready for Legislative process 15-May Submit draft President presents draft budget 21 budget to to the Legislature legislature May 22 - ministries and agencies develop 22 31 prioritized draft cash plans based on appropriations Cash planning proposed to legislature Cash planning M o F and CMCo Secretariat consolidate and match June 1- prioritized draft cash plans 23 15 from ministries and agencies and develop consolidated prioritized government cash plan Anticipated Legislature adopts the 24 legislative FY2007/08 National Budget June 21 approval Finalize Ministries and agencies adjust 25 prioritized cash cash plans in accordance with June 21 - plan the Legislature's appropriations 30 Ministries and agencies submit 25 Issue first and BOBapprove fist set of July 6 allotments allotments - 156 - I rA rA -a c cd OD C .d Q r: 0 Q r A 4 E 03 2 0 a, $ 14 a, 6E -0 c cd .i E aii a E e .- .3 Y8 Y a, $ s 2 i .-0 i-' cd m Ea, .e a, OD id 0 C i .d U0 B0 s0 .-bij c ; a E e .- .e Y8 Y a, Q -2 u .-0 Y c rd 0..