Udenrigsudvalget 2012-13
URU Alm.del Bilag 208
Offentligt
2012:8Mary BetleyAndrew BirdAdom Ghartey
Joint Evaluation
Evaluation of PublicFinancial Management Reformin Ghana, 2001–2010Final Country Case Study Report
Joint Evaluation 2012:8
Evaluation of Public FinancialManagement Reformin Ghana 2001–2010Final Country Case Study Report
Mary BetleyAndrew BirdAdom Ghartey
Submitted by fiscus Public finance consultants and Mokoro ltdtotheEvaluation Management groupJune 2012
Authors:Mary Betley, Andrew Bird, Adom GharteyThe views and interpretations expressed in this report are the author and do notnecessarily reflect those ofthe commissioning agencies, Sida, Danida and AfDB.
Joint Evaluation 2012:8Commissioned by Sida, Danida and AfDBDate of final report:June 2012Published by:Sida, 2012Copyright:Mary Betley, Andrew Bird, Adom GharteyDigital edition published by:Sida, 2012Layout and print:Citat/Edita 2012Art.no.SIDA61499enurn:nbn:se:sida-61499enISBN:978-91-586-4196-9This publication can be downloaded/ordered from: www.Sida.se/publications
Acronyms and AbbreviationsAAPADMDAfDBAFROSAIARICBDUBMZBPEMSCABRICAGDCEPSCFAACHRAJCIDACPIACSODACDANIDADCDFIDDPECERPFMEUFAAFARFCFMISGASGBS(HIPC) Annual Assessment & Action PlanAid and Debt Management DivisionAfrican Development BankAfrica Organisation of Supreme Audit InstitutionsAudit Recommendation Implementation CommitteesBudget Development UnitGerman Ministry of Cooperation and DevelopmentBudget and Public Expenditure Management SystemCollaborative African Budget Reform InitiativeController and Accountant General’s DepartmentCustoms, Excise and Preventative ServiceCountry Financial Accountability AssessmentCommission on Human Rights and Administrative JusticeCanadian International Development AgencyCountry Policy & Institutional AssessmentCivil Society OrganisationsDevelopment Assistance Committee (of the OECD)Danish International Development AssistanceDeveloping CountryDepartment for International Development of the UKDevelopment PartnerEuropean CommissionExternal Review of the Public Finance ManagementEuropean UnionFinancial Administration ActFinancial Administration RegulationsFinance CommitteeFinancial Management SystemsGhana Audit ServiceGeneral Budget Support5
ACroNymS AND ABBrEvIAtIoNS
GDPGIFMISG-JASGNIGoGGPRSGRAGSGDAGTZHDIHIPCIAAIAUICRIDAIEGIFMSIMFINTOSAIIPPDLGAMDAsMDBSMDBS-FMDGMMDAsMoFEPMoLGRDMPsMTEFNAONDCNDPC6
Gross Domestic ProductGhana Integrated Financial Management Information SystemGhana Joint Assistance StrategyGross National IncomeGovernment of GhanaGrowth and Poverty Reduction StrategyGhana Revenue AuthorityGhana Shared Growth and Development AgendaGerman Technical CooperationHuman Development IndexHighly Indebted Poor CountriesInternal Audit AgencyInternal Audit UnitImplementation Completion ReportInternational Development AssistanceIndependent Evaluation Group (World Bank)Integrated Financial Management SystemInternational Monetary FundInternational Organisation of Supreme Audit InstitutionsIntegrated Personnel and Payroll Database systemLocal Government AuthorityMinistries, Departments and AgenciesMulti-Donor Budget SupportMulti-Donor Budget Support FrameworkMillennium Development GoalMetropolitan, Municipal, and District AssembliesMinistry for Financial and Economic PlanningMinistry of Local Government and Rural DevelopmentMembers of ParliamentMedium Term Expenditure FrameworkNational Audit OfficeNational Democratic CongressNational Development Planning Commission
ACroNymS AND ABBrEvIAtIoNS
NPMNPPODAOECDOHCSPACPAFPBAsPDPEFAPERPETSPFMPIUPPBPPPPSCRAGBSAPSBSSECSECOSMTAPSSAST/MTAPTATINTSAUNDPUNICEFUSDWB
New Public ManagementNew Patriotic PartyOfficial Development AssistanceOrganisation for Economic Cooperation & DevelopmentOffice of the Head of Civil ServicePublic Accounts CommitteeProgress Assessment FrameworkProgramme Based ApproachesParis DeclarationPublic Expenditure & Financial AccountabilityPublic Expenditure ReviewPublic Expenditure Tracking SurveyPublic Finance ManagementProject Implementation UnitPublic Procurement BoardPurchasing Power ParityPublic Service CommissionRevenue Agency Government BoardStrategic Action PlanSector Budget SupportState Enterprises CommissionSwiss – State Secretariat for Economic AffairsShort and Medium Term Action Plan for PFMSub-Saharan AfricaShort & Medium Term Action Plan (for PFM Reform)Technical AssistanceTaxpayer Identification NumberTreasury Single AccountUnited Nations Development ProgrammeUnited Nations Children’s FundUnited States DollarWorld Bank7
PUFMARP Public Financial Management Reform Programme
Table of ContentsExecutive Summary ...................................................................................101Summary of objectives and Approach .............................................131.1 Objectives of the evaluation ..............................................................131.2 Evaluation approach .........................................................................131.3 Evaluation questions.........................................................................171.4 Fieldwork process ............................................................................181.5 Report Structure ...............................................................................19Inputs and Context: the design of PFm reform............................... 202.1 The Reform context...........................................................................202.2 PFM reform baseline........................................................................252.3 Direct Reform Inputs ........................................................................272.4 Structures to design and manage PFM reform inputs .................312.5 Complementary Donor inputs to PFM reform ...............................332.6 Civil Society pressure for PFM reform ...........................................352.7 Relevance of PFM Reform Inputs ....................................................36outputs: the Delivery of PFm reform .............................................. 433.1 Key reform outputs ..........................................................................433.2 Efficiency and coordination of reform outputs ..............................503.3 External drivers and blockers of change .......................................52outcomes and overall Assessment ................................................ 564.1 Changes in PEFA scores ..................................................................564.2 Relevance of PFM improvements for servicedelivery, especially for women ........................................................ 614.3 The effectiveness of PFM reforms: from inputsto intermediate outcomes ...............................................................63Conclusions and Wider Lessons ......................................................675.1 Summary of key conclusions ...........................................................675.2 Wider lessons....................................................................................685.3 Recommendations for the Design and Managementof PFM reform processes ................................................................71
2
3
4
5
Annex A: Summary matrices of responses to Evaluation Questions forCountry and Component Case Histories.................................. 77Annex B: List of Persons Consulted........................................................117Annex C: List of references ....................................................................120Annex D: Summary of PFm reform inputs, outputsand intermediate outcomes.....................................................124Annex E: terms of reference ..................................................................1308
Joint Evaluations ......................................................................................142
tABLE oF CoNtENtS
table of Figures & tablesFigure 1.The Evaluation Framework and the place of the EvaluationQuestions within the Intervention Logic ........................................16table 1:Overall Budgetary Trends, 2001-2010 .............................................22table 2:DP allocations to PFM reform activities ..........................................28table 3:Overview of Development Partner fundingfor PFM reform ..................................................................................28table 4:Total MDBS Commitments and Disbursements2003-2010 ...........................................................................................29table 5:Poverty-Reducing Expenditures...................................................... 61Box 1:Evaluation questions ............................................................................17Box 2:Summary of Key PFM Events, Late 1990s-present...........................24Box 3:Summary of PUFMARP 1997-2003.....................................................26Box 4:Summary of GoG’s PFM Short and MediumTerm Action Plan 2006-2009 ...............................................................27Box 5:Comparison of 2006-09 Short & MediumTerm Action Plan with identified weaknesses ...................................37Box 6:Alignment of 2006-2009 Short & MediumTerm Action Plan with PEFA Results..................................................40Box 7:Overview of PFM Reform Outputs and DP Contributions .................46Box 8:Summary of changes in intermediate outcomesbased on PEFA assessments ..............................................................58Box 9:Summary of PFM reform inputs, outputs, and intermediateoutcomes ...............................................................................................73
9
Executive Summaryoverview of findingsThis Country Report has been prepared by Fiscus Limited, UK, in collabora-tion with Mokoro Ltd, Oxford, as one of three country reports in the JointEvaluation of Public Financial Management Reform, managed by the Afri-can Development Bank, Denmark and Sweden. The evaluation looked at twomain questions: (i) where and why do Public Finance Management (PFM)reforms deliver results and (ii) where and how does donor support to PFMreform efforts contribute most effectively to results? Our overall conclusionsare presented in the Summary Matrix contained in Annex A, which summa-rises, for the overall Ghana PFM reform programme 2001 -2010 and for fourcomponent areas within the reform programme, our answers to the 12 Evalu-ation Questions, posed within the study.The study’s main conclusion is that, relative to the significant funds expendedon Public Finance Management (PFM) reform over the study period, successhas been largely disappointing. The most substantial progress was found in astronger legislative base. However, similarly to other countries, Governmentof Ghana (GoG) has experienced significant challenges in implementing thenew laws. Otherwise, the most effective reforms appear to have been the rev-enue management activities, as they have led to a sustained output in the formof changed processes and a significant increase in revenues as a share of GrossDomestic Product (GDP) during the period studied.During the period studied, PFM outcomes deteriorated in a number of keyareas, including budget credibility, the build-up of expenditure arrears, andcompliance with expenditure controls. The deterioration in outcomes islargely a consequence of the failure to achieve a number of desired PFMreform outputs, including a fully-functioning Medium Term ExpenditureFramework (MTEF), a workable Financial Management Information System,and greater efficiency in resource flows to Ministries, Departments and Agen-cies (MDAs) and MMDAs. The limited improvements in intermediate out-comes were largely independent of PFM reform actions.On the other hand, Ghana has experienced significant improvements inregard to external oversight. The most notable achievements have been theclearance of the backlog of audits, the introduction of open Public AccountsCommittee (PAC) hearings on audit reports, and the timely submission of cen-tral government audit reports to Parliament.The degree of political commitment and leadership was found to be the mainbinding constraint to the relative success of PFM reform. Commitment tend-10
ExECutIvE SummAry
ed to be relatively strong at the beginning of reforms, which enabled legisla-tion to be enacted and reform programmes to be initiated, but commitmenthas tended to wane over time as other political priorities have taken prece-dence, thus hampering the completion of reforms under way and the imple-mentation of the new legislation. Political incentives for reform were found tobe strongest at the beginning of a new government. However, progress was attimes hampered by disruptions to the continuity of senior management per-sonnel following electoral change. The political cycle also led to a recurringpattern of rapid fiscal expansion, followed by fiscal consolidation and theimposition of stronger expenditure controls. As a consequence, for two yearseither side of each election both administrative effort and political attentionhave been repeatedly diverted away from the implementation of reforms.In terms of the role of DPs in the success, or otherwise, of reforms, it did notappear that Development Partner (DP) support was a key positive factor inachieving PFM reform progress. While the total amount of resourcesappeared to be sufficient, and disbursement delays were not a critical hin-drance to reform, the effectiveness of DP contributions was undermined bypolicy space constraints, particularly in terms of the appropriate design of spe-cific reforms. In particular, there has been a tendency to focus on technologi-cal solutions rather than changes to the underlying processes. There has alsobeen a failure to adapt reform designs in the light of implementation results.External support has therefore appeared to have greater traction at the earlystages of reforms, in facilitating design and the initial implementation, thansubsequently in sustaining or deepening reform initiatives, which would clear-ly call for adaptation of initial designs and the application of the lessons ofimplementation experience. Exceptions to this have occurred where therehave been sustained DP contributions over a longer period of time, with sup-port demand-driven, targeted to specific needs, and adaptable in terms ofdesign and re-design. Reforms to revenue administration are the most salientexample.
implicationsAs Ghana, a newly middle-income country enters a phase characterised bylikely changes in its partnership with DPs, including in terms of the MDBS, itsgovernment and senior management are likely to exercise greater active con-trol of its reform programme, including the provision of resources. The studyhighlighted the following areas of potential risk management for both GoGand DPs in its future PFM reforms.• GoG and DPs should recognise the importance of continuity in reform initiatives and consider longer-term and flexible approaches to supportthat can take advantage of windows of opportunity (e.g. political space)while allowing for scaling back when conditions for making progressare less favourable.
11
ExECutIvE SummAry
• Explicit analyses should be undertaken of stakeholder readiness for reform, particularly for Information Technology (IT) projects andreforms involving functions to be devolved. A technological (IT) solu-tion may not always be appropriate, particularly if the underlyingmanual processes (e.g. internal controls) are inadequate.• Identify demand and management factors that are likely to facilitate successful reform, including the role of Civil Society Organisations(CSOs) and peer-to-peer learning opportunities.• Allow adequate time to plan and sequence reforms, including plan-ning, designing, testing (piloting), reviewing, and taking action basedon reviews of the testing, and completing reforms. Successful reformprogrammes should not be led by systems (and technology).• Consider the provision of training in leadership for senior political and administrative levels, particularly for those in newly-appointed posi-tions.• Improve incentives for active senior management linked to reform per-formance.• In light of the impact of the electoral cycle, build into the reform pro-gramme periods of review or a pause in the timetable for bringing onboard new stakeholders.
12
1. Summary of Objectivesand ApproachThis Country Report has been prepared by Fiscus Limited, UK, in collabora-tion with Mokoro Ltd, Oxford, as one of three country reports in the JointEvaluation of Public Financial Management Reform, managed by the Afri-can Development Bank, Denmark and Sweden. It incorporates the commentsreceived on the first draft by the Management Group and the peer reviewer aswell as those of the in-country stakeholders, who attended the presentation ofthe draft report in Accra during January 2012. This report, together with theevaluation of PFM Reforms in Malawi and Burkina Faso, has been incorpo-rated into an accompanying synthesis report and will be the subject of a widerprocess of dissemination within the Africa region and beyond.
1.1 ObJECTivES OF ThE EvAluATiOnThe evaluation aims to address two core questions presented in the Terms ofReference:a) Where and why do PFM reforms deliver results (i.e. improvements in thequality of budget systems)?;andb) Where and how does donor support to PFM reform efforts contribute mosteffectively to results?It is thus a dual evaluation, involving first an evaluation of the overall pro-grammes of PFM reform conducted over 2001 to 2010 in Ghana, and second-ly, an evaluation of the external support provided to these reforms by donors.
1.2 EvAluATiOn APPROAChThe evaluation framework has been defined in the Inception Report for the3-country study1. It utilises the Organisation for Economic Development(OECD-DAC) evaluation criteria and thus assesses the relevance, efficiency,effectiveness and sustainability of both the overall programme of PFM reformand the external support provided. The overall goal is to draw lessons on (i)the types of PFM reform pursued and their interaction with the Ghana con-text and (ii) the mechanisms of external support that most contributed to theirsuccess. Success is associated with improvements in the quality of budget sys-tems, as measured primarily by changes in the Public Expenditure and Finan-cial Accountability (PEFA) assessment framework indicators and the narrative1Lawson, A. (2011). Joint Evaluation of PFM Reforms in Burkina Faso, Ghana andMalawi: Inception Report. Oxford, Fiscus Limited, Mokoro Limited.13
1. SummAry oF oBJECtIvES AND APProACH
PEFA reports. The evaluation framework characterises these changes as inter-mediate outcomes in a ‘PFM Theory of Change Framework’, which underliesthe evaluation.The Framework (presented in Figure 1 below) requires the detailing ofPFM Reform inputs, outputs and intermediate outcomes and the examinationof the relationship between them. In addition, it allows for the analysis of howexternal constraints – conceptualised as political, financial and policy spaceconstraints – impact on the translation of reform inputs into outputs.Inputsare defined as the resources and other inputs provided in order topromote PFM reform. These are divided between direct funding by govern-ments to internal PFM reform efforts, external funding by Donors of PFMreform efforts and complementary inputs by Donors, aimed at facilitating bet-ter PFM through the use of country systems and the provision of budget sup-port, or improving the design and implementation of PFM reforms throughpolicy dialogue and external monitoring (often linked to budget support or toIMF supported arrangements).Outputsare defined as the immediate changes in the architecture andsubstance of the PFM system generated by the combined set of inputs. Theseare categorised into four groups: i) Changes in human resource endowments(people and skills); ii) Changes in laws, procedures and rules; iii) Changes insystems and business processes; and iv) Changes in Organisational factors (thequality of management, the work culture, the degree of organisational devel-opment).Intermediate Outcomesare the changes generated in the quality ofthe PFM system, as measured by changes in the quality of:i) Strategic budgeting;ii) Budget Preparation (including budget deliberation by the Legislature);iii) Resource management (covering both inflows and outflows);iv) Internal controls, audit and monitoring;v) Accounting and reporting; andvi) External Accountability.The framework utilises the PEFA assessment framework to measure changesin each of these clusters of functions, based on a categorisation of the sub-dimensions of the PEFA indicators between each of them. The categorisationis based on Andrews (2010) and is also applied by De Renzio et al (2010). Thecharacteristics of (a) transparency and comprehensiveness, (b) the quality oflinks to policies and plans, and (c) the efficiency and effectiveness of control,oversight accountability are subsumed within this categorisation.External constraints are seen to impinge on the PFM reform ‘productionfunction’, in other words with the capacity of PFM reform inputs to generatethe planned outputs. Key constraints are “political constraints” related to thedegree of ownership and support for given PFM reforms, “financial con-straints” related to the ability to finance PFM reforms in the face of compet-ing priorities, and “policy space constraints”, related to the nature of policyideas which might potentially be considered in designing PFM reforms.14
1. SummAry oF oBJECtIvES AND APProACH
A key task for the evaluation of PFM reforms in Ghana was therefore toexamine whether there have been external constraints which have preventedor slowed down the translation of reform inputs into reform outputs and whichof the three types of constraint have been most significant in this respect. Thepurpose has been to identify what have been the binding constraints on differ-ent types of reforms and to reach a judgement on whether reform challengeswere due to reform models which lay beyond the prevailing “production pos-sibility frontier”, examining also the role of donor support in this process.
15
16
figure 1. the Evaluation framework and the place of the Evaluation Questions within the intervention logicOUTPUTSINTERMEDIATE OUTCOMESFINAL OUTCOMES
INPUTS
PFm reform Intervention LogicPeople &Skills:Strategic BudgetingEQ 9
• Government-funded inputs
Government PFm reformInputsBudget PreparationresourcemanagementEQ 11
Fiscal Discipline:
EQ 1
EQ 2
• umbers of PFM professionals N(auditors, etc.)• evpt of specific skills, D(incl. reform management)
• ulfillment of planned fiscal Ftargets• aintenance of Sustainable MDeficit
DP funded support to PFmreform
1. SummAry oF oBJECtIvES AND APProACH
(delivered in aharmonised&aligned manner):• Institutional strengthening• Advisory TA• Diagnostca Work
EQ 7
• hanges in Laws. Rules &CProceduresEQ 6
Laws, rules &procedures
Strategic Allocationofresources:• onsistency of executed C&approved aggregate /departmental budgets• onsistency of revenue Ctargets &coIIecfions
Systems &BusinessProcesses:Internal control.Audit &monitoringAccounting & reportingEQ 10
EQ 3
Complementary DevelopmentPartner inputs:External Accountability
• omputer systemsC• pecific approaches to budget-Sing, accounting addit, treasurymanagement, etc.
• Use of Country Systems• Provision of Budget SupportP• olicy Dialogue &monitoring ofPFM reform
• mproved managementI• rganisational developmentO• mproved work culture.IEQ 8
organisational factors:
• volution of Unit costs ofpublic EservicesEQ 12
operational Efficiencyin Public Spending
Public &Civil Society pressurefor improved PFm:
EQ 5
• Voting preferences• Political lobbying• Research &advocacy• Regional/ international norms
EQ 4
• olitical Constraints:Degree of leadership /ownership of PFM reform at Administratve and Politcal levels.P• Financial Constraints:Availability &timeliness of funding for PFM reform, presence.' absence of economic crisis.• olicy Constraints:Openness of the policy reform agenda, receptivity to new ideas. policy space for long-term reforms.P
External Constraints on the PFm reform “production possibility frontier”
1. SummAry oF oBJECtIvES AND APProACH
1.3 EvAluATiOn quESTiOnSbox 1 : Evaluation questionsA Inputs & context: the design of PFm reformWhat has been the nature and the scale of PFM reform inputs pro-vided by Government and by Donors?What types of structures have been used for the design and man-agement of these reform inputs? Have these structures served toprovide a coordinated and harmonised delivery framework?What types of complementary actions have Donors taken to sup-port PFM reforms and what has been their significance? Have theyhad any influence on the external constraints to reform?To what extent has there been domestic public pressure or re-gional institutional pressure in support of PFM reform and whathas been the influence on the external constraints to reform?How relevant was the PFM reform programme to the needs andthe institutional context? Was donor support consistent with na-tional priorities? To what extent were adaptations made in re-sponse to the context and the changing national priorities?
EQ. 1:EQ. 2:EQ. 3:EQ. 4:EQ. 5:
B outputs: the delivery of PFm reformEQ. 6:What have been the outputs of the PFM reform process and towhat extent has direct donor support contributed to these out-puts?EQ. 7:How efficiently were these outputs generated? Was the pacingand sequencing of reforms appropriate and cost-effective? Wasthe cost per output acceptable?EQ. 8:What have been the binding external constraints on the deliveryof PFM reform outputs: political, financing or policy factors? Howhas this varied across different PFM reform components?C outcomes: overall assessment of PFm reform & of donorsupport to PFm reformEQ. 9:What have been the intermediate outcomes of PFM reforms,in terms of changes in the quality of PFM systems?EQ. 10:To what extent have the outcomes generated been relevant toimprovements in the quality of service delivery, particularly forwomen and vulnerable groups?EQ. 11:Have reform efforts been effective? If not, why not? If yes, towhat extent PFM reform outputs been a causal factor in thechanges identified in intermediate outcomes?EQ. 12:To what extent do the gains identified at the Intermediate Out-come levels appear sustainable? Is the process of PFM reformsustainable?17
1. SummAry oF oBJECtIvES AND APProACH
The Ghana study was based on twelve evaluation questions common to thethree country cases (see Box 1 above and the fuller presentation of responsesto evaluation questions and corresponding judgement criteria in Annex A).The Evaluation Questions are structured so as to provide a standardisedframework for assembling evidence, so that the results of the country studiescan be easily synthesised to provide answers to the overall high-level ques-tions, which the evaluation addresses. The questions marry core OECD DACevaluation questions with the concerns specific to the study.
1.4 FiEldwORk PROCESSThe research for the country study was undertaken in two phases. The firstphase, an initial desk phase, involved reviewing background documents andgathering published data on the Ghana context and the specific PFM reformscarried out from the late 1990s to the present. The output from this phasewas a desk report, whose findings have been included in the current report.2The study’s second phase, the field visit, involved extensive interviews andfocus group meetings with stakeholders in Accra and in one district adminis-tration. An Aide-Memoire setting out the findings was circulated followingthe field visit.3While each of the three country case studies examines PFM reforms dur-ing the 10-year period from 2001 through 2010, the Ghana case study includ-ed the late 1990s in its review. This extended time period was considered rel-evant in order to include the whole of the PUFMARP reforms, which beganin 1998, and which provided the foundation for many of the reforms, whichtook place during the study period. In addition, the study makes references inpassing to recent reforms that were initiated during 2010 (e.g. in budget for-mulation [programme-based budgets] and the GIFMIS) in order to commenton the extent to which they respond to previous reform experience, since it istoo early to evaluate the performance of these reforms.In Ghana, as in the other two country studies, an assessment was made ofthe overall progress of the PFM reform programme and of a number of spe-cific reforms, as case histories. The case histories focussed on four specificPFM reforms, specifically, (i) Financial Management Information Systems(FMIS), including Budget Planning and Expenditure Management System(BPEMS) and the Integrated Payroll and Personnel Database (IPPD2); (ii) theMedium Term Expenditure Framework (MTEF); (iii) revenue management;and (iv) internal audit. Together, these components cover the majority (interms of monetary value spent) of the reforms that were carried out duringthe period studied. At the same time, they provide for comparative results tobe studied, with two of these reforms (FMIS and MTEF) enjoying significantDP support, whilst a third (internal audit) had very limited external support.2318
Gordon, A, and M Betley (2011), Evaluation of Public Financial Management Reformin Ghana 2001-2010: Desk Report, Fiscus/Mokoro: Oxford.Betley, M, A Bird, and A. B. Ghartey (2011), Joint Evaluation of Public FinancialManagement Reform in Burkina Faso, Ghana & Malawi: Ghana Country Case Study– Field Visit: Aide-Mémoire, Fiscus/Mokoro: Oxford.
1. SummAry oF oBJECtIvES AND APProACH
The fourth reform, revenue management, is common to the three case studycountries and provides a contrast in terms of being relatively technical innature and relatively self-contained (i.e. predominantly managed by a singleinstitution).Reviews of documentation and data, stakeholder interviews and focusgroups provided the main evidence for the analyses against the evaluationquestions. The stakeholders consulted are shown in Annex B, and the docu-ments examined during the study are listed in Annex C. The former includedrepresentatives from central agencies (specifically, Ministry of Finance andEconomic Planning [MoFEP], Controller and Accountant-General’s Depart-ment [CAGD], and Ghana Revenue Authority [GRA]), as well as sector min-istries (Ministry of Health and Ministry of Tourism), and one local govern-ment unit (a district assembly, known collectively in Ghana as Metropolitan,Municipal, and District Assemblies [MMDAs]).While every attempt has been made to ensure that the responses to theevaluation questions are based on the best evidence available, the lack of rel-evant, consistent and comprehensive data represents a significant shortcomingto the analysis. This is particularly true for actual GoG expenditures forPFM-related activities, as well as for disaggregated data on DP PFM commit-ments and disbursements. Furthermore, the change in government in 2009,following a period of 8 years, meant that many of those who were in leader-ship or management positions during the implementation of the reforms beingstudied were no longer in post. This was also true for DP officials involved inPFM reforms, many of whom moved on to other positions outside of Ghana.Nonetheless, the study team managed to consult with a number of formerGoG officials who had been involved in the implementation of the reformsunder the previous government.
1.5 REPORT STRuCTuREThe report follows the standardised structure for the three country studies. Inaddition to this chapter on the Study objectives and approach, it comprises (i)a chapter describing the context and evaluating the inputs to PFM reforms inGhana; (ii) a chapter on the planned and actual outputs; (iii) a chapter discuss-ing the intermediate outcomes; and iv) a chapter providing conclusions andwider lessons.A series of annexes contain summary matrices of the responses to the 12Evaluation Questions for Ghana as a whole and for the four PFM reform“case histories” (Annex A), a list of those consulted during the study (AnnexB), bibliographic references (Annex C), and an overall summary of PFMreform inputs, outputs and intermediate outcomes over the 2001 – 2010 evalu-ation period (Annex D).
19
2. inputs and Context: the designof PFM reform2.1 ThE REFORM COnTExTGhana has a population of 23 million people, of whom around 51% live inrural areas. Per capita gross national income (GNI) was US$1,530 in 2009.4The country is divided into 10 administrative regions (Ashanti, Brong Ahafo,Central, Eastern, Greater Accra, Northern, Upper West, Upper East, Voltaand Western). Ghana has seen progress in reducing poverty in recent yearsalthough there are considerable differences in socio-indicators between thenorth and the south, with relatively greater poverty in the northern regions.The absolute number of poor declined sharply in the south between 1992 and2006 (2.5 million fewer poor), while it increased in the north (0.9 million morepoor).5In 2010 the country’s ranking in the United Nations Development Pro-gramme’s (UNDP) human development index (HDI) was 130 out of 182 coun-tries. Ghana became the first African country to reach the first MillenniumDevelopment Goal (MDG) on halving its poverty and hunger rates before2015.
overview of economic and fiscal performanceWhen the New Patriotic Party (NPP) won the elections in 2000, it inheritedan economy that had suffered a macroeconomic shock in 1999. Inflation wasrunning at 41%. The budget deficit and external and domestic debts were atunsustainable levels and this inevitably had a negative impact on expenditurein social sectors and poverty reduction activities as the Government struggledto service its debt and stabilise public finances.However, the anti-inflationary monetary policy and the fiscal consolida-tion that was implemented by GoG facilitated a subsequent period of positivegrowth that was also helped by strong commodity prices. In 2007, the annualreal growth rate reached 6.3%; only the second consecutive year that thegrowth rate had been in excess of 6% since the 1980s. Between 2000 and2007, the average real gross domestic product (GDP) growth rate was 5.2%, afigure higher than the sub-Saharan average of 4.8% in the same period.6In2008 growth rates reached their highest for two decades rising to 7.3%.7Morerecently economic growth has slowed in Ghana as it has globally. In Ghanaeconomic growth in 2009 dropped to 4.7% – the lowest since 2002. Economic456720
PPP, current international US$ – World Bank (2011)World Bank, 2011Allsop et al, 2009.Using rebased GDP, this figure would be 8.4%.
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
growth began to recover in 2010 and is projected to accelerate to almost 14%in 2011 on the back of global recovery, exceptional public investment in therising oil sector, and revenues from anticipated new oil discoveries.8Ghana’s macroeconomic situation is considered to be “delicate” by theWorld Bank.9This is in part due to its reliance on primary products but alsoto instability in the region, which has led to Ghana’s hosting refugees fromother countries. Agriculture accounts for about a third of GDP, while theindustrial sector contributes 28%. Ghana continues to be overly dependent ona few primary commodities. A narrow range of exports constitutes a signifi-cant part of Ghana’s GDP, specifically, gold (42%) and cocoa (30%), whichtogether accounted for over 70% of exports in 2009. Despite government poli-cies over a number of years, which aimed to encourage industrialisation, man-ufacturing accounts for just 9% of exports.10The discovery of substantialreserves of oil and gas in Ghana will provide a new source of revenue but onewhich is vulnerable to shocks in global oil prices.The global financial crisis has had an impact on the Ghanaian economy,with lower export values, a fall in commodity prices, less and more expensiveforeign capital, lower remittances and fewer tourists. This in turn has had animpact on income growth, job losses and budgetary pressures, leading toreduce government spending on social protection systems. However, with thestart of oil production in late 2010, GDP per capita is projected to exceed$1,400 in 2011.11For 2010, the Ghana Statistical Services projected a 6.6%growth in real GDP.12The impact of oil-related investment expenditures (e.g.construction, information and communication technologies, hotels, financialintermediation) and continuously favourable climatic and terms of trade con-ditions were seen to trigger a slight increase in economic growth in the firsthalf of 2010. This was reinforced by the moderate rebound in private sectorcredit growth which has occurred since June 2009.The Government’s stabilisation policies led to improved fiscal perfor-mance post-2001 (Table 1). Revenue generation has been stronger, assisted inpart by improved tax administration, greater collection of internally generat-ed funds (IGFs), and higher levels of remitted profits, while expenditures havebeen contained, partly through reductions in debt servicing requirements.Domestic revenues increased to nearly 24% of GDP by 2005. This revenuemobilisation effort, supported by HIPC debt relief, allowed Government bothto reduce its reliance on domestic financing of the deficit and to increasedomestically financed primary expenditure to just fewer than 30% of GDP in2005, up from 23% in 2002. Fiscal performance has also benefited from lowerdomestic interest rates.
891011
African Economic Outlook, 2011World Bank, 2010.African Economic Outlook, 2011This figure differs from the GNI figure above as the latter also includes its incomereceived from other countries (notably interest and dividends), less similar paymentsmade to other countries.12 World Bank, 2010
21
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
table 1: overall budgetary trends, 2001-2010% of GDPActual Actual Actual Actual Actual Actual Actual Actual Actual2Actual2
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010
Total25.0 21.1 25.5 30.2 28.3 27.3 28.8 27.5 16.4 17.4Revenues1Total32.7 26.1 29.0 33.3 30.7 34.3 37.3 41.0 20.4 23.5Expendi-turesAggregate(7.7) (5.0) (3.6) (3.1) (2.5) (7.0) (8.5) (13.5) (4.0) (6.1)Deficit11Including grants2Figures shown are as % of non-oil GDP (re-based)Source: IMF, WB
Political contextThe current political structure in Ghana is framed by the 1992 Constitution.This established the fourth republic, multi party elections, and a return toconstitutional rule with an elected administration. The period covered by thisstudy has seen a consolidation of democratic rule in Ghana. The NationalDemocratic Congress (NDC) party’s rule came to an end in 2000, and theNPP government took over for 8 years, after which NDC regained power.There has been evidence over this period of a deepening of democracy,including a strengthened role for Parliament, and a more active civil societyand media. In 2002, the then-President inaugurated a reconciliation commis-sion to look into human rights violations that had occurred during the mili-tary rule.The NPP Government focused on poverty reduction and, in 2003, the firstGhana Poverty Reduction Strategy Paper (GPRS I) was published, covering2003-2005. This was to be the framework for co-ordinating social and eco-nomic development in Ghana. It was to inform government programmes inall sectors, and set out a number of priorities for government. The strategyhad a strong emphasis on poverty reduction and improving living standards.In December 2004, national elections with a strong turn-out (85%) wereheld, and the incumbent President was re-elected. In 2006 the new Govern-ment published the second GPRS (2006-2009), entitled the “Growth and Pov-erty Reduction Strategy”, marking its emphasis on economic growth and itsaim of facilitating Ghana’s promotion to the status of a middle-income coun-try by 2015, a goal which was achieved in July 2011, following the rebasing ofGDP.13
22
13 As indicated by the World Bank.
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
Following close elections (including a run-off election) in December 2008, agovernment led by the NDC was elected and took office in January 2009.14The new government prepared its national development strategy, known asthe Ghana Shared Growth and Development Agenda (GSGDA), covering2010-2013. The NDC Government also prepared a longer-term strategy doc-ument, “An Agenda for Shared Growth and Accelerated Development for aBetter Ghana (2010 – 2016) which, as per Ghanaian Constitution requires “...within two years after assuming office, the President shall present to Parlia-ment a coordinated programme of economic and social development policies,including agricultural and industrial programmes at all levels and in allregions of Ghana”. The implementation of the Coordinated Programme isplanned to be undertaken through Medium-Term National Development Pol-icy Frameworks. The Ghana Shared Growth and Development Agenda(GSGDA), 2010 to 2013, covers the first phase of this Coordinated Pro-gramme. It is expected to form the basis for the preparation of developmentplans and annual budgets at the sector and district levels throughout the coun-try. The Coordinated Programme is expected to be read alongside the GhanaShared Growth and Development Agenda (GSGDA), 2010 to 2013 and othersector specific policy documents mentioned elsewhere in the CoordinatedProgramme document.Box 2 summarises key PFM events from the late 1990s, with the full chro-nology set out in Annex A of the Ghana Desk Report.
14 It has been noted that the 2008 election provided a test of the country’s democraticstrength; in spite of a very close margin of votes between the two dominant parties,state institutions, in particular the judiciary and the electoral commission, withstoodsignificant tension (Allsop et al, 2009).
23
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
box 2: Summary of Key PfM Events, late 1990s-presentF rom late 1990s-early 2000s••Introduction of new classification system, which simplified thecategorisation of economic expendituresRemoving dual budgeting
2003-2005•••••Greater macro-fiscal stability achievedFirst GPRS, 2003-2005Greater links between planning and budgeting process and GPRS inthe form of MTEF workshops with MoFEP and MDAs to determine thelinkage of policy proposals in budget submissions to the GPRSAPR process (first APR in 2003)Achieving of HIPC completion point, leading to greater availability ofresources, to be targeted to investment; included in medium-termmacro-fiscal framework
2005-2006•••••2006 budget was first one passed before beginning of the comingbudget yearGPRSII 2006-2009STAP, emphasising the strengthening of MTEFGreater PAC involvement in budget processIncorporation of MDRI resources in medium term macro-fiscalframework
2007-2008••••••Greater analysis of links between budgets and GPRSChanges in classification; streamlining of MTEF activitiesFirst introduction of policy hearingsGreater focus on technical hearings: attempts to place greateremphasis on justification of budget submissions for allocation ofadditional resourcesPreparation of initial MTEF submissions by pilot MMDAsMTEF training
2009-2010•••••Ghana Revenue Authority Act passedProcess of initiating Ghana Revenue Authority beginsWork on GIFMIS (to replace BPEMS) beginsPlans for the introduction of programme budgeting under wayPreparation work on Petroleum Revenue Management Bill begins
24
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
2.2 PFM REFORM bASElinEDuring the last 15 years, there have been two main PFM reform agendas, (i)the Public Financial Management Reform Programme (PUFMARP), from1997-2003; and (ii) GoG’s Short and Medium Term Action Plan (ST/MTAP),covering 2006-2009. The suite of reforms associated with the Ghana Integrat-ed Financial Management Information System (GIFMIS), begun in 2010,may be considered to represent a significant part of the current PFM reformprogramme; however, the study focuses on the reforms pre-2009/10, as thesehave been under way for a significant enough amount of time to review theireffect.15These PFM reform programmes have been set in the overall contextof GoG’s national medium-term development plans, the GPRS I, GPRS II,and the GSGDA. Overall, in comparison with other countries in the region,Ghana was an early adopter in sub-Saharan Africa of key PFM reforms, suchas the Medium Term Expenditure Framework (MTEF).The main central GoG agencies relating to PFM are the Ministry ofFinance and Economic Planning (MoFEP), the Controller and AccountantGeneral’s Department (CAGD)16, the Public Services Commission (PSC), theOffice of the Head of the Civil Service (OHCS), the State Enterprises Com-mission (SEC), and the National Development Planning Commission(NDPC). The Bank of Ghana is the Government’s banker; GoG also operatesaccounts at commercial banks. Also part of the central management frame-work, and critical for PFM, are the Public Procurement Authority (PPA)17, theIAA, and the Ghana Revenue Authority (GRA); these agencies operate underseparate Acts and have a statutory mandate. External scrutiny agenciesinclude the Ghana Audit Service (GAS), headed by the Auditor-General, andtwo Parliamentary committees, the Finance Committee (FC), and the PublicAccounts Committee (PAC).
15 This is for practical reasons but is not to give the impression that PFM reforms havestalled under the new government or that the evaluation has not been balanced in itsreview.16 The Controller and Accountant General reports to the Minister of Finance.17 Formerly, the Public Procurement Board
25
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
box 3: Summary of PufMaRP 1997-2003the Public Financial management reform Programme (PuFmArP)was a 6-year multi-component GoG programme to strengthen PFM. TheProgramme, which was under way between 1997 and 2003 was supportedmainly by funding from IDA, with co-financing provided by DFID, CIDA, andthe EU. The Programme’s components included:1. Budget preparation – introduction of an MTEF;2. BPEMS, an integrated financial management information system;3. Cash management – introduction of a modern cash managementsystem;4. Aid and debt management – improving data on aid and debt managementand the links with CAGD and BoG;5. evenue management – introduction of VAT, unique Taxpayer RIdentification Number, IT system for tax assessment, collection andreporting, and strategy for managing customs data;6. Procurement – formulation of national procurement code anddevelopment of mechanisms for compliance with code;7. Auditing – development of national audit standards, specification of auditreports, and introduction of value-for-money audits;8. Legal framework – review of legislative framework and development ofrevised financial rules and regulations for Parliamentary approval;9. Human resources development – training for staff in programmecomponent areas.Components in italics indicate those where the scope was scaled down during projectimplementation, due in part to make way for the additional resource requirements ofBPEMS.
26
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
box 4: Summary of gog’s PfM Short and Medium-term actionPlan 2006-2009At the beginning of 2006, MoFEP published its three-year strategic plan andits short and medium-term PFM Action Plan (ST/MTAP), covering the period2006-2009, in line with GPRSII. The Government updated the Action Planin August 2006, following the PEFA assessment. Theshort-term ActionPlansets out reforms currently being introduced, which focus on improvingthe efficiency of resource and information flows through the system. Thespecific measures in the short-term Action Plan include:• The on-going building of an improved Integrated Personnel and Payroll Database system (IPPD2);• The Government’s integrated financial management system: the Budget and Public Expenditure Management System (BPEMS);• The introduction of a decentralised payment system, beginning initially in pilot ministries, including the Ministry of Education;• Strengthening of internal audit and procurement processes, focusing mainly on establishing effective internal audit and procurementinstitutions and processes in MDAs and MMDAs and accompanying stafftraining.In addition to these short-term measures, the Government’smedium-termaction planconsists of a matrix of reforms centred on 9 focal areas. Withineach focal area, output targets are given, the main agencies responsibleas well as other agencies involved are named, activities to be undertakenare detailed, and the risks are identified. The reforms are comprehensiveand cover most areas of the PEFA framework. The action plan was revisedfollowing the 2006 PEFA assessment, and the short and medium-termmeasures were prioritised and sequenced and a rough estimate of costswas added.
2.3 diRECT REFORM inPuTSEQ 1:What has been the nature and scale of PFM reform inputs providedby Government and Donors?
country overviewTable 2 summarises DP allocations to PFM reform activities from 1998 to2010. DPs represented the bulk of funding for PFM reform activities. Themajority of funds have been spent on the FMIS reforms (BPEMS, IPPD2 andGIFMIS), followed by strategic budgeting (specifically, MTEF-related activi-ties). Significantly fewer amounts were spent on resource management (reve-nue, expenditure, aid, and debt), and audit. Additionally, limited diagnosticwork has been undertaken by both the World Bank and the IMF, as well as byother DPs in preparation for the provision of their programmatic support.27
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
table 2: DP allocations to PfM reform activitiesPFm Support by main DPWorld BankEUBMZOthertotal12
Estimated specific PFm projectsupport, 1998-2010 (mn uS$)118.411.210.8210.751.1
Data represent best estimates, based on, in some cases, very limited available data.€10.5 million.Sources: WB, EC, BMZ, DFID, MoFEP
With the exception of FMIS (where the majority of funds were spent on hard-ware and software), the bulk of the funds on PFM reforms have been spent ontechnical assistance. GoG support has been spent primarily on training staffon existing reform initiatives under way or rolling out such reforms.18Themain DPs supporting PFM reforms (by value) during the study period includeDFID, the EC, BMZ/GTZ, the IMF, and the World Bank. Others includeDanida, CIDA, Japan, KFW, Switzerland, and UN.Table 3 summarises the funding from the main DPs on PFM activitiesduring the study period. PFM-specific DP support was in the form of multi-component programmes, managed by either a single DP (e.g. Good FinancialGovernance), or a pooled-funding arrangement (e.g. GIFMIS).19
table 3: overview of Development Partner funding for PfM reformPFm Focal AreaStrategic budgeting/budgetpreparationResource managementInternal controls, audit,monitoringAccounting, reporting (BPEMS)External accountabilitytotal1
Estimated specific PFm projectsupport, 1998-2010 (mn uS$)14.619.00.715.611.251.1
Data represent best estimates, based on, in some cases, very limited available data.
28
18 It is to be noted that the GoG expenditure estimates given in this section are specifi-cally related to reform activities (i.e. excluding day-to-day activities of relevant institu-tions) and do not include personnel-related expenditures.19 GIFMIS is a pooled fund with a joint steering committee and common funding ofagreed activities. Some harmonisation of M&E has taken place in order to ensure onereporting standard; this was funded by DFID.
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
There was a marked reduction in DP funding for PFM activities between2004 and 2006. This coincided with the commencement of a programme ofmulti-donor budget support (MDBS), which provided significant funds to theConsolidated Fund (annual average of just over US$300 mn since 2003)(Table 4). These funds were available to support GoG’s budget, but were notaimed specifically at PFM reform activities, nor can it be assumed that theywere used as such. A recent study20of the benefits of MDBS examined wheth-er or not the shift to MDBS: (i) resulted in increased DP project funding forPFM reform (i.e. to support MDBS); and/or (ii) facilitated greater governmentspending on PFM reform through a substitution effect. The study found noevidence to support either proposition.
table 4: total MDbS commitments and Disbursements 2003-20102003 2004 2005 2006 2007 2008 20092
2010
Commitments (mn US$)1281.4 302.2 285.3 372.4 319.6 347.9 601.1 451.5Actual disbursements(mn US$)MDBS (actual) as %of dev. assistanceMDBS (actual) as %of gov’t expend.12
277.9 309.0 281.9 312.2 316.6 368.1 525.2 403.030.0% 26.7% 29.3% 33.0% 26.5% 25.7% 34.6%N/A 12.7% 10.2% 8.3% 7.7% 8.3% 9.3%––
Data refer to pledges, as officially recorded by MoFEP.Commitments and disbursements reflect funding provided in response to the impact onthe economy of the global financial crisis.Sources: MoFEP, GAS (audited accounts), BoG (exchange rates)
component case historiesThe FMIS case history includes two main FMIS-related reforms, the intro-duction of the Budget Planning and Expenditure Management System(BPEMS) under PUFMARP, and the Integrated Personnel and Payroll Data-base (IPPD2). For BPEMS, external support was provided mainly between1997 and 2003 under PUFMARP, with minimal external amounts providedthereafter. The main sources of funding under PUFMARP were IDA(US$15.3 mn), and GoG ($4.6 mn). Following the conclusion of PUFMARP,GoG continued providing funding to BPEMS from its annual budget. It isnotable that the final expenditures under PUFMARP were 50% higher thanoriginally programmed; this additional expenditure was financed by transfer-ring part of the budget originally intended for the revenue management com-ponents (specifically, covering support to Customs, Excise and PreventativeService (CEPS)). Inputs centred on TA for system design (re-engineering busi-ness processes), the provision of hardware (including networks), and financialmanagement software (based on Oracle Financials).For IPPD2, DP support was provided by DFID, largely for system specifi-cation, hardware, application licence, customisation costs and training staff inMDAs.20 Betley and Burton, 2011.29
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
MtEfThe majority of funding for the MTEF came from PUFMARP; total fundingunder the programme for the MTEF was estimated at US$4.58 million(external) and US$ 0.18 million (GoG). The reform inputs, which werefocused primarily on MoFEP and MDAs, were provided mainly as technicalassistance. They involved two main areas of support: (i) the development andspecification of the MTEF at MDA level (mainly activity-based), and (ii)improvements to the formulation of the annual budget, including budget cir-cular, installation of budget preparation software in MDAs and staff training.Following PUFMARP, some limited financing was available from CIDA(long-term advisory support), GTZ (MTEF training), and UNICEF (interna-tional site visits and advisory support).Revenue managementUnder the PUFMARP programme, disbursed IDA funding amounted toUS$0.95 mn. Under the subsequent Good Financial Governance programme,supported by BMZ, disbursed funding for each phase was: (i) Phase I (9/2003– 8/2006): €2.8 mn; (ii) Phase II (9/2006 – 3/2010): €5.0 mn; Phase III(4/2010-13) €9 mn. Co-financing of around €2.6 mn has also been providedby Switzerland (SECO). Otherwise, during 2005/06, DFID provided an advi-sor on improving non-tax revenue collection in the natural resources sector.internal auditThe introduction of a formal internal audit function began with the passage ofthe Internal Audit Agency Act in 2003. Previously, internal audit had beenfocussed largely on compliance checking of payment vouchers (pre-audit),rather than on systemic reviews on behalf of management of internal controlsystems.Very limited DP funding was spent during the study period, mainly ontraining. This included EC support (estimated at less than €0.5 mn)21that wasprovided towards the end of the study period (in 2008) to train staff in MDAs’Internal Audit Units (IAUs). Some limited support for training and ICTequipment was also provided by the World Bank.GoG funding during the period in question was focussed on getting theIAA operational, and initiating an internal audit function in MDAs andMMDAs, including staff familiarisation and training. GoG funds spent onintroducing internal audit amounted to an estimated �12 mn per year.22
30
21 Study team estimates based on project documentation.22 Based on estimates by the IAA.
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
2.4 STRuCTuRES TO dESiGn And MAnAGEPFM REFORM inPuTSEQ 2:What type of structures has been used for the design andmanagementof these reform inputs? Have these structures served to providea coordinated and harmonised delivery framework?
country overviewAs described above, the main reform activities during the period studied werethose under the PUFMARP reform programme, the GTZ/BMZ-support forresource management, the EC support to external audit, and wholly-GoG-financed reform activities. As such, the institutional structures involved inPFM reform followed those of the relevant reform programmes.For the PUFMARP, the governance structure involved ultimate oversightby a Government Steering Committee, chaired by the Deputy Minister ofFinance, and representatives of senior management of MoFEP, including theChief Director, CAGD (Deputy Controller), as well as representatives of majorspending MDAs. Day-to-day functions were the responsibility of the PUF-MARP Implementation Team, headed by the Project Manager.As indicated below, the PUFMARP Project Completion Report noted sig-nificant concerns about the relevance of the design of different aspects ofPUFMARP, including mis-specification and over-estimation of the readinessof the implementing agencies for the reforms.Immediately following the PUFMARP, solely GoG-financed reformactivities were co-ordinated by the individual divisions concerned, largely inMoFEP. Following the publication of its ST/MTAP, MoFEP created a specif-ic unit in 2007, the Budget Development Unit (BDU), under the Budget Divi-sion, to facilitate and drive forward budget reforms. From 2001 through 2008,the overall public sector reform programme was co-ordinated by the Ministryof Public Sector Reform.With a limited number of PFM reforms during the study period, andresources channelled through large PFM programmes, donor harmonisationof these programmes was less of an issue. As indicated above, from 2003, asignificant proportion of DPs’ annual resources to GoG (around 30%) havebeen channelled through MDBS. One of the benefits of MDBS was greaterDP harmonisation of its support, including on PFM, with GoG’s priorities andprogrammes.23
23 See Betley and Burton (2011), and Betley (2008).
31
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
component case historiesfMiSThere were marked differences in the design and management arrangementsfor BPEMS and IPPD2 that reflected differences in the scope and complexityof the respective reform initiatives:• BPEMS. There were fundamental weaknesses in the technical designand management arrangements for BPEMS. It was conceptualised pri-marily as technology driven reform with insufficient attention given tochanges in PFM processes and procedures that should have precededthe reform, to change management, and to the assessment of capacityissues and training requirements. Furthermore, BPEMS was initiallymanaged through the PUFMARP project implementation unit, whichhad no functional or operational responsibility for the reform. Thisresulted in implementation of BPEMS being distanced from its two cli-ent departments (CAGD and the Budget Division), contributing to alack of ownership for the reform. It was only following completion ofPUFMARP in 2003 that the responsibility for BPEMS was transferredto the CAGD. However, coordination between the CAGD and theBudget Department remained weak, particularly as by then BudgetDivision had developed its own software application for budget plan-ning and managing budget releases.• IPPD2. By comparison, IPPD2 was a more narrowly specified andsimpler reform, involved only a limited number of users of the sys-tem, and presented fewer change management issues. It was man-aged by the CAGD and implemented through a contract with a con-sultancy firm.MtEfLike BPEMS, the MTEF reforms were financed through PUFMARP,although through parallel funding provided by DFID. The PUFMARPappraisal document provided only a very general specification of the reformwhich emphasised its technical features rather than its institutional and pro-cess aspects. The reform was implemented independently from the other com-ponents of PUFMARP through a separate MTEF Project Unit which subse-quently became the Budget Development Unit that has continued to beresponsible for overseeing the preparation of the Budget and overseeing fur-ther development of the reform.Although these arrangements facilitated strong ownership of the reformwithin the Budget Division, they may have undermined coordination andharmonisation between the MTEF and the other elements of the PFMreforms being supported through PUFMARP. For example, the Budget Divi-sion developed its own ACTIVATE software application for budget prepara-tion rather than using the budget preparation module of the BPEMS system.Delays in the implementation of the BPEMS meant that the performance ele-ment of the chart of accounts that was meant to support the MTEF budgetingreforms was not implemented.
32
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
The design of the MTEF process allows for a review of the experience fromthe most recent MTEF/budget planning cycle to be fed into the planning ofthe subsequent MTEF. A number of external reviews of the MTEF reformhave been undertaken including a 2003 case study undertaken by the Over-seas Development Institute as part of wider study of the design and applica-tion of MTEFs as a tool for poverty reduction, the 2004 PUFMARP projectcompletion report, and the 2009 External Review of PFM.
Revenue managementThe revenue management component of PUFMARP was co-ordinated andmanaged through the programme’s Steering Group and the PIU. The moresignificant BMZ-funded support was provided directly to the relevant revenueagencies, specifically the Revenue Agencies Governing Board (RAGB), theInternal Revenue Service, and the VAT Service. Subsequently, the Secretariatcovering the Taxpayer Identification Numbering (TIN) programme, whichwas previously managed by the PUFMARP PIU, was brought under theRAGB. With only one DP providing support at a time, harmonisation amongDPs was not an issue in practice.internal auditAn IAA steering committee was established to provide oversight for thereform, with the Auditor General as the Chair and the Director General ofthe IAA as the Vice Chair. A formal memorandum of understanding betweenGAS and IAA spells out the operational relationships between the two organ-isations, including the sharing of information, reports, resources, and train-ing. Internal audit standards from International Organisation of SupremeAudit Institutions (INTOSAI) and African Organisation of Supreme AuditCommission (AFROSAI-E) were adopted.However, the reforms have left the IAs confused as to where they belong.They have dual reporting responsibilities at MDA and MMDA levels betweenthe IAA and MDA and between IAA and MMDA. The IAA posts the Inter-nal Auditors to MDAs and MMDAs and, although they report directly to theIAA, they are expected to be part of the MDA/MMDA. This dual allegiancehas created tensions and suspicion among the IAUs, MDAs and MMDAs.
2.5 COMPlEMEnTARy dOnOR inPuTSTO PFM REFORMEQ 3:What types of complementary actions have Donors taken tosupport PFM reforms and what has been their significance? Havethey had any influence on the external constraints to reform?
country overviewThe primary complementary action undertaken by DPs was the Multi-DonorBudget Support (MDBS) Framework. In particular, the introduction ofMDBS in 2003 provided a useful forum for policy dialogue around reform in
33
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
general, but particularly in terms of PFM systems, as well as issues around ser-vice delivery.24It provided a forum for shared PFM assessments, with themeetings for the annual External Reviews of Public Financial Managementand the initial PEFA assessment being open for all stakeholders. The MDBSframework also facilitated information sharing and harmonisation of DPreform activities.However, the more direct way in which the MDBS framework sought toinfluence reform actions was primarily through the MDBS triggers. All of thePFM MDBS triggers were met or declared to be met25, although it may beargued that there was some interpretation in the achievement of some of thetriggers, particularly with BPEMS.26Ironically, the MDBS policy dialogue was least successful in dealing withPFM outcomes. The fiscal crisis of 2008, and the subsequent political fall-out,had an impact on the collective nature of MDBS dialogue, with the WorldBank opting to undertake its own adjustment programme with GoG outsideof the MDBS framework. This, together with strong future economic growthprospects, and with the rebased GDP, and DPs’ own fiscal pressures, can beexpected to have an impact on future levels of MDBS.27In terms of the extent of aid-on-budget for funding for PFM reform meas-ures, as the majority of funding for the PUFMARP reforms was from an IDAcredit, the resources were on-budget. However, since they used World Bankprocedures, they were not on-Treasury or on-accounting. Reporting on theuse of funds was also separate. Grant-funded programme support tended touse the DPs’ own procedures28, and was not systematically on budget.29Underthe MDBS framework, KFW, CIDA and Danida sponsored a number of par-tial audits of MDAs and MMDAs, known as selected flows audits. Overall,less than 50% of all external resource flows use government procedures.30One potential external constraint to PFM reform was the predictability ofexternal disbursements, on which the PEFA reports indicate that there wereweaknesses in the early part of the period studied, but that predictability hadimproved after 2004.31
component case historiesfMiSBPEMS was initially implemented as part of PUFMARP, which supported arelatively comprehensive programme of PFM reforms. In practice, the242526272829See Betley and Burton, 2011.See the Ghana desk report for this study (Gordon and Betley, 2011).See Betley and Burton, 2011.On the other hand, Japan joined the MDBS group.With some exceptions (e.g. Danida)The Budget Statement contains a table of estimated current year-to-date spend on ex-ternally financed projects, but this is not comprehensive, and the information is oftendifferent to that held by the MDAs and the DPs.30 A D score was recorded for the D-3 indicator in both the 2006 and 2009 PEFAs.31 A C score was achieved in the 2006 PEFA for D-2(i), but this had improved to an Ascore in the 2009 PEFA.
34
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
reforms were not well sequenced and were implemented independently ofeach other. This meant that the potential benefits of managing the reformswithin a wider reform agenda were not realised.The subsequent MDBS dialogue had minimal influence on the FMISreforms although two triggers were linked to FMIS implementation. The firstrequired complete deployment of all six BPEMS modules in eight pilot MDAsby 2006. This was not achieved but was “declared met” by the 2007 MDBSReview even though it was only partially met. The second involved the inte-gration of 50% of subverted agencies into the IPPD2 by May 2008 was met inAugust 2008.
MtEfFollowing the completion of the DFID assistance in 2003, the government hascontinued to finance and sustain the MTEF reforms. This has been comple-mented by limited assistance from CIDA (long-term advisory support), BMZ(training and short-term advisory support), and UNICEF (MoFEP interna-tional experience familiarisation and short-term advisory support). Therehave been no MDBS triggers relating to the MTEF, and there is no evidencethat the MDBS dialogue has influenced the evolution of the MTEF initiative.
2.6 Civil SOCiETy PRESSuRE FOR PFMREFORMEQ 4:To what extent have there been domestic public pressure orregional institutional pressure in support of PFM reform and whathas been the influence on the external constraints to reform?
country overviewThere appears to have been limited domestic public pressure for PFM reforms,partly due to the perceived technical nature of such reforms. There are fewCSOs undertaking budget analysis and advocacy and little fiscal and budget-ary analysis undertaken by academia. Non-availability of budget executiondata remains a major constraint to independent budget analysis, and MoFEPhas not actively facilitated such analysis. However, the recent opening of PAChearings on external audit reports to the public and the televising of the hear-ings has increased public awareness of the role of the PAC and led directly topressure on the PAC to increase its technical capacities and improve the qual-ity and timeliness of its scrutiny of GAS reports.32Regional (African) peer-to-peer experience-sharing is cited as an impor-tant factor in enabling GoG officials to learn about international experience.For the Budget Division in MoFEP, CABRI is seen as having been particular-ly influential. In addition, MoFEP has a programme of enabling staff to gaininternational experience (through international degree programmes) andpotentially job-swaps or job placements.32 We note the request by PAC for support to the Secretariat on analysing external auditreports.35
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
component case historiesfMiSThere was no discernible engagement with civil society, academia or themedia over the implementation of BPEMS and IPPD2. This was consistentwith the view of the FMIS as technological reform. Similarly there has beenlittle pressure from regional and international organisation in support of thereforms, although MoFEP increasingly recognises that the experience withBPEMS compares unfavourably with the implementation of FMIS initiativesin other African countries.MtEfUnlike some other countries where similar reforms have been introduced,Ghana’s MTEF has not involved specific measures to involve civil society,academics and the media in the budget process. MoFEP organises consulta-tions with CSOs in October each year, but these take place too late in thebudget process to influence resource allocations. Prior to the consultations,there is a call for submissions from CSOs that is published in the newspapers.Similarly the MTEF reform has not led to any strengthening of parliamentaryinvolvement in the budget process, which remains limited to the period fol-lowing presentation of the draft budget.There are few CSOs undertaking budget analysis and advocacy and lim-ited fiscal and budgetary analysis undertaken by academia. Non-availabilityof budget execution data remains a major constraint to independent budgetanalysis, and MoFEP has done little to encourage such analysis.There is evidence that regional PFM fora have begun to influence theMTEF reform. For example, a meeting of Collaborative African BudgetReform Initiative (CABRI) in the second quarter of 2010 was influential inconvincing officials in the Budget Division to adopt a more strategic pro-gramme-based approach to budget planning.
2.7 RElEvAnCE OF PFM REFORM inPuTSEQ 5:How relevant was the PFM reform programme to the needs and theinstitutional context? Was donor support consistent with nationalpriorities? To what extent were adaptations made in response tothe context and the changing national priorities?
country overviewAs indicated above, there were a few formal domestic PFM reform strategydocuments during the period studied. It is notable that the ST/MTAP wasrevised and re-released in early 2007 following the publication of the firstPEFA assessment; the revised version reflected GoG’s revised policy objectivesin response to weaknesses identified in the PEFA. Specifically, the changesinclude: a reprioritisation of activities based on the PEFA assessment; explicitsections added to explain the prioritisation and the sequencing of reforms;estimates of the likely cost of activities were included, and for each focal area,
36
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
the likely outcome (mainly, expected changes in processes) in 2009 (the finalyear of the Action Plan) was indicated. However, there was no indication ofthe potential impact on service delivery – both original and revised STAPsfocussed more on actions and outputs rather than on outcomes. At the sametime, the issue of how change was to be managed was not addressed, specifi-cally, the organisational development, capacity development and motivationalinitiatives needed to drive each objective.
box 5: comparison of 2006-09 Short & Medium term action Planwith identified weaknessesInclusion inWeakness2006-09 reformIdentifiedProgramme(2001 HIPC,(St=Short-term2004 CFAA,Action; mt=med.-2006/09 PEFA) term Action)DP SupportHIPCHIPC, CFAASTMTDFID, DE,UNDP
Elements of PFmSystema. cross-cuttingLegal FrameworkCapacity buildingOrganisationalstructure/functionalreviewsIFMSCo-ordinationacross PFM systemsBudget policymanagementb. budget credibilityBudget deviations inaggregateBudget deviationsby MDARevenue projec-tions/outturnsExpenditure arrears
HIPC
STMT
WB, DFID
PEFA(-/C)PEFA(D/C)HIPCMTMT
c. comprehensiveness/transparencyBudgetclassificationComprehensivenessof budget docsHIPC, CFAA,PEFA(-/C)MTMT37
WB,IMF
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
Elements of PFmSystemBudget comprehen-siveness & unre-ported opsInter-governmentalfiscal managementAggregate fiscalrisk/SOEsPublic access tofiscal info
Inclusion inWeakness2006-09 reformIdentifiedProgramme(2001 HIPC,(St=Short-term2004 CFAA,Action; mt=med.-2006/09 PEFA) term Action)DP SupportCFAA,PEFA(C/-)PEFA(-/D+)PEFA(C/D+)MTMTMTUNDP
D. Policy-based budgetingStructure of budgetprocessMTEF/multi-yearbudgetingE. budget executionTax administrationreformDomestic resourcemobilisationPEFA(C/C)PEFA(C/-)MTMTMTMTCFAAHIPCCFAA,PEFA(C/D+)HIPC, CFAA,PEFA(D+/D+)HIPC, CFAA,PEFA(C/C)MTMTMTMTEC, WBGIZGIZ, DFIDUSHIPC, CFAAHIPC, CFAAPEFA(C/-)ST, MTDFID, CIDA,GIZ,UN
Expend. Predictabil-HIPC,ity/commitmentPEFA(C/D+)controlsDebt, cash manage-mentPayroll controlsProcurementInternal controlsInternal audit
IMF,USDFIDWB, DFID.GIZ
f. accounting, recording and reporting38
Reconciliationof accounts
MT
IMF
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
Elements of PFmSystemResources rec’d byservice deliveryunitsIn-year budgetreportsAnnual financialstatementsExternal auditLegis. scrutiny ofannual budget law
Inclusion inWeakness2006-09 reformIdentifiedProgramme(2001 HIPC,(St=Short-term2004 CFAA,Action; mt=med.-2006/09 PEFA) term Action)DP SupportPEFA(D)CFAAHIPC, CFAAHIPC, CFAAHIPC;PEFA(-/D+)DFIDMTMTMTMTEC
g. External scrutiny/audit
Legis. scrutiny of ex-HIPC (-/D+)ternal audit reports12
HIPC refers to 2001 HIPC action plan recommendations.CFAA refers to recommendations in the 2004 CFAA Action Plan.3PEFA refers to 2006 and 2009 PEFA scores C or D (central government). 2006 & 2009scores separated by “/”.4DP support may not be comprehensive due to lack of available information.
The alignment between the PFM reforms and Government needs (i.e. PFMweaknesses) may be analysed in two ways: firstly, whether or not the reformsundertaken adequately reflected PFM weaknesses, and secondly whether ornot DPs’ contributions were appropriately targeted. Box 5 shows the align-ment between the weaknesses identified (in the 2001 HIPC survey, 2004CFAA, and 2006 PEFA), their inclusion in the GoG’s 2006-09 PFM reformprogramme and the availability of DP support for reform.In terms of the alignment of DP support with PFM weaknesses, the analy-sis indicates that, in general, DP support was focussed on those areas with thegreatest needs. Based on the link between DP support on the one hand andGoG’s PFM needs on the other, as measured by HIPC, CFAA assessmentsand the scores of Cs and Ds in the 2006 PEFA assessment, there is a reason-able correspondence between DP support and those PFM areas assessed asrequiring strengthening. DP support is concentrated disproportionately – butfor good reasons – on budget execution, followed by the MTEF.DP support has not been provided to any area, which was assessed as beingreasonably strong, although there are relatively few of those. One of the areasnot identified as requiring assistance and without DP support is the establish-ment of a system to identify and monitor expenditure arrears, although DPsare reported to have consistently raised this issue with GoG over the past yearsbut GoG did not request support for such a system during the period studied.
39
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
The lack of such a system enabled GoG to build up a large stock of newarrears during 2007 and 2008, which reached an estimated 4% of GDP.33Box 6 below examines GoG’s reform programme in more detail andassesses how close it is in practice to the PFM weaknesses as indicated by thePEFA and to GoG’s PFM medium-term Action Plan. The analysis indicatesthat nearly all PEFA indicators are included in the Plan. This was partly bydesign, as the Action Plan was revised following the 2006 PEFA assessment.However, the measures are not prioritised, which undermines the Plan’s use-fulness. It is interesting to note, however, that the reform programme is inreality a MoFEP Action Plan since those aspects of PFM undertaken by otheragencies (e.g. internal audit, external audit) largely do not feature (except rev-enue administration measures).
box 6: alignment of 2006-2009 Short & Medium term action Planwith PEfa ResultsSt/mtAP FocalAreas1/Key objectives• To improve fiscalresourcemobilisation• Formulate andimplement soundmacro-economicpoliciesrelated2006 PEFAIndicator/resultsPI-14 (C)PI-16 (C)PI-8 (C)PI-23 (D)PI-2 (C)
Selected WorkPlan Activity• Linking/integrating revenuesystems• Improving monitoring ofexpenditure commitments• Consolidating of fiscal data• Public Expenditure TrackingSurveys• More accurate wage billplanning• Improving MTEF throughcapacity development• Harmonising central/localclassification systems• Facilitating SOE inputs into thebudget• Budget classification• Budget reporting• Cash releases• Donor harmonisation
1. fiscal Policy Management – Macro-Economic Stability
2. Strengthen budget formulation/Preparation• Allocate andmanage financialresourcesefficiently,effectively andrationallyPI-12 (C)PI-8 (C)PI-9 (C)PI-5 (B)
3. Strengthen budget implementation• Improve publicexpendituremanagementPI-24 (C+)PI-16 (C)D-2 (C)
40
33 As estimated in the IMF Letter of Assessment, March 2010.
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
St/mtAP FocalAreas1/Key objectives
Selected WorkPlan Activity• Variance between planned and actual revenues• Inventory control for revenueagencies)• Bank reconciliation• Comprehensiveness of reportedgovt. operations
related2006 PEFAIndicator/resultsPI-3 (A)PI-15 (C)PI-22 (C)PI-7 (A)
4. financial Regulatory and Management Environment• Account for allpublic financesproperly• Development of asset register• Production of timely accounts• Financial instructions• Preparation of statutoryaccounts• Activities for implementingPublic Procurement Act• Activities for implementing IAAAct• Improvement of capacity ofAudit Service, follow-up• Implement IPPDPI-6 (C)PI-24(C+),PI-25(C+)PI-20 (C)PI-7 (A)PI-19 (C+)PI-21 (C)PI-26 (C+)
5. integrated Payroll and Personnel System• Improve thehuman resourceand institutionalmanagementcapacity• Reduce andrestructuredomestic debtPI-18 (C)
6. External Resource Mobilisation, aid and Debt Management• Improve data & reporting on useof ext. assistance• Improve quality of external anddomestic debt data• Debt sustainability analysis andmanagement• Fiscal information on externalloans• Revenue database interface,improved rev admin• Revenue arrears• Analysis of revenues againsttargetsD-1 (C+),D-2 (C)PI-17 (B)PI-12 (C)PI-I7 (B)
7. Revenue Management• Improve fiscalresourcemobilisationPI-14 (C)PI-15 (C)PI-3 (A)
41
2. INPutS AND CoNtExt: tHE DESIGN oF PFm rEForm
component case historiesfMiSFMIS modernisation was and continues to be a central pillar of the govern-ment’s reform effort. However, the design and implementation of the donorassistance failed to synchronise related process and procedural reforms andchange management issues. Due to implementation delays and difficulties,donor support for both BPEMS and IPPD2 ended before new systems wereoperational. Thereafter, the government continued funding the implementationof the reforms, although in the case of IPPD2 this was after a gap of some years.Since the NDC government came to office in January 2009 the FMIS reformshave been more strongly championed by political leadership in MoFEP.MtEfInitially there was strong political level support for the more strategic objec-tive-led approach to budgeting that was to be introduced through the MTEF.In recent years the focus of political support has shifted away from the MTEFreforms towards improving the presentation of government’s fiscal and budgetpolicies and strategies in the Budget Statement. This shift in emphasis mayreflect the perception that the MTEF reforms have not delivered against theiroriginal objectives.The front-loading of donor assistance under PUFMARP facilitated a veryrapid roll-out of the MTEF reforms. However the PUFMARP implementa-tion report noted that this had contributed to the reforms being superficial ina number of key respects. Following the completion of the PUFMARP, thecontinuation of the MTEF reforms relied primarily on domestic financing.However, during this period the focus was on sustaining the reforms that hadbeen introduced, and little was done to address the fundamental weaknessesin the design of the MTEF reforms that had become apparent by the end ofPUFMARP.Revenue managementThe GTZ-funded reforms were designed on the basis of an assessment of theweaknesses in revenue administration. The phased-approach to the pro-gramme, with a pause after each phase for feeding the results of the evaluationof that phase into the design of the new phase, seems to have enabled thereforms to adapt to current circumstances.
42
3. Outputs: the delivery of PFMreform3.1 kEy REFORM OuTPuTS34
EQ 6:What have been the outputs of the PFM reform process and to whatextent has direct donor support contributed to these outputs?34
country overviewOutputs achieved during the study period relating to strategic budgeting andbudget preparation include:• Promulgation of a new Organic Budget Law, the Financial Adminis-tration Act (FAA), 2003.• Since 1999, the Budget Statement has provided 3-year forward esti-mates by MDA for the Consolidated Fund, although the credibility ofthese estimates is very weak.• Establishment of activity-based budgeting process (for service andinvestment Consolidated Fund expenditures) in all MDAs, includingpreparation of detailed MDA activity-based budget documents.• The 2006 Budget Statement was the first to be presented to, and thefirst Appropriations Act to be promulgated by, Parliament before thebeginning of the budget year to which it related. Each Budget State-ment since then has been presented to Parliament and promulgatedbefore the beginning of the relevant budget year.• In addition, the Budget Statement has improved the amount of infor-mation it contains (e.g. on retained internally-generated funds [IGFs],Statutory Funds, and DP funding).35• Greater stakeholder input into the budget process.• Preparation of Citizen’s Guide to the Budget (2007 and 2008), as wellas a serialised publication of the budget in the national dailynewspapers.• Piloting of composite budgets, which aimed to show budgetary infor-mation for districts and regions, by combining central and local expen-ditures in each district and region.Legislation and regulation related to resource management was strengthenedduring the study period. In addition to the Financial Administration Act,2003, a number of revisions of the Financial Administration Regulations(FAR) have been prepared (2004, 2009, and 2011). Strengthening of publicprocurement procedures began with the promulgation of the Public Procure-34 The Inception Report (op cit) defines outputs as “the immediate changes in the archi-tecture and substance of the PFM system generated by the combined set of inputs”.35 It is noted that the issue of comprehensive reporting of “aid on budget” is outstanding.
43
3. outPutS: tHE DELIvEry oF PFm rEForm
ment Act, 2003. More recently, legislation relating to revenue administration(the Ghana Revenue Agency [GRA] Act, 2009) was promulgated.In addition, the following outputs have been achieved in relation to resourcemanagement:• introduction of Tax Identification Numbering System;• introduction of a reworked indirect tax regime (VAT);• establishment of Tax Policy Unit in MoFEP;• establishment of the GRA;• the closing of a significant number of government bank accounts(including redundant or unauthorised accounts) (intended to be inpreparation for a Single Treasury Account (STA);• The elaboration of guidelines for a strengthened commitment controlsystem;• The establishment of an Expenditure Management Committee toassist with cash management;• A harmonised chart of accounts, covering MDAs and MMDAs, wasagreed;• An integrated payroll and personnel database (IPPD2) became opera-tional;• A number of decentralised treasuries were established and becameoperational. These treasuries were intended to facilitate more timelyreleases of funds to cost centres in critical areas, such as education.In terms of internal control, audit and monitoring, the main reform output isrelated to the passage of legislation for internal audit. The Internal AuditAgency Act (IAA) was passed in 2003, which established the Internal AuditAgency, and the process of establishing IAUs in MDAs and MMDAs begun.The operational framework of the internal audit units was enhanced by theadoption of public sector internal audit regulations and standards and the useof an internal audit manual and programmes. The operational efficiency ofthe MDAs and MMDAs has been measured and monitored by the IAA byreviewing the submissions received relating to the: (i) internal audit charters;(ii) annual audit plans and (iii) quarterly audit reports. The quarterly internalaudit reports are subjected to scrutiny for adherence to quality standards andcompliance with statutory regulations and procedures. Specifically, 342 IAUsand ARICs in MDAs and MMDAs have been successfully established,although not all are currently fully functional.Limited progress on accounting and reporting was made during the peri-od studied. Some monthly budget execution reports were made available tothe public through MoFEP’s website, and information on retained IGFs beganto be included in budget execution reports.For external accountability, the promulgation of the External Audit Agen-cy Act, 2000 was a key output. In addition, clearance of backlogs of externalaudit reports by GAS (Consolidated Fund financial statements) occurred intwo batches, 2005 and 2009, both of which occurred one year after elections(see below for more on this issue). Linked to this, timeliness of submission ofexternal audit reports to Parliament has improved in recent years.44
3. outPutS: tHE DELIvEry oF PFm rEForm
In terms of strategic budgeting and budget planning, disappointing progresswas made in the following areas:• monitoring and evaluation of the GPRS;• translating the broader policy objectives in the GPRS into sector strat-egy documents – experience is mixed;• translating sector policy objectives into budget priorities (making aclear link between policy priorities and budgetary allocations);• Making the budget documents, and the trade-offs therein, clear andmeaningful.Whilst efforts were made to improve the budget execution process, disap-pointing progress was made in the following areas:• the effective operation of the planned commitment control system;• the introduction of a Single Treasury Account;• expected improvements in timely expenditure authorisation (includingfor commitments) through the implementation of the planned Finan-cial Management Information System (FMIS) were not forthcoming,cf. BPEMS;• Monthly CAGD budget execution reports have not been made publicsystematically; there can be a significant lag (the most recent reportavailable is for March 2007);• procurement rules and procedures were not followed consistently, con-tributing to the re-accumulation of arrears after 2006;• Being able to ensure that appropriated resources were provided for in atimely fashion and in line with budgetary allocations.• Little progress has been made in incorporating externally-financedproject expenditures in CAGD reports.In terms of external accountability, there was less progress on the following:• Follow-up actions taken by MDAs on audit recommendations; as indi-cated above, an attempt was made to improve follow-up through theestablishment of Audit Recommendation Implementation Committees(ARIC) in budget entities, but these are reported to be ineffective inmany cases.• Whilst audit reports have been presented to Parliament in a timeliermanner recently, challenges remain, particularly in terms of their con-sideration by PAC, and follow-up actions to PAC recommendationsundertaken by the Executive.Box 7 summarises the discussion above and assesses the relative contributionof DP support to the outputs achieved; the analysis only includes those out-puts, which have been sustained, and thus not all PFM reform measures areshown. In classifying the role of DP contributions, the table distinguishesamongst DP support whose attributions are: (i) direct; (ii) indirect; (iii) unclear;and (iv) none, or not applicable. This classification for each area is shown inthe third column of the table.
45
46
box 7: overview of PfM Reform outputs and DP contributionsDP contribution – direct,indirect, unclear, N/AremarksSignificant IDA, and DFID, contribu-tions, under PUFMARP:directCIDA-funded MTEF advisor (1 year):unclearGoGBMZdirect(DFID in earlier period)GoG:indirect(linked to MDBStrigger)Unification of revenue agencieswithin GRA still work in progressNonetheless, significant GoG fundsare still to be found in commercialbanks.Related reform, Single Treasury Ac-count, proving difficult to introduceCIDA advisor’s contract was notextended by mutual agreement
PFm Area
outputs (type of output)
1
Strategic budg-eting/ budgetpreparation
MTEF documents prepared by allMDAs (systems)
3. outPutS: tHE DELIvEry oF PFm rEForm
More financial information in Budg-et Statement (systems)
Resource man-agement – rev-enue
Introduction of VAT (systems)Introduction of TIN (systems)GRA Act (laws)
Resource man-agement –Treasury
Closing of numerous bank accounts(organisation)
More regular reporting on aid anddebt management (systems)
WB with Commonwealth Secretari-atdirect
PFm AreaIMF support:direct(limited)
outputs (type of output)
1
DP contribution – direct,indirect, unclear, N/AremarksCommitment control system notsustained
Resourcemanagement –expendituresUS Treasury long-term advisorysupport:direct (limited)WB, GoG:direct,thenindirectsub-sequently(linked to MDBS trigger)GoG (ST/MTAP), DFID, France:indirect
Introduction of commitment controlsystem (systems)
Introduction of cash forecasting andmanagement process (systems)
Cash management committees(involving MoFEP, CAGD, BoG, andMDAs) not sustainedWas not operational and wasultimately scrappedAudit reports indicate issues withdata accuracy and internal controlproblems (e.g. ghost workers).Not all staff from sub-vented agen-cies are on the systemInitially helped improve fund flows tohealth and education but initial suc-cesses undermined by non-inclu-sion of all funds and failure to dealwith underlying causes of delays inreleasing fundsProblems have been experiencedwith implementation of legislation
Introduction of BPEMS (systems)
Introduction of payroll and person-nel database (systems)
Introduction of decentralised Treas-uries in some MDAs (referred to in-country as Treasury realignment)(systems)
GoG (ST/MTAP), DFID:indirect(linked to MDBS trigger)
3. outPutS: tHE DELIvEry oF PFm rEForm
Procurement reform: New procure- World Bank, DFID and GIZ:directment legislation promulgated (laws)
47
48
PFm Area
outputs (type of output)
1
DP contribution – direct,indirect, unclear, N/Aremarks
Internal control, IAA Act passed (laws)GoGaudit and moni- IAUs being established in MDAs and,toringto lesser extent, in MMDAs (organi-sation)GoGDFID-sponsored:directBMZ provided technical support toPAC:direct (limited)Not consistent or necessarily up-to-dateThe hearings have been well re-ceived and appear to form an impor-tant part of the accountability cycle.
Accounting andreporting
Limited public access to budget ex-ecution reports (systems)
External ac-countability
PAC hearings open to the public(rules)
3. outPutS: tHE DELIvEry oF PFm rEForm
External Agency Act passed (laws)
GoGSustainability of stronger capacitiesnot clear
Capacity-building on internal, exter- EC, Danida, CIDA, KfW, (selectednal audit capacity (HR)flows audit):unclear
1
As described in the Inception Report, outputs have been classified as: (i) changes in human resource endowments (people and skills); (ii) changes in laws, proce-dures and rules; (iii) changes in systems and business processes; or (iv) changes in organisational factors.
3. outPutS: tHE DELIvEry oF PFm rEForm
component case historiesfMiSIn practice, there have been limited outputs to show for the investment under-taken on BPEMS, estimated at around US$ 22.5 million since 1997. By theend of 2010, the system was not fully operational in any of the 8 pilot MDAs,with parallel systems still being run alongside, and it was therefore scrapped.For the IPPD2, the story is marginally more positive, with the system beingused for payroll management36, although the human resource managementmodule has yet to be fully implemented. Furthermore, implementation ofIPPD2 took place several years after donor funding had ceased when the gov-ernment actions to strengthen payroll management made the implementationof an improved payroll management system a priority.MtEfRelative to what was expected to be achieved under PUFMARP; the impactsof the MTEF reforms have been limited. Activity-based budget documents,prepared by all MDAs, are excessively detailed, are not strategically-focussedand do not include all expenditures, notably they exclude personnel-relatedexpenditures, which represent the vast majority of GoG Consolidated Fundexpenditures. There is no strategic framework within which the MDAs’ sub-missions are prepared, nor are there credible medium-term ceilings. MDAs’detailed forward estimates for activities continue to be unrealistic. The inclu-sion of personal emoluments (PE) in the MTEF had been intended as part ofthe design to be carried out at some stage; the initial stages of MTEF imple-mentation concentrated on items 3 (goods and services) and 4 (capital). How-ever, including PE was a much bigger challenge, as it would have entailed sep-arating personnel time amongst individual activities, and MTEF implementa-tion stalled at the initial stage (stopping with items 3 and 4). Weaknesses in theMTEF approach that had been identified by the end of PUFMARP have notyet been effectively addressed.Revenue management:During the earlier (PUFMARP) period, the outputs were focussed on imple-menting specific systems and processes, including the introduction of a TaxIdentification Numbering System, covering all types of tax and a new indirecttax regime (VAT). Later on, the reform turned to more organisational out-puts, including: the establishment of Tax Policy Unit in MoFEP, and automa-tion of operations in six IRS pilot offices. Changes in human resource endow-ment followed, with the training and capacity development of staff, includingin the Tax Policy Unit in MoFEP. Most recently, there has been a legislativeoutput, with the passage of the GRA Act (2009). Currently, the establishmentof the GRA as a unified organisation is under way.
36 Nonetheless, the system continues to experience problems with its implementation,as noted in external audit reports.
49
3. outPutS: tHE DELIvEry oF PFm rEForm
internal auditThe initiating output for a formal internal audit function was the promulga-tion of the Internal Audit Agency Act (2003). Thereafter, GoG concentratedon establishing the organisational structure as set out by the Act, includingthat of the IAA itself, as well as Internal Audit Units (IAUs) in each MDA andMMDA, with a signed charter. By the end of 2010, IAA documentation37sug-gests that 146 Internal Audit Units had been established in MDAs, represent-ing around 85% of all MDAs.38In MMDAs, 155 IAUs had been established,representing around 90% of local government; overall, 88% of IAUs had beenestablished, at least on paper. Signed charters had been completed for 67% ofIAUs in MDAs and MMDAs.In order to address concerns over limited follow-up to audit recommenda-tions (both external and internal audit),39GoG specified that Audit Recom-mendation Implementation Committees (ARICs) should be established inMDAs and MMDAs. As such, by the end of 2010, ARICs had been estab-lished in 135 MDAs (around 78% of MDAs) and 124 MMDAs (73%). In termsof the preparation and submission of reports, during the most recent year ofimplementation, annual audit plans were submitted by around 72% of IAUs.However, IAUs submitted fewer than 30% of their required quarterly reports.Interviews with stakeholders indicate that the mere establishment of anIAU or ARIC does not guarantee their effective functioning. In particular,the study team found that in a number of cases ARICs did not meet regularly,or that IAUs continued to carry out their pre-reform functions. The limitationof resources (primarily, the lack of appropriate skills) appears to hinder theestablishment of fully functioning IAUs and ARICs in all MDAs andMMDAs.
3.2 EFFiCiEnCy And COORdinATiOn OFREFORM OuTPuTSEQ 7:How efficiently were these outputs generated? Was the pacing andsequencing of reforms appropriate and cost-effective? Was thecost per output acceptable?
country overviewThe efficiency of PFM reforms overall may be analysed as a spectrum, look-ing at the relationship between outputs and inputs for the PFM areas with themost outputs and those areas with the greatest amount spent. Of the five PFMfocal areas, taking the information provided above on outputs and inputs, thearea with the highest efficiency level appears to be resource management (rev-enue), whilst that with the lowest would appear to be accounting and report-37 The IAA’s Annual Report for 2010.38 In 21 ministries, and 125 departments and agencies, out of 23 ministriesand 149 departments and agencies.39 IAA estimates a response rate to the follow-up of audit recommendationsof around 34%.
50
3. outPutS: tHE DELIvEry oF PFm rEForm
ing. In the middle lie external accountability and internal control, audit andmonitoring, whilst strategic budgeting and resource management (expendi-tures) would lie at the lower end of the middle range.
component case historiesfMiSImplementation delays and the limited roll-out of the BPEMS meant that theFMIS investment had very little impact on PFM (ultimately leading theBPEMS system to be scrapped), and its efficiency was therefore low. Implemen-tation of IPPD2, although limited in scope, was relatively more successful after2005,40from when it enjoyed stronger political support. A number of factorsexplain the poor performance of the FMIS reforms including: (i) weak owner-ship of the reforms; (ii) seeing BPEMS as primarily a technological reformrather than a PFM reform; (iii) insufficient technical capacities to implementand maintain the systems; (iv) over-optimistic scheduling of implementation;(v) weak procurement management; and (vi) insufficient attention paid to theunderlying business processes in system design and implementation.MtEfIn a narrow sense the outputs of the MTEF reforms, measured against the ini-tial specification, were to a considerable extent met, and after initial donorfunding was finished have subsequently been sustained. However, looked atmore broadly, the outputs measured against the requirement to establish amore realistic and strategic medium-term budgeting process has not beenachieved. These weaknesses were identified relatively early on but were neversatisfactorily addressed. (Most notably the core activity-based budgetingreform that characterises Ghana’s MTEF continued largely unchangeddespite being widely regarded as having been ineffective, while subsequentsteps to introduce a more strategic approach such as the introduction of MDApolicy hearings have not been sustained.) Overall, the efficiency of the reforminvestment should be considered as having been low.Revenue managementIt may be argued that the efficiency of investment is relatively high, given thata relatively modest investment has led to outputs that have been sustained,particularly in terms of VAT (which had experienced a false start in the mid-1990s) and the Taxpayer Identification Number.internal auditWith relatively low investment in internal audit reforms (from either GoG orDPs) and an ambitious reform by nature41, it is perhaps not surprising thatthere has been relatively low impact. It could be argued that it is relatively too40 Though on-going problems with implementation continue.41 The introduction of an internal audit function throughout government is often foundto be a challenge, due to scale and a lack of understanding of the scope and nature ofinternal audit.
51
3. outPutS: tHE DELIvEry oF PFm rEForm
early to expect a fully operational internal audit function to be in place fromscratch; the 2006 PEFA report did not score the internal audit indicator forthat reason.
3.3 ExTERnAl dRivERS And blOCkERSOF ChAnGEEQ 8:What have been the binding external constraints on the delivery ofPFM reform: political, financing or policy factors? How has thisvaried across the different PFM reform components?
financial contextThis study found that neither insufficient resources (either GoG or DP), northe timeliness of disbursements were a major impediment to greater progresson PFM reform. This finding was corroborated by stakeholders associatedwith all of the major reforms studied. Indeed, a number of reforms were intro-duced using GoG’s own resources, including IPPD2, which DPs chose not tocontinue to support following problems with the predecessor system, IPPD1.42GoG also continued to fund BPEMS after the end of DP support; the bindingconstraint on BPEMS was political, as discussed below. For the MTEF, whileonly limited additional DP resources were made available after the end ofPUFMARP, the binding constraint to further reform, or to fix the problemsfound during implementation, was political in nature, rather than financial. Asimilar story may be told for both the internal and external audit reforms.The study period coincided with the introduction of MDBS, which pro-vided a significant amount of additional resources to the Consolidated Fund(representing around 30% of total development assistance or 6% of GDP).While MDBS was not directed explicitly at reform of PFM systems, some trig-gers were applicable to such systems. However, the MDBS triggers were large-ly ineffective at facilitating the success of PFM reforms and could be viewed asa missed opportunity. While GoG met all of the PFM MDBS triggers, manyof them dealt with processes rather than outputs or outcomes (e.g. “continue toensure implementation of the Public Procurement Agency”), or they wereselected to facilitate achievement rather than push progress (e.g. “completedeployment of BPEMS modules in pilot MDAs”, rather than e.g. “BPEMS isthe singular computerised system used to implement, account for, and recordbudgetary expenditures”), or their progress was interpreted flexibly by DPs(i.e. where the spirit of the reform had been taken forward even if the letter ofa particular reform had not been achieved, they were considered to have been
52
42 IPPD1 was a French supported system that was operational in the 1990s. By the later1990s the limitations of IPPD1 had become apparent and DFID agreed to supportthe design of a new system IPPD2 and the project began in 1999. However, the DFIDsupport was halted in January 2001 as the project was seen as going nowhere. It wasthen picked up again by with GoG funding in 2005.
3. outPutS: tHE DELIvEry oF PFm rEForm
achieved).43MDBS as a missed opportunity was highlighted during discus-sions on internal audit reforms. It may be argued that MDBS triggers wererelatively more successful when they focussed on reform initiatives, whichwere already prioritised by government. Triggers appear to have been lesssuccessful when they were used to push for reforms which were not high onGoG’s agenda.Examination of the revenue management reforms identified a number offacilitating factors for DP support, including continuity and flexibility of sup-port, long-term commitment, support which is demand-driven in design, andthe ability to adapt the nature or scope of support to reflect changes in politi-cal or other contexts. Nonetheless, these lessons were applicable to that partic-ular context and would not necessarily translate to other reforms, since not allDPs are able to operate in the way that BMZ does, or to partner a single agen-cy (the Revenue Agencies Governing Board) in that case.The relatively strong and largely consistent GDP growth experienced byGhana during the period studied, led by favourable commodity prices, ena-bled a modest increase in domestic revenues as a share of GDP, and facilitatedreal increases in wages for public sector workers. External finance increasedby around 50% over the period44, and public expenditures rose to a peak of41% of GDP in 2008 compared to 28% in 1999. The absence of macro crisesafter 1999, debt relief from HIPC and MDRI, and the relatively limited effectof the global financial crisis, not to mention the recent resource flows from oil,has stood Ghana in good stead in terms of Treasury resources.
Policy spaceLimited policy space appeared to have been a binding constraint for both theMTEF and BPEMS. At the time, the MTEF was a relatively new as a conceptapplied to Africa, and there was limited information on experiences else-where. The specific design chosen was based on the advice of an external con-sultancy team, and focused on the bottom-up (MDA) aspect of budget plan-ning and preparation. Little attention was given to the top-down elements (themedium-term fiscal framework and strategic resource allocation), which arenow generally seen as the initial stage of an MTEF reform.).Given the technical nature of the reforms, it would have been difficult forGoG to debate the pros and cons of different approaches. It can be arguedthat Ghana was a testing ground for an MTEF and that GoG was open tobeing a pioneer, as it was viewed internationally as a leading reformer in Afri-ca. Internally, the process was widely discussed, at both political and seniormanagement levels throughout government, and this helped to cement thepolitical commitment required (as discussed above).As an integrated financial management information system, the design ofthe BPEMS was relatively ambitious. The choice of IT platform, a leadinginternational financial management software application, to some extent43 These points were made by both Betley and Burton, 2011 and the 2010 IEG evalua-tion of PRSC/MDBS.44 During the 2000s, excluding the outlier year of 2009 (where DPs provided additionalfunds to support GoG during the global financial crisis)
53
3. outPutS: tHE DELIvEry oF PFm rEForm
reflected GoG’s perspective as a leading reformer and stated ambition ofbecoming a middle income country within 15 years. However it brought chal-lenges of needing to customise the system for use in a government rather thana corporate context and raised problems of IT infrastructure demands andconnectivity. It is likely that there was insufficient understanding amongsthigh-level stakeholders of these implications and also of the requirements forre-engineering PFM systems and processes.For the revenue management reforms, the relative openness of the policydialogue facilitated the reforms. With a single DP providing support (first,DFID and then GIZ), there was greater flexibility and more opportunities forcollaboration in the design of the reforms. The VAT re-design took intoaccount the mistakes of the previous failed introduction and drew on interna-tional lessons of experience. The scheduling of the reform provided for naturalbreaks in implementation that allowed for review and redesign, and in somecases halting, of individual elements of reform. The support from the DP wasalso sufficiently flexible to take into account the impact of the electoral cycleon implementation and, due to longer-term DP commitment, to respond tomore politically driven elements of the reform when the opportunity arose(e.g. merging of the revenue agencies).Looking at the role of the policy dialogue as part of MDBS, GoG officialsreported that they found MDBS to be a valuable forum for discussing PFMreform issues, including technical ones, for requesting external TA and projectassistance, and in some cases also for keeping up pressure on reform, andarguably accelerating the pace of some reforms.45A distinction is madebetween the overall indirect effect of the MDBS dialogue and the direct effectof conditionality (and the related performance incentive effect of the variabletranche). Whilst the financial penalty was relatively small for the government,the reputational incentives for the government appear to have been taken veryseriously by MoFEP at least.
Political contextThe degree of GoG political commitment to reform has been the main bind-ing constraint to the success of reforms. GoG showed strong initial commit-ment to the MTEF, reforms in revenue management, external audit, internalaudit and procurement. This commitment facilitated the programme of PFMlegislative reform in each of these areas during the past decade. However, con-tinued high-level commitment to these reforms, including to the implementa-tion of legislation, has been more problematical.For the MTEF, high-level political (Cabinet-level and Parliament) commit-ment, as well as that of MDA Chief Directors, was actively sought throughoutits introduction, and this enabled such an MDA-wide systemic change to beimplemented. Political stakeholders were supportive initially as they saw thereform as being able to facilitate implementation of the policies of the relative-ly new government (in 1997).45 See Betley, Burton, 2011. The positive impact of the MDBS on the pace of reform wasalso cited by the IEG’s 2010 evaluation of PRSC/MDBS.
54
3. outPutS: tHE DELIvEry oF PFm rEForm
It may be argued that the MTEF is an example of a process that has been sus-tained despite the severe limitations of its design.46This suggests the effective-ness of the original change management process, which involved the activeinvolvement of Chief Directors, and regular communication and informationexchange across stakeholders. Arguably the use of the MTEF software to con-trol fund releases to ministries at the level of individual activities has helped toinstitutionalise ownership of the reform within MoFEP.By contrast, BPEMS did not have anywhere near the same level of politi-cal or administrative commitment as the MTEF. It was seen as a technicalreform, something that didn’t require political or high-level support, and thusits original design didn’t actively seek to engage high-level stakeholders. Thefailure of BPEMS to move beyond the pilot stage was also influenced by thesenior administration level’s (Chief Directors) not being convinced about thebenefits of change, and having little incentive to implement the required sys-temic changes in business processes. In the design of the new GIFMIS initia-tive, much greater attention has been given to involving Cabinet and Parlia-ment and to addressing change management issues.The experience with MTEF and BPEMS suggests that GoG found it dif-ficult to halt reforms that had already absorbed a significant amount ofresources, even when it was clear that these reforms had failed or were notmeeting their objectives. Senior ministry management was either not suffi-ciently engaged or was unwilling to take decisions that would have reflectedthe failure of a reform. It took a change of government for the decision to betaken to abandon BPEMS and to undertake the assessment of requirements forupdating financial management systems that led to the new GIFMIS reform.The revenue management reforms also had significant political commit-ment at the start of the reforms (e.g. to implement the VAT),47but the reformprogress stalled in the mid-2000s as a result of faltering commitment and lead-ership for the subsequent reform step of merging the separate revenue agen-cies. It took a change in government, supported by a committed DP, tounblock the reforms.Implementation of internal audit reform was hampered by the failure tosecure the effective support of senior management in many MDAs. Similarly,the political leadership was slow to understand the intent of the IAA law,which affected the pace of reform.All of the PFM reforms have been affected by disruptions in the tenure ofsenior management, particularly in MoFEP (e.g. Chief Director and DirectorLevels), following a change in political leadership. Most recently, following the2008 elections, MoFEP did not have a permanently-appointed Chief Directoruntil July 2011, 30 months after the new government came into power, and 27months after the previous (long-running) Chief Director had left the post. TheBudget Director was replaced, and there were changes in personnel withinother divisions. Previous changes to senior political leadership in MoFEP hadan adverse effect on BPEMS “progress”.46 MDAs continue to this day to produce detailed activity-based budgets.47 It is to be noted that earlier VAT reforms (prior to the period being studied here) weresupported by DFID.
55
4. Outcomes and OverallAssessment4.1 ChAnGES in PEFA SCORESEQ 9:What have been the intermediate outcomes of PFM reforms, interms of changes in the quality of PFM systems?The study assessed PFM intermediate outcomes in terms of changes over theperiod in HIPC AAP indicators and the PEFA indicators. However, weak-nesses in the quality of data for the HIPC assessment meant that this studyrelies for its analysis on the PEFA assessments undertaken in 2006 and 2009,which provide a more comprehensive assessment of PFM systems.48The maintrends in PFM performance across the eight PFM focal areas of the study areset out in Box 8.Overall, there were limited improvements in PFM intermediate outcomesover the period. A total of 7 PEFA indicators registered an improvementbetween 2006 and 2009, while 11 showed no change and 8 recorded somedeterioration. This is broadly consistent with the assessment of PFM reformimplementation contained in the previous two chapters. Specifically:• There was a small improvement in the indicators for: (i) multi-yearbudgeting (PI-12), largely due to a debt sustainability analysis havingbeing undertaken; (ii) revenue management (PI-15), reflecting improve-ments in tax collections; and (iii) resources received by service deliveryunits, largely due to PETS analyses having been undertaken. Addition-ally, during the early part of the period, considerable progress wasmade in strengthening the legal framework for PFM. However, imple-mentation of the new legislative and regulatory provision proceededslowly and as a result did not significantly impact on the indicatorscores.•Regarding budget preparation there was some improvement in thebudget calendar sub-indicator (PI-11) following a change that resultedin the Appropriation Bill now being tabled and approved by Parlia-ment before the beginning of the fiscal year. While this should havebrought greater credibility and transparency to budget preparation,the reform was undermined by weakening fiscal discipline, an increasein the incidence of unauthorised spending, and the failure to maintaina bridge table showing expenditure by function. As a result overallthere was some deterioration in the performance indicators relating tobudget preparation.
56
48 Progress and outcomes in the most recent period (2009 and 2010) are not covered dueto lack of PEFA data for these years.
4. outComES AND ovErALL ASSESSmENt
•
Outcomes relating to budget execution and accounting were unsatis-factory. Issues included the level of expenditure arrears (PI-4), themanagement of budget and cash releases, the control of commitments(PI-16) and the implementation of internal expenditure controls (PI-20). In each of these areas, the PEFA indicators either showed noimprovement or deteriorated between 2006 and 2009. Continuedproblems and delays with accounts reconciliation, in-year budgetreporting and annual financial reporting were reflected in no changein the relevant indicators (PI-22 and PI-24). The lack of progress inthese reform areas was consistent with the failure of the BPEMSreform.In addition, Parliament’s timely review and follow-up of external auditreports deteriorated, in part due to limited capacities in the PublicAccounts Committee (PAC). The support provided to the PAC byDFID over 2007 to 2009 does seem to have paid dividends in the sub-sequent period, however. Notably, progress has been recorded in termsof timely review of audit reports, with PAC reviewing the 2007 and2008 audit reports in early 2011. In addition, training of the media inthe role of the PAC and the broadcasting of PAC hearings on radio andTV served to generate significant public interest and, consequently, toraise the standard of the proceedings.
•
57
58
box 8: Summary of changes in intermediate outcomes based on PEfa assessmentsDeterioration inintermediate outcome1Improvement largely due toDebt Sustainability Analysishaving been undertakenBudget credibility indicatorsdeteriorated from B (agg. ex-pend) and A (rev) to C and BPI-5 [budget classification]deteriorated from a B to CPI-27 [legislative budgetscrutiny] deteriorated from aC+ to D+éin PI-11[budget process]largely due to passageof Appropriation Act be-fore fiscal year.êin budget credibility dueto slippage in fiscal dis-ciplineêin PI-27 [leg. scrutiny ofbudget] due to slippagein unauthorised ex-penditure during yrêin PI-5 [budget classifi-cation] due to not main-taining bridge table forfn classificationPI-14éin PI-15 due to improve-ments in the clearanceof tax arrearsNo change in inter-mediate outcome1remarks
PFm Focal Area
Improvement inintermediate outcome1
Strategic budg-eting
PI-12 [multi-year budgeting]improved from C to C+
4. outComES AND ovErALL ASSESSmENt
Budget prepara-tion
PI-11 [budget process] im-proved from B to A
Resourcemanagement –revenue
PI-15 [tax collections]improved from C to C+
PI-13 [taxpayer liabilities]shown to deteriorate, butinconsistent interpretationof scoring
PFm Focal AreaPI-16 [commitment controls]deteriorated from C to D+PI-17 [cash management] de-teriorated from B to D+
Improvement inintermediate outcome1
Deterioration inintermediate outcome1
No change in inter-mediate outcome1remarksêin PI-16 [commitmentcontrols] due to slip-page in predictability ofresource flows to MDAsêin PI-17 [cash manage-ment] due to slippage incash managementPI-18 [payroll con-trols]Due to lack of reliable data,PI-19 [procurement] was notscored originallyêin PI-20 [non-wage in-ternal controls] due toslippage in commitmentcontrolsIn the first PEFA, PI-21 [in-ternal audit] was not scoredas it was judged too early toassess as the internal auditfunction was in the earlystages of being-established
Resource man-agement –Treasury, aid anddebt manage-ment
Resource man-agement – ex-pendituresPI-20 [non-wage internalcontrols] deteriorated from Cto D+
4. outComES AND ovErALL ASSESSmENt
Internal control,audit and moni-toring
59
60
PFm Focal AreaPI-22 [accountsreconciliation]PI-24 [in-yearbudget reporting]PI-25 [annual finan-cial reporting]PI-28 [leg. scrutiny of ext.audit reports] deterioratedfrom C to D+PI-26 [external au-dit]Continued weaknesses intimeliness and accuracy ofin-year and end-year budgetexecution reports
Improvement inintermediate outcome1
Deterioration inintermediate outcome1
No change in inter-mediate outcome1remarks
Accountingand reporting
Externalaccountability
ê in PI-28 [leg. scrutiny ofext. audit reports] due to theimpact of a political disputewith Auditor-General
4. outComES AND ovErALL ASSESSmENt
1
Based on PEFA results in 2006 compared to 2009.
4. outComES AND ovErALL ASSESSmENt
4.2 RElEvAnCE OF PFM iMPROvEMEnTSFOR SERviCE dElivERy, ESPECiAllyFOR wOMEnEQ 10:To what extent are the outcomes generated relevant toimprovements in the quality of service delivery, particularly forwomen and vulnerable groups?It is difficult to obtain reliable data on service delivery for women and vulner-able groups. The annual budget allocations in 2011 to the Ministry ofEmployment and Social Welfare and the Ministry of Women and Childrenrepresented less than 0.5% of the total budget.49While the share of discretion-ary (MDA) resources flowing to MDG-related sectors (particularly health andeducation) increased relative to other sectors over the past decade, most of thisincrease went to pay for significant increases in wage rates; flows relating toother line items either stagnated or decreased.Given the weaknesses in GoG budget documentation and subsequentreporting, it is very difficult to determine the extent to which expenditureswere spent on improved service delivery for women and vulnerable groups, orthe degree to which such expenditures in these areas are prioritised and pro-tected during times of fiscal consolidation. Expenditures from the Consolidat-ed Fund on a defined set of “poverty-reducing” activities are able to betracked. However, in practice, these activities are not necessarily prioritised(Table 5).50
table 5: Poverty-Reducing Expenditures(GHC mn unlessotherwise specified)total expenditures220042,196380.5162.924.7%20052,516493.4195.427.4%20063,489618.6199.820073,964711.8183.720085,38520098,081
Of which:Pro-poor expendi-tures – MDAs onlyHIPCPro-poor(MDAs+HIPC)as % of total exp1
1,108.7 1,274.5187.524.1%549.922.6%
23.5% 22.6%
Data were not collected prior to 2004, and final figures for 2010 were not yet availableat the time of the review.2Total Consolidated Fund expenditures, including both discretionary andnon-discretionary expendituresSource: Audited Accounts, Ghana Audit Service
49 Figures are all for the Consolidated Fund (GoG discretionary allocations).50 See World Bank, External Reviews of Public Expenditure Management, e.g. 2009.
61
4. outComES AND ovErALL ASSESSmENt
In terms of planned expenditures, while the very detailed MTEF documenta-tion provides information down to the activity level, it is far too detailed andthe activities insufficiently strategic, with many of the activities essentiallyinputs, such as the purchase of vehicles or the carrying out of seminars, to pro-vide useful information on their linkage to the delivery of services. At thesame time, the lack of comprehensiveness of the MTEF51means that it is notpossible to have information even on the total planned allocations in particu-lar sectors.Furthermore, due to differences in the Charts of Accounts used betweenMoFEP (responsible for budgeting) and CAGD (responsible for accounting,recording and reporting on actual expenditures), information on actualexpenditures against budgeted appropriations by activity cannot be generat-ed.52The Annual Financial Statements produced by CAGD often have signifi-cant errors, which usually remain even after audit. Limited information isavailable on resources received by service delivery units. While PETS analy-ses were carried out in the health and education sectors in 2007, they only cov-ered part of resources received, and there were serious concerns about thequality of the output. The lack of reliable information on actual expendituresaffects the work of civil society organisations trying to hold the government toaccount for its policy objectives.Expenditure commitment controls worsened during the period studied asa result of the rapid expansion of budgetary expenditures, particularly ininvestment spending. This led to a ballooning of the fiscal deficit and a signifi-cant accumulation of payment arrears. As discussed above, the most recentperiod of expansionary fiscal policy occurred during both the build-up to thelast elections and the global financial crisis. Existing weaknesses in PFM sys-tems and processes to direct resources for service provision were put undergreater pressure by reduced fiscal discipline in the latter part of the periodstudied.Thus, there is little evidence that the PFM reforms contributed to theimproved delivery and to better targeting of services on women and vulnera-ble groups. Fundamentally this reflected: (i) the failure of the MTEF reform toresult in a more strategic and better targeted approach to budget planning;and (ii) the lack of progress in strengthening budget execution and monitoringsystems which was linked to the failure of the BPEMS reform.
62
51 Since the activity basis excludes personnel-related expenditures and those financedby sources other than the Consolidated Fund, such as the Statutory Funds (e.g. theNational Health Insurance Fund).52 CAGD had announced that from end-2011 the new Chart of Accounts and the firstphase of GIFMIS would ensure that compatible budget and expenditure figures wereavailable for 2011.
4. outComES AND ovErALL ASSESSmENt
4.3 ThE EFFECTivEnESS OF PFM REFORMS:FROM inPuTS TO inTERMEdiATEOuTCOMESEQ 11:Have reform efforts been effective? If not, why not? If yes, to whatextent have PFM reform outputs been a causal factor in thechanges identified in intermediate outcomes?Overall, there has been mixed success with PFM reform efforts. The mostsubstantial success has been with passing legislation but GoG has faced signif-icant challenges to implement the new laws. Otherwise, the most effectivereforms appear to have been the revenue management activities, as they haveled to a sustained output in the form of changed processes (successful introduc-tion of VAT, and the introduction of the TIN), and there has been a signifi-cant increase in revenues as a share of GDP during the period studied.Large and complex projects have been less successful than more narrowly-targeted initiatives. The two major reforms introduced under PUFMARPinvolving the introduction of a more policy-led medium-term budget processand the introduction of a financial management information system to sup-port budget executions, have not contributed to improved intermediate out-comes, nor prevented the deterioration in such outcomes. Indeed, theimprovements in intermediate outcomes (as measured by PEFA) were largelyindependent of the main PFM reform actions.The study attempted to explore the reasons why some reforms were suc-cessful while others were not. It has identified five main factors.The electoral cycle had a negative influence on consistency ofimplementation of the PFM reform agenda.During each of the elec-toral cycles in the period studied, there was a marked deterioration in fiscaldiscipline in the 18 months leading up to the elections. This resulted in thebuild-up of arrears and a weakening of commitment controls and other inter-nal controls (e.g. payroll). In the aftermath of each election, the new govern-ment undertook clearance of arrears and tightening up of internal controls,before the following electoral cycle resulted in a repeat of the cycle. This hasresulted in the deterioration in a number of PFM intermediate outcomes asmeasured by the PEFA assessment criteria. It has also diverted administrativeeffort and political attention away from the implementation of reforms. Politi-cisations of appointments to senior ministry and administrative positions hascreated uncertainty and disrupted the implementation of some reforms.Technical design flaws – due perhaps to policy space con-straints – afflicted several of the PFM reforms:• BPEMS was over-ambitious in its conceptualisation, aiming to imple-ment an integrated system (covering strategic budget preparation, pro-curement, execution, accounting, reporting, recording, and HR) to berolled out to all MDAs and MMDAs (312 entities at that time).5353 Relatively early on, the budget preparation and HR components were dropped fromthe design.63
4. outComES AND ovErALL ASSESSmENt
The design of the reform failed to take account of the limited IT famil-iarity and significant constraints to IT connectivity that until now havenot been fully resolved. It also emphasised a technical, IT-led approachthat failed to address adequately the requirements for underlyingreforms to PFM processes and procedures. The management of thereforms, under a separate and relatively independent PIU, reinforcedthese weaknesses by marginalising key stakeholders, principally theCAGD and Budget Division, while insufficient attention was given tocapacity issues and training requirements.•IPPD2, the other main IT reform, was relatively more successful partlybecause it was much simpler in design, with its implementation focusedon establishing a payroll processing and personnel management data-base for MDAs. As payroll processing is a central function, the new ITsystem involved only limited numbers of users of the system, whichwere concentrated at CAGD. It did not involve as many changes tounderlying processes and hence posed fewer change managementissues. Nevertheless, it also took close to a decade to implement at rela-tively high cost and in its current, operational form only really com-prises half of the system originally envisaged, providing a payroll pro-cessing function only, with no corresponding personnel data-base. Aswith BPEMS, weak structures for monitoring progress and coordinat-ing implementation contributed significantly to these failings.Weaknesses in the initial design of the MTEF directly contributed toits failure to establish a more strategic policy-led approach to budget-ing. The focus of the reform was on activity based costing and budget-ing and little attention was give to the top-down strategic elements ofthe reform (the macro-fiscal framework, expenditure policy and sectorpriorities) and to the role of Cabinet in the budget decision-makingprocess. Furthermore the new procedures were not applied to person-nel expenditure despite the fact that these represented the largest shareof most MDA budgets. The result was an excessively detailed budgetplanning process that was not comprehensive and which lacked real-ism, with MDAs continuing to submit budget requests substantially inexcess of the available resource framework.
•
64
Insufficient articulation of the PFM reform vision and strategycontributed to ephemeral political ownership of PFM reforms.There was no overriding GoG prepared strategy to guide the PFM reformprocess. PFM diagnostic studies were largely carried out by the DPs, often inresponse to their own programming agendas. PFM reforms were articulatedthrough a mixture of DP-prepared reform initiatives (e.g. PUFMARP), shortand medium-term action plans prepared by MoFEP, and PFM actions listedin the annual Budget Statements. The lack of a clear GoG-owned and priori-tised PFM reform strategy, meant reforms often lacked impetus to addressproblems as they arose. As is not unusual, the focus of political leaders shiftedwith regularity in response to political needs and did not necessarily follow a
4. outComES AND ovErALL ASSESSmENt
logical pattern. This meant that reforms that may have begun with initialhigh-level political and administrative support were left to struggle when suchsupport waned; the continuation of BPEMS after DP support ended is a goodexample. It also led to reform initiatives becoming compartmentalised withina single department without linkage to other elements of the PFM system. Forexample, although activity-based budgeting reforms were introduced in thelate 1990s there is still no capability within the PFM system to account forexpenditure by budget activity.Deficiencies in bureaucratic and management structuresadversely affected the capacity of MoFEP to achieve the success-ful realisation of PFM reforms.A range of factors can be identified.These include weaknesses in the high-level management and co-ordination ofthe reforms, departmental rivalries, and insufficient attention to change man-agement requirements, failure to address human resources management andcapacity issues, and ineffective monitoring of progress. The result was a ten-dency for reforms to continue at the own pace and a failure to take timelyaction to redesign or halt reforms that were clearly not working. For example,although problems with the MTEF design were identified early on, theMTEF process continued largely unchanged despite it being widely seen ashaving failed. Similarly, although issues with BPEMS were identified duringimplementation and various solutions were attempted, leading to projectexpenditures mushrooming, it took a change in government to halt BPEMScompletely. Weaknesses in management decision-taking processes were alsoreflected in the tendency for reforms to get stuck at the pilot stage.Sustained political commitment has been a key factor to the success-ful implementation of PFM reforms. Both the revenue reforms and IPPD2received strong political level support, including at Cabinet level, whichfacilitated their implementation. By contrast where such support has notbeen present (e.g. external audit and arguably BPEMS), reform progress hasbeen slower.The study has also examined the role of DPs and the extent to which thisinfluenced the success of PFM reforms. DPs played a key role in supportingthe reform process and financing the implementation of the reforms.54Thelevel of DP funding appeared to be sufficient, and DPs were responsive toGoG requests, particularly at times of economic and fiscal difficulty.55DP dis-bursements were broadly in line with commitments in aggregate, although thetiming did not always match GoG’s desired timing, which delayed some activ-ities.56Nonetheless, disbursement delays were not a critical hindrance toreform.The way in which DP support for PFM reform was provided also had animpact on its success. Specifically:• Donor support that spanned a relatively long time horizon and whichprovided a flexible support framework was better able to respond to54 See analysis summarised in Section 3, Box 7 above.55 See Betley, Burton, 2011.56 This has been the case with GIFMIS, for example, and previously with the develop-ment and testing of the MTEF.
65
4. outComES AND ovErALL ASSESSmENt
••
reform opportunities as they arose, which was often linked to the politi-cal cycle. The revenue reforms supported by BMZ are a good exampleof where a flexible approach to support facilitated the reform.Effective co-ordination among DPs also played an important role inreducing transaction costs and avoiding overlapping initiatives.DP support that was tightly focused and limited in scope was oftenmore effective than large ambitious projects, such as PUFMARP,which covered several component reforms that proved unwieldy tomanage.
EQ 12:To what extent do the gains identified at the Intermediate Outcomelevels appear sustainable? Is the process of PFM reformsustainable?The most sustainable improvements in intermediate outcomes appear to bethose in revenue management, including the introduction of the TIN, and of afunctioning VAT. The latter is notable as it followed an earlier failed attemptat introducing VAT in the mid-1990s. In the more recent introduction, greatersuccess has been attributed to better public education in advance of the re-introduction and to the establishment of cross-party support for the re-intro-duction of VAT. In the first attempt at introducing VAT, a number of tacticalmistakes were made by the NDC Government, which were quickly seizedupon by the opposition. The fact that VAT was re-introduced at a lower rate(10 %), with rates being increased later once systems were established andpublic acceptance well embedded, was also pertinent.The government which took power in 2009 signalled a renewed commit-ment to PFM reform through the high priority that it gave to establishing afunctioning financial management information system, its support to therestructuring of the MTEF with adoption of a more strategic programme-based budgeting approach, the amalgamation of three revenue agencies, thedevelopment with the assistance of the IMF of a programme for deficit reduc-tion and the elimination of arrears, and the reestablishment of payment con-trols. It also initiated the restructuring of MoFEP, creating a new unit respon-sible for PFM reform, answering to senior management, and facilitated anemphasis on building PFM capacity (including through the efforts of theBudget Development Unit) through PFM-related study tours and/or interna-tional or regional PFM fora (e.g. CABRI) to learn about different PFM sys-tems. Direct exposure to reforms elsewhere has generated demand for specificreforms, including recently the introduction of programme-based budgeting.While it is too early to determine the extent to which these reforms arelikely to result in sustainable improvements in PFM outcomes, the recognitionthat PFM reforms have had limited success, the comprehensive approach thatis being taken to tackling key weaknesses in PFM systems and the strongerdemand from parliament to show improvements in PFM outcomes suggeststhat there are grounds for optimism.66
5. Conclusions and wider lessons5.1 SuMMARy OF kEy COnCluSiOnSThe study posed the following questions at the beginning: (i) where and whydo PFM reforms deliver results (i.e. improvements in the quality of budget sys-tems?); and (ii) where and how does donor support to PFM reform efforts con-tribute most effectively to results? The analysis in Box 9 summarises the rela-tionship between inputs, outputs and intermediate outcomes and the relatedbenefitting or inhibiting factors.The study’s main conclusion is that, relative to the significant fundsexpended on PFM reform over the study period, success has been largely dis-appointing. The most substantial progress was made with strengthening thelegislative base but, in common with other countries, GoG has experiencedsignificant challenges to implement the new laws. Otherwise, the most effec-tive reforms appear to have been the revenue management activities, as theyhave led to a sustained output in the form of changed processes (successfulintroduction of VAT, and the introduction of the TIN), and there has been asignificant increase in revenues as a share of GDP during the period studied.As VAT is a major source of revenue, it is likely that the successful implemen-tation of the reform has been an important factor.The most ambitious reform initiatives have generally been the leastsuccessful,57with the exception of the revenue management reforms, whichcould be considered to be a succession of smaller initiatives. By contrast theMTEF reform aimed at improving the linkage between policy objectives andbudgetary expenditures and the budget execution reforms that focussed onimproving financial management information systems have not contributed toimproved intermediate outcomes, nor prevented the deterioration in such out-comes. Indeed, the limited improvements in intermediate outcomes wereindependent of PFM reform actions.The main binding constraint was the extent and continuity of politicalcommitment and leadership (including the impact of the political cycle), fol-lowing by issues of policy space. Factors of inter-agency and intra-agency co-operation (relative incentives), weak bureaucratic management, and an incon-sistent strategic vision were also features affecting reform progress. The avail-ability of financial resources, either DP or GoG, does not appear to have beena constraining external factor. Physical outputs (e.g. legislation, regulations,standards, documents) appear to be more numerous and more sustainablethan systems changes or the implementation of legislation. Reforms and
57 This is not to say that it is necessarily the size itself which is the sole factor responsiblefor the relative lack of success.
67
5. CoNCLuSIoNS AND WIDEr LESSoNS
changes in intermediate outcomes appear to be more effective with theupstream elements of PFM than on the downstream ones.Reforms were hindered where design was inappropriate, scope was over-ambitious and where there was too much of an emphasis on technologicalchange (relative to addressing existing weaknesses in underlying business pro-cesses), and underplaying or under-estimating the required organisationalchanges.Initial commitment to reform (demand from GoG due to fiscal pressuresand DPs, and strong political level leadership) at the beginning of period(PUFMARP) facilitated early reforms. This commitment was difficult to sus-tain in the longer term, and hence reform progress flagged, affecting thosereforms being implemented but not yet completed. One response to challengesin implementing reforms, piloting, sometimes became an end in itself (ratherthan a means to an end), to be replaced by something new, rather than learn-ing and adapting, based on the initial implementation lessons.The role of the wider economy, in particular strong economic growth, andthe level of commitment to fiscal discipline appeared to have a more signifi-cant effect on intermediate outcomes (positively or negatively) than did specif-ic PFM reform activities.DP contributions have had both direct and indirect effects, but their effec-tiveness has been undermined by large set-piece and time-bound projects thatfocussed on technological solutions rather than changes to the underlying pro-cesses.58External support has had greater traction at the beginning ofreforms, to facilitate the design and the initial implementation, over those forsustaining or deepening reforms. Exceptions to this have occurred wherethere has been sustained DP contributions over a longer period of time, withsupport demand-driven, targeted to specific needs, adaptable in terms ofdesign and re-design (willingness to pause), and commitment signalled inadvance. The main benefit from MDBS was in the GoG-DP (overall) policydialogue, rather than from the external triggers.There was pressure from both DPs and GoG for new and ambitiousreforms to be implemented as a response to waning or failing reforms. DP sup-port emerging from this pressure tended to be inefficient and to lead to poorly-sustained results.
5.2 widER lESSOnSThe study identified a number of factors which appeared to facilitate PFMreforms (though not necessarily ensure their sustainability). These positivedrivers of change include:•Demand for reform– demand for PFM reforms in Ghana has come fromthree main sources: (i) internally (GoG)-generated demand, as theresult of fiscal pressures, internal/external assessments of PFM systems(e.g. PEFA assessments) or peer-to-peer learning; (ii) public (citizenry)pressure; and (iii) dialogue with DPs, particularly through the MDBS.68
58 This is not to say whose initiative (i.e. GoG, DP, or both) these set-pieces projects were.
5. CoNCLuSIoNS AND WIDEr LESSoNS
•
•
•
•
•
Role of media– with an active and open media and a citizenry who areactively engaged with the media (particularly, radio and print media,but also television), there is evidence of pressure being brought to bearon public officials.59One example is the greater attention placed on thePAC’s reviews of external audit reports through the hearings being tel-evised. However, the ability of the public to hold officials to account isundermined by the lack of timely financial information available to thepublic (e.g. in-year fiscal reports), the lack of comprehensibility of keybudget documents (e.g. the Medium Term Expenditure Framework[MTEF]), and limited expertise on budgetary and financial matters bysome officials, CSOs and the general public.Peer-to-peer learning– learning from the experience of relevant peers, e.g.through the Collaborative African Budget Reform Initiative (CABRI),has been cited as an important factor for generating internal demandby technical staff for specific types of reforms. Seeing how reformswork in practice (e.g. programme-based budgeting in South Africa) hasprovided an impetus and guidance for change. However, a sense ofcompetition amongst different countries can lead to demand forreforms that are inappropriate to the local context and/or overambi-tious (e.g. the latest PFM trend) in order to be “first”. MoFEP middle-level management (particularly in the Budget Division) have also pri-oritised the training of their staff through relevant mid-length overseasPFM programmes, which is serving to provide staff with exposure toalternative PFM processes and to generate a cadre of core staff withsuch experience. However, in the absence of progress on reforms, thereis a risk that such training becomes quickly degraded.DP support which is flexible, demand-driven, involves a longer-term commitment,and which explicitly involves the opportunity to design or (re)design the content asthe project goes along (based on the current institutional and policy environment andan evaluation of the previous phase of support)– this is the basis for the BMZwork with the Revenue Agencies.MDBS dialogue– a recent study of MDBS60found that the policy dia-logue as part of the joint GoG-DP MDBS framework facilitated GoG’spolicy- and decision-making process through keeping up pressure onthe issues and arguably enabling GoG to accelerate some reform pro-cesses. However, in terms of leading to measurable outputs or out-comes, these pressures appeared to have greater effect on areas otherthan PFM (e.g. related to MDGs).Involvement of the political level early on,including Cabinet, and Parliament,appears to have been an important facilitating factor in the reformprocess.
59 This pressure may be necessary but not sufficient for there to be action (e.g. an investi-gation or further consequences) taken.60 Betley M. and Burton J., (March 2011), Ghana, Assessing the Benefits of Multi-DonorBudget Support, Mokoro Ltd.
69
5. CoNCLuSIoNS AND WIDEr LESSoNS
70
The study also identified a number of obstacles to change, including:•Tendency for overly ambitious reforms– this has led to the inappropriatedesign of some reforms, increased the risk of reforms, and led to unre-alistic expectations of the timetable for undertaking the reforms andthe outcomes of the reforms. The reforms carried out in Ghana illus-trate the tension between a desire for ambitious (e.g. multi-component)reforms and the riskiness of doing so.•A focus on technolog y without addressing underlying processes and incentives– anumber of the reforms in Ghana have been seen largely in technologi-cal terms, without first addressing the underlying PFM reform issuesand incentives. This has led to a tendency to move (too) quickly to atechnological solution without first tackling the requirement for chang-es to business processes. This tendency can have three main effects,namely: (i) ignoring the wider political economy of the reforms andthereby failing to address the needs of all stakeholders and thus bringthem on board; (ii) raising (often unrealistic) expectations for what thereforms can achieve; and (iii) automating the underlying weaknessesrather than addressing them.•Desire for piloting but no appetite for moving to the next stage based on lessonslearnt– during the period studied, there were a number reforms thatwere piloted for many years (up to 8 years in some cases) without beingrolled out; examples include composite budgeting, certain modules inthe Budget Planning and Expenditure Management System (BPEMS)in pilot MDAs, and the realigned Treasuries, which were all piloted atthe same time, often with the same set of stakeholders. This served tocause significant confusion for those stakeholders undertaking thepilots. The failure to move the reforms forward suggests that they hadnot been sufficiently planned and tested and that there was insufficientcapacity (or will) to address issues identified during the pilot phase.•Pressure to introduce new reforms before current ones are embedded –not onlywere the pilots not rolled out (or stopped) but they were followed bynew (different) reforms; an example was the composite budgetingreforms, which began being piloted in 2003 and which were supposedto be combined with the introduction of MTEFs at the MMDA level,but which (as of 2011) are now being superseded by the introduction ofprogramme-budgets and the Ghana Integrated Financial Manage-ment Information System (GIFMIS). This pressure has led to furtherconfusion and may potentially have affected full engagement with thenew reforms.•Tendency for departmental silo workingmakes PFM reforms more difficult,particularly with reforms which involve specific MDAs or parts ofMDAs, possibly to the exclusion of others. These may exacerbate exist-ing inter-MDA or intra-MDA institutional tensions.•Insufficient time to plan reforms and explore alternatives.Reform planning,including planning how reforms will affect different stakeholders andhow these changes will be addressed, is likely to take considerable time.Not exploring alternative methodologies may lead to inappropriately-specified reforms. The MTEF is a relevant example in Ghana.
5. CoNCLuSIoNS AND WIDEr LESSoNS
•
Political economy– reforms which potentially work against formal orinformal institutional interests have been harder to achieve. The com-pletion of the transfer (referred to as “migration” in the Ghana context)of the subvented agencies on to the central payroll system is one suchexample.
5.3 RECOMMEndATiOnS FOR ThE dESiGnAnd MAnAGEMEnT OF PFM REFORMPROCESSESThe observations outlined above suggest a number of implications for bothGoG and DPs. Any PFM reform involves risk; it is critical for reform plannersand managers to manage the risk, and the points below indicate areas ofpotential risk management.1.Reform of PFM is a process– there is no “beginning” or “end” to reform.Both GoG and DPs should recognise the importance of continuity ofreform, which may not follow a linear path. Longer-term support to thebroad direction of reform (e.g. as with revenue administration reforms)can provide helpful continuity of support and allow reform to takeadvantage of windows of opportunity, e.g. political space.2.Understand the risks applicable to each type of reform and explicitly factor in riskmanagement.One of the aspects to be addressed during the planningstage should include an analysis of readiness for reform (e.g. large ITprojects should review the existing extent of IT culture in the organisa-tions involved; the more limited the existing IT culture, the more riskyis the reform).61In planning reforms where functions are to bedevolved, MoFEP should first test the preparedness and competence ofthe agencies involved before functions are devolved.623.Less ambitious and stepped reform approaches set within a broader reform frame-work are likely to be more sustainable and involve less risk.Start out with lim-ited changes that address a particular issue. Then incorporate theexperience from implementation into the design of the next step in thereform process.4.Identify demand and management factors that are likely to facilitate successfulreform:• a strong demand for change and a common understanding of what is needed, including the sharing of international experiences withpeers in other countries to build and strengthen a common knowl-edge base;
61 It is also worth noting, as above, that a technological (IT) solution may not always beappropriate, particularly if the underlying manual processes (e.g. internal controls) areinadequate.62 This implication may be relevant to the current on-going process of fiscal decentrali-zation and devolution of fiscal resources to MMDAs.
71
5. CoNCLuSIoNS AND WIDEr LESSoNS
5.
6.
7.
8.
72
• clearly-defined (and understood) roles and responsibilities played by external stakeholders, particularly where the reforms involve agen-cies outside of MoFEP;• strong accountability mechanisms in place, including externally with external scrutiny bodies and CSOs and internally with projectmanagement;• an appreciation by stakeholders that the reforms are not only about technical changes; institutional arrangements need to be clear andfacilitate discipline and decision-making and scrutiny of those deci-sions;• strong, clear and evidenced leadership of the reform;• Effective and respected arrangements for management are in place.Reforms need to be planned and sequenced.Appropriately planned andsequenced reforms should recognise that planning, design and testingtakes (sometimes significant) time; this step should not be rushed. Sys-tems (and technology) should support reform and not vice versa. Insti-tutional and organisational issues (both within and across organisa-tions) must be addressed as part of reform planning. There should be ashort (time-bound) pilot phase included after testing but only when thereform is ready to roll out. Finally, it is important to be realistic, bear-ing in mind the PFM calendar and capacity demands. Too manyreform initiatives may distract staff from their day-to-day tasks; man-agers should set priorities and manage the sequencing of reforms sothat the budget process continues in a timely fashion.Ensure that there is sufficiently strong, active and appropriately dispersed leader-ship.The role of leadership of reform is most effective when it is visible,consistent, at the appropriate level, and actively engaged with theappropriate level in stakeholder institutions. Effective PFM leadersmanage upwards (to the political level), downwards (to the manage-ment level), and outwards (both across government, and to thoseresponsible for external accountability, including Parliament and civilsociety). Consideration may be given to providing training in leader-ship for senior political and administrative levels, particularly for thosein newly-appointed positions.Ensure appropriate management of reformat different levels – effective man-agers (i) ensure that there is a clear plan to get from where an institu-tion is currently to the next step(s), (ii) take responsibility for deliveringon the work plan, and (iii) delegate to staff and motivate them to deliv-er, giving appropriate recognition for activities done well and in a time-ly fashion, as well as enforcing sanctions when justified. Current effortsto introduce more performance elements into senior management, par-ticularly for Chief Directors, are important.Review and evaluate reform progress and be prepared to take action if things gowrong.In other words, GoG and partners should be ready to admitwhen reforms are failing and address weaknesses promptly, or activelytake the decision to halt the reforms. Incentives to improve senior man-agers’ performance may facilitate in this process.
box 9: Summary of PfM reform inputs, outputs, and intermediate outcomesoutputs2Partial (e.g. excludessalary-related costs)activity-based MDAbudgetsPilot phase has lastednearly 10 yearsDifferences in classifi-cation systems betweencentral, local not re-solved prior to pilotingIndirect (policy dia-logue)Weaknesses in, & am-bitiousness of, MTEFdesign. Pressure to rollout early.Direct (design, imple-mentation)Causal factors in posi-tive/negative progressrole of DPContributions3,4IntermediateoutcomesDeterioration in cred-ibility of multi-yearframework
PFm focalarea
Inputs1
Specific reform
Strategicbudgeting
MTEF
Support to MTEF: TA,design of MTEF, soft-ware. Funding of $4.6mn (WB) (and GoG of$0.2 mn). Co-financ-ing from DFID andCIDA.More comprehensivebudget info, moretimely budget calen-dar: solely GoG inputsFinance and Adminis-tration ActFinancial Administra-tion RegulationsMore comprehensivebudget informationApprop. Act passed be-fore beginning of fiscalyearCitizens’ GuidesPre-budget consulta-tions w/ publicBudget information onwebsiteCommitment by GoGCommitment by GoGCommitment by GoGCommitment by GoGN/A/indirect (policy dia-logue)N/A/indirect (policy dia-logue)N/A/indirect (policy dia-logue)N/A
Compositebudgets
Budgetpreparation
GoG inputs
Legislative ba-sis for PFM
Budget contents
Improvement in budgetcalendar,More transparency inbudget information(greater accountabilityto the public)
Timing of budg-et submission toParliament
5. CoNCLuSIoNS AND WIDEr LESSoNS
More communi-cation with pub-lic on budget
73
74
PFm focalareaTIN introducedVAT re-introducedGRA ActTax Policy Unit estab-lished in MoFEPBeginning the processof integrating revenueservicesRegular reports on amt,composition of publicdebt, guarantees of ext.loans to parastatalsNew commitment con-trol system introduced(not sustained)Establishment of sepa-rate Treasuries in someMDAs, regions (im-provements in cashflows not comprehen-sive or sustained)Closing of a number ofredundant bank ac-countsGoG-DP policy dialogueCommitment/Institu-tional factors/ inter-agency co-operationGoG-DP policy dialogueCommitment/inter-agency co-operationGoG-DP policy dialogueCommitment/Institu-tional factors/ inter-agency co-operation (tosustain)Commitment by GoGunder PUFMARP re-formsExternal software sup-portDirect (design, soft-ware)Commitment by GoG,with DP supportDirect (design, imple-mentation)Commitment by GoG,with DP supportDirect (design, imple-mentation)Commitment by GoG,with DP supportDirect (design, imple-mentation)
Inputs1
Specific reform
outputs2
Causal factors in posi-tive/negative progress
role of DPContributions3,4
Intermediateoutcomes
Direct taxes
Resourcemanage-ment – rev-enues
TA, training, software,hardwareFunding of €6.8 mn(BMZ) + €2 mn (SECO).Prior funding fromDFID.
Indirect taxes
Significant increase inrevenues, not clear towhat extent this wasdue to improved effi-ciency
Greater admin-istrative effi-ciency
5. CoNCLuSIoNS AND WIDEr LESSoNS
Resourcemanage-ment –Treasury,aid & debt
Aid, debt man-agement
Slippage in fiscal disci-pline (weaknesses incommitment controls)
Advisory support,softwareFunding of $0.2 mn(WB) + $0.03 mn (GoG)(debt management),$0.2 mn (WB) + $0.03mn (GoG) (cash mng-mt). Other fundingfrom US Treasury,WB, and IMF
Commitmentcontrols
Direct (limited, not sus-tained)
Treasury rea-lignment
N/A/indirect (policy di-alog)
Single Treasuryaccount
Direct (but not fully ef-fective), indirect (policydialogue)
PFm focalareaAttempts to improve in-formation on cash flowsto/from MDAs not com-prehensive. or sus-tainedProcurement Actpassed; challenges withimplementationIPPD2 being used forministries, depts. butonly some agenciesInternal Audit AgencyAct passedInternal Audit AgencyestablishedStandards, manual pro-ducedIAUs establishedARICs establishedImplementation in 6 pi-lot MDAs not sustained.System abandoned.Ambitiousness of de-sign, insuff. high-levelcommitment in prac-tice, seen primarily astechnol. reform, man-agement problems(agency ownership),technical problemsGoG-DP policy dialogue,DP triggerUnder PUFMARP re-formsCommitment by GoG,GoG-DP policy dialogue,DP triggerIndirect (policy dia-logue)N/A, indirect (policy dia-logue)GoG-DP policy dialogue,DP triggerInstitutional factors(implementation)Direct (design), indirect(implementation)GoG-DP policy dialogue,DP triggerCommitment/Institu-tional factors/ inter-agency co-operationDirect (but not fully ef-fective), indirect (policydialogue)
Inputs1
Specific reform
outputs2
Causal factors in posi-tive/negative progress
role of DPContributions3,4
IntermediateoutcomesSlippage in fiscal disci-pline
Resourcemanage-ment -ex-penditures
Advisory supportFunding from CIDA,WB, IMF, DFID, DKK,and UNICEF.
Cash manage-ment
Procurement
Internalcontrols,audit, &monitoring
TrainingFunding of €<0.5 mn(EU). Other fundingfrom DFID and France.
Payroll controls
Internal audit functioninitiated
Internal audit
Accountingand report-ing
Funding of $15.3 mn(WB)+ $4.6 mn (GoG)(BPEMS)
FMIS/BPEMS
Direct (under PUF-MARP), thenindirect (policy dia-logue)
Slippage in fiscal disci-pline
5. CoNCLuSIoNS AND WIDEr LESSoNS
Timeliness ofin-year budgetexecution re-ports
More timely gazetting
N/A/ indirect (policy di-alogue)
75
76
PFm focalareaExternal Audit AgencyActStandards, manual pro-ducedPerformance, VfM au-dits begunClearance of auditbacklogMore timely PAC reviewof audit reportsPublic access to PACdebatesCommitment by GoG,public pressureN/A/indirect (policy dia-logue)Commitment by GoGunder PUFMARP re-forms, with DP supportGoG-DP policy dialogue,DP triggerDirect (design, TA sup-port)Improved accountabilityto the public for publicfinances
Inputs1
Specific reform
outputs2
Causal factors in posi-tive/negative progress
role of DPContributions3,4
Intermediateoutcomes
Externalaccount-ability
TA/capacity building(GAS), training, auditsof selected flows.Funding of €8 mn (EU)(External audit), €0.5mn/yr (BMZ) (PAC),DFID (GAS, PAC) £0.4mn, KfW, CIDA, DA-NIDA (DKK 0.5 mn/yr)
External audit
Parliamentaryscrutiny of pub-lic finance
1
5. CoNCLuSIoNS AND WIDEr LESSoNS
2
DP funding only. GoG funding not available to match this breakdown.Main outputs shown. Does not include staff training, considered an input. Where outputs have not been sustained, this has been noted.3Analysis from Box 7 above.4DP support shown may not be comprehensive due to non-availability of data.
Annex A: Summary Matrices ofResponses to Evaluationquestions for Country andComponent Case histories
77
78
PfM Reform component/initiative: financial Management information SystemsFindings• Two major financial management information systems (FMIS) initiatives were launched in thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemswere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• Donor support for both BPEMS and IPPD2 had finished by 2003 and minimal donor financing wasprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January2006. A follow-on donor funded investment operation was approved in 2010 and involvesimprovement and updating of the BPEMS system and its transformation into the GhanaIntegrated Financial Management Information System (GFIMIS).• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and toa lesser extent the Budget Division in MoFEP. Roll out involved initially the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.BPEmS• The original estimated of BPEMS was USD 11.95 million to be financed under a PUFMARPthrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68million, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.• The reform inputs covered four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMprocesses and procedures and the functional design of the system;– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAsites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.IPPD2• IPPD2 is a human resources management and payroll system that uses the Oracle HMRSapplication. Inputs covered the system specification, cost of IT hardware, the application licence,
Evaluation Question
Judgement criteria/Possible indicators
a inputs & context: the design of PfM reform
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
A.1. What has been the na-ture and scale of PFmreform inputs providedby Government andDonors?
• Government funds forPFM reforms committedand actually disbursed byyear over the evaluationperiod.• Donor funds for PFMreforms committed andactually disbursed by yearover the evaluation period• Nature of supportprovided to PFM reformefforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)
level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)
Evaluation Question
Judgement criteria/Possible indicators
million, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.• The reform inputs covered four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMprocesses and procedures and the functional design of the system;– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAsites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensiveFindingscustomisation was required.
a inputs & context: the design of PfM reform
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
79
– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs (FMIS) initiatives were launched in the• Two major financial management information systemswere less than originally envisaged.• Government funds forA.1 What has been the na-– At1990s: (i)the Budget Planning MoFEP decided to continue theSystem (BPEMS)of the BPEMSlate the completion of PUFMARP, and Expenditure Management implementation which wasPFM reforms committedture and scale of PFmusing its own resources. The annual cost of Reform Program and support and (ii)thefunded under the Public Finance Managementthe Oracle licence (PUFMARP),for the applicationand actually disbursed byreform inputs providedis around USD 180,000.Integrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenyear over the evaluationby Government andintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemsperiod.Donors?IPPD2linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.were• Donor funds for PFM• IPPD2 is a human both BPEMS and IPPD2 and payroll by 2003 and minimal donor HMRS• Donor support forresources managementhad finishedsystem that uses the Oraclefinancing wasreforms committed andapplication. Inputs covered further development of BPEMS IT hardware, the application licence,provided thereafter and thethe system specification, cost of and IPPD2 was largely funded fromactually disbursed by yearcustomisation costs and training of staff in line profiled asdomestic sources. Both reforms were strongly ministries. key elements of the PFM reform effortover the evaluation period• under the 3project began inMedium-Term Action Plan that wasto replace by MoFEP inIPPD1The IPPD2 Year Short and 1999, when the decision was taken prepared the existing January• Nature of supportsystem when it became funded investment operation was longer be guaranteed. Funding2006. A follow-on donor clear that vendor support could noapproved in 2010 and involves forprovided to PFM reformIPPD2 was provided by DFID but was halted in January 2001 due to unsatisfactory progress. Theimprovement and updating of the BPEMS system and its transformation into the Ghanaefforts (equipment,project resumed in 2005 with GoG Information the going live in June 2006 with pilots in theIntegrated Financial Management funding with System (GFIMIS).training, TA, diagnosticMinistry of Health, Audit Service Controller and Accountant General’s Department (CAGD) and• The reforms were centred in the and the pensions office. In October 2006 the system was rolledtowork.)out to all other the Budget Division in MoFEP. Roll out involved initially the central governmenta lesser extentMDAs.• Focal areas for reforms byministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.function (based on PFMA. 2.What type of structures• Structure of design teamBPEmS“clusters” as in Andrewsand related consultationhas been used for the• There were significant limitations in the conceptualisation and design of BPEMS. It was seen2010)process: Balance of inputsBPEmSdesign and manage-primarily as a technology driven reform with insufficient attention being given to the changes in• Focal areasGovt/by Donors/ for reforms by • The original estimated of BPEMS was USD 11.95 million preceded or been undertaken in parallelment of these reformPFM processes and procedures that should either have to be financed under a PUFMARPthrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68organisational location/Consultants; Extent ofinputs? Have thesewith BPEMS. The institutional analysis underlying BPEMS and PUFMARP was weak withmillion, of which USDgiven to change management issuesUSDthe assessment of capacity issueslevel of governmentinput/ consultation withstructures served toinsufficient attention 15.28 million was funded by IDA and and 4.60 million funded by GoG. Thisrepresentedrequirements. PUFMARP was a complex project with 11 major components63(Ministry of Finance,end users of PFM systemprovide a coordinatedand training a 50% cost overrun compared with the original estimates.• The reform inputs coveredcomponents (MTEF, BPEMS and Audit) accounted for 87.6% ofSector ministries, Local(Sector ministries, LGsand harmonised deliv-although the three largest four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMGovernments, Parliament,and service institutions).ery framework?expenditure under the project. The implementation completion report (ICR) noted the size andprocesses PUFMARP presented a functional design of the system;Management• CSOs, etc.) & co-complexity ofand procedures and the “serious challenge to government in terms of coordination– Technicalmanagement”64– purchase of hardware systems, installation of LANs in at 56 MDAordination structure forand project infrastructure .sites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISPFM reforms (Ad hoc• BPEMS was initially managed through the PUFMARP project implementation unit (PIU) which63 The 10 components of the PUFMARP listed inNormalwere:hadBudget preparation and MTEF (fundedFinancials withBudget implementation/BPEMS;(i)software. The system chosenor operational responsibility for the reform. This resulted in theProject Units vs.the ICRno functional responsibility was Oracleby DFID); (ii)six modules purchased: (i)generalledger; (ii)purchase order; the client departments (CAGD and the2008 to(v)public sector(iii) Cash management; (iv) Aidmanagement structures;and debt management system;project being distanced from (iii)accounts payable; (iv)cash management; Division). The mid-(v) revenue management (removed from the project in NovemberBudgetcover the increasedbudgeting; and (vi)accounts receivable. Because Oracle Financials had responsibility for thedesigned forGovt-controlled(vii) externalterm reviewaudit (funded byrecommended that transfer ofdevelopment;beenlegal frameworkcost of the BPEMS software); (vi) procurement;vs. Sharedandcorporateof the PUFMARPthebeen(viii) humanin a government environment, extensiveinternalmarket and had notEC);widely usedresourcemanagement(ix)Donor-Govt(xi) project management.but this was not implemented. It was only when PUFMARP came to an end thatBPEMS,revision; (x) communications strategy; andmanagement;customisation BPEMS was transferred to the BPEMS Secretariat, which was established as aUse of consultants forresponsibility for was required.64 World Bank (May 2004), Implementation Completion Reportdivision within MoFEP and located in CAGD. BPEMSFinance Management Technicaltheon a Credit to the Republic of Ghana for a Publicrelied heavily on consultants forAssistancemanagerial or purely– Training – inputs for training ofthe development of supporting technical manuals and provisionProject, World Bank Report No. 28089-GH, p6.advisory roles)customisation of the system and staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– training. Because of PUFMARP, MoFEP decided to continue the implementation of the BPEMSof At the completion the consultants reported to the PIU, the relationship and accountability to• Arrangements forusing its own resources. The annual cost of At the conclusion and support it was applicationCAGD and the Budget Division was weakened. the Oracle licenceof PUFMARP,for the decided thatmonitoring & evaluation.is around USD 180,000.the further development and roll out of BPEMS should be led by the BPEMS Secretariat with• Level of harmonisationminimal reliance on external consultants.and alignment of differentIPPD2• Monitoring of the BPEMS was initially undertaken through PUFMARP. This included World Bankdonor contributions.
80
has been used for thedesign and manage-ment of these reforminputs? Have thesestructures served toprovide a coordinatedand harmonised deliv-ery framework?• There were significant limitations in the conceptualisation and design of BPEMS. It was seenprimarily as a technology driven reform with insufficient attention being given to the changes inPFM processes and procedures that should either have preceded or been undertaken in parallelwith BPEMS. The institutional analysis underlying BPEMS and PUFMARP was weak withinsufficient attention given to change management issues and the assessment of capacity issuesand training requirements. PUFMARP was a complex project with 11 major components63although the three largest components (MTEF, BPEMS and Audit) accounted for 87.6% ofexpenditure under the project. The implementation completion report (ICR) noted the size andcomplexity of PUFMARP presented a “serious challenge to government in terms of coordinationFindingsand project management”64.ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE• BPEMS was initially managed through the PUFMARP project implementation unit (PIU) whichhad major financial management operational responsibility for the reform. This resulted the• Two no functional responsibility orinformation systems (FMIS) initiatives were launched in in theproject being distanced from the client departments (CAGD and the Budget Division). The mid-late 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasterm review of the PUFMARP recommended that transfer of management responsibilityfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)the forBPEMS, but this was not implemented. It was only when PUFMARP came to an that hadIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system end that beenresponsibility for BPEMS was transferred to the BPEMS Secretariat, which was established asintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemsadivision within MoFEP and located in CAGD. IPPD2 would become the HR module of BPEMS.were linked in that it was envisaged that the BPEMS relied heavily on consultants for thecustomisation for both BPEMS and IPPD2 had finished by 2003 technical manuals and provision• Donor support of the system and the development of supportingand minimal donor financing wasof training. Because the the further development of PIU, the relationship and accountability toprovided thereafter and consultants reported to theBPEMS and IPPD2 was largely funded fromCAGD and the Budget Division was weakened. At the conclusion of PUFMARP, PFM decided thatdomestic sources. Both reforms were strongly profiled as key elements of theit wasreform effortthe further Year Short and Medium-Term Action Plan be was prepared by Secretariat withunder the 3development and roll out of BPEMS shouldthatled by the BPEMS MoFEP in Januaryminimal reliance donor funded investment2006. A follow-onon external consultants. operation was approved in 2010 and involves• improvement the BPEMS was the BPEMS system and its transformation into the GhanaMonitoring of and updating of initially undertaken through PUFMARP. This included World Bankled supervisions and a mid-term review. In the System (GFIMIS).Integrated Financial Management Information years following PUFMARP no special monitoringarrangements were put in in the Controller and Accountant General’s Department highlighted in• The reforms were centred place although implementation progress and issues are(CAGD) and tothe external review of PFM that were MoFEP. Roll out involved and 2009.a lesser extent the Budget Division in undertaken in 2006, 2008initially the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.IPPD2• The IPPD2 project was managed by the CAGD and implemented through a contract with aBPEmSconsultancy firm. It was a much simpler initiative than BPEMS with its under a PUFMARP• The original estimated of BPEMS was USD 11.95 million to be financed implementation focusedon establishing a payroll databases and management system in the MDAs. It involved onlythrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68limited of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thismillion,number of users of the system and raised fewer systemic change management issues.represented a 50% cost overrun compared with the original estimates.• The IDA funding for the BPEMS was both on plan and on budget. It was administered using World• The reform inputs covered four main areas:Bank procedures, rather than government systems. The DFID funding for IPPD2 was on plan and– Requirement definition and design – technical assistance for re-engineering of PFMadministered through a service provider.processes and procedures and the functional design of the system;• The MDBS dialogue and triggers had very limited influence on the implementation of the FMIS– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAreforms and since 2006 only two triggers have been related to FMIS implementation. The firstsites and linking of 18 sites over a wireless loop.required complete deployment of all six BPEMS module in the eight pilot MDAs by December– Financial management software – purchase, customisation and installation of the FMIS2006. This was not achieved but was “declared met” by the 2007 review, presumably because itsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalwas considered that MoFEP was not in a position to meet the trigger. The second requiredledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorintegration of 50% of subverted agencies into IPPD2, and conducting a payroll audit to reconcilebudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thethe IPPD2 payroll with the establishment list. This had not been met at the time of the May 2008corporate market and had not been widely used in a government environment, extensiveannual review, but was subsequently met by August 2008. There were no MDBS triggers relatedcustomisation was required.to IPPD2.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.IPPD2• IPPD2 is a human resources management and payroll system that uses the Oracle HMRSapplication. Inputs covered the system specification, cost of IT hardware, the application licence,customisation costs and training of staff in line ministries.
Evaluation Question
and related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system(Sector ministries, LGsand service institutions).Judgement criteria/• Management & co-Possible indicatorsforordination structure
a inputs & context: the design of PfM reformPFM reforms (Ad hocProject Units vs. Normal• Government funds forA.1 What has been the na-management structures;PFM reforms committedture and scale of PFmGovt-controlled vs. Sharedand actually disbursed byreform inputs providedDonor-Govt management;year over the evaluationby Government andUse of consultants forperiod.Donors?• managerial or purelyDonor funds for PFMadvisory roles)reforms committed andactually disbursed by year• over the evaluation periodArrangements formonitoring & evaluation.• Nature of support• provided to PFM reformLevel of harmonisationand alignment of differentefforts (equipment,donor contributions.training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,• Use of country systems:A.3. What types of comple-Sector ministries, Localextent to which externalmentary actions haveGovernments, Parliament,aid is ‘on plan’, ‘on budget’Donors taken to supportCSOs, etc.)and ‘on treasury’?PFm reforms and what• Ranking of donor efforts tohas been their signifi-support national systemscance? Have they hadas expressed in Parisany influence on the ex-Declaration monitoringternal constraints to re-reports and PEFAform?indicators D2 and D3.• Scale and evolution overthe evaluation period of aid
flows provided as GBS,SBS or debt relief andrelative contribution toreduction of treasurymanagement costs and todiscretionary resourcesavailable to Government,(including effect onfinancing constraints forPFM reform outputs).
on establishing a payroll databases and management system in the MDAs. It involved onlylimited number of users of the system and raised fewer systemic change management issues.
A.3. What types of comple-mentary actions haveDonors taken to supportPFm reforms and whathas been their signifi-cance? Have they hadany influence on the ex-ternal constraints to re-form?• The IDA funding for the BPEMS was both on plan and on budget. It was administered using WorldBank procedures, rather than government systems. The DFID funding for IPPD2 was on plan andadministered through a service provider.• The MDBS dialogue and triggers had very limited influence on the implementation of the FMISreforms and since 2006 only two triggers have been related to FMIS implementation. The firstrequired complete deployment of all six BPEMS module in the eight pilot MDAs by December2006. This was not achieved but was “declared met” by the 2007 review, presumably because itwas considered that MoFEP was not in a position to meet the trigger. The second requiredintegration of 50% of subverted agencies into IPPD2, and conducting a payroll audit to reconcilethe IPPD2 payroll with the establishment list. This had not been met at the time of the May 2008annual review, but was subsequently met by August 2008. There were no MDBS triggers relatedto IPPD2.Findings• Two major financial management information systems (FMIS) initiatives were launched in thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemswere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• Donor support for both BPEMS and IPPD2 had finished by 2003 and minimal donor financing wasprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January2006. A follow-on donor funded investment operation was approved in 2010 and involvesimprovement and updating of the BPEMS system and its transformation into the GhanaIntegrated Financial Management Information System (GFIMIS).• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and toa lesser extent the Budget Division in MoFEP. Roll out involved initially the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.BPEmS• The original estimated of BPEMS was USD 11.95 million to be financed under a PUFMARPthrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68million, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.• The reform inputs covered four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMprocesses and procedures and the functional design of the system;– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAsites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.
Evaluation Question
• Use of country systems:extent to which externalaid is ‘on plan’, ‘on budget’and ‘on treasury’?• Ranking of donor efforts tosupport national systemsas expressed in ParisDeclaration monitoringreports and PEFAindicators D2 and D3.• Scale and evolutionJudgement criteria/overthe evaluation period of aidPossible indicators
a inputs & context: the design of PfM reformflows provided as GBS,SBS or debt funds for• Governmentrelief andA.1 What has been the na-relative contribution toPFM reforms committedture and scale of PFmreduction of disbursedand actually treasury byreform inputs providedmanagement costs andyear over the evaluation toby Government anddiscretionary resourcesperiod.Donors?• available to Government,Donor funds for PFM(including effect onreforms committed andfinancing constraints foractually disbursed by yearPFM reform outputs).over the evaluation period• Nature of support over• Nature and evolutiontime of dialogue reformprovided to PFMon PFMwithin (equipment,effortsGBS/PRSC, SBSstructures or similartraining, TA, diagnosticdonor-govt fora, includingwork.)• use of PFM reformFocal areas for reforms byconditions in GBS/ PFMfunction (based on SBSdisbursement conditions.“clusters” as in Andrews• 2010)Influence of dialogue onthe “political” and “policy• Focal areas for reforms byspace” constraintsorganisational location/restricting choice oflevel of government PFMreform outputs.(Ministry of Finance,(Consistency of actual• Sector ministries, Localchanges with PFM reformGovernments, Parliament,triggers and/or points ofCSOs, etc.)emphasis in GBS/ PRSCdialogue)• Consideration of counter-factual:• Outputs: would PFMreforms have evolveddifferently in absence of
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
81
GBS and related policydialogue structures?• Outcomes: Wouldintermediate outcomeshave been different if useof country systems hadbeen lower?
82
Evaluation Question
• (Consistency of actualchanges with PFM reformtriggers and/or points ofemphasis in GBS/ PRSCdialogue)• Consideration of counter-factual:• Outputs: would PFMreforms have evolvedJudgement criteria/differently in absence ofPossible indicatorsFindings
a inputs & context: the design of PfM reformGBS and related policydialogue structures?• Government funds forA.1 What has been the na-• PFM reforms committedOutcomes: Wouldture and scale of PFmintermediate outcomesand actually disbursed byreform inputs providedhave been different if useyear over the evaluationby Government andof countryperiod. systems hadDonors?• been lower? for PFMDonor funds
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
A.4. to what extent hasthere been domesticpublic pressure or re-gional institutionalpressure in support ofPFm reform and whathas been the influenceon the external con-straints to reform?
• reforms committed andNature of engagementactually disbursed by yearwith PFM reform issuesover the evaluation periodwithin civil society,• Nature of support media.academia and the• provided to PFM reformExtent of participation ofefforts (equipment,CSOs in policy dialogue ontraining, TA, diagnosticPFM reform.• work.)Presence/ absence of• specific research and byFocal areas for reformsfunction (based onon PFMadvocacy by CSOs PFM“clusters” as in Andrewsreform issues.BPEmS• 2010)Relevance of PFM issues• Focal areas for reforms by • The original estimated of BPEMS was USD 11.95 million to be financed under a PUFMARPto political outcomes:through an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68• organisational location/Discussion of PFM issuesmillion, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thislevel of governmentwithin election campaignsrepresented a 50% cost overrun compared with the original estimates.(Ministry of of serviceFinance,• Discussion• The reform inputs covered four main areas:Sector ministries, Localdelivery issues related to– Requirement definition and design – technical assistance for re-engineering of PFMGovernments, Parliament,PFM in election campaignsprocesses and procedures and the functional design of the system;CSOs, etc.)• Voter concerns during – Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAelections (from Afrosites and linking of 18 sites over a wireless loop.Barometer and other– Financial management software – purchase, customisation and installation of the FMISrelevant publications)software. The system chosen was Oracle Financials with six modules purchased: (i)general• Existence of regional orledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorinternational bodies (e.g.budgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for theWAEMU, EITI) promotingcorporate market and had not been widely used in a government environment, extensivespecific norms on PFMcustomisation was required.issues, and their relative– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentinfluence on domesticof the BPEMS system meant that the training inputs were less than originally envisaged.political discourse.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMS• Evolution in the quality ofusing its own resources. The annual cost of the Oracle licence and support for the applicationpublic/ CSO engagementis around USD 180,000.with PFM issues over timeand its influence on PFMIPPD2reform outputs.• IPPD2 is a human resources management and payroll system that uses the Oracle HMRS• Consideration of counter-application. Inputs covered the system specification, cost of IT hardware, the application licence,factual: would PFM reformcustomisation costs and training of staff in line ministries.outputs have evolved
• Two major financial management information systems (FMIS) initiatives were launched in thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemswere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• Donor support part of the overall PFM review) finished by 2003 and minimal donor financing was• (Addressed as for both BPEMS and IPPD2 hadprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January2006. A follow-on donor funded investment operation was approved in 2010 and involvesimprovement and updating of the BPEMS system and its transformation into the GhanaIntegrated Financial Management Information System (GFIMIS).• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and toa lesser extent the Budget Division in MoFEP. Roll out involved initially the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.
Evaluation Question
• Relevance of PFM issuesto political outcomes:• Discussion of PFM issueswithin election campaigns• Discussion of servicedelivery issues related toPFM in election campaigns• Voter concerns during elections (from AfroBarometer and otherrelevant publications)• Existence of regional orinternational bodies (e.g.WAEMU,criteria/JudgementEITI) promotingspecific norms onPossible indicatorsPFMFindings
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
a inputs & context: the design of PfM reformtheir relativeissues, andinfluence on funds for• Government domesticA.1 What has been the na-political discourse.PFM reforms committedture and scale of PFm• and actuallythe quality ofEvolution in disbursed byreform inputs providedpublic/ CSO engagementyear over the evaluationby Government andwith PFMperiod. issues over timeDonors?• and its influencePFMDonor funds for on PFMreform outputs.reforms committed and• actually disbursed by yearConsideration of counter-factual: evaluation reformover thewould PFMperiodoutputs have evolved• Nature of supportdifferently PFM reformprovided toin absence ofdomestic public pressureefforts (equipment,or pressure diagnostictraining, TA,from regional/international institutions?work.)• Scale and focus reforms by• Focal areas for of supportA.5. How relevant was thefunction (based on PFMin relation to identifiedPFm reform pro-“clusters” as in AndrewsPFM weaknesses at startgramme to the needs2010)and during implemen-and the institutional• Focal areas for reforms bytation of reforms (inclu-context? Was donororganisational location/ding weaknesses in HRsupport consistent withlevel of government ofendowments, qualitynational priorities? to(Ministry of Finance,institutions & rules,what extent were adap-Sector ministries, Localquality of systems &tations made in re-Governments, Parliament,business processes, andsponse to the contextCSOs, etc.)quality of organisations).and the changing na-• Consistency of donor-tional priorities?funded support withGovernment PFM reformefforts, with ongoingpublic administrationreforms and overall Govtpolicies.
• Two major financial management information systems (FMIS) initiatives were launched in thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemswere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• Donor support for both BPEMS and IPPD2 had finished by 2003 and minimal donor financing wasprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January2006. A follow-on donor funded investment operation was approved in 2010 and involvesimprovement and updating of the BPEMS system and its transformation into the GhanaIntegrated Financial Management Information System (GFIMIS).• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and to• a lesser extent the Budget DivisionIPPD2 wasRoll out involved initially design and customisationThe donor support for BPEMS and in MoFEP. focused on the systems the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.and the provision of supporting IT equipment. Due to the implementation delays and difficultiesin rolling out the new systems the donor support ended before the new systems wereBPEmSoperational. However, the government continued funding the implementation of the reforms,• The originalthe case of IPPD2 this was USD 11.95 million to be financed under a PUFMARPalthough in estimated of BPEMS occurred after a gap of four years.through an IDA credit. The final expenditure spread overthe government’s PFM reform effort.• FMIS modernisation was and remains a central pillar of the period 1997-2003 was USD 19.68million, of whichin the15.28 million was funded by IDA and USD 4.60 million fundedin the face ofThis is reflected USD continuation of the reforms after donor funding ceased and by GoG. Thisrepresented a 50% cost rolling out the new systems.originalthe NDC government that came intocontinued difficulties in overrun compared with the Under estimates.• The reform inputs covered four main areas: on the FMIS reforms which have been stronglyoffice in 2008 there was a renewed emphasis– Requirementthe Deputyand design – technical assistance for re-engineering of PFMchampioned by definition Minister in MoFEP. New donor financing has been secured totallingprocesses and procedures to establish a countrywide GIFMIS. TheUSD 55.8 million in June 2010 and the functional design of the system;project which is funded by– Technical infrastructure – purchase of hardware upgrade of BPEMS and its extension toIDA, DFID, Danida and the EU involves a substantial systems, installation of LANs in at 56 MDAsites and linking of 18 sitesregional and district levels. over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.
83
• Extent to which scope andfocus of support wereadapted to the context,especially to the level ofownership and thecapacity for reform
tations made in re-sponse to the contextand the changing na-tional priorities?office in 2008 there was a renewed emphasis on the FMIS reforms which have been stronglychampioned by the Deputy Minister in MoFEP. New donor financing has been secured totallingUSD 55.8 million in June 2010 to establish a countrywide GIFMIS. The project which is funded byIDA, DFID, Danida and the EU involves a substantial upgrade of BPEMS and its extension toregional and district levels.
84
Evaluation Question• Two major financial management information systems (FMIS) initiatives were launched in thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which wasfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemswere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• Donor support for both BPEMS and IPPD2 had finished by 2003 and minimal donor financing wasprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January2006. A follow-on donor funded investment operation was approved in 2010 and involvesimprovement and updating of the BPEMS system and its transformation into the GhanaIntegrated Financial Management Information System (GFIMIS).• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and toa lesser extent the Budget Division in MoFEP. Roll out involved initially the central governmentministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.
quality of systems &business processes, andquality of organisations).• Consistency of donor-funded support withGovernment PFM reformefforts, with ongoingpublic administrationJudgement criteria/Govtreforms and overallPossible indicatorspolicies.ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASEFindings
a inputs & context: the design of PfM reform• Extent to which scope andfocus of support were• Government funds forA.1 What has been the na-adapted to the context,PFM reforms committedture and scale of PFmespecially to the level ofand actually disbursed byreform inputs providedownership and theyear over the evaluationby Government andcapacityperiod. for reformDonors?• implementation.PFMDonor funds forConsistency & coherence• reforms committed andwith wider Donor policiesactually disbursed by yearin countryover the evaluation period• Nature of supportthe ParisConsistency withDeclaration objectives asprovided to PFM reformregards ownership,efforts (equipment,harmonisation andtraining, TA, diagnosticalignment.work.)• Flexibility in design of byFocal areas for reformssupport (based on PFMfunctionand existence ofadequate as in Andrews“clusters”feed-back/learning2010) mechanisms inorder to adapt the design• Focal areas for reforms byand implementation oforganisational location/support to changinglevel of governmentneeds.(Ministry of Finance,Sector ministries, Localoutputs: the delivery of PfM reformsGovernments, Parliament,Documentation of outputsA.6. What have been the out-• CSOs, etc.):puts of the PFm reformprocess and to what ex-• by PFM function(according to PFM clusterstent has direct donorin Andrews 2010) andsupport contributed to• by type of output (People &these outputs?Skills; Laws & Rules;
Systems & Businessprocesses; Organisationalchanges)• Matching of outputs toPFM reform inputs,differentiating if possibleGovernment & Donorinputs.• Analysis of chronology ofevents to determinecausal links betweendonor support to reform
BPEmS• The original estimated of BPEMS was USD 11.95 million to be financed under a PUFMARPthrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68million, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.• The reform inputs covered four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMprocesses and of the two FMIS the functional design BPEMS, has been considered• The performanceprocedures and reforms, particularly of the system;– Technical infrastructure – purchase of hardwarecomprehensive functioning FMISat 56 MDA inunsatisfactory. After 15 years of implementation, a systems, installation of LANs in is still notsites and linking 2010, the over a wireless not been fully implemented in any of the eight pilotplace. By the end ofof 18 sitesBPEMS had stillloop.– Financialand was still running alongside the existing manually based systems. IPPD2 wasministries, management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with not interfaced with BPEMS. Sincebeing used for HR and payroll management system, but wassix modules purchased: (i)generalledger; (ii)purchase development and roll-out has been funded with minimal donor support2003 spending on FMISorder; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because FMIS Financialsand the government has remained committed to theOraclereforms. had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.BPEmS• The original timetable for the implementation of BPEMS envisaged pilot implementation by June– Training – inputs for training of1999 budget and full operation for the 2000 in the development1998, its partial operation for the staff in MoF, CGAD and pilot MDAs. Delays budget. By the endof the BPEMS system meant as follows: (i)requirement definition and design envisaged.of 2003 the achievements were that the training inputs were less than originally had been largely– At the completion of PUFMARP, MoFEP decided tousers; (ii)much of the technical the BPEMScompleted and considered adequate by the potential continue the implementation ofusing its own resources. The annual cost installation licence and support for the applicationinfrastructure including hardware systems,of the Oracleof MDA LANs was in place; (iii)a new 41is chart of accounts haddigit around USD 180,000. been implemented; (iv) the FMIS software had been installed and wasin partial use in CAGD although in parallel with the older systems; and (v) training had takenIPPD2However, the system had yet to be piloted in any ministry. The ICR for PUFMARPplace.• IPPD2 is a human resources management and could be achieved by 2007. Oracle HMRSconsidered that full implementation of BPEMS payroll system that uses theapplication.subsequent implementation of BPEMS continued to be dogged byapplication licence,• In practice, Inputs covered the system specification, cost of IT hardware, the delays. A report ofcustomisation costs and training of staffthat partial implementation of the BPEMS modulesa donor field visit in April 2008 recorded in line ministries.
learning mechanisms inorder to adapt the designand implementation ofsupport to changingneeds.• The performance of the two FMIS reforms, particularly BPEMS, has been consideredunsatisfactory. After 15 years of implementation, a comprehensive functioning FMIS is still not inplace. By the end of 2010, the BPEMS had still not been fully implemented in any of the eight pilotministries, and was still running alongside the existing manually based systems. IPPD2 wasbeing used for HR and payroll management system, but was not interfaced with BPEMS. Since2003 spending on FMIS development and roll-out has been funded with minimal donor supportFindingsgovernment has remained committed to the FMIS reforms.and the
outputs: the delivery of PfM reforms
A.6. What have been the out-puts of the PFm reformprocess and to what ex-tent has direct donorsupport contributed tothese outputs?Evaluation Question
• Documentation of outputs:• by PFM function(according to PFM clustersin Andrews 2010) andJudgement criteria/• by type of output (People &PossibleLaws & Rules;Skills;indicators
A.1 What has been the na-ture and scale of PFmreform inputs providedby Government andDonors?
a inputs & context: the design of PfM reformBusinessSystems &
processes; Organisational• Government funds forchanges)PFM reforms committed• and actuallyoutputs to byMatching of disbursedPFM reform evaluationyear over theinputs,differentiating if possibleperiod.• Government & DonorDonor funds for PFMinputs.reforms committed and• actually disbursed by yearAnalysis of chronology ofevents to determineover the evaluation periodcausal of support• Naturelinks betweendonor support to reformprovided to PFM reformand observed changesefforts (equipment, atoutput level.training, TA, diagnostic• work.)Consideration of• counterfactual: reforms byFocal areas for wouldthese outputs have beenfunction (based on PFMgenerated in in Andrews“clusters” as absence ofdonor2010) support to PFMreform?• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
85
BPEmSThe major financial management information systems (FMIS) initiatives were launched in the• Two original timetable for the implementation of BPEMS envisaged pilot implementation by June1998, its partial operation for the 1999 Expenditure Management System (BPEMS) which waslate 1990s: (i)the Budget Planning andbudget and full operation for the 2000 budget. By the endof 2003 the achievements were as follows: (i)requirement definition and design had beenfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)the largelycompleted and considered adequate by the 2 (IPPD2) which replaced a system that had beenIntegrated Personnel and Payroll Databasepotential users; (ii)much of the technicalinfrastructure including hardware systems, installation of MDA LANs was in place; (iii)a new 41introduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemsdigit linked accounts had been implemented; (iv) would become the HR been installed andwerechart ofin that it was envisaged that the IPPD2the FMIS software hadmodule of BPEMS. wasin partial use in CAGD although in parallel had the older systems; and (v) training financing• Donor support for both BPEMS and IPPD2 withfinished by 2003 and minimal donorhad taken wasplace. However, the and the further development of BPEMS and IPPD2 was PUFMARPprovided thereafter system had yet to be piloted in any ministry. The ICR for largely funded fromconsidered that full implementation of BPEMS profiled achieved by 2007.domestic sources. Both reforms were stronglycould be as key elements of the PFM reform effort• under the 3 subsequent implementation of BPEMS continued to be dogged by delays. A report ofIn practice, Year Short and Medium-Term Action Plan that was prepared by MoFEP in Januarya donor field visit donor 2008 recorded that partial implementation of the BPEMS modules2006. A follow-onin Aprilfunded investment operation was approved in 2010 and involvesincluding the and updating of the BPEMS system and its transformation into the Ghanaimprovementgeneral ledger and accounts payable modules had been achieved in CAGD and65eight pilot Financial Management Information ministries it wasIntegratedministries. . Within these eight pilot System (GFIMIS).estimated that only 6 per cent ofinvoices had were centred in through BPEMS in the previous year. It also noted that inadequate• The reforms been processed the Controller and Accountant General’s Department (CAGD) and totraining extent the Budget and the in MoFEP. Roll out involved initially to central governmenta lesser had been providedDivision ability of staff in the pilot ministriesthework with BPEMS waslow. Two years later when the GIFMIS project was being prepared the sub-vented agencies.ministries, departments and agencies, and additionally for IPPD2 , the situation had changedlittle. There had been no further roll out of the system, and little further deepening in the use ofthe system. Despite the implementation of a fibre WAN linking MDAs in Accra, connectivityBPEmSremained a estimated of BPEMS was USD 11.95 million to be financed under a PUFMARP• The originalproblem. Thus by 2010, the BPEMS had still not achieved the level of implementationthat had originally been planned for 2000. The pilot over the period 1997-2003 on USD 19.68through an IDA credit. The final expenditure spreadministries continued to relywastheir manualaccounting systems, 15.28 million was funded by utilise USD 4.60 million funded by GoG. Thismillion, of which USDwhile CAGD had continued toIDA andits ACCPAC system for the producingthe Public Accounts, and its monthly and annual reports.represented a 50% cost overrun compared with the original estimates.• The reform inputs coveredof Public Expenditure and Financial Management noted that “at the2009 External Review four main areas:current roll out rate, BPEMS will be obsolete andassistance for re-engineering of PFM full– Requirement definition and design – technical not maintainable by the time it reachescoverage” and considered that ”the main option facing thethe system; is relaunching BPEMSprocesses and procedures and the functional design of Governmentusing an updated version with standardof hardware systems, installation ofthat would include a– Technical infrastructure – purchase functionalities of Oracle Financials LANs in at 56 MDAhuman resource module to managewireless loop.sites and linking of 18 sites over a the payroll”. These requirements were subsequentlyreflected in the design of the GIFMIS–project for which financing was secured inof the FMIS– Financial management software purchase, customisation and installation mid 2010.software. The system chosen was Oracle Financials with six modules purchased: (i)generalIPPD2ledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sector• IPPD2 was operational across all MDAs byBecause Oracle Financials had been designed for thebudgeting; and (vi)accounts receivable. the end of 2006. The system is run by CAGD and themajor MDAs are connected to the central server inin a government environment, extensivecorporate market and had not been widely used CAGD. Data for smaller ministries continues65 Van der Helm, A., Le Mounier, X., Nehmeyer-Srocke, I., Ostler,be entered at CAGDrequired.G., Owusu-Ayim,using information provided by the MDAs. By August 2008 IPPD2 had alsoto customisation wasJ. (May 2008), Report on BPEMS and IPPD2 Field Visits.been rolled out to over 50% of subvented agencies, although at the end of 2010 a number of majorsubvented – inputs for training of universities, had still to implement IPPD2. In the absence of– Trainingagencies, including thestaff in MoF, CGAD and pilot MDAs. Delays in the developmentthe full roll out of IPPD2,meant that the training inputs were less than originally envisaged.of the BPEMS system it is not possible for CAGD to produce a comprehensive governmentpayroll. completion of PUFMARP, MoFEP and controls required implementation of the BPEMS– At theMoreover, the manual proceduresdecided to continue thebefore new employees can beentered into the resources. The continue to generate significant delays in the payments of theirusing its own payroll system annual cost of the Oracle licence and support for the applicationsalaries, in some180,000. up to two years. Overall, CAGD considers that the IPPD2 operatesis around USD cases ofeffectively although with some limitations.
reform?
86
Evaluation QuestionIPPD2IPPD2 was operational across all information end of 2006. The system is run launched in the• Two major financial managementMDAs by the systems (FMIS) initiatives wereby CAGD andthemajor MDAs are Budget Planning and Expenditure Management System (BPEMS) which waslate 1990s: (i)theconnected to the central server in CAGD. Data for smaller ministries continuesto be entered the Public Finance Management Reform Program (PUFMARP), and (ii)thefunded underat CAGD using information provided by the MDAs. By August 2008 IPPD2 had alsobeen rolled out to over 50% of subvented agencies, although at the end system number of majorIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced aof 2010 athat had beensubvented in the early 1990s (IPPD1). Although implemented as separate projects, absence ofintroducedagencies, including the universities, had still to implement IPPD2. In the the systemsthe full roll in of IPPD2, envisaged that the CAGD to produce a comprehensive governmentwere linkedoutthat it was it is not possible forIPPD2 would become the HR module of BPEMS.payroll. Moreover, the manual and IPPD2 and controls required before new employees can be• Donor support for both BPEMSprocedureshad finished by 2003 and minimal donor financing wasentered into the payroll the further development of BPEMS and IPPD2 in the payments of fromprovided thereafter andsystem continue to generate significant delays was largely fundedtheirsalaries, sources. Both reforms were strongly profiled as key elements of the PFM reform effortdomesticin some cases of up to two years. Overall, CAGD considers that the IPPD2 operateseffectively although with some limitations.under the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January• 2006. A follow-on donor funded investment operationhuman resources functionalities that areIPPD2 was supplied and configured with a number of was approved in 2010 and involvescurrently not and updating of the BPEMS system and module, a non-monetary Ghanaimprovementbeing used. These include a recruitmentits transformation into thebudgetingmodule (manpower Management Information System (GFIMIS).Integrated Financialceilings, headcount etc.), an absence management module, and a changemanagement module. In practice the larger MDAs have had to invest in their own systems and• The reforms were centred in the Controller and Accountant General’s Department (CAGD)to tocarry out these functions. Division in inefficient and reduces the level of integration in HR anda lesser extent the BudgetThis is bothMoFEP. Roll out involved initially the central governmentpayroll management.ministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.
Judgement criteria/Possible indicators
remained a problem. Thus by 2010, the BPEMS had still not achieved the level of implementationthat had originally been planned for 2000. The pilot ministries continued to rely on their manualaccounting systems, while CAGD had continued to utilise its ACCPAC system for the producingthe Public Accounts, and its monthly and annual reports.• The 2009 External Review of Public Expenditure and Financial Management noted that “at thecurrent roll out rate, BPEMS will be obsolete and not maintainable by the time it reaches fullcoverage” and considered that ”the main option facing the Government is relaunching BPEMSusing an updated version with standard functionalities of Oracle Financials that would include ahuman resource module to manage the payroll”. These requirements were subsequentlyFindingsin the design of the GIFMIS project for which financing was secured in mid 2010.reflected
a inputs & context: the design of PfM reform
A.1 What has been the na-ture and scale of PFmreform inputs providedby Government andDonors?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• Implementation delays and the low-level of roll out of BPEMs that has been achieved means thatBPEmSthe original estimated systems to date has had million to be financed under a PUFMARP• Theinvestment in FMISof BPEMS was USD 11.95very little impact on public finance management.This was despite an estimated investment in BPEMS under PUFMARP of USD was USD 19.68through an IDA credit. The final expenditure spread over the period 1997-200319.9 million and66subsequent GoG financed expenditure funded by that is probably in excess of USD 2.6 millionmillion, of which USD 15.28 million wassince 2003 IDA and USD 4.60 million funded by GoG. This.Implementation of IPPD2 has been more successfuloriginal estimates. implemented as arepresented a 50% cost overrun compared with the both because it wassimpler and more covered four main areas:• The reform inputs focused reform, and because from 2005 there was strong political support formeasures to strengthen the management of the assistance for re-engineering of PFM– Requirement definition and design – technicalgovernment payroll and wage bill. A number offactors have beenprocedures and the failure of the BPEMSthe system;processes and cited to explain the functional design of reform:– Technical infrastructure – purchase of hardware systems, installation of LANs inlevel and in– Weak ownership of the reforms. This applied at political and senior management at 56 MDAthe primary stakeholders in the reform (CAGD andsites and linking of 18 sites over a wireless loop. the Budget Division). It reflectedweaknesses in the design of the project, principally (i)inadequate arrangements for– Financial management software – purchase, customisation and installation of the FMIScoordination with and chosen was with the executive level and among the main stakeholders;software. The system involvementOracle Financials with six modules purchased: (i)general(ii)management responsibility under PUFMARP (iv)cash management; which was distancedledger; (ii)purchase order; (iii)accounts payable;being assigned to a PIU (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thefrom the primary and had not been widely used in a government environment, management• Appropriate to degree ofcorporate marketstakeholders in the project; and (iii)failure to address changeextensive66 This assumes Oracle licensing and supportnature of1.3 million over theinvolved in introducing an IT based system.issuesperiodwas required.ownership/costs of $customisation2004-10 and an equivalent expenditure on personnel, training and roll-out costs.– Seeing BPEMS as primarily a technological reform rather than a PFM reform. The provisionpolitical & admin supportunder PUFMARP in training both resources and time pilot MDAs. Delays in the processesfor reform?– Training – inputs forterms of of staff in MoF, CGAD andfor the review of business developmentand the reform design and consultation processes was inadequate. This contributed to the• Relative cost of outputs inof the BPEMS system meant that the training inputs were less than originally envisaged.lack of ownership of the changes to be introduced continue the implementation of they wererelation to budgeted costs.– At the completion of PUFMARP, MoFEP decided towith BPEMS, particularly where the BPEMSseen its own resources. The annual cost of the Oracle licence and support for the application• Influence of externalusingas potentially threatening entrenched roles and responsibilities. The choice of asophisticated 180,000.support to PFM reformis around USDfinancial management software system designed for private sectororganisations, necessitated extensive customisation which may have distracted attentionupon pacing andfrom the requirements for re-engineering business processes.sequencing.IPPD2– Insufficient technical capacities to implement and maintain the system. Issues ofHMRS• Counterfactual: would• IPPD2 is a human resources management and payroll system that uses the Oracle systemavailability, unreliability the system specification, cost of between the central servers andpacing & sequencing ofapplication. Inputs coveredof power supply, and connectivityIT hardware, the application licence,MDAs have costs and training of staff in line ministries.outputs have beencustomisationnever been satisfactorily resolved. Staffs in the MDAs are reported as lacking
A. 7. How efficiently werethese outputs gener-ated? Was the pacingand sequencing of re-forms appropriate andcost-effective? Was thecost per output accept-able?
• Government funds forPFM reforms committedand actually disbursed byyear over the evaluationperiod.• Donor funds for PFMreforms committed andactually disbursed by yearover the evaluation period• Nature of supportprovided to PFM reformefforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews• Consistency of outputs2010)• produced with planned byFocal areas for reformsprogramme (quantity andorganisational location/timing) governmentlevel of• Extent of coordination(Ministry of Finance,between outputs LocalSector ministries,• Quality of pacing andGovernments, Parliament,sequencing of outputCSOs, etc.)production:• internally coherent/efficient?• Consistent with capacityendowments?
A. 7. How efficiently werethese outputs gener-ated? Was the pacingand sequencing of re-forms appropriate andcost-effective? Was thecost per output accept-able?
Evaluation Question
• Consistency of outputsproduced with plannedprogramme (quantity andtiming)• Extent of coordinationbetween outputs• Quality of pacing andsequencing of outputproduction:• internally coherent/efficient?• Consistent with capacityendowments?• Appropriate to degree ofJudgement criteria/ownership/ naturePossible indicatorsof• Implementation delays and the low-level of roll out of BPEMs that has been achieved means thatthe investment in FMIS systems to date has had very little impact on public finance management.This was despite an estimated investment in BPEMS under PUFMARP of USD 19.9 million andsubsequent GoG financed expenditure since 2003 that is probably in excess of USD 2.6 million66.Implementation of IPPD2 has been more successful both because it was implemented as asimpler and more focused reform, and because from 2005 there was strong political support formeasures to strengthen the management of the government payroll and wage bill. A number offactors have been cited to explain the failure of the BPEMS reform:– Weak ownership of the reforms. This applied at political and senior management level and inthe primary stakeholders in the reform (CAGD and the Budget Division). It reflectedweaknesses in the design of the project, principally (i)inadequate arrangements forcoordination with and involvement with the executive level and among the main stakeholders;(ii)management responsibility under PUFMARP being assigned to a PIU which was distancedfrom the primary stakeholders in the project; and (iii)failure to address change managementissuesFindingsinvolved in introducing an IT based system.
a inputs & context: the design of PfM reformpolitical & admin supportfor reform?• Government funds forA.1 What has been the na-• PFM reformsof outputs inRelative cost committedture and scale of PFmrelation to budgeted costs.and actually disbursed byreform inputs providedInfluence of external• year over the evaluationby Government andsupportperiod. to PFM reformDonors?• upon pacing and PFMDonor funds forsequencing.reforms committed andCounterfactual: would• actually disbursed by yearpacing & sequencing ofover the evaluation periodoutputs have been• Nature of supportdifferent to PFM reformprovided in absence ofexternal PFM support?efforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
87
8. What have been the bind-ing external constraints onthe delivery of PFm reform:
• Degree of ownership ofreforms at theadministrative and
– Seeing BPEMS as primarily a technological reform rather than a PFM reform. The provisionunder PUFMARP in terms of both resources and time for the review of business processes• Two major financial management information systems (FMIS) initiatives were launched in theand the reform design and consultation processes was inadequate. This contributed to thelate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which waslack under the Public Finance Management Reform Program (PUFMARP), where theyfunded of ownership of the changes to be introduced with BPEMS, particularlyand (ii)the wereseen as Personnel and Payroll Database 2 roles and responsibilities. The choice of aIntegratedpotentially threatening entrenched(IPPD2) which replaced a system that had beensophisticated early 1990s (IPPD1). Although implemented as separate projects, theintroduced in thefinancial management software system designed for private sector systemsorganisations, necessitated extensive customisation which may have distracted BPEMS.were linked in that it was envisaged that the IPPD2 would become the HR module ofattentionfrom the requirements for re-engineering business by 2003 and• Donor support for both BPEMS and IPPD2 had finishedprocesses. minimal donor financing was– Insufficient technical capacities to implement and maintain the system. Issues of systemprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromavailability, unreliability of power supply, and connectivity between the central servers anddomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortMDAs have never been satisfactorily resolved. Staffs in was prepared by MoFEP in Januaryunder the 3 Year Short and Medium-Term Action Plan that the MDAs are reported as lackingadequate IT skills and confidence operate the system effectively, in 2010 and involves2006. A follow-on donor funded investment operation was approvedreflecting both inadequateand inappropriately timed training.improvement and updating of the BPEMS system and its transformation into the Ghana– Overoptimistic timetabling. ThisInformation System (GFIMIS).Integrated Financial Management remained a problem throughout the period. The initialtimetabling envisaged that BPEMS would be Accountant General’s 3-4 years. This reinforced• The reforms were centred in the Controller andfully operational withinDepartment (CAGD) and tothe focus on the Budget Division in MoFEP. Roll out involved initially the central governmenta lesser extent the technological aspects of the FMIS and resulted in less attention being given togiven to departments and agencies, and additionally for IPPD2 the sub-vented agencies.ministries,the PFM elements. Subsequently it resulted in the system being piloted before it hadbeen properly tested, connectivity issues addressed and staff adequately trained. As a resulttheBPEmSpiloting phase, instead of lasting for a few months, has been on-going in some ministries,for example Health, of BPEMS was USD 11.95 million to be financed under a PUFMARP• The original estimatedsince 2004.– Procurement management. The splitting of procurement period 1997-2003 was USD 19.68through an IDA credit. The final expenditure spread over the for IT systems, network systemsand connectivity, and IT hardware between different suppliers resulted in no single vendormillion, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisbeing responsible and fully accountable with the original estimates.represented a 50% cost overrun comparedfor the implementation of BPEMS. This contributed toproblems in the coordination and phasing• The reform inputs covered four main areas: of procurement with for example IT equipmentarriving years in advance of it being required for much delayed the roll out BPEMS. As– Requirement definition and design – technical assistance for re-engineering of PFM resultprocurement for BPEMS involved considerable inefficiency and excessive waste.processes and procedures and the functional design of the system;The design of the new donor purchase of hardware launched in 2010 to support the upgrading• – Technical infrastructure –funded project that wassystems, installation of LANs in at 56 MDA ofBPEMSand linking of 18 sites over a wireless loop.sites and it’s relaunching as GIFMIS, has taken account of many of the lessons from failure toimplement BPEMS. Management arrangements for the new project emphasise the involvement– Financial management software – purchase, customisation and installation of the FMISof the key stakeholders in specialist team to manage different elements of the GIFMIS project.software. The system chosen was Oracle Financials with six modules purchased: (i)generalThere is a (ii)purchase order; (iii)accounts payable;businessmanagement; (v)public sector ofledger; greater emphasis on re-engineering key (iv)cash processes covering the reviewPFM legislation and legislation, the introduction of aOracle Financials had beenbudget for thebudgeting; and (vi)accounts receivable. Because new chart of accounts and designedclassification, and the strengthening internalused in a government environment, extensivecorporate market and had not been widely expenditure control practices. There is also acustomisation was required.strong emphasis on addressing change management issues and on strengthening PFMfunctions in inputs for training of staff in whether issues around the Delays in the development– Training –MDAs. However, it less clear MoF, CGAD and pilot MDAs.realistic timetabling andappropriate sequencing meantreforms have been fully were less than originally envisaged.of the BPEMS system of the that the training inputs addressed.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the and• The overriding reasons for the failure to establish an effective FMIS related to the designBPEMSusing its own the reforms. The role of external Oracle licence and support was probably notmanagement of resources. The annual cost of theconstraints is less clear andfor the applicationis around strong. However:particularly USD 180,000.
88
Evaluation Questionstrong emphasis on addressing change management issues and on strengthening PFMfunctions in MDAs. management information systems (FMIS) initiatives were launched in the• Two major financialHowever, it less clear whether issues around the realistic timetabling andappropriate sequencing Planning and have been fully addressed.late 1990s: (i)the Budgetof the reformsExpenditure Management System (BPEMS) which was
Judgement criteria/Possible indicators
problems in the coordination and phasing of procurement with for example IT equipmentarriving years in advance of it being required for much delayed the roll out BPEMS. As resultprocurement for BPEMS involved considerable inefficiency and excessive waste.• The design of the new donor funded project that was launched in 2010 to support the upgrading ofBPEMS and it’s relaunching as GIFMIS, has taken account of many of the lessons from failure toimplement BPEMS. Management arrangements for the new project emphasise the involvementof the key stakeholders in specialist team to manage different elements of the GIFMIS project.There is a greater emphasis on re-engineering key business processes covering the review ofPFM legislation and legislation, the introduction of a new chart of accounts and budgetclassification, and the strengthening internal expenditure control practices. There is also aFindingsANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
a inputs & context: the design of PfM reform
A.1 What has been the na-ture and scale of PFmreform inputs providedA.8. What have been theby Government andbinding external con-Donors? on the deliverystraintsof PFm reform: politi-cal, financing or policyfactors? How has thisvaried across the differ-ent PFm reform compo-nents?
• Government funds forPFM reforms committedand actually disbursed• Degree of ownership ofbyyear over the evaluationreforms at theperiod.administrative and• political levels (narrow vs.Donor funds for PFMreforms committed andbroad; depth ofactually disbursed by yearcommitment to reformover the evaluation thisand motivations forperiod• commitment).Nature of supportprovided interaction• Quality ofto PFM reformefforts (equipment,between administrativetraining, TA, cadres.and political diagnosticwork.)• Extent and nature of• political accountability byFocal areas for reformsfunction (based on PFM(within ruling party, vis-à-“clusters” as in vis-à-visvis Legislature, Andrews2010)Electorate & Civil Society)• and extent to which this isFocal areas for reforms byorganisationalpatronage or location/level of governmentperformance-based.(Ministry of Finance,• Economic context and itsSector ministries, Localinfluence on financing ofGovernments, Parliament,PFM reform: per capitaCSOs, etc.)GDP and % growth inevaluation period,significance of domesticrevenue, significance ofaid, level of total anddiscretionary publicspending; absence/presence of macro crises.
• funded under the Publicfor the failure to establish an effective FMIS related to and design andThe overriding reasons Finance Management Reform Program (PUFMARP), the (ii)theIntegrated Personnel and Payrollrole of external constraints is less clear and was probably notmanagement of the reforms. The Database 2 (IPPD2) which replaced a system that had beenintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemsparticularly strong. However:were linked in that it was envisaged that the the FMIS reforms maythe HR module of BPEMS.– The absence of external championing for IPPD2 would become have contributed to the failure• Donor support for both BPEMS and IPPD2arose. In the start up of GIFMIS a workshop briefingto address management issues as they had finished by 2003 and minimal donor financing wasprovided thereafter and the further development of BPEMS and IPPD2 was largely fundedlevelwas organised for parliamentarians at which MoFEP officials were challenged over the fromdomestic sources. Both reforms were strongly profiled as key elements of thein the reformofof government commitment to addressing some of the issues that resulted PFM failure effortunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in JanuaryBPEMS.2006. A follow-on donor funded investment operation was approved in 2010 and involves– Ghana’s relatively strong economic and fiscal performance during through the mid 2000s mayimprovement and updating of the BPEMSof measuresits strengthen PFMinto the Ghana an FMIS.have affected the priority and urgency system and to transformation and introduceIntegrated Financial Management Information System (GFIMIS).The more challenging economic and fiscal environment that faced the NDC government that• The reforms werein 2009 coincided with a renewed emphasis on strengthening PFM systems intocame to power centred in the Controller and Accountant General’s Department (CAGD) anda lesser extent the Budget was seen as a priority. out involved initially the central governmentwhich an effective FMIS Division in MoFEP. Rollministries, departmentsfell sharply following the completion of the PUFMARP in 2003 and– Financing for BPEMS and agencies, and additionally for IPPD2 the sub-vented agencies.subsequent funding was probably inadequate to support the further development and roll outBPEmSBEPEMS. However, it seems that this reflected a lack of management commitment to andof• The original estimated of BPEMS was USDoverriding funding financed under a PUFMARPconfidence in the reform rather than an 11.95 million to be constraint .through an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68million, of which USD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.• The reform inputs covered four main areas:– Requirement definition and design – technical assistance for re-engineering of PFMprocesses and procedures and the functional design of the system;– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAsites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.IPPD2• IPPD2 is a human resources management and payroll system that uses the Oracle HMRSapplication. Inputs covered the system specification, cost of IT hardware, the application licence,customisation costs and training of staff in line ministries.
• Has timeliness of fundingbeen an issue?• Nature of “conventionalwisdom” on PFM reforms:does prevailing thinkingexclude certain reformoptions?• Has policy discussion onPFM reform been open interms of range ofparticipants and range of
of BEPEMS. However, it seems that this reflected a lack of management commitment to andconfidence in the reform rather than an overriding funding constraint .
Evaluation Question
Electorate & Civil Society)and extent to which this ispatronage orperformance-based.• Economic context and itsinfluence on financing ofPFM reform: per capitaGDP and % growth inevaluation period,significance of domesticrevenue, significance ofaid, level of total anddiscretionary publicJudgement criteria/spending; absence/Possible indicatorscrises.presence of macroFindings
a inputs & context: the design of PfM reform• Has timeliness of funding
A.1 What has been the na-ture and scale of PFmreform inputs providedby Government andDonors?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
been an issue?• Two major financial management information systems (FMIS) initiatives were launched in the• Government funds for• PFM reforms committedNature of “conventionallate 1990s: (i)the Budget Planning and Expenditure Management System (BPEMS) which waswisdom” on disbursed byfunded under the Public Finance Management Reform Program (PUFMARP), and (ii)theand actuallyPFM reforms:does prevailing thinkingIntegrated Personnel and Payroll Database 2 (IPPD2) which replaced a system that had beenyear over the evaluationexcludeintroduced in the early 1990s (IPPD1). Although implemented as separate projects, the systemsperiod. certain reformwere linked in that it was envisaged that the IPPD2 would become the HR module of BPEMS.• options?Donor funds for PFMHas policy discussion on• reforms committed and• Donor support for both BPEMS and IPPD2 had finished by 2003 and minimal donor financing wasPFM reform been open inprovided thereafter and the further development of BPEMS and IPPD2 was largely funded fromactually disbursed by yearterms of range ofdomestic sources. Both reforms were strongly profiled as key elements of the PFM reform effortover the evaluation periodparticipants and range ofunder the 3 Year Short and Medium-Term Action Plan that was prepared by MoFEP in January• Nature of supportideas?2006. A follow-on donor funded investment operation was approved in 2010 and involvesprovided to PFM reform• efforts (equipment,Overall, what have beenimprovement and updating of the BPEMS system and its transformation into the Ghanathe binding constraintsIntegrated Financial Management Information System (GFIMIS).training, TA, diagnostic onthe PFM• The reforms were centred in the Controller and Accountant General’s Department (CAGD) and towork.) productiona lesser extent the Budget Division in MoFEP. Roll out involved initially the central government• possibility frontier?Focal areas for reforms by• function (based on PFMHow has this variedministries, departments and agencies, and additionally for IPPD2 the sub-vented agencies.between different reform“clusters” as in Andrewscomponents?BPEmS2010)• Focal areas for reforms by • The original estimated of BPEMS was USD 11.95 million to be financed under a PUFMARPc. outcomes & overall assessment of PfM reform and of donor support to PfM reformthrough an IDA credit. The final expenditure spread over the period 1997-2003 was USD 19.68organisational location/million, of which of the overall PFM review)level of governmentC.1 – C.4(Addressed as partUSD 15.28 million was funded by IDA and USD 4.60 million funded by GoG. Thisrepresented a 50% cost overrun compared with the original estimates.(Ministry of Finance,• The reform inputs covered four main areas:Sector ministries, Local– Requirement definition and design – technical assistance for re-engineering of PFMGovernments, Parliament,processes and procedures and the functional design of the system;CSOs, etc.)– Technical infrastructure – purchase of hardware systems, installation of LANs in at 56 MDAsites and linking of 18 sites over a wireless loop.– Financial management software – purchase, customisation and installation of the FMISsoftware. The system chosen was Oracle Financials with six modules purchased: (i)generalledger; (ii)purchase order; (iii)accounts payable; (iv)cash management; (v)public sectorbudgeting; and (vi)accounts receivable. Because Oracle Financials had been designed for thecorporate market and had not been widely used in a government environment, extensivecustomisation was required.– Training – inputs for training of staff in MoF, CGAD and pilot MDAs. Delays in the developmentof the BPEMS system meant that the training inputs were less than originally envisaged.– At the completion of PUFMARP, MoFEP decided to continue the implementation of the BPEMSusing its own resources. The annual cost of the Oracle licence and support for the applicationis around USD 180,000.
89
90
PfM Reform component/initiative: Medium-term Expenditure frameworkFindings• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms tostrategic budgeting and budget preparation processes. The objective in introducing the MTEFwas “to improve the planning and budgeting of public expenditures and thus contribute tostrengthened fiscal policy formulation and implementation.”67.• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which wasprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARPwas estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.• The reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly astechnical assistance. They involved two main areas of support:– The development and specification of the MTEF reform covering: (i)formulation of theprogram categories for planning, budgeting and accounting; (ii)preparing the expenditureprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating aid financed projects and programs and preparing computer software for thispurpose.– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelbudget circular that included budget ceilings; (ii)the installation of budget software in theMDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and theinformation system in its preparation.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further rdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the MTEF reform and related training activities.
Evaluation Question
Judgement criteria/Possible indicators
a inputs & context: the design of PfM reform
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?
• Government funds forPFM reforms committedand actually disbursed byyear over the evaluationperiod.• Donor funds for PFMreforms committed andactually disbursed by yearover the evaluation period• Nature of supportprovided to PFM reformefforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)
There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for67 World Bank (October 1996), Republic of Ghana Public Financial Management Technical Assistance Report: Technical Annex,the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’World Bank Report No T-6977-GH.policy areas was close to US$1.5 million. This figure related to operating expenses and capitalexpenditure only and did not include personnel expenses and allowances which are not brokendown by policy area.• The PUFMARP appraisal document provided only a very general specification of the MTEFreform that emphasised the more technical aspects of the reform (formulating programcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetsoftware). There was little mention of the more strategic elements of an MTEF reform coveringthe development of the medium-term macro-fiscal strategy and framework, and the review ofsector programs and priorities and their implications for the expenditure prioritisation andbudget planning. In practice, the detailed design of the reform was undertaken following projectstart-up with the arrival of the DFID funded consultants in September 1997. However, its
A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-ordinated and harmo-
• Structure of design teamand related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system(Sector ministries, LGs
Evaluation Question•
efforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)Judgement criteria/Possible indicatorsprogram categories for planning, budgeting and accounting; (ii)preparing the expenditureprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating aid financed projects and programs and preparing computer software for thispurpose.– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelbudget circular that included budget ceilings; (ii)the installation of budget software in theMDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and theinformation system in its preparation.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further rdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF,Findingsto support the further development of the MTEF reform and related training activities.
a inputs & context: the design of PfM reform
• • •
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-ordinated and harmo-nised delivery frame-work?•
• •
• Government funds forPFM reforms committedand actually disbursed byyear over the evaluationperiod.• Structure of design team• Donor funds for PFMreforms committed andand related consultationactually disbursed by yearprocess: Balance of inputsover the evaluation periodby Donors/ Govt/• Nature of support ofConsultants; Extentprovided to PFM reforminput/ consultation withefforts (equipment,end users of PFM systemtraining, TA, diagnostic(Sector ministries, LGswork.)and service institutions).• Management & co-• Focal areas for reforms byfunction (based on PFMordination structure for“clusters” as in AndrewsPFM reforms (Ad hoc2010) Units vs. NormalProject• Focal areas for reforms bymanagement structures;organisational location/Govt-controlled vs. Sharedlevel of governmentDonor-Govt management;(Ministry of Finance,Use of consultants forSector ministries, Localmanagerial or purelyGovernments, Parliament,advisory roles)Arrangements for• CSOs, etc.)monitoring & evaluation.• Level of harmonisationand alignment of differentdonor contributions.• • •
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
91
A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-
• Structure of design teamand related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system
There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation was initiated in 1997 under theGhana’s medium-term expenditure framework (MTEF) initiative for the Budget Division relatingto the improved public expenditure management’ area and ‘human resources reforms toPublic‘ Finance Management Reform Program (PUFMARP). The MTEF involved development’policy areas was close to US$1.5 million. This figure related to operating expenses the MTEFstrategic budgeting and budget preparation processes. The objective in introducingand capitalexpenditure only and did not include personnel expenses and allowances which are not brokenwas “to improve the planning and budgeting of public expenditures and thus contribute todown by policy area.strengthened fiscal policy formulation and implementation.”67.The funding requirement document provided only appraisal was USD 4.6 million which wasThe PUFMARP appraisal estimated at PUFMARP a very general specification of the MTEFprovided as emphasised the more technical aspects of the reform (formulating programreform that project co-financing by DFID. The actual financing provided by the end of PUFMARPwas estimated at USD 4.58 millionand pricefinancing and USD 0.18 million of GoG financing.categories, expenditure profiling, of donor basis of forecasts, model budget circular, and budgetThe reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly assoftware). There was little mention of the more strategic elements of an MTEF reform coveringtechnical assistance. They involved two main areas of support: framework, and the review ofthe development of the medium-term macro-fiscal strategy and– The development and specification of the MTEF reformthe expenditure prioritisation andsector programs and priorities and their implications for covering: (i)formulation of theprogram categories for planning, budgeting and accounting; (ii)preparing the expenditurebudget planning. In practice, the detailed design of the reform was undertaken following projectprofile for the arrival of the DFID funded consultants in September 1997. However, itsstart-up witheach MDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating aid financed that of the appraisal document.emphasis continued to reflect projects and programs and preparing computer software for thispurpose.The support for the introduction of the MTEF effectively operated as a separate sub-project– Improvements to the formulation of the annual budget covering: (i)the preparation of aDFID,within PUFMARP under an MTEF Project Unit. Because funding was provided directly by modelbudget component operated budget ceilings; (ii)the installation of budget software in thethe MTEF circular that includedrelatively independently of the components managed by theMDAs and (iii)the training of budget and finance Project Unit subsequently became and thePUFMARP project implementation unit. The MTEFstaff in the preparation of the MTEFthe Budgetinformation system has preparation.Development Unit thatin itscontinued to be responsible for overseeing the preparation of theAt the completion of the PUFMARP, MoFEP continued the implementation and furtherMTEF/budget and further development of the reform.development of the MTEF reforms. A number of specific improvements were introduced aimedWhile the management arrangements for the MTEF facilitated strong ownership of the reformsat improving the policy focus and transparency of the budget process, while further r betweenwithin the Budget Division they may have undermined coordination and harmonisationdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF and other components of PUFMARP. For example although BPEMS was specified tothe MTEFbudget preparation module,introduced under PUFMARP remained largely unchanged.include a reform and the procedures this was never used and instead the Budget DivisionVery limited short-term assistance was provided by donors after 2003. This included BMZ and developed its own ACTIVATE software application for budget preparation and subsequently for UNICEF, to support the further development of the MTEF reform and related training activities.managing budget releases. The delays in implementation and roll-out of BPEMS meant that theThere is no data on actual budgetary spending on the MTEF following the completion ofperformance segment of the Chart of Accounts that was to support the MTEF budgeting reformsPUFMARP. However, in the result although ministries have been preparing their budgets usingwas not implemented. As a 2007 Budget the spending allocation for the Budget Division relatingto objective, output and activity classification since 1999 it has remained impossible to monitoranthe ‘ improved public expenditure management’ area and ‘human resources development’policy areas for expenditure against this classification.and account was close to US$1.5 million. This figure related to operating expenses and capitalexpenditure only and did not include personnel expenses and allowances which are not brokendown by policy area.Monitoring of the implementation of the MTEF reforms was initially undertaken within theframework for monitoring and supervision of onlywider PUFMARP project. The experience fromThe PUFMARP appraisal document provided the a very general specification of the MTEFthe latest MTEF/Budget cycle was technical aspectsinto the preparation of the following cycle.reform that emphasised the more supposed to feed of the reform (formulating programThe 2009 External Review of Public Financial Management (ERPFM) contained an assessment ofcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetthe role MTEF reformlittle mention greater effectivenesselements of anin public spending68.software). There was in promoting of the more strategic and efficiency MTEF reform coveringGhana was also case study in the 2003macro-fiscal undertaken by the Overseas Developmentthe development of the medium-term assessment strategy and framework, and the review ofInstitute of the designpriorities and theirthe MTEF as tool the poverty reduction69.sector programs and and application of implications for for expenditure prioritisation andbudget planning. In practice, the detailed design of the reform was undertaken following project
92
Evaluation Question
Donor-Govt management;Use of consultants formanagerial or purelyadvisory roles)• Arrangements formonitoring & evaluation.• Level of harmonisationand alignment of differentdonor contributions.Judgement criteria/Possible indicatorsANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• While the management arrangements for the MTEF facilitated strong ownership of the reformswithin the Budget Division they may have undermined coordination and harmonisation betweenthe MTEF and other components of PUFMARP. For example although BPEMS was specified toinclude a budget preparation module, this was never used and instead the Budget Divisiondeveloped its own ACTIVATE software application for budget preparation and subsequently for managing budget releases. The delays in implementation and roll-out of BPEMS meant that theperformance segment of the Chart of Accounts that was to support the MTEF budgeting reformswas not implemented. As a result although ministries have been preparing their budgets usingan objective, output and activity classification since 1999 it has remained impossible to monitorand account for expenditure against this classification.Findings
• Monitoring of the implementation of the MTEF reforms was initially undertaken within theframework for monitoring and supervision of the wider initiative was initiated in 1997 under the• Ghana’s medium-term expenditure framework (MTEF) PUFMARP project. The experience from• Government funds forA.1.What has been the na-the latest MTEF/Budget cycle was supposed to feed into the The MTEF involved reforms cycle.Public Finance Management Reform Program (PUFMARP). preparation of the following toPFM reforms committedture and scale of PFmThe 2009 External Review of Public Financial Management objective in introducing the MTEFstrategic budgeting and budget preparation processes. The(ERPFM) contained an assessment ofand actually disbursed byreform inputs providedthe “to MTEF reform in promoting greater effectiveness and efficiency in public spending68wasroleimprove the planning and budgeting of public expenditures and thus contribute to .year over the evaluationby Government and Do-Ghana was also case study in the 2003 and implementation.”67.strengthened fiscal policy formulation assessment undertaken by the Overseas Developmentperiod.nors?69Institute of the design and application PUFMARP as tool for poverty 4.6 million• Donor funds for PFM• The funding requirement estimated atof the MTEFappraisal was USDreduction .which was• reforms committed andUse of country systems:A.3. What types of comple-• provided as project co-financing by DFID. The actual financing provided by the end of PUFMARPThe external aid provided by DFID and BMZ was provided as direct technical assistance and theactually disbursed by yearwas estimated atnot recorded in the government budget. USD 0.18 million of GoG financing.wereextent to which externalmentary actions haveexpenditure was USD 4.58 million of donor financing and During the period 2006-09, thereover the evaluation period • The reform inputs, which were focusedreforms. TheMoFEPto which thewere provided mainly asaid is ‘on plan’, ‘on budget’Donors taken to supportno MDBS triggers relating to the MTEF primarily on extent and MDAs, MDBS policy dialogue• Nature of supporttechnical assistance. They involved two MTEFareas of support:and ‘on treasury’?PFm reforms and whatinfluenced the later development of the main reforms is no clear.providedof donorreform toto PFM efforts– The development and specification of the MTEF reform covering: (i)formulation of the• Rankinghas been their signifi-efforts (equipment,program categories for planning, budgeting and accounting; (ii)preparing the expendituresupport national systemscance? Have they hadtraining, TA, diagnosticprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;as expressed in Parisany influence on the ex-work.)(iv)integrating aid financed projects and programs and preparing computer software for thisDeclaration monitoringternal constraints to re-• reports and PEFAFocal areas for reforms bypurpose.form?function (based on PFM– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelindicators D2 and D3.budget circular that included budget ceilings; (ii)the installation of budget software in the• “clusters”evolution overScale and as in Andrews2010)MDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and thethe evaluation period of aid• Focal provided as GBS, byinformation system in its preparation.flows areas for reformsorganisational location/• At the completion of the PUFMARP, MoFEP continued the implementation and furtherSBS or debt relief andlevel of government todevelopment of the MTEF reforms. A number of specific improvements were introduced aimedrelative contribution(Ministry of Finance,at improving the policy focus and transparency of the budget process, while further rreduction of treasurySector ministries, Localtodevelopment of the ACTIVATE software application took place. However, the main elements of management costs andGovernments, Parliament,the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.discretionary resourcesCSOs, etc.) Government,• Very limited short-term assistance was provided by donors after 2003. This included BMZ and available toUNICEF, to support the further development of the MTEF reform and related training activities.(including effect onThere is no data on actual budgetary spending on the MTEF following the completion offinancing constraints forPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’PFM reform outputs).68 World Bank, Danida, DFID, KFW (May 2009), Ghana, 2009policy areas was close to US$1.5 million. This figure– Volumeoperating expenses and capitalExternal Review of Public Financial Managementrelated toII, The Medium-Term• Nature and evolution overexpenditure only and did not include personnel expenses and allowances which are not brokentime ofReporton47639-GH.Expenditure Framework, World BankdialogueNo.PFMdown by policy area.69 Short, J. (May 2003), Countrywithin GBS/PRSC, SBSof the MTEF in Ghana, Overseas Development Institute, London.Case Study 4: Assessment• The PUFMARP appraisal document provided only a very general specification of the MTEF• structures or similarStructure of design teamA. 2.What types of struc-donor-govt fora, includingreform that emphasised the more technical aspects of the reform (formulating programand related consultationtures have been useduse of PFM reform inputscategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetprocess: Balance offor the design and man-conditions in GBS/ SBSsoftware). There was little mention of the more strategic elements of an MTEF reform coveringby Donors/ Govt/agement of these re-disbursement conditions.the development of the medium-term macro-fiscal strategy and framework, and the review ofConsultants; Extent ofform inputs? Have• Influence of dialogue onsector programs and priorities and their implications for the expenditure prioritisation andinput/ consultation withthese structuresthe “political” andsystembudget planning. In practice, the detailed design of the reform was undertaken following projectend users of PFM “policyserved to provide a co-space” constraints LGsstart-up with the arrival of the DFID funded consultants in September 1997. However, its(Sector ministries,ordinated and harmo-restricting choice of PFMemphasis continued to reflect that of the appraisal document.and service institutions).nised delivery frame-• The support for the introduction of the MTEF effectively operated as a separate sub-project• reform outputs.co-Management &work?• (Consistency of actualwithin PUFMARP under an MTEF Project Unit. Because funding was provided directly by DFID,ordination structure forchanges with PFM reform
a inputs & context: the design of PfM reform
any influence on the ex-ternal constraints to re-form?
Evaluation Question
as expressed in ParisDeclaration monitoringreports and PEFAindicators D2 and D3.• Scale and evolution overthe evaluation period of aidflows provided as GBS,SBS or debt relief andrelative contribution toreduction of treasurymanagement costs and todiscretionary resourcesavailable to Government,(including effect onJudgement criteria/financing constraints forPossible indicatorsFindings
a inputs & context: the design of PfM reformoutputs).PFM reform• Government funds for• Nature and evolution overA.1.What has been the na-time reforms committedPFM of dialogue on PFMture and scale of PFmwithin GBS/PRSC, SBSand actually disbursed byreform inputs providedstructures or similaryear over the evaluationby Government and Do-donor-govt fora, includingperiod.nors?• use of PFM reformDonor funds for PFMconditions in GBS/ SBSreforms committed anddisbursement conditions.actually disbursed by year• over the evaluation periodInfluence of dialogue onthe “political” and• Nature of support “policyspace” constraintsprovided to PFM reformrestricting choice ofefforts (equipment, PFMreform outputs.training, TA, diagnostic• work.)(Consistency of actual• changes with PFM reformFocal areas for reforms bytriggers (based on PFMfunction and/or points ofemphasis as in Andrews“clusters”in GBS/ PRSCdialogue)2010)• Focal areas forof counter-• Consideration reforms byfactual:organisational location/• level of governmentOutputs: would PFMreforms have evolved(Ministry of Finance,differently in absence ofSector ministries, LocalGBS and related policyGovernments, Parliament,dialogue structures?CSOs, etc.)• Outcomes: Wouldintermediate outcomeshave been different if useof country systems hadbeen lower?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
93
A.4. to what extent hasthere been domesticA. 2.publictypes of struc-What pressure or re-gionalhave been usedtures institutionalpressure in support offor the design and man-PFm reformthese re-agement of and whathas been the influenceform inputs? Haveon the structuresthese external con-straintsto provide a co-served to reform?
• Nature of engagementwith PFM reform issues• within civilof design teamStructure society,academia and the media.and related consultation• Extent of Balance of inputsprocess: participation ofCSOs in policy dialogue onby Donors/ Govt/PFM reform. Extent ofConsultants;• Presence/ absencewithinput/ consultation ofspecific research andend users of PFM system
• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms tostrategic budgeting and budget preparation processes. The objective in introducing the MTEFwas “to improve the planning and budgeting of public expenditures and thus contribute tostrengthened fiscal policy formulation and implementation.”67.• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which wasprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARPwas estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.• The reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly astechnical assistance. They involved two main areas of support:– The development and specification of the MTEF reform covering: (i)formulation of theprogram categories for planning, budgeting and accounting; (ii)preparing the expenditureprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating aid financed projects and programs and preparing computer software for thispurpose.– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelbudget circular that included budget ceilings; (ii)the installation of budget software in theMDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and theinformation system in its preparation.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further rdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the MTEF reform and related training activities.There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’policy areas was close to US$1.5 million. This figure related to operating expenses and capitalexpenditure only and not involved specific measures to involve civil society, academia and the• The MTEF reform hasdid not include personnel expenses and allowances which are not brokendown in the MTEF andmediaby policy area. budget process. MoFEP organises consultations with CSOs in Octobereach year prior appraisal document provided only a very general specificationtakes place too late• The PUFMARP to the presentation of the budget in Parliament. However, this of the MTEFto influence emphasised the more technical aspects of the reform (formulating programreform that budget allocations.• The MTEF reform has not resulted in aprice basis of forecasts, model budget circular, and budgetcategories, expenditure profiling, and significant increase in engagement with Parliament in thebudget process. Unlike South Africa and Uganda, the MTEF has notof an MTEF reform coveringsoftware). There was little mention of the more strategic elements involved the preparation pre-budget report that is presented to Parliament setting out the and framework, and the review ofthe development of the medium-term macro-fiscal strategy framework, policies and prioritieswithin which the forthcoming budget is toimplications for the expenditureno mid-year budgetsector programs and priorities and their be prepared. Similarly there is prioritisation andimplementation review that is presented to Parliament.budget planning. In practice, the detailed design of the reform was undertaken following project
94
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
Evaluation Question• The MTEF reform has not involved specific measures to involve civil society, academia and themedia in the MTEF and budget process. MoFEP organises consultations with CSOs in October• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under theeach year prior to the presentation of the budget in Parliament. However, this takes place too latePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms toto influence budget allocations.strategic budgeting and budget preparation processes. The objective in introducing the MTEF• The MTEF reform has not resulted in a significant increase in engagement with Parliament in thewas “to improve the planning and budgeting of public expenditures and thus contribute tobudget process. Unlike South Africa and Uganda, the MTEF has not involved the preparation pre-strengthened fiscal policy formulation and implementation.”67.budget report that is presented to Parliament setting out the framework, policies and priorities• The funding requirement estimated at PUFMARP appraisal was USD 4.6 no mid-year budgetwithin which the forthcoming budget is to be prepared. Similarly there is million which wasprovided as project co-financing by DFID.to Parliament.implementation review that is presented The actual financing provided by the end of PUFMARPwas estimated at USD 4.58 million of donor budget analysis and advocacy. There is minimal• There are relatively few CSOs undertaking financing and USD 0.18 million of GoG financing.• The reformfiscal and budgetaryfocusedundertaken by the academic community. Non-availabilityanalysis of inputs, which were issues primarily on MoFEP and MDAs, were provided mainly astechnical assistance. They involveda major constraint support:of budget execution data is seen as two main areas of to independent budget analysis. The– The development and specification of the MTEF reform covering: (i)formulation of the GhanaIntegrated Social Development Centre (ISODEC) has been responsible for preparing theprogram categories for planning, budgeting and Budget Survey that is prepared by thereport that is part of the biennial international Openaccounting; (ii)preparing the expenditureprofile for each International Budget Partnership.Washington basedMDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating that regional PFM fora are now influencing the PFM reform agenda. A for this• There is evidenceaid financed projects and programs and preparing computer software meetingpurpose.of the Collaborative Africa Budget Reform Initiative (CABRI) in the second quarter of 2010 was– Improvements to the formulation of Budget Division that activity basedpreparation of a modelinfluential in convincing officials in the the annual budget covering: (i)the budgeting needed to bebudget circular that included budget ceilings; (ii)the installation of budget software in thereplaced by a more strategic program based approach, and that Ghana should access relevantMDAs and (iii)the training of budget Africa and staff in theexperience with such reforms in Southand financeMauritius. preparation of the MTEF and theinformation system in its preparation.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further rdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the MTEF reform and related training activities.There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’policy areas was close to US$1.5 million. This figure related to operating expenses and capitalexpenditure only and did not include personnel expenses and allowances which are not brokendown by policy area.• The PUFMARP appraisal document provided only a very general specification of the MTEFreform that emphasised the more technical aspects of the reform (formulating programcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetsoftware). There was little mention of the more strategic elements of an MTEF reform coveringthe development of the medium-term macro-fiscal strategy and framework, and the review ofsector programs and priorities and their implications for the expenditure prioritisation andbudget planning. In practice, the detailed design of the reform was undertaken following projectstart-up with the arrival of the DFID funded consultants in September 1997. However, itsemphasis continued to reflect that of the appraisal document.• Initially, there was strong political level support for the more strategic objective-led approach to• The support for the introduction of the MTEF effectively operated as a separate sub-projectbudgeting that was to be introduced through the MTEF, with the Minister and Deputy Minister forwithin PUFMARP under an MTEF Project Unit. Because funding was provided directly by DFID,Finance involved in opening workshops organised to inform MDAs about the reforms and the
• Outputs: would PFMreforms have evolveddifferently in absence ofGBS and related policydialogue structures?• Outcomes: Wouldintermediate outcomeshave been different if useJudgement criteria/hadof country systemsPossible indicatorsbeen lower?Findings
a inputs & context: the design of PfM reform• Nature of engagementA.4. to what extent haswith PFM reform issuesthere been domestic• Government funds forA.1.What has been the na-within civil society,public pressure or re-PFM reforms committedture and scale of PFmacademia and the media.gional institutionalreform inputs provided• and actually disbursedofExtent of participation bypressure in support ofyear over the evaluationonby Government and Do-CSOs in policy dialoguePFm reform and whatperiod.nors?PFM reform.has been the influence• Presence/ absence of• Donor funds for PFMon the external con-reformsresearch andspecific committed andstraints to reform?actually disbursedon PFMadvocacy by CSOs by yearover the evaluation periodreform issues.• Nature of support issues• Relevance of PFMprovided tooutcomes:to political PFM reformefforts (equipment,• Discussion of PFM issuestraining, TA, diagnosticwithin election campaigns• work.)Discussion of service• delivery issues related toFocal areas for reforms byfunction (based campaignsPFM in election on PFM“clusters” as in Andrews• Voter concerns during 2010)elections (from Afro• Focal areasand reforms byBarometer for otherorganisational location/relevant publications)Existence of regional• level of government or(Ministry of Finance, (e.g.international bodiesSector ministries, LocalWAEMU, EITI) promotingGovernments, Parliament,specific norms on PFMCSOs, etc.) their relativeissues, andinfluence on domesticpolitical discourse.• Evolution in the quality ofpublic/ CSO engagementwith PFM issues over timeand its influence on PFMreform outputs.• Structure of design teamA. 2.What types of struc-• Consideration of counter-and related consultationtures have been usedfactual: would PFM reformprocess: Balance of inputsfor the design and man-outputs have evolvedby Donors/ Govt/agement of these re-differently in absence ofConsultants; Extent ofform inputs? Havedomestic public pressureinput/ consultation withthese structuresor pressure from regional/end users of PFM systemserved to provide a co-international institutions?(Sector ministries, LGsordinated and harmo-and service institutions).nised delivery frame-• Scale and focus of supportA.5. How relevant was the• Management & co-work?in relation to identifiedPFm reform pro-ordination structure forPFM weaknesses at startgramme to the needs
Evaluation Question
Barometer and otherrelevant publications)• Existence of regional orinternational bodies (e.g.WAEMU, EITI) promotingspecific norms on PFMissues, and their relativeinfluence on domesticpolitical discourse.• Evolution in the quality ofpublic/ CSO engagementwith PFM issues over timeand its influence onJudgement criteria/PFMreform outputs.Possible indicatorsFindings
a inputs & context: the design of PfM reformof counter-• Considerationfactual: would PFM reform• Government funds forA.1.What has been the na-outputs have evolvedPFM reformsabsence ofture and scale of PFmdifferently in committedand actually disbursed byreform inputs provideddomestic public pressureyear over the evaluationby Government and Do-or pressure from regional/period.nors?international institutions?• Donor funds for PFM• reforms committed andScale and focus of supportA.5. How relevant was theactually disbursed by yearin relation to identifiedPFm reform pro-over the evaluation periodPFM weaknesses at startgramme to the needs• Nature of supportand duringand the institutionalprovided to PFM ofimplementation reformcontext? Was donorefforts (equipment,reforms (includingsupport consistent withtraining, TA, diagnosticweaknesses in HRnational priorities? towork.)endowments, quality ofwhat extent were adap-• institutions & rules,Focal areas for reforms bytations made in re-function (based on &quality of systems PFMsponse to the context“clusters” as in Andrewsbusiness processes, andand the changing na-2010) of organisations).qualitytional priorities?• Focal areas fordonor- byConsistency of reformsorganisational location/funded support withlevel of governmentGovernment PFM reform(Ministry of Finance,efforts, with ongoingSectoradministrationpublic ministries, LocalGovernments, Parliament,reforms and overall GovtCSOs, etc.)policies.• Extent to which scope andfocus of support were
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
95
• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms tostrategic budgeting and budget preparation processes. The objective in introducing the MTEFwas “to improve the planning and budgeting of public expenditures and thus contribute tostrengthened fiscal policy formulation and implementation.”67.• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which was• provided as project co-financing by DFID. The actual financing provided by the end ofapproach toInitially, there was strong political level support for the more strategic objective-led PUFMARPwas estimated was to be introduced through the MTEF, with the Minister and GoG financing. forbudgeting that at USD 4.58 million of donor financing and USD 0.18 million of Deputy Minister• The reform inputs, which were focused primarily on MoFEP MDAs about the reforms and the asFinance involved in opening workshops organised to inform and MDAs, were provided mainlytechnicalto the budget process. Thetwo mainwas also support:involved in chairing meetings onchanges assistance. They involved Minister areas of actively– The development and specification of the MTEF reform covering: (i)formulation of thethe MTEF reforms within MoFEP. In recent years the focus has shifted away from the MTEFprogram categories for planning, budgeting and accounting; (ii)preparing the expenditurebudgeting reforms towards improving the presentation of the government’s fiscal and budgetprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;policies and strategies in the Budget Statement.(iv)integrating of financed assistance facilitated a very rapid roll out of the software for this• The front-loadingaidthe donorprojects and programs and preparing computerMTEF reformspurpose.that was initially seen as having very impressive results. The first review of the MTEF held in– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelMarch 1999 noted that “so far what has been achieved has been extraordinary....the first yearbudget circular that change, budget ceilings; (ii)the installation of budget in such a shortbudget has produced aincluded which has not been seen in any other country,software in theMDAs . However, training of budget and finance staff in the preparation of the MTEF andperiod”70and (iii)theat conclusion of PUFMARP, the MTEF reform component was rated theinformation system in its and the Implementation Completion Report noting that “the attempt‘moderately unsatisfactory’preparation.• to put all elements of an PUFMARP, MoFEP continued the became a single and further inAt the completion of the MTEF in place in what effectively implementation cycle resulteddevelopment in a number of key respects”71. It might have been more appropriate for a moresuperficiality of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focusbeentransparency of the budget process, while for donor funding tomeasured approach to have and taken to the introduction of the MTEF, and further rdevelopment of the ACTIVATE software application took place. However, the main elements of have been extended over a longer period.• the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.Within the context PUFMARP there was also a lack of complementarities in the donor support• Very limited short-term assistance was provided by donors after 2003. This included BMZ and that was provided. Thus while the World Bank was funding the BPEMS which included a budgetUNICEF, to support the further development of the MTEF reform and related training activities.preparation module, DFID was separately funding the ACTIVATE budget preparation application There is no data on to the budgetary spending on the MTEFthrough its supportactual MTEF component of PUFMARP. following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’adapted to the context,policy areas was close to US$1.5 million. This figure related to operating expenses and capitalespecially to the level of70 Short, J. (May 2003) p17.ownership and the71 World Bank (May 2004), Implementation Completion Reportexpenditureto the Republicinclude personnel expenses and allowances which are not brokenon a Creditonly and did notof Ghana for a Public Finance Management Technical Assistancedown by policy area.capacity for reformProject, World Bank Report No. 28089-GH, p8.implementation.• The PUFMARP appraisal document provided only a very general specification of the MTEF• Structure of & coherenceA. 2.What types of struc-• Consistency design teamreform that emphasised the more technical aspects of the reform (formulating programand related consultationtures have been usedwith wider Donor policiescategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetprocess: Balance of inputsfor the design and man-in countrysoftware). There was little mention of the more strategic elements of an MTEF reform coveringby Donors/ Govt/ the Parisagement of these re-• Consistency withthe development of the medium-term macro-fiscal strategy and framework, and the review ofConsultants; Extent of asform inputs? HaveDeclaration objectivessector programs and priorities and their implications for the expenditure prioritisation andinput/ consultation withthese structuresregards ownership,budget planning. In practice, the detailed design of the reform was undertaken following projectend users of PFM systemserved to provide a co-
96
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
Evaluation Question• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms tostrategic budgeting and budget preparation processes. The objective in introducing the MTEFwas “to improve the planning and budgeting of public expenditures and thus contribute tostrengthened fiscal policy formulation and implementation.”67.• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which wasprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARPwas estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.• The reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly astechnical assistance. They involved two main areas of support:– The development and specification of the MTEF reform covering: (i)formulation of theprogram categories for planning, budgeting and accounting; (ii)preparing the expenditureprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;(iv)integrating aid financed projects and programs and preparing computer software for thispurpose.– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelbudget circular that included budget ceilings; (ii)the installation of budget software in theMDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and theinformation system in its preparation.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further r• development of the ACTIVATE software application took place. However, the main elements of Assessments of the outputs of the MTEF reforms are provided by the 2004 PUFMARPimplementation completion report which focuses on the progress remained largely MTEFthe MTEF reform and the procedures introduced under PUFMARPin establishing theunchanged.process and the 2009 EPFM which assesses how the MTEF reforms are operating. The different• Very limited short-term assistance was provided by donors after 2003. This included BMZ and focus of these two assessments reflects the respective stage in the institutionalisation of theUNICEF, to support the further development of the MTEF reform and related training activities.MTEF reforms. on actual budgetary spending on the MTEF following the completion ofThere is no data• The PUFMARP implementation completion report assessed the for the Budgethad beenrelatingPUFMARP. However, in the 2007 Budget the spending allocation progress that Division madewith MTEF reforms by 2003 in termsmanagement’ area and ‘human resources development’to the ‘ improved public expenditure of five outputs:– Formulation ofclose to US$1.5 million. This figure related to operating expenses and capitalofpolicy areas was the annual budget within a complete rolling MTEF. While the main elementsan MTEF had been did in place their implementation had been variable. A number ofexpenditure only and put not include personnel expenses and allowances which are not brokenweaknesses in thedown by policy area. process were identified First, insufficient attention had been given to theupstream tasks of macro-fiscal fiscal planning and forecasting and setting realistic resource• The PUFMARP appraisal document provided only a very general specification of the MTEFreform that emphasised the more technical aspects of the reform made through the MTEFceilings for budget preparation. The ICR noted that “allocations (formulating programcategories,were largely irrelevantand price basis of forecasts, model budget circular, and budgetprocess expenditure profiling, as resources actually available to sector managementsoftware). There was little mention of the more strategic elementscash releases”. Second, theredepended on unpredictable and opaque in-year decisions about of an MTEF reform coveringthe development of thereview process macro-fiscal strategy and framework, and the reviewofwas no expenditure medium-term in place either at the strategic level involving analysis ofsector programs and priorities and their implications for the expenditure prioritisation andexpenditure patterns, issues and priorities, or in assessing on-going spending programbudget planning. In practice, the detailed design of the reform was treated as a ministry overheadperformance and achievements. Third, staff related costs were undertaken following projectstart-up with the arrival of the DFID funded consultants in September 1997. However,meant thatexpense and excluded from the MTEF budget process. In the education sector this itsemphasis 75% of expenditure fell outsideappraisal document. procedures. Fourth, thearound continued to reflect that of the of the MTEF budgeting• The support forof activity based budgeting ran ahead of the accounting reforms sub-projectintroduction the introduction of the MTEF effectively operated as a separate with the resultwithin PUFMARP under an MTEF Project Unit. Because funding wasagainst the new MTEFthat expenditures were not recorded in the government accounts provided directly by DFID,the MTEF classification, making it impossible to undertakethe components managed by thebudget component operated relatively independently of meaningful expenditure analysis.
• Consistency of donor-funded support withGovernment PFM reformefforts, with ongoingpublic administrationreforms and overall Govtpolicies.• Extent to which scope andJudgement criteria/focus of support werePossible indicators‘moderately unsatisfactory’ and the Implementation Completion Report noting that “the attemptto put all elements of an MTEF in place in what effectively became a single cycle resulted insuperficiality in a number of key respects”71. It might have been more appropriate for a moremeasured approach to have been taken to the introduction of the MTEF, and for donor funding tohave been extended over a longer period.• Within the context PUFMARP there was also a lack of complementarities in the donor supportthat was provided. Thus while the World Bank was funding the BPEMS which included a budgetpreparation module, DFID was separately funding the ACTIVATE budget preparation application through its support to the MTEF component of PUFMARP.Findings
a inputs & context: the design of PfM reformthe context,adapted toespecially to the level of• Government funds forA.1.What has been the na-ownership and thePFM reforms committedture and scale of PFmcapacity for reformand actually disbursed byreform inputs providedimplementation.by Government and Do-• year over the& coherenceConsistency evaluationperiod.nors?with wider Donor policies• in country for PFMDonor fundsreforms committed and• Consistency with the Parisactually disbursed by yearDeclaration objectives asover the ownership,periodregards evaluation• Nature of supportharmonisation andprovided to PFM reformalignment.• efforts (equipment, ofFlexibility in designtraining,and existence ofsupport TA, diagnosticwork.) feed-back/adequate• learning mechanisms in byFocal areas for reformsfunctionadapt the designorder to (based on PFM“clusters” as in Andrewsand implementation of2010) to changingsupport• Focal areas for reforms byneeds.organisational location/leveloutputs: the delivery of PfM reformsof government(Ministry of Finance,A.6. What have been the out-Documentation of outputs :Sector ministries, Local• Governments, Parliament,by PFM functionputs of the PFm reform(accordingprocess and to what ex-CSOs, etc.) to PFM clustersin Andrews 2010) andtent has direct donor• by type of output (People &support contributed toSkills; Laws & Rules;these outputs?Systems & Businessprocesses; Organisationalchanges)• Matching of outputs toPFM reform inputs,• Structure of design teamA. 2.What types of struc-and related consultationtures have been useddifferentiating if possibleprocess: Balance of inputsfor the design and man-Government & Donorby Donors/ Govt/agement of these re-inputs.Consultants; Extent ofform inputs? Have• Analysis of chronology ofinput/ to determinethese structureseventsconsultation withend users of PFM systemserved to provide a co-causal links between(Sector ministries, LGsordinated and harmo-donor support to reformservice institutions).nised delivery frame-and observed changes at• output level. & co-Managementwork?ordination structure for• Consideration ofPFM reforms (Ad hoccounterfactual: would
needs.
outputs: the delivery of PfM reforms
A.6. What have been the out-puts of the PFm reformprocess and to what ex-tent has direct donorsupport contributed tothese outputs?
Evaluation Question•
Documentation of outputs :• by PFM function(according to PFM clustersin Andrews 2010) and• by type of output (People &Skills; Laws & Rules;Systems & Businessprocesses; Organisationalchanges)• Matching of outputsJudgement criteria/toPFM reform inputs,Possible indicators• Assessments of the outputs of the MTEF reforms are provided by the 2004 PUFMARPimplementation completion report which focuses on the progress in establishing the MTEFprocess and the 2009 EPFM which assesses how the MTEF reforms are operating. The differentfocus of these two assessments reflects the respective stage in the institutionalisation of theMTEF reforms.• The PUFMARP implementation completion report assessed the progress that had been madewith MTEF reforms by 2003 in terms of five outputs:– Formulation of the annual budget within a complete rolling MTEF. While the main elements ofan MTEF had been put in place their implementation had been variable. A number ofweaknesses in the process were identified First, insufficient attention had been given to theupstream tasks of macro-fiscal fiscal planning and forecasting and setting realistic resourceFindings
• •
a inputs & context: the design of PfM reformdifferentiating if possibleGovernment & Donor• Government funds forA.1.What has been the na-inputs.PFM reforms committedture and scale of PFm• and actually disbursed ofAnalysis of chronology byreform inputs providedevents to the evaluationyear over determineby Government and Do-causalperiod.links betweennors?• donor support toPFMDonor funds for reformand observed changes atreforms committed andoutput level.actually disbursed by year• over the evaluation periodConsideration ofcounterfactual: would• Nature of supportthese outputs have beenprovided to PFM reformgenerated in absenceefforts (equipment, ofdonor support to PFMtraining, TA, diagnosticreform?work.)
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)•
ceilings for budget preparation. The ICR noted that “allocations made through the MTEFprocess were largely irrelevant as resources actually available to initiated in 1997 underGhana’s medium-term expenditure framework (MTEF) initiative wassector management thedepended on unpredictable and opaque in-year decisions about cash releases”. Second,Public Finance Management Reform Program (PUFMARP). The MTEF involved reforms to therewas no expenditure review process in place either at The objective in introducing the MTEFstrategic budgeting and budget preparation processes.the strategic level involving analysis ofexpenditure patterns, issues and priorities, public expenditures and spending programwas “to improve the planning and budgeting of or in assessing on-going thus contribute toperformance and achievements. Third, staff related costs were treated as a ministry overheadstrengthened fiscal policy formulation and implementation.”67.expense and excluded estimated at PUFMARP appraisal was USD 4.6 million this meantThe funding requirementfrom the MTEF budget process. In the education sectorwhich was thataround 75% of expenditure fell outside The actual financing provided by the end of theprovided as project co-financing by DFID. of the MTEF budgeting procedures. Fourth,PUFMARPintroduction of activity based budgeting ran ahead and USD 0.18 million of GoG financing.was estimated at USD 4.58 million of donor financing of the accounting reforms with the resultthat expenditures were not recorded in the government accounts against the new MTEFThe reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly asbudget classification, making it impossible to undertake meaningful expenditure analysis.technical assistance. They involved two main areas of support:– The development and specification of the MTEF reform covering: (i)formulation of the– A comprehensive broad based budget. This had only been partially achieved with a significantshare of categories for planning, budgeting and accounting; (ii)preparing the expenditureprogramdonor funded project expenditure and spending financed from internally generatedfunds still not being in line with existing policies; (iii)specifying the price basis for forecasts;profile for each MDAcaptured in the budget.– (iv)integrating aid financed projects and programs and preparing computer software for thisComputerised budget preparation process. The delays in implementing BPEMS meant led tothe Budget Division developing its own budget preparation software application (ACTIVATE) purpose.rather than utilise the BPEMS budget preparation module.– Improvements to the formulation of the annual budget covering: (i)the preparation of a model– budget circularon an MTEF basis. The way in which activity based budgeting had been thePrinted budget that included budget ceilings; (ii)the installation of budget software inimplemented meant that of budget estimates document became excessively detailed withMDAs and (iii)the trainingthebudget and finance staff in the preparation of the MTEF and the aresulting loss of strategic focus.information system in its preparation.– Budget guidelinesthe PUFMARP, MoFEPIt was not clearimplementation and furtherinvolved inAt the completion of approved by Cabinet. continued the to what extent Cabinet wasthe approval the MTEF reforms. A and budget ceilings.development ofof the budget circular number of specific improvements were introduced aimed
97
A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-
• Structure of design teamand related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system
at improving the that although the MTEF budgeting reforms process, while further rThe ICR consideredpolicy focus and transparency of the budget had not achieved their planneddevelopment of the ACTIVATE software application took place. However, the main elements of outputs, it was likely that the progress made would be consolidated and that a fully fledged MTEFthe be achieved.wouldMTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and • In the years following 2003 the MTEF process a number of improvements were introduced intoUNICEF, to support thethe budget has been passedMTEF reform and related trainingthe financialthe budget. Since 2006 further development of the by Parliament before the start of activities.There is no wereon actualstrengthen spending on the MTEF following theof sector policyyear. Steps data taken to budgetary policy linkage with the introduction completion ofPUFMARP. However,budget calendar, and the preparation of a Citizens’ Budget Division relatinghearings early in the in the 2007 Budget the spending allocation for the Guide to the Budgetto the ‘ improved publicand 2008 budgets. The MTEF budgeting reforms were also rolled out toStatement for the 2007 expenditure management’ area and ‘human resources development’policy areas was close to US$1.5 million.2009 ERPFM, manyto operating expenses and capitalpilot MMDAs. However, at the time of the This figure related of the shortcomings that had beenexpenditurethe ICR had not been addressed. expenses and allowances which are not brokenidentified in only and did not include personneldown by policy area.• – There had been improvements in macro-fiscal forecasting with budget planning was based onThe PUFMARP appraisal document provided only a very general specification of the MTEFa reasonably accurate macro-fiscal framework with actual annual revenues within 3%reform that emphasised the more technical aspects of the reform (formulating program ofbudget forecasts during the period price basis of forecasts, model budget circular, and budgetcategories, expenditure profiling, and2005-07. However, this did not translate into a crediblebudget There was little There remained with large differences of an MTEF aggregatesoftware).ceilings process.mention of the more strategic elementsbetween thereform coveringindicative ceiling and the final appropriation contained in the budget (averaging 10.7% over thethe development of the medium-term macro-fiscal strategy and framework, and the review ofperiod 2005-07). Ceilings continued to be provided for the expenditure prioritisationsector programs and priorities and their implicationsto ministries only for services andandinvestment costs, and not the detailed design and allowances. Compliance with ceilings wasbudget planning. In practice,for personnel costs of the reform was undertaken following project
98
Evaluation Question•
Judgement criteria/Possible indicators
The ICR considered that although the MTEF budgeting reforms had not achieved their plannedoutputs, it was likely that the progress made would be consolidated and that a fully fledged MTEFwould be achieved.• In the years following 2003 the MTEF process a number of improvements were introduced intothe budget. Since 2006 the budget has been passed by Parliament before the start of the financialyear. Steps were taken to strengthen policy linkage with the introduction of sector policyhearings early in the budget calendar, and the preparation of a Citizens’ Guide to the BudgetStatement for the 2007 and 2008 budgets. The MTEF budgeting reforms were also rolled out topilot MMDAs. However, at the time of the 2009 ERPFM, many of the shortcomings that had beenidentifiedFindingsin the ICR had not been addressed.
a inputs & context: the design of PfM reform
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?• •
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• Government funds forPFM reforms committedand actually disbursed byyear over the evaluationperiod.• Donor funds for PFMreforms committed andactually disbursed by yearover the evaluation period• Nature of supportprovided to PFM reformefforts (equipment,training, TA, diagnosticwork.)• Focal areas for reforms byfunction (based on PFM“clusters” as in Andrews2010)• Focal areas for reforms byorganisational location/level of government(Ministry of Finance,Sector ministries, LocalGovernments, Parliament,CSOs, etc.)• • •
– There had been improvements in macro-fiscal forecasting with budget planning was based ona reasonably accurate macro-fiscal framework with initiative was revenues 1997 underGhana’s medium-term expenditure framework (MTEF) actual annualinitiated inwithin 3% of thebudget forecasts during the period 2005-07. However, this did not translate into a crediblePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms tobudget ceilings process. There remained processes. The objective in introducing the MTEFstrategic budgeting and budget preparation with large differences between the aggregateindicative ceiling planning and budgeting of contained in the budget thus contribute towas “to improve theand the final appropriation public expenditures and(averaging 10.7% over theperiod 2005-07). policy formulation and provided to ministries only for services andstrengthened fiscalCeilings continued to beimplementation.”67.investment costs, and estimated at PUFMARP appraisal was USD 4.6 million which wasThe funding requirementnot for personnel costs and allowances. Compliance with ceilings wasweak with ministry budget requests for the 2008 budget exceeding the indicative ceilings byprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARP24%.was estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.– Despite the introduction of sector policy hearings,MoFEP and MDAs, were provided mainly asThe reform inputs, which were focused primarily on the MTEF process still did not facilitate aclear link between They involved areas and budget support:technical assistance. priority policy two main areas of allocations. There was no systematicanalysis and overview of the strategic policy assumptions and spending choices across– The development and specification of the MTEF reform covering: (i)formulation of thesectors to support the sector and budgeting and accounting; The very detailed presentationprogram categories for planning, MDA resource allocations. (ii)preparing the expenditure ofMDA budgets means that the strategies policies; (iii)specifying the price basis are not clear.profile for each MDA in line with existingand priorities underlying MDA budgetsfor forecasts;The continued aid financed projects and programs and preparing computer software for(iv)integrating absence of comprehensive and reliable data on budget execution makes it thisdifficult topurpose. undertake the analysis of budgetary trends and performance that should informand guide resource allocation decisions.– Improvements to the formulation of the annual budget covering: (i)the preparation of a model– budget circular that included budget ceilings; to be undermined by shortfalls and delays inThe credibility of MDA budgets had continued (ii)the installation of budget software in thebudget releases against the budget and finance staff factor was weak payroll planning whichMDAs and (iii)the training of planned budget. A majorin the preparation of the MTEF and theresulted in high levels of preparation.information system in itsvariance in MDA expenditure against their budget (which averagedAt 26.5% over the 2003-2007 budgets). Within MDA budgets difference between budgetthe completion of the PUFMARP, MoFEP continued the implementation and furtherallocations and releases were particularly of specific improvements were introduceddevelopment of the MTEF reforms. A number large for services and investment, the two aimedcategories of policy focus and transparency based budgeting). The chart of accountsat improving theexpenditure covered by activity of the budget process, while further r still didnot include codes for the MTEF activity classification with the result that it continued to bedevelopment of the ACTIVATE software application took place. However, the main elements of impossible to report expenditure against MTEF budget allocations.the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.The study team found that situation has changed little subsequent to the 2009 ERPFM. Indeed theVery limited short-term assistance was provided by donors after 2003. This included BMZ and Citizen’s Guide to the Budget and the MTEF policy hearingreform and related training activities.UNICEF, to support the further development of the MTEF had lapsed, and the start of the MTEF/Budget preparation cycle significantly delayed. The guidelines for the the completion of budgetThere is no data on actual budgetary spending on the MTEF following preparation 2011were not issued until September,Budget the spending allocation for the Budget budget ceilings.PUFMARP. However, in the 2007 with ministries no longer being provided with Division relatingHowever, there was increasing recognition of the need for a thorough re-engineering of theto the ‘ improved public expenditure management’ area and ‘human resources development’MTEF process, with the decision taken in 2010 figure related to operating expenses and capitalpolicy areas was close to US$1.5 million. This to replace activity based budgeting with a morestrategic program-based budgeting personnel expenses and allowances which are not brokenexpenditure only and did not include approach.
A.7. How efficiently werethese types of struc-A. 2.What outputs generat-ed? Was thebeen usedtures have pacing andsequencing of reformsfor the design and man-appropriate these re-agement of and cost-effective? Was the costform inputs? Haveper output acceptable?these structures
served to provide a co-ordinated and harmo-nised delivery frame-work?
• Consistency of outputs• produced with plannedStructure of design teamprogrammeconsultationand related (quantity andtiming) Balance of inputsprocess:• Extent of coordinationby Donors/ Govt/between outputs ofConsultants; Extent• Quality of pacing andinput/ consultation withsequencing of outputend users of PFM systemproduction:(Sector ministries, LGs• internally coherent/and service institutions).• efficient?Management & co-• Consistentstructure forordination with capacityendowments?
down by above, the• As notedpolicy area.original specification of the MTEF reforms emphasised the “bottom-up”elements of an MTEF (program-based budgeting, a very general specification of the MTEF• The PUFMARP appraisal document provided only consistency between MDA budgets andpolicies, detailed costingthe more technicalaid financed projects). It barely mentioned thereform that emphasised and integration of aspects of the reform (formulating programupstream elements of an MTEF (macro-fiscal framework, policy and expenditure review) thatcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetare conventionally seen as mention of the more strategic elements of an MTEF reform coveringsoftware). There was little the initial priorities in establishing a robust MTEF process.Implementation of the MTEF reform largely reflected this specification, except that the modelthe development of the medium-term macro-fiscal strategy and framework, and the review ofchosenprograms and priorities and their implications for the expenditure prioritisation andsector emphasised a more detailed activity based budgeting approach.• In a narrow sense the outputs of reform, measuredthe reform was undertaken following to abudget planning. In practice, the detailed design of against the initial specification, were projectconsiderable the arrival of the DFID funded consultants in September 1997. However, its beenstart-up with extent met, and after the initial donor funding was finished have subsequentlysustained.continued to broader sense the outputs, measured against the requirement toemphasis However, in reflect that of the appraisal document.establish a more realistic and strategic medium-term budgeting process have not been• The support for the introduction of the MTEF effectively operated as a separate sub-projectachieved. Moreover, the weaknesses in the approach had been identified earlydirectlyelaboratedwithin PUFMARP under an MTEF Project Unit. Because funding was provided on and by DFID,in both the ICR for PUFMARP, and the 2003 ODI case study of Ghana’s MTEF. It is perhaps
Evaluation Question
Judgement criteria/Possible indicators
– The credibility of MDA budgets had continued to be undermined by shortfalls and delays inbudget releases against the planned budget. A major factor was weak payroll planning whichresulted in high levels of variance in MDA expenditure against their budget (which averaged26.5% over the 2003-2007 budgets). Within MDA budgets difference between budgetallocations and releases were particularly large for services and investment, the twocategories of expenditure covered by activity based budgeting). The chart of accounts still didnot include codes for the MTEF activity classification with the result that it continued to beimpossible to report expenditure against MTEF budget allocations.• The study team found that situation has changed little subsequent to the 2009 ERPFM. Indeed theCitizen’s Guide to the Budget and the MTEF policy hearing had lapsed, and the start of the MTEF/Budget preparation cycle significantly delayed. The guidelines for the preparation 2011 budgetwere not issued until September, with ministries no longer being provided with budget ceilings.However, there was increasing recognition of the need for a thorough re-engineering of theMTEF process, with the decision taken in 2010 to replace activity based budgeting with a moreFindingsprogram-based budgeting approach.strategic• As noted above, the original specification of the MTEF reforms emphasised the “bottom-up”elements of an MTEF (program-based budgeting, consistency between MDA budgets and• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thepolicies, detailed costing and integration of aid financed projects). It barely mentioned thePublic Finance Management Reform Programframework, policyMTEF involved reforms tothatupstream elements of an MTEF (macro-fiscal (PUFMARP). The and expenditure review)strategic budgeting and as the initial priorities in establishing a robustin introducing the MTEFare conventionally seen budget preparation processes. The objective MTEF process.was “to improve the planningreform largely reflectedexpenditures and thus contribute modelImplementation of the MTEF and budgeting of public this specification, except that the tostrengthened fiscal policy formulation and implementation.”67.chosen emphasised a more detailed activity based budgeting approach.• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which was• In a narrow sense the outputs of reform, measured against the initial specification, were to aprovided as project met, and after DFID. The actual financing provided have end of PUFMARPconsiderable extentco-financing bythe initial donor funding was finishedby thesubsequently beenwas estimated at USD broader sense the outputs, measured 0.18 million of GoG financing.sustained. However, in4.58 million of donor financing and USDagainst the requirement to• establish a more realisticwere focused primarily on MoFEP and MDAs, were provided mainly asThe reform inputs, which and strategic medium-term budgeting process have not beentechnical Moreover, the weaknesses in main areas of support:achieved. assistance. They involved two the approach had been identified early on and elaborated– The the ICR for PUFMARP, and the of the MTEF reform of Ghana’s MTEF. It is perhapsin bothdevelopment and specification 2003 ODI case studycovering: (i)formulation of theprogram categories for planning, budgeting and accounting; (ii)preparing the expendituresurprising that the core activity based budgeting reform continued largely unchanged and thatprofile for each MDA in line with existing more strategic approach, price basis for forecasts;measures taken subsequently to support a policies; (iii)specifying thesuch as the introduction of(iv)integrating aid were not sustained.MDA policy hearings,financed projects and programs and preparing computer software for thispurpose.• In the view study team a major reason was that with activity based budgeting having been fully– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelimplemented across all MDAs and significant investment made in the ACTIVATE budget budget MoFEP that included budget ceilings; to introduce further budget software in thesoftware, circularwas understandably reluctant (ii)the installation of substantial changes.MDAs and the buy-in to a of budget and finance staff in the insufficient of MoFEP to forceFurthermore(iii)the training wider MTEF reform process was preparationforthe MTEF and theinformation system lack preparation.through changes. This in its of buy-in would appear to reflect two main factors:• – There was tendency for “silo-working” whereby a reform initiative is seen as belonging to aAt the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopmentdepartment within MoFEP whichof specific improvements were with limitedaimedparticular of the MTEF reforms. A number then proceeds to implement it introduced inputsat improving the policy focus and transparency of the budget process, whilethe Economic Policyfrom other departments. For example, it should have been expected that further rdevelopment of the ACTIVATE software application took place. However, the main elements of and Forecasting Division in MoFEP should have played a much stronger role in the MTEFthe MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.reforms.• – The skills and capacities in the Budget Division did not fully adjust to accommodate the moreVery limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the primarily a policy-led planning function.strategic under an MTEF that sees budgeting as MTEF reform and related training activities.There is no data on actual budgetary spending on the MTEF following of former sector planningElsewhere MTEF reforms have been accompanied by the merging the completion ofPUFMARP. However, in the 2007planning and budgeting department. The establishment of adepartments into a combined Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditureExpenditure Monitoring ‘human resources development’Budget Development Unit and an management’ area and Unit in the mid-2000s recognisespolicy areas was close to US$1.5 million. Thisbuilding more integrated planning and budgetingthis requirement, although it stops short of figure related to operating expenses and capitalexpenditure only and did notlevel. personnel expenses and allowances which are not brokencapacities at sector desk includedown by policy area.• – Over the period 2001-10, it can be estimated that around USD6specificationspent on the MTEFThe PUFMARP appraisal document provided only a very general million was of the MTEFrelated reforms72. Since most of the weaknesses of the reform (formulating programreform that emphasised the more technical aspectsin the MTEFP that were identified at theconclusion of PUFMARP were never effectively addressed, model budget circular, and budgetcategories, expenditure profiling, and price basis of forecasts,the efficiency of this investmentmust be considered low. The initial the more on establishing activity MTEF reform coveringsoftware). There was little mention ofemphasisstrategic elements of anbased budgeting wasprobably a mistake and a more comprehensive and strategically focused approach tothe development of the medium-term macro-fiscal strategy and framework, and the review ofintroducing an and priorities and their implications allowed the reform to evolve and addresssector programs MTEF might have worked better andfor the expenditure prioritisation andissues as they In practice, the detailed design of the reform was undertaken following projectbudget planning.arose.
a inputs & context: the design of PfM reformof outputs• ConsistencyA.7. How efficiently wereproduced with plannedthese outputs generat-• Government(quantity andA.1.What has been the na-programme funds fored? Was the pacing andPFM reforms committedture and scale reformsof PFmtiming)sequencing ofreform inputs provided• and actually disbursed byExtent of coordinationappropriate and cost-year overoutputsby Governmentthe costbetween the evaluationeffective? Was and Do-period. of pacing andnors?• Qualityper output acceptable?• sequencing of outputDonor funds for PFMreforms committed andproduction:• actually disbursed by yearinternally coherent/over the evaluation periodefficient?• Nature of support• Consistent with capacityprovided to PFM reformendowments?• efforts (equipment, ofAppropriate to degreetraining, TA,nature ofownership/ diagnosticwork.) & admin supportpolitical• for reform? for reforms byFocal areasfunction cost of outputs in• Relative (based on PFM“clusters” as in Andrewsrelation to budgeted costs.Influence of external• 2010)• Focal areas for reforms bysupport to PFM reformorganisational location/upon pacing andlevel of governmentsequencing.Counterfactual: would• (Ministry of Finance,Sector ministries, Localpacing & sequencing ofGovernments, Parliament,outputs have beenCSOs, etc.) absence ofdifferent inexternal PFM support?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
99
A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-
• Structure of design teamand related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system
100
Evaluation Question
Judgement criteria/Possible indicators
Budget Development Unit and an Expenditure Monitoring Unit in the mid-2000s recognisesthis requirement, although it stops short of building more integrated planning and budgetingcapacities at sector desk level.– Over the period 2001-10, it can be estimated that around USD6 million was spent on the MTEFrelated reforms72. Since most of the weaknesses in the MTEFP that were identified at theconclusion of PUFMARP were never effectively addressed, the efficiency of this investmentmust be considered low. The initial emphasis on establishing activity based budgeting wasprobably a mistake and a more comprehensive and strategically focused approach tointroducing an MTEF might have worked better and allowed the reform to evolve and addressFindingsas they arose.issuesANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• The MTEF reforms in Ghana have yet to establish a realistic policy-led, medium-term budgetingprocess. The primary reasons for this relate to weaknesses in the original design and roll-out of• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under thereform and the failure to address the problems with the approach when they were identified. ThePublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms torole of external constraints is less clear:strategic budgeting and budget preparation processes. The objective in introducing the MTEF– The initial strong political level support for the reform weakened as time went on. This maywas “to improve the planning and budgeting of public expenditures and thus contribute tohave reflected the perception the MTEF process had been effectively institutionalised and wasstrengthened fiscal policy formulation and implementation.”67.no longer seen as a critical PFM reform. Within the donor community there appears to have a• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which wasbeen a reluctance to emphasise the MTEF in the context of the MDBS dialogue when otherprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARPPFM reforms, particularly those relating to BPEMS were seen as more pressing priority.was estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.– Similarly, the interest of external stakeholders and CSOs in the MTEF reforms has been• The reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly aslimited. The dialogue between GoG and the CSOs on PFM reforms continues to be quitetechnical assistance. They involved two main areas of support:limited, and focused the transparency of the budget process, more accessible budget– The development and specification of the MTEF reform covering: (i)formulation of thedocumentation, and improving the availability of budget data.program categories for planning, budgeting and accounting; (ii)preparing the expenditure– The MTEF reform has not been associated with a clear policy imperative (policy space).profile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;Relatively strong economic and fiscal performance through the mid-2000s weakened the(iv)integrating aid financed projects and programs and preparing computer software for thispressure for expending planning and budget prioritisation reform. At the same time thepurpose.absence of reliable budget execution data remained a more fundamental constraint to– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelimproved expenditure analysis and budget planning.budget circular that included budget ceilings; (ii)the installation of budget software in the– Financing for the MTEF reforms does not appear to have been a significant constraint. TheMDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and thereform was well funded through PUFMARP and subsequently the budget of the Budgetinformation system in its preparation.Division in MoFEP has included substantial funding for PFM reform.• At the completion of the PUFMARP, MoFEP continued the implementation and furtherdevelopment of the MTEF reforms. A number of specific improvements were introduced aimedat improving the policy focus and transparency of the budget process, while further rdevelopment of the ACTIVATE software application took place. However, the main elements of the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the MTEF reform and related training activities.There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’policy areas was close to US$1.5 million. This figure related to operating expenses and capital• Has timeliness of fundingexpenditure only and did notPUFMARP during the 2001 and 2002, and subsequentbroken72 This is based on assumed expenditure of USD 1.9 million under the MTEF component ofinclude personnel expenses and allowances which are notGoGbeen an issue?down by policy area.• NatureUSD 0.5 millionfinanced spending assumed at aroundof “conventionalannually.• wisdom” on PFM reforms: • The PUFMARP appraisal document provided only a very general specification of the MTEFStructure of design teamA. 2.What types of struc-does prevailing thinkingreform that emphasised the more technical aspects of the reform (formulating programand related consultationtures have been usedexclude certain reformcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetprocess: Balance of inputsfor the design and man-options? Govt/software). There was little mention of the more strategic elements of an MTEF reform coveringby Donors/agement of these re-• Has policy discussion onthe development of the medium-term macro-fiscal strategy and framework, and the review ofConsultants; Extent ofform inputs? HavePFM reform been open insector programs and priorities and their implications for the expenditure prioritisation andinput/ consultation withthese structuresterms of range of systembudget planning. In practice, the detailed design of the reform was undertaken following projectend users of PFMserved to provide a co-participants and range ofstart-up with the arrival of the DFID funded consultants in September 1997. However, its(Sector ministries, LGsordinated and harmo-ideas?emphasis continued to reflect that of the appraisal document.and service institutions).nised delivery frame-• Overall, what & co-been• The support for the introduction of the MTEF effectively operated as a separate sub-projectManagement havework?the bindingstructure foronwithin PUFMARP under an MTEF Project Unit. Because funding was provided directly by DFID,ordination constraintsthe PFM production
a inputs &have been the• Degree of ownership ofA.8. Whatcontext: the design of PfM reformreforms at thebinding external con-• Government funds forA.1.What has been the na-administrative andstraints on the deliveryPFM reforms committedture and scale of PFmpolitical levels (narrow vs.of PFm reform: politi-and actually disbursed byreform inputs providedbroad; depth ofcal, financing or policyyear over the evaluationby Government and Do-commitment to reformfactors? How has thisperiod.nors?and motivations for thisvaried across the differ-• Donor funds for PFMcommitment).ent PFm reform compo-reforms committed and• Quality of interactionnents?actually disbursed by yearbetween administrativeover the evaluation periodand political cadres.• Nature of support• Extent and nature ofprovided to PFM reformpolitical accountabilityefforts (equipment,(within ruling party, vis-à-training, TA, diagnosticvis Legislature, vis-à-viswork.)Electorate & Civil Society)• Focal areas for reforms byand extent to which this isfunction (based on PFMpatronage or“clusters” as in Andrewsperformance-based.2010)• Economic context and its• Focal areas for reforms byinfluence on financing oforganisational location/PFM reform: per capitalevel of governmentGDP and % growth in(Ministry of Finance,evaluation period,Sector ministries, Localsignificance of domesticGovernments, Parliament,revenue, significance ofCSOs, etc.)aid, level of total anddiscretionary publicspending; absence/presence of macro crises.
absence of reliable budget execution data remained a more fundamental constraint toimproved expenditure analysis and budget planning.– Financing for the MTEF reforms does not appear to have been a significant constraint. Thereform was well funded through PUFMARP and subsequently the budget of the BudgetDivision in MoFEP has included substantial funding for PFM reform.
Evaluation Question
and extent to which this ispatronage orperformance-based.• Economic context and itsinfluence on financing ofPFM reform: per capitaGDP and % growth inevaluation period,significance of domesticrevenue, significance ofaid, level of total anddiscretionary publicspending; absence/Judgement criteria/presence of macroPossible indicatorscrises.Findings
a inputs & context: the design of PfM reform• Has timeliness of fundingbeen an issue?• Ghana’s medium-term expenditure framework (MTEF) initiative was initiated in 1997 under the• Government funds forA.1.What has been the na-• Nature of “conventionalPublic Finance Management Reform Program (PUFMARP). The MTEF involved reforms toPFM reforms committedture and scale of PFmwisdom” on PFM reforms:strategic budgeting and budget preparation processes. The objective in introducing the MTEFand actually disbursed byreform inputs provideddoes prevailing thinkingwas “to improve the planning and budgeting of public expenditures and thus contribute toyear over the evaluationby Government and Do-exclude certain reformstrengthened fiscal policy formulation and implementation.”67.period.nors?options?• Has policy discussion on• The funding requirement estimated at PUFMARP appraisal was USD 4.6 million which was• Donor funds for PFMreforms committed andinprovided as project co-financing by DFID. The actual financing provided by the end of PUFMARPPFM reform been openactually disbursed by yearwas estimated at USD 4.58 million of donor financing and USD 0.18 million of GoG financing.terms of range ofover the evaluation period • The reform inputs, which were focused primarily on MoFEP and MDAs, were provided mainly asparticipants and range of• Nature of supporttechnical assistance. They involved two main areas of support:ideas?providedwhat have beento PFM reform– The development and specification of the MTEF reform covering: (i)formulation of the• Overall,efforts (equipment,program categories for planning, budgeting and accounting; (ii)preparing the expenditurethe binding constraints ontraining, productionTA, diagnosticprofile for each MDA in line with existing policies; (iii)specifying the price basis for forecasts;the PFMwork.)(iv)integrating aid financed projects and programs and preparing computer software for thispossibility frontier?• How has thisfor reforms byFocal areas variedpurpose.• function (based onreformPFM– Improvements to the formulation of the annual budget covering: (i)the preparation of a modelbetween different“clusters” as in Andrewsbudget circular that included budget ceilings; (ii)the installation of budget software in thecomponents?2010)MDAs and (iii)the training of budget and finance staff in the preparation of the MTEF and the• Focal areas for reforms byinformation systemc. outcomes & overall assessment of PfM reform and of donor support to PfM reformin its preparation.organisational location/• At the completion of the PUFMARP, MoFEP continued the implementation and further(Addressed as part of the reforms. A numberC.1 -C.4• development of the MTEFoverall PFM review)of specific improvements were introduced aimedlevel of government(Ministry of Finance,at improving the policy focus and transparency of the budget process, while further rSector ministries, Localdevelopment of the ACTIVATE software application took place. However, the main elements of Governments, Parliament,the MTEF reform and the procedures introduced under PUFMARP remained largely unchanged.CSOs, etc.)• Very limited short-term assistance was provided by donors after 2003. This included BMZ and UNICEF, to support the further development of the MTEF reform and related training activities.There is no data on actual budgetary spending on the MTEF following the completion ofPUFMARP. However, in the 2007 Budget the spending allocation for the Budget Division relatingto the ‘ improved public expenditure management’ area and ‘human resources development’policy areas was close to US$1.5 million. This figure related to operating expenses and capitalexpenditure only and did not include personnel expenses and allowances which are not brokendown by policy area.• The PUFMARP appraisal document provided only a very general specification of the MTEFreform that emphasised the more technical aspects of the reform (formulating programcategories, expenditure profiling, and price basis of forecasts, model budget circular, and budgetsoftware). There was little mention of the more strategic elements of an MTEF reform coveringthe development of the medium-term macro-fiscal strategy and framework, and the review ofsector programs and priorities and their implications for the expenditure prioritisation andbudget planning. In practice, the detailed design of the reform was undertaken following project
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
101
A. 2.What types of struc-tures have been usedfor the design and man-agement of these re-form inputs? Havethese structuresserved to provide a co-
• Structure of design teamand related consultationprocess: Balance of inputsby Donors/ Govt/Consultants; Extent ofinput/ consultation withend users of PFM system
102
PfM Reform component/initiative: Revenue ManagementFindingsPuFmArP (1997-2003)• During the period studied, reform of revenue management was part ofthe PUFMARP suite of reforms. Reforms were concentrated on: (i)improving income tax administration through introducing a uniqueincome taxpayer code; and (ii) improving the efficiency of theinstitutional structure administering revenue collections.• The main problem that the reforms attempted to solve was the fact thatthere was a significant tax gap, i.e. that actual revenue collections (e.g.as a % of GDP) were lower than their potential. This was due to the factthat: (i) taxpayer compliance (particularly, income tax, but alsocustoms) was considered to be low; and (ii) efficiency of collections wasweak, as was revenue reporting and forecasting.• Whilst the reforms were originally targeted at both income tax andcustoms duties, and institutional support to the different revenueservices, the PUFMARP project subsequently concentrated its supporton income tax and indirect tax measures, and reallocated otheroriginally-planned project resources to BPEMS.• Specifically, the revenue management reforms during this periodcovered: (i) introduction of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since 1995; and (iii) staff training.• DP support under PUFMARP included a limited amount of IDA funding,which took the form of: technical assistance, training support, andlimited equipment.reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughthe provision of capacity building, equipment and analytical software;(ii) improving the efficiency of operations of the Internal RevenueService through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and trainingsupport.Introduction of GrA (2010-2013)
Evaluation Question
Judgement criteria/ Possible indicators
a inputs & context: the design of PfM reform
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?
• Government funds for PFM reforms committed andactually disbursed by year over the evaluation period.• Donor funds for PFM reforms committed andactually disbursed by year over the evaluation period• Nature of support provided to PFM reform efforts(equipment, training, TA, diagnostic work.)• Focal areas for reforms by function (based on PFM“clusters” as in Andrews 2010)• Focal areas for reforms by organisational location/level of government (Ministry of Finance, Sectorministries, Local Governments, Parliament, CSOs,etc.)
covered: (i) introduction of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since 1995; and (iii) staff training.• DP support under PUFMARP included a limited amount of IDA funding,which took the form of: technical assistance, training support, andlimited equipment.
Evaluation Question
Judgement criteria/ Possible indicators
reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughFindingsthe provision of capacity building, equipment and analytical software;
a inputs & context: the design of PfM reform
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?
• Government funds for PFM reforms committed andactually disbursed by year over the evaluation period.• Donor funds for PFM reforms committed andactually disbursed by year over the evaluation period• Nature of support provided to PFM reform efforts(equipment, training, TA, diagnostic work.)• Focal areas for reforms by function (based on PFM“clusters” as in Andrews 2010)• Focal areas for reforms by organisational location/level of government (Ministry of Finance, Sectorministries, Local Governments, Parliament, CSOs,etc.)
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
A.2.What type of struc-tures have been usedfor the design andmanagement of thesereform inputs? Havethese structuresserved to provide a co-ordinated and harmo-nised delivery frame-work?
• Structure of design team and related consultationprocess: Balance of inputs by Donors/ Govt/Consultants; Extent of input/ consultation with endusers of PFM system (Sector ministries, LGs andservice institutions).• Management & co-ordination structure for PFMreforms (Ad hoc Project Units vs. Normalmanagement structures; Govt-controlled vs. SharedDonor-Govt management; Use of consultants formanagerial or purely advisory roles)• Arrangements for monitoring & evaluation.• Level of harmonisation and alignment of differentdonor contributions.
(ii) improving the efficiency of operations of the Internal RevenueService(1997-2003)PuFmArPthrough computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified part of• During the period studied, reform of revenue management was GhanaRevenue Authority, of reforms. Reforms were concentrated on: (i)the PUFMARP suiteincluding organisational development and HRchanges (on-going).improving income tax administration through introducing a unique• income taxpayer code; and(in two 3-year phases, 2003-2005, and 2006-Total GTZ funding support (ii) improving the efficiency of the2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).institutional structure administering revenue collections.• The main problem that the reforms attemptedassistance and trainingsupport mainly took the form of technical to solve was the fact thatsupport. a significant tax gap, i.e. that actual revenue collections (e.g.there wasas a % of GDP) were lower than their potential. This was due to the factIntroduction of GrA (2010-2013)that: (i) taxpayer compliance (particularly, income tax, but also• BMZ continued to support the RAGB (and subsequent form the GRA)customs) was considered to be low; and (ii) efficiency of collections wasthrough the continuation of its Good Financial Governance programme.weak, as was revenue reporting and forecasting.Total support for the current phase is €9 mn.• Whilst the reforms were originally targeted at both income tax and• The Swiss State Secretariat for Economic Affairs (SECO) is providingcustoms duties, and institutional support to the different revenuesupporting assistance of €2.096 mn for 2010-2013.services, the PUFMARP project subsequently concentrated its supporton income tax and indirect tax measures, and reallocatedrevenue management reforms supported by PuFmArP:other• originally-planned the appraisal document highlighted the revenueIn terms of design, project resources to BPEMS.• Specifically, the revenue management reforms during this periodmanagement reform activities that would be undertaken, coveringcovered:in income tax, indirect tax, and customs, as well as support toreforms (i) introduction of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since the National Revenue Secretariat (whose duties were subsequently1995; and (iii) staff training. Agencies Governing Board). The scope ofsubsumed into the Revenue• DP supportactivities was veryincluded a limited amount of IDA duringthe reform under PUFMARP ambitious, and was scaled back funding,which took the form of: technicalon measures for income and indirectproject implementation to focus assistance, training support, andlimited equipment.taxes.• As with other PUFMARP areas, the revenue reforms operated as areforms to rAGB (2003-2008)arrangements for responsibility fordistinct sub-project. Detailed• Following PUFMARP, reforms to revenue administration continuedmanagement and monitoring of the revenue reform activities are notunder GoG support and GTZin the project document. In practice, as theclearly spelt out separately (BMZ), as part of the latter’s Good FinancialGovernance programme.specific individual revenue servicesreforms are focussed on• These were focussed on the following measures: (i) establishment and(specifically, the Internal Revenue Service (IRS), and the VAT Service defining the strategic operations of a Tax Policy Unit in MoFEP through(VATS)), it was senior management in these agencies who had the provision of capacity building, equipment and analytical software;responsibility for managing and monitoring the reform activities.(ii) improving the efficiency of operations of the Internal Revenue• There were no specific consultancy arrangements for management ofService through computerisation and organisational changes; and (iii)revenue reform activities, and, beyond the overall PUFMARP projectrestructuring of the 3 separate revenues services into a unified Ghanamanagement unit, there did not appear to be a specific project unitRevenue Authority, including organisational development and HRcovering revenue reforms.changes (on-going).• Harmonisation of reforms amongst DPs was simplified by the• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-delineation in responsibility amongst the two DPs (WB and ODA/DFID)2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).involved in supporting the revenue activities (direct and indirect tax,• The support mainly took the form of technical assistance and training
103
nised delivery frame-work?
Donor-Govt management; Use of consultants formanagerial or purely advisory roles)• Arrangements for monitoring & evaluation.• Level of harmonisation and alignment of differentdonor contributions.
104
Evaluation Question
Judgement criteria/ Possible indicators
taxes.• As with other PUFMARP areas, the revenue reforms operated as adistinct sub-project. Detailed arrangements for responsibility formanagement and monitoring of the revenue reform activities are notclearly spelt out separately in the project document. In practice, as thereforms are focussed on specific individual revenue services(specifically, the Internal Revenue Service (IRS), and the VAT Service (VATS)), it was senior management in these agencies who had responsibility for managing and monitoring the reform activities.Findings
a inputs & context: the design of PfM reform
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?
• Government funds for PFM reforms committed andactually disbursed by year over the evaluation period.• Donor funds for PFM reforms committed andactually disbursed by year over the evaluation period• Nature of support provided to PFM reform efforts(equipment, training, TA, diagnostic work.)• Focal areas for reforms by function (based on PFM“clusters” as in Andrews 2010)• Focal areas for reforms by organisational location/level of government (Ministry of Finance, Sectorministries, Local Governments, Parliament, CSOs,etc.)
A.3. What types of comple-mentary actions haveDonors taken to sup-port PFm reforms andwhat has been theirsignificance? Havethey had any influenceon the external con-straints to reform?
• There were no specific consultancy arrangements for management ofrevenue reform activities, and, beyond the overall PUFMARP projectPuFmArP (1997-2003)management unit, there did not appear to be management was part• During the period studied, reform of revenuea specific project unit ofcovering revenue reforms.the PUFMARP suite of reforms. Reforms were concentrated on: (i)• improving incomereforms amongst DPs was simplified bya uniqueHarmonisation of tax administration through introducing thedelineation in responsibility amongst the two DPs (WB and ODA/DFID)income taxpayer code; and (ii) improving the efficiency of theinvolved in supporting administering revenue collections.institutional structure the revenue activities (direct and indirect tax,respectively).• The main problem that the reforms attempted to solve was the fact thatthere was a significant tax gap, i.e. that actual revenue collections (e.g.revenue management reforms supported by GFG programme:• as a % of GDP) were lower thanmeasures wereThis was dueBMZ staff inRevenue management reform their potential. designed by to the factthat: (i) taxpayer compliance (particularly, income tax, but alsoconsultation with GoG.• customs) was considered to be staff and (ii)RAGB, andof collections wasThe reforms were managed by low; of the efficiency subsequently theweak, as was revenue reporting and forecasting.GRA.• Whilst the reforms were originally targeted at both income tax and• customs PUFMARP and GFG programmes weredifferent revenueBoth the duties, and institutional support to the designed andservices, the PUFMARP project subsequently concentrated itsofimplemented as external project-type interventions; thus, use supporton income tax and indirect tax measures, and reallocated othercountry systems was limited.• originally-planned project resources to BPEMS.The revenue reform measures were largely self-contained, and• Specifically, the revenue management reforms during this periodfollowed one after the other (i.e. the BMZ-supported reforms followedcovered: onintroduction the PUFMARP reforms had ended).logically (i) from where of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since • There have been no specific revenue-related MDBS triggers or targets.• 1995; and (iii) staff training. this question are dealt with in the discussionThe other issues related to• DP support under PUFMARP included a limited amount of IDA funding,of PFM as a whole.which took the form of: technical assistance, training support, andlimited equipment.reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughthe provision of capacity building, equipment and analytical software;
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• Use of country systems: extent to which external aidis ‘on plan’, ‘on budget’ and ‘on treasury’?• Ranking of donor efforts to support national systemsas expressed in Paris Declaration monitoringreports and PEFA indicators D2 and D3.• Scale and evolution over the evaluation period of aidflows provided as GBS, SBS or debt relief and relativecontribution to reduction of treasury managementcosts and to discretionary resources available toGovernment, (including effect on financingconstraints for PFM reform outputs).• Nature and evolution over time of dialogue on PFMwithin GBS/PRSC, SBS structures or similar donor-govt fora, including use of PFM reform conditions inGBS/ SBS disbursement conditions.• Influence of dialogue on the “political” and “policyspace” constraints restricting choice of PFM reformoutputs.• (Consistency of actual changes with PFM reformtriggers and/or points of emphasis in GBS/ PRSCdialogue).
• Consideration of counter-factual:• Outputs: would PFM reforms have evolveddifferently in absence of GBS and related policydialogue structures?• Outcomes: Would intermediate outcomes have beendifferent if use of country systems had been lower?
A.4. to what extent hasthere been domesticpublic pressure or re-
• Nature of engagement with PFM reform issueswithin civil society, academia and the media.• Extent of participation of CSOs in policy dialogue on
(ii) improving the efficiency of operations of the Internal RevenueService through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and trainingsupport.• The revenue reforms under PUFMARP did not involve explicitIntroduction ofor specific activities with civil society. This was probablyinteractionsGrA (2010-2013)• BMZ continued to support thereforms to tax administration. GRA)due to the technical nature of RAGB (and subsequent form thethrough the continuation of its Good Financial Governance programme.
on the external con-straints to reform?• The other issues related to this question are dealt with in the discussionof PFM as a whole.
Evaluation Question
contribution to reduction of treasury managementcosts and to discretionary resources available toGovernment, (including effect on financingconstraints for PFM reform outputs).• Nature and evolution over time of dialogue on PFMwithin GBS/PRSC, SBS structures or similar donor-govt fora, including use of PFM reform conditions inGBS/ SBS disbursement conditions.• Influence of dialogue on the “political” and “policyspace” constraints restricting choice of PFM reformoutputs.• (Consistency of actual changes with PFM reformtriggers and/or points of emphasis in GBS/ PRSCdialogue).Judgement criteria/ Possible indicatorsFindingsPuFmArP (1997-2003)• During the period studied, reform of revenue management was part ofthe PUFMARP suite of reforms. Reforms were concentrated on: (i)improving income tax administration through introducing a uniqueincome taxpayer code; and (ii) improving the efficiency of theThe revenue reforms under PUFMARP did not involve explicit• institutional structure administering revenue collections.• The main problem that the reforms attempted to solve was the fact thatinteractions or specific activities with civil society. This was probablythere was technical naturegap, i.e. thatto tax administration.due to the a significant tax of reforms actual revenue collections (e.g.ascontrast to other PFM areas, their potential.limited numbers of civil• In a % of GDP) were lower than there are very This was due to the factthat: (i) taxpayer compliance (particularly, income tax, but alsosociety organisations or academic institutions who deal with revenuecustoms) was considered tomostly concerned with revenue as part wasissues; the ones that do are be low; and (ii) efficiency of collections ofweak, as was analysis.reporting and forecasting. the implementationoverall fiscal revenue This will likely change with• Whilst the and the greater emphasis on oil revenue,income tax a greaterof the EITI reforms were originally targeted at both as there is andcustoms duties, and institutional supportoil the different revenueemphasis on accountability for the use of to revenue.services, the PUFMARP II of the BMZ support to the revenue its support• The evaluation of Phase project subsequently concentrated agencieson income tax and indirect tax measures, and reallocated other incited exposure to “international best practice” as being importantoriginally-planned project resources to BPEMS.influencing Ghana’s tax administration reforms.• Specifically, the revenue management reforms during this period• The remaining issues related to this question are dealt with in thecovered: (i) of PFM as a whole.discussion introduction of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since 1995; and (iii) staff training.• DP support under PUFMARP included a limited amount of IDA funding,which took the form of: technical assistance, training support, andlimited equipment.
a inputs & context: the design• PfM reformof counter-factual:ofConsideration• Outputs: would PFM reforms have evolved• Government funds for PFM reforms committed andA.1.What has been the na-differently in absence of GBS and related policyactually disbursed by year over the evaluation period.ture and scale of PFmDonor funds for PFMreform inputs provided• dialogue structures? reforms committed andOutcomes: Would by year over outcomes have beenby Government and Do-• actually disbursedintermediate the evaluation perioddifferent support provided to PFM reform efforts• Nature ofif use of country systems had been lower?nors?Nature of engagement with PFM reform issues• (equipment, training, TA, diagnostic work.)A.4. to what extent has• within areas for reforms by function (based on PFMFocal civil society, academia and the media.there been domestic“clusters” as in Andrews CSOs in policy dialogue onpublic pressure or re-• Extent of participation of 2010)• Focal areas for reforms by organisational location/PFM reform.gional institutionalPresence/ absence (Ministry research andpressure in support of• level of governmentof specificof Finance, Sectorministries, Local Governments, Parliament, CSOs,advocacy by CSOs on PFM reform issues.PFm reform and whatRelevance of PFM issues to political outcomes:has been the influence• etc.)– Discussion of PFM issues within electionon the external con-campaignsstraints to reform?– Discussion of service delivery issues related toPFM in election campaigns– Voter concerns during elections (from Afro Barometer and other relevant publications)– Existence of regional or international bodies (e.g.WAEMU, EITI) promoting specific norms on PFMissues, and their relative influence on domesticpolitical discourse.– Evolution in the quality of public/ CSO engagementwith PFM issues over time and its influence onPFM reform outputs.• Consideration of counter-factual: would PFM reformoutputs have evolved differently in absence ofdomestic public pressure or pressure from regional/international institutions?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
105
A.5. How relevant was thePFm reform pro-gramme to the needsand the institutionalcontext? Was donorsupport consistentwith national priori-ties? to what extentwere adaptationsmade in response to
• Scale and focus of support in relation to identifiedPFM weaknesses at start and during implementationof reforms (including weaknesses in HRendowments, quality of institutions & rules, qualityof systems & business processes, and quality oforganisations).• Consistency of donor-funded support withGovernment PFM reform efforts, with ongoing publicadministration reforms and overall Govt policies.• Extent to which scope and focus of support were
reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment and• defining the strategicformal GoG of a Tax Policydocument outlining GoGIn the absence of any operations PFM strategy Unit in MoFEP throughthe provision of period studied prior to the Growthanalytical software;priorities in the capacity building, equipment and and PovertyReduction Strategy (GPRS I and II) and the Short-Term/Medium-term(ii) improving the efficiency of operations of the Internal RevenueAction Plan (ST/MTAP), in practice, the DP PFM programme documentsService through computerisation and the GTZ/BMZ Good Financial (iii)(specifically, those for PUFMARP and organisational changes; andrestructuring of the 3 separate revenues services into a unified GhanaGovernance [GFG] programme), prepared in consultation with GoG,Revenue Authority, including organisational development and HRserved to proxy as such.• changes (on-going).The PUFMARP programme document provides a relevant, if limited,• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-analysis of the institutional and operational challenges facing the2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO). the programmerevenue agencies but there is less explanation of how• The support mainly took the form of technical assistance and training
106
Evaluation Question• In the absence of any formal GoG PFM strategy document outlining GoGpriorities in the periodPuFmArP (1997-2003)studied prior to the Growth and PovertyReduction period studied, I and II) of revenue management was part of• During theStrategy (GPRS reform and the Short-Term/Medium-termAction Plan (ST/MTAP), in practice, the DP PFM programme on: (i)the PUFMARP suite of reforms. Reforms were concentrateddocuments(specifically, those for PUFMARP and the GTZ/BMZ Good a uniqueimproving income tax administration through introducingFinancialGovernance [GFG] programme), prepared in consultation withincome taxpayer code; and (ii) improving the efficiency of the GoG,served to proxy as such.institutional structure administering revenue collections.• The main problem that the reforms attempted to solve was if limited,• The PUFMARP programme document provides a relevant, the fact thatanalysis of the institutional and operational challenges facing thethere was a significant tax gap, i.e. that actual revenue collections (e.g.revenue GDP) were lower than their potential. of how the programmeas a % ofagencies but there is less explanationThis was due to the factmeasures will address these challenges.that: (i) taxpayer compliance (particularly, income tax, but also• customs) was considered to be low; and (ii) efficiencybasis of a needsThe GTZ/BMZ GFG programme was prepared on the of collections wasanalysis was revenue reporting and forecasting.weak, asundertaken with GTZ support.• Whilst the reforms were originally measures for revenue management.• The main change to the PUFMARP targeted at both income tax andIn practice, external institutional support to the different revenuecustoms duties, and and GoG support PUFMARP measures was verylimited, as resources originally subsequently concentrated its supportservices, the PUFMARP projectintended for revenue managementwere diverted to support BPEMS.on income tax and indirect tax measures, and reallocated other• originally-planned project resources to BPEMS.Within the long-term support committed by GTZ/BMZ (covering 2003-2015), each 3-year phasemanagement reforms during this period• Specifically, the revenue ended with an evaluation, leading to thedesign of the next 3-year phase. This was appreciated by GoGcovered: (i) introduction of a unique Taxpayer Identification Number; (ii)stakeholders.the re-introduction of VAT, whose introduction had been stalled since In terms of the consistency• 1995; and (iii) staff training. of revenue measures with GPRS and ST/MTAP, the GPRS documents are insufficiently specific on revenue• DP support under PUFMARP included a limited amount of IDA funding,management.which took the form of: technical assistance, training support, andlimited equipment.
issues, and their relative influence on domesticpolitical discourse.– Evolution in the quality of public/ CSO engagementwith PFM issues over time and its influence onPFM reform outputs.• Consideration of counter-factual: would PFM reformoutputs have evolved differently in absence ofdomestic public pressure or pressure from regional/Judgement criteria/ Possible indicatorsinternational institutions?Findings
a inputsrelevant wasdesign• PfM reformofScale and focus of support in relation to identifiedA.5. How& context: thethePFM weaknesses at start and during implementationPFm reform pro- na-• Government funds for PFM reforms committed andA.1.What has been theof reforms (including year over the evaluation period.grammescale of PFmto the needsactually disbursed by weaknesses in HRture andand theinputs provided• endowments, quality of institutions & rules, qualityinstitutionalDonor funds for PFM reforms committed andreformof systems & business processes, evaluation periodcontext? Was donoractually disbursed by year over theand quality ofby Government and Do-organisations).support• Nature of support provided to PFM reform effortsnors? consistentConsistency of donor-funded support with• (equipment, training, TA, diagnostic work.)with national priori-Governmentfor reforms by function (based on PFMties? to what extent• Focal areas PFM reform efforts, with ongoing publicadministration Andrews 2010)were adaptations“clusters” as inreforms and overall Govt policies.• Focal areas for reforms byfocus of support weremade in response to• Extent to which scope and organisational location/adapted to the context, especially to the level ofthe context and thelevel of government (Ministry of Finance, Sectorownership Local Governments, Parliament, CSOs,changing national pri-ministries, and the capacity for reformimplementation.orities?etc.)• Consistency & coherence with wider Donor policiesin country• Consistency with the Paris Declaration objectives asregards ownership, harmonisation and alignment.• Flexibility in design of support and existence ofadequate feed-back/ learning mechanisms in orderto adapt the design and implementation of support tochanging needs.
b. outputs: the delivery of PfM reforms
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
B.1. What have been theoutputs of the PFm re-form process and towhat extent has directdonor support contrib-uted to these outputs?
• Documentation of outputs :• by PFM function (according to PFM clusters inAndrews 2010) and• by type of output (People & Skills; Laws & Rules;Systems & Business processes; Organisationalchanges)• Matching of outputs to PFM reform inputs,differentiating if possible Government & Donorinputs.• Analysis of chronology of events to determine causallinks between donor support to reform and observedchanges at output level.
• Consideration of counterfactual: would theseoutputs have been generated in absence of donorsupport to PFM reform?
B.2. How efficiently werethese outputs gener-ated? Was the pacingand sequencing of re-
• Consistency of outputs produced with plannedprogramme (quantity and timing)• Extent of coordination between outputs• Quality of pacing and sequencing of output
reforms to rAGBTax Identification Numbering System, covering all types• Introduction of(2003-2008)• Following PUFMARP, reforms to revenue administration continuedof taxIntroduction of a new indirect tax regime (VAT)• under GoG support and GTZ (BMZ), as part of the latter’s Good Financial• Governance programme.Establishment of Large Taxpayer Unit in MoFEP (originally, in IRS?)• These were focussedPolicy Unit in MoFEP• Establishment of Tax on the following measures: (i) establishment andAutomation strategic operations pilot offices• defining the of operations in 6 IRSof a Tax Policy Unit in MoFEP through• the provision of capacity building, equipment and analytical software;Training/capacity development of staff• The establishment of the GRA• (ii) improving the efficiency of operations of the Internal RevenuePassage of the GRA Act, 2009Service through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and trainingsupport.• It may be argued that the efficiency of investment is relatively high,Introductiona relatively modest investment has led to outputs that havegiven thatof GrA (2010-2013)• been sustained, particularly in terms of VAT (which had experienced a BMZ continued to support the RAGB (and subsequent form the GRA)through the continuation of its Good Taxpayer Identification Number.false start in the mid-1990s) and the Financial Governance programme.
B.1. What have been theoutputs of the PFm re-form process and towhat extent has directdonor support contrib-uted to these outputs?• Introduction of Tax Identification Numbering System, covering all typesof tax• Introduction of a new indirect tax regime (VAT)• Establishment of Large Taxpayer Unit in MoFEP (originally, in IRS?)• Establishment of Tax Policy Unit in MoFEP• Automation of operations in 6 IRS pilot offices• Training/capacity development of staff• The establishment of the GRA• Passage of the GRA Act, 2009FindingsPuFmArP (1997-2003)• During the period studied, reform of revenue management was part of• the PUFMARP suite of reforms. Reforms were concentrated on: (i)It may be argued that the efficiency of investment is relatively high,improvinga relatively modest investment has led to outputs uniquegiven that income tax administration through introducing a that haveincome taxpayer code; and (ii) improving the efficiency of thebeen sustained, particularly in terms of VAT (which had experienced a institutional structure administering revenue collections. Number.false start in the mid-1990s) and the Taxpayer Identification• The main problem that the reforms attempted to solve was the fact thatthere was a significant tax gap, i.e. that actual revenue collections (e.g.as a % of GDP) were lower than their potential. This was due to the factthat: (i) taxpayer compliance (particularly, income tax, but alsocustoms) was considered to be low; and (ii) efficiency of collections wasweak, as was revenue reporting and forecasting.• Whilst the reforms were originally targeted at both income tax andcustoms duties, and institutional support to the different revenueservices, the PUFMARP project subsequently concentrated its supporton income tax and indirect tax measures, and reallocated otheroriginally-planned project resources to BPEMS.• Specifically, the revenue management reforms during this period• covered: (i) introduction of a unique Taxpayer Identification Number; (ii)The binding external constraint was primarily political commitment,particularly in terms of organisational changes related to thethe re-introduction of VAT, whose introduction had been stalled since integration of the revenue agencies.1995; and (iii) staff training.The change in government may have facilitated amount of IDA funding,• DP support under PUFMARP included a limited the demonstration ofpolitical commitment and thereby energised the reforms. The waywhich took the form of: technical assistance, training support, and thatthe reforms were designed, such that each phase was designed takinglimited equipment.into account the current situation and the lessons learnt from thereforms tophase,(2003-2008)previousrAGBhelped to enable the reforms to monitor and respond to• Following PUFMARP,policy space.availability (or not) of reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughthe provision of capacity building, equipment and analytical software;
Evaluation Question
• Documentation of outputs :• by PFM function (according to PFM clusters inAndrews 2010) and• by type of output (People & Skills; Laws & Rules;Systems & Business processes; Organisationalchanges)• Matching of outputs to PFM reform inputs,differentiating if possible Government & Donorinputs.• Analysis of chronology of events to determine causallinks between donor support to reform and observedchangescriteria/ Possible indicatorsJudgementat output level.
a inputs & context: the design• PfM reformof counterfactual: would theseofConsiderationoutputs have been for PFM reforms committed and• Government fundsgenerated in absence of donorA.1.What has been the na-support disbursed by yearactually to PFM reform? over the evaluation period.ture and scale of PFm• Donor funds for PFM reforms committed andB.2.reform inputs provided• Consistency of outputs produced with plannedHow efficiently wereactually disbursed by yeartiming) evaluation periodby Government and Do-programme (quantity and over thethese outputs gener-• Natureof coordination between outputs effortsnors? Was the pacing• Extent of support provided to PFM reformated?• (equipment, training,sequencing of outputQuality of pacing and TA, diagnostic work.)and sequencing of re-• production: for reforms by function (based on PFMFocal areasforms appropriate and“clusters”coherent/ efficient?• internally as in Andrews 2010)cost-effective? Was• Focal areas for reforms by organisational location/the cost per output ac-• Consistent with capacity endowments?Appropriate to degree of ownership/ nature of• level of government (Ministry of Finance, Sectorceptable?ministries,admin Governments, Parliament, CSOs,political & Local support for reform?• etc.)Relative cost of outputs in relation to budgeted costs.• Influence of external support to PFM reform uponpacing and sequencing.• Counterfactual: would pacing & sequencing ofoutputs have been different in absence of externalPFM support?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
• Degree of ownership of reforms at the administrativeB.3. What have been theand political levels (narrow vs. broad; depth ofbinding external con-commitment to reform and motivations for thisstraints on the deliverycommitment).of PFm reform: politi-cal, financing or policy• Quality of interaction between administrative andpolitical cadres.factors? How has this• Extent and nature of political accountability (withinvaried across the dif-ruling party, vis-à-vis Legislature, vis-à-visferent PFm reformElectorate & Civil Society) and extent to which this iscomponents?patronage or performance-based.• Economic context and its influence on financing ofPFM reform: per capita GDP and % growth inevaluation period, significance of domestic revenue,significance of aid, level of total and discretionarypublic spending; absence/ presence of macro crises.• Has timeliness of funding been an issue?
107
• Nature of “conventional wisdom” on PFM reforms:does prevailing thinking exclude certain reformoptions?• Has policy discussion on PFM reform been open interms of range of participants and range of ideas?• Overall, what have been the binding constraints onthe PFM production possibility frontier?
(ii) improving the efficiency of operations of the Internal RevenueService through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and trainingsupport.
ferent PFm reformcomponents?previous phase, helped to enable the reforms to monitor and respond toavailability (or not) of policy space.
108
Evaluation QuestionPuFmArP (1997-2003)• During the period studied, reform of revenue management was part ofthe PUFMARP suite of reforms. Reforms were concentrated on: (i)improving income tax administration through introducing a uniqueincome taxpayer code; and (ii) improving the efficiency of theinstitutional structure administering revenue collections.• The main problem that the reforms attempted to solve was the fact thatthere was a significant tax gap, i.e. that actual revenue collections (e.g.as a % of GDP) were lower than their potential. This was due to the factthat: (i) taxpayer compliance (particularly, income tax, but also• customs) was considered toto increase the efficiency of collections wasThe reforms were intended be low; and (ii)revenuecollection. The reforms enabled a significant increase in revenuesweak, as was revenue reporting and forecasting.duringperiod.• Whilst the reforms were originally targeted at both income tax andIn terms duties, and whilst there are three PEFA indicators (including• customs of systems, institutional support to the different revenuetotal 9 dimensions) on tax administration, in reality, many of itsservices, the PUFMARP project subsequently concentrated thesupportdimensions refer indirect tax measures, and basis, or to otheron income tax andto the underlying legislative reallocatedother areasthat are affected by the type of administrative reforms undertaken inoriginally-planned project resources to BPEMS.Ghana.• Specifically, the revenue management reforms during this periodcovered: (i) introduction of a unique Taxpayer Identification Number; (ii)• the re-introduction of VAT, whose introduction had been stalled since As the reforms relate to improving the efficiency of revenueadministration, and thus on1995; and (iii) staff training. revenue collections, they provide themeans with which MDAs and included a limited their service funding,• DP support under PUFMARPMMDAs undertakeamount of IDAprovision.which took the form of: technical assistance, training support, andlimited equipment.reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughthe provision of capacity building, equipment and analytical software;(ii) improving the efficiency of operations of the Internal RevenueService through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and trainingsupport.
ruling party, vis-à-vis Legislature, vis-à-visElectorate & Civil Society) and extent to which this ispatronage or performance-based.• Economic context and its influence on financing ofPFM reform: per capita GDP and % growth inevaluation period, significance of domestic revenue,significance of aid, level of total and discretionarypublic spending; absence/ presence of macro crises.• Has timeliness of funding been an issue?Judgement criteria/ Possible indicatorsFindings
a inputs & context: the design• PfM reformofNature of “conventional wisdom” on PFM reforms:does prevailing thinking exclude certain reform• Government funds for PFM reforms committed andA.1.What has been the na-options?actually disbursed by year over the evaluation period.ture and scale of PFm• Has policy discussion reformsreform been open inreform inputs provided• Donor funds for PFM on PFM committed andterms of range of participants and evaluation periodactually disbursed by year over therange of ideas?by Government and Do-• Nature of support provided to PFM reform efforts• Overall, what have been the binding constraints onnors?the PFM production possibility frontier?(equipment, training, TA, diagnostic work.)• How has thisfor reforms by function (based on PFM• Focal areas varied between different reformcomponents?“clusters” as in Andrews 2010)• Focal areas for reformsdonor support tolocation/c. outcomes & overall assessment of PfM reform and ofby organisationalPfM reformlevel of government (Ministry of Finance, SectorChanges in performance of PFM system over periodC.1. What have been the in-• ministries, Local Governments, Parliament, CSOs,as measured by HIPC AAP indicators and PEFAtermediate outcomesetc.)indicators.of PFm reforms, interms of changes in the• Nature of PFM system changes identified:quality of PFm systems• By PFM function (according to PFM cluster)• By nature of change (Upstream, De jure,?Concentrated vs. Downstream, De facto,Deconcentrated).
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
C.2. to what extent are theoutcomes generatedrelevant to improve-ments in the quality ofservice delivery, par-ticularly for womenand vulnerablegroups?
• Changes in quality of PFM system directly relevant toservice delivery, especially for women & vulnerablegroups:• Classification of the budget (PI-5): does this permitprotection/ prioritisation of specific functions orprogrammes?• Information in Budget documentation (PI-6): doesthis include performance data or targets for priorityprogrammes?• Does the quality of in-year reports and final accounts(PI-24 & 25) and the access of the public (PI-10)permit analysis and discussion of spending &performance for priority programmes?• Is the predictability of funds for commitment ofexpenditure (PI-16) improving?
• Is the quality of information on resources received byservice delivery units (PI-23) improving?• Extent to which improvements may be attributed tothe PFM reform programme, and particularly toexternal support?• Consider counterfactual: would these changes havehappened in the absence of the reform programme/the externally supported component of the reformprogramme?
C.3. Have reform efforts
• Causes of changes, specifically contribution of
Introduction of GrA (2010-2013)• BMZ continued to support the RAGB (and subsequent form the GRA)• Effectiveness is measured in terms of whether or not the reforms werethrough the continuation of its Good Financial Governance programme.
outcomes generatedrelevant to improve-ments in the quality ofservice delivery, par-ticularly for womenand vulnerablegroups?administration, and thus on revenue collections, they provide themeans with which MDAs and MMDAs undertake their service provision.
Evaluation Question
service delivery, especially for women & vulnerablegroups:• Classification of the budget (PI-5): does this permitprotection/ prioritisation of specific functions orprogrammes?• Information in Budget documentation (PI-6): doesthis include performance data or targets for priorityprogrammes?• Does the quality of in-year reports and final accounts(PI-24 & 25) and the access of the public (PI-10)permit analysis and discussion of spending &performance for priority programmes?• Is the predictability of funds for commitment ofexpenditure (PI-16) improving?Judgement criteria/ Possible indicatorsFindings
a inputs & context: the design• PfM reformof information on resources received byofIs the qualityservice delivery units PFM reforms committed and• Government funds for(PI-23) improving?A.1.What has been the na-• actually disbursed by year overmayevaluation period.Extent to which improvements the be attributed toture and scale of PFmthe PFM reform programme, and particularly toreform inputs provided• Donor funds for PFM reforms committed andexternal support?actually disbursed by year over the evaluation periodby Government and Do-• Nature of counterfactual: would these changes have• Consider support provided to PFM reform effortsnors?happened in the absence of the reform programme/(equipment, training, TA, diagnostic work.)• the externally supportedby function (based reformFocal areas for reforms component of the on PFMprogramme? in Andrews 2010)“clusters” as• Focal areas for reforms by organisational location/• Causes of changes, specifically contribution ofC.3. Have reform effortslevel of government (Ministry of Finance, Sectoridentified PFM reform outputs.been effective? If not,Consistency of actual changes with objectives andwhy not? If yes, to what• ministries, Local Governments, Parliament, CSOs,etc.)outcome targets of PFM reform programme.extent PFm reform• Extent to which external constraints (political,outputs been a causalfinancial, policy) have undermined effectiveness offactor in the changesPFM reform.identified in intermedi-• Consider counterfactual: what change could haveate outcomes?been expected at Intermediate Outcome level inabsence of PFM reform as a whole, and in absence ofexternal support to PFM reform?
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
C.4. to what extent do thegains identified at theIntermediate out-come levels appearsustainable? Is theprocess of PFm re-form sustainable?
• Recent trends in Outcomes: do these suggest pastgains will be sustained?• Is there a commitment (at political andadministrative levels) to continue PFM reforms?• Are there organisational structures in place (onsupply & demand sides) to sustain PFM reforms?• Is there the financial and technical capacity withinGovernment to sustain PFM reform in the absence ofexternal support?• If not, is there a framework in place for continuingexternal support while building local capacity?
PuFmArP (1997-2003)• During the period studied, reform of revenue management was part ofthe PUFMARP suite of reforms. Reforms were concentrated on: (i)improving income tax administration through introducing a uniqueincome taxpayer code; and (ii) improving the efficiency of theinstitutional structure administering revenue collections.• The main problem that the reforms attempted to solve was the fact thatthere was a significant tax gap, i.e. that actual revenue collections (e.g.• as a % of GDP)is measuredthan theirof whetherThis was due to the wereEffectiveness were lower in terms potential. or not the reforms factthat: (i) taxpayer compliance (particularly, income tax, but alsoimplemented as planned. By this measure, the outputs achievedcustoms)those expected; nevertheless, given greaterof collections wasmatched was considered to be low; and (ii) efficiency politicalweak, as was revenue reporting and forecasting.commitment, the speed of implementation of organisational reform in• Whilst the may have been faster. targeted at both income tax andparticular reforms were originallycustoms duties, and institutional support to the different revenueservices, the PUFMARP project subsequently concentrated its supporton income tax and indirect tax measures, and reallocated otheroriginally-planned project resources to BPEMS.• Specifically, the revenue management reforms during this periodcovered: (i) introduction of a unique Taxpayer Identification Number; (ii)the re-introduction of VAT, whose introduction had been stalled since The fact that many of the reforms (e.g. TIN, functional VAT) have been • 1995; and (iii) staff training.• DP support under PUFMARPof sustainability. This is particularly thesustained indicates the level included a limited amount of IDA funding,which took the form of: technical assistance, training support, andcase with the latter, since an earlier (mid-1990s) introduction of VAT limited successful.was notequipment. Changes in the specification of the system(including a more appropriate VAT rate) enabled it to gain traction.reforms to rAGB (2003-2008)• Following PUFMARP, reforms to revenue administration continuedunder GoG support and GTZ (BMZ), as part of the latter’s Good FinancialGovernance programme.• These were focussed on the following measures: (i) establishment anddefining the strategic operations of a Tax Policy Unit in MoFEP throughthe provision of capacity building, equipment and analytical software;
109
(ii) improving the efficiency of operations of the Internal RevenueService through computerisation and organisational changes; and (iii)restructuring of the 3 separate revenues services into a unified GhanaRevenue Authority, including organisational development and HRchanges (on-going).• Total GTZ funding support (in two 3-year phases, 2003-2005, and 2006-2008) amounted to €€6.8 mn (BMZ) + €2 mn (SECO).• The support mainly took the form of technical assistance and training
110
PfM Reform component/initiative:Findings• Inputs were primarily GoG-based, at approximately �12 mn annually.External funding of less than an estimated €<0.5 mn was provided bythe EU, mainly for on-site training.• IAA provided capacity building to IAU staff, including preparation ofaudit manuals, training, on-site support programmes – going to thefield to help internal auditors, do risk assessment, plan and implementmanagement observations and response to the observations.
internal audit
Evaluation Question
Judgement criteria/ Possible indicators
a inputs & context: the design of PfM reform
A.1.What has been the na-ture and scale of PFmreform inputs providedby Government and Do-nors?
• Government funds for PFM reforms committed andactually disbursed by year over the evaluation period.• Donor funds for PFM reforms committed andactually disbursed by year over the evaluation period• Nature of support provided to PFM reform efforts(equipment, training, TA, diagnostic work.)• Focal areas for reforms by function (based on PFM“clusters” as in Andrews 2010)• Focal areas for reforms by organisational location/level of government (Ministry of Finance, Sectorministries, Local Governments, Parliament, CSOs,etc.)• The introduction of the internal audit function was designed andmanaged by the IAU with little external support.
A.2.What type of struc-tures have been usedfor the design andmanagement of thesereform inputs? Havethese structuresserved to provide a co-ordinated and harmo-nised delivery frame-work?
• Structure of design team and related consultationprocess: Balance of inputs by Donors/ Govt/Consultants; Extent of input/ consultation with endusers of PFM system (Sector ministries, LGs andservice institutions).• Management & co-ordination structure for PFMreforms (Ad hoc Project Units vs. Normalmanagement structures; Govt-controlled vs. SharedDonor-Govt management; Use of consultants formanagerial or purely advisory roles)• Arrangements for monitoring & evaluation.• Level of harmonisation and alignment of differentdonor contributions.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
A.3. What types of comple-mentary actions haveDonors taken to sup-port PFm reforms andwhat has been their
• Use of country systems: extent to which external aidis ‘on plan’, ‘on budget’ and ‘on treasury’?• Ranking of donor efforts to support national systemsas expressed in Paris Declaration monitoringreports and PEFA indicators D2 and D3.
• There were little/no complementary actions taken by DPs.
significance? Havethey had any influenceon the external con-straints to reform?
• Scale and evolution over the evaluation period of aidflows provided as GBS, SBS or debt relief and relativecontribution to reduction of treasury managementcosts and to discretionary resources available toGovernment, (including effect on financingconstraints for PFM reform outputs).• Nature and evolution over time of dialogue on PFMwithin GBS/PRSC, SBS structures or similar donor-govt fora, including use of PFM reform conditions inGBS/ SBS disbursement conditions.• Influence of dialogue on the “political” and “policy
these structuresserved to provide a co-ordinated and harmo-nised delivery frame-work?
• Management & co-ordination structure for PFMreforms (Ad hoc Project Units vs. Normalmanagement structures; Govt-controlled vs. SharedDonor-Govt management; Use of consultants formanagerial or purely advisory roles)• Arrangements for monitoring & evaluation.• Level of harmonisation and alignment of differentdonor contributions.• There were little/no complementary actions taken by DPs.
A.3. What types of comple-mentary actions haveDonors taken to sup-port PFm reforms andwhat has been theirEvaluation QuestionFindings• Inputs were primarily GoG-based, at approximately �12 mn annually.External funding of less than an estimated €<0.5 mn was provided bythe EU, mainly for on-site training.• IAA provided capacity building to IAU staff, including preparation ofaudit manuals, training, on-site support programmes – going to thefield to help internal auditors, do risk assessment, plan and implementmanagement observations and response to the observations.
• Use of country systems: extent to which external aidis ‘on plan’, ‘on budget’ and ‘on treasury’?• Ranking of donor efforts to support national systemsas expressed in Paris Declaration monitoringreports and PEFAPossibleD2 and D3.Judgement criteria/indicatorsindicators
• The introduction of the internal audit function was designed andmanaged by the IAU with little external support.
• There was little or no external or regional pressure for reform.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
• There were little/no complementary actions taken by DPs.
111
a inputs & context: the design• PfM reformofScale and evolution over the evaluation period of aidsignificance? Havethey had any influenceflows provided as GBS, SBS or debt relief and relative• Government funds for PFM reforms committed andA.1.What has been the na-on the external con-contribution to reduction of treasury managementactually disbursed by year over the evaluation period.ture and scale of PFmstraints to reform?costs and to discretionary resources available toreform inputs provided• Donor funds for PFM reforms committed andGovernment, (including effect on financingactually disbursed by year over the evaluation periodby Government and Do-constraints for PFM reform outputs).• Nature of support provided to PFM reform effortsnors?• Nature and evolution over time of dialogue on PFM(equipment, training, TA, diagnostic work.)within GBS/PRSC, SBS structures or similar donor-• Focal areas for reforms by function (based on PFMgovt fora, including use of PFM reform conditions in“clusters” as in Andrews 2010)GBS/ SBS disbursement conditions.• Focal areas for reforms by organisational location/• Influence of dialogue on the “political” and “policylevel of government (Ministry of Finance, Sectorspace” constraints restricting choice of PFM reformministries, Local Governments, Parliament, CSOs,outputs.etc.)• (Consistency of actual changes with PFM reformtriggers and/or points of and related GBS/ PRSC• Structure of design teamemphasis in consultationA.2.What type of struc-dialogue)process: Balance of inputs by Donors/ Govt/tures have been used• Consultants; Extent of input/ consultation with endConsideration of counter-factual:for the design and• users of PFM system reformsministries, LGs andOutputs: would PFM (Sector have evolvedmanagement of thesedifferently in absenceservice institutions). of GBS and related policyreform inputs? Havedialogue structures?• Management & co-ordination structure for PFMthese structuresOutcomes: Would intermediate outcomesserved to provide a co-• reforms (Ad hoc Project Units vs. Normal have beendifferent if use of country Govt-controlled vs. Sharedmanagement structures;systems had been lower?ordinated and harmo-Donor-Govt management; Use of consultants fornised delivery frame-• managerial or purely advisory roles)Nature of engagement with PFM reform issuesA.4. work? extent hasto whatwithin civil society, academia and the media.there been domestic• Arrangements for monitoring & evaluation.• Extent of participation of CSOs in policy dialogue onpublic pressure or re-• Level of harmonisation and alignment of differentPFM reform.gional institutionaldonor contributions.pressure in support of• Presence/ absence of specific research andadvocacy by CSOs on PFM reform issues.PFm types of comple-• Use of country systems: extent to which external aidA.3. Whatreform and whatRelevance ‘on budget’ and political outcomes:has beenactions havethe influence• is ‘on plan’,of PFM issues to‘on treasury’?mentary– Discussion of PFM issues within nationalon the external con-• Ranking of donor efforts to supportelection systemsDonors taken to sup-straints to reform?as campaigns in Paris Declaration monitoringexpressedport PFm reforms and– Discussion of service delivery issuesreports and PEFA indicators D2 and D3. related towhat has been theirPFM in election campaigns• Scale and evolution over the evaluation period of aidsignificance? Have– Voter concerns during elections (from Afro flows provided as GBS, SBS or debt relief and relativethey had any influenceBarometer and other relevant publications)contribution to reduction of treasury managementon the external con-– Existence of regional or international bodies (e.g.costs and to discretionary resources available tostraints to reform?WAEMU, EITI) promoting specific normsGovernment, (including effect on financing on PFMissues, and their relative outputs).constraints for PFM reforminfluence on domesticpolitical evolution• Nature anddiscourse.over time of dialogue on PFM– Evolution in the SBS structures CSO engagementwithin GBS/PRSC,quality of public/or similar donor-
112
public pressure or re-gional institutionalpressure in support ofPFm reform and whathas been the influenceon the external con-straints to reform?Findings• Inputs were primarily GoG-based, at approximately �12 mn annually.External funding of less than an estimated €<0.5 mn was provided bythe EU, mainly for on-site training.• IAA provided capacity building to IAU staff, including preparation ofaudit manuals, training, on-site support programmes – going to thefield to help internal auditors, do risk assessment, plan and implementmanagement observations and response to the observations.
Evaluation Question
• Extent of participation of CSOs in policy dialogue onPFM reform.• Presence/ absence of specific research andadvocacy by CSOs on PFM reform issues.• Relevance of PFM issues to political outcomes:– Discussion of PFM issues within electioncampaigns– Discussion of service delivery issues related toPFM in electionPossible indicatorsJudgement criteria/campaigns
• Internal audit was a new function. Prior to the reform, internal auditing• The introduction of the internal audit function was designed andwas restricted to compliance external support.managed by the IAU with littlechecking of payment vouchers (pre-audit), rather than on systemic reviews on behalf of management ofinternal control systems. The reporting relationship was with theController and Accountant General. Under PUFMARP, the auditfunction was separated into internal and external audits.• The Internal Audit Agency Act specified the new function in broadterms, and this guided the design of the reforms.• With little DP support, there were no DP harmonisation issues.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
• There were little/no complementary actions taken by DPs.
a inputs & context: the design of PfM reform– Voter concerns during elections (from Afro Barometer and other relevant publications)• Government funds for PFM reforms committed andA.1.What has been the na-– Existence of regional or international bodies (e.g.actually disbursed by year over the evaluation period.ture and scale of PFmWAEMU, EITI) promoting specific norms onreform inputs provided• Donor funds for PFM reforms committed and PFMissues, and their relative influence on domesticactually disbursed by year over the evaluation periodby Government and Do-political discourse.• Nature of support provided to PFM reform effortsnors?– Evolution in the qualitydiagnosticCSO engagement(equipment, training, TA, of public/ work.)with PFM for reforms by function influence PFM• Focal areas issues over time and its (based on onPFM reform Andrews“clusters” as inoutputs. 2010)• Focal areas forof counter-factual: would PFM reformConsideration reforms by organisational location/• outputs have evolved differently in absence oflevel of government (Ministry of Finance, Sectordomestic public pressure or pressure from regional/ministries, Local Governments, Parliament, CSOs,international institutions?etc.)• Scale and of design team and related to identifiedA.5. What type of struc-• Structure focus of support in relation consultationA.2.How relevant was thePFM weaknesses at start and during implementationPFm reform pro-process: Balance of inputs by Donors/ Govt/tures have been usedof reforms (including weaknesses in HRgramme to the needsConsultants; Extent of input/ consultation with endfor the design andendowments, quality of institutions & rules, qualityand the institutionalusers of PFM system (Sector ministries, LGs andmanagement of theseof systems & businesscontext? Was donorservice institutions). processes, and quality ofreform inputs? Haveorganisations).support consistent• Management & co-ordination structure for PFMthese structuresConsistency of donor-funded vs. Normalwith national priori-served to provide a co-• reforms (Ad hoc Project Unitssupport withGovernment PFM reform Govt-controlled vs. Sharedties? to what extentmanagement structures; efforts, with ongoing publicordinated and harmo-administration reforms and overall Govt policies.were adaptationsDonor-Govt management; Use of consultants fornised delivery frame-Extent to which scope and focus of support were• managerial or purely advisory roles)made in response towork?adapted to the for monitoring & evaluation.the context and the• Arrangementscontext, especially to the level ofownership and the capacity for reformchanging national pri-• Level of harmonisation and alignment of differentimplementation.orities?donor contributions.• Consistency & coherence with wider Donor policiesA.3. What types of comple-• Use of country systems: extent to which external aidin countrymentary actions haveConsistency with the Paris ‘on treasury’?• is ‘on plan’, ‘on budget’ and Declaration objectives as• Rankingownership, harmonisation and alignment.Donors taken to sup-regards of donor efforts to support national systemsport PFm reforms and• as expressed in Paris Declaration monitoringFlexibility in design of support and existence ofreports and PEFA indicators D2 and D3.what has been theiradequate feed-back/ learning mechanisms in orderto adapt the design and implementation of support to• Scale andneeds. over the evaluation period of aidsignificance? Havechanging evolutionflows provided as GBS, SBS or debt relief and relativethey had any influenceb. outputs:external con-PfM reformscontribution to reduction of treasury managementon thethe delivery ofstraints to been thereform?• costs and to discretionary resources available toDocumentation of outputs :B.1. What have• Government, (including effect PFM clusters inby PFM function (according to on financingoutputs of the PFm re-constraints for PFM reform outputs).Andrews 2010) andform process and to• Nature of output (People &time of dialogue on PFM• by type and evolution over Skills; Laws & Rules;what extent has directwithin GBS/PRSC, SBS structures or similar donor-Systems & Business processes; Organisationaldonor support contrib-govt fora, including use of PFM reform conditions inchanges)uted to these outputs?• GBS/ SBSof outputs to PFM reform inputs,Matching disbursement conditions.• Influence of dialogue on the “political” and “policydifferentiating if possible Government & Donorspace” constraints restricting choice of PFM reforminputs.outputs.The main outputs during the study period include:• Passage of the Internal Audit Agency Act, 2003• Establishment of the Internal Audit Agency (IAA) to co-ordinate,facilitate and provide quality assurance for internal audit activitieswithin MDAs and MMDAs. The operational framework of the internalaudit units had been enhanced by the adoption of public sector internalaudit regulations and standards and the use of an internal audit manualand programmes.• Internal audit standards adopted from INTOSAI and AFROSAI E.
ties? to what extentwere adaptationsmade in response tothe context and thechanging national pri-orities?
terms, and this guided the design of the reforms.• With little DP support, there were no DP harmonisation issues.
Evaluation Question
Government PFM reform efforts, with ongoing publicadministration reforms and overall Govt policies.• Extent to which scope and focus of support wereadapted to the context, especially to the level ofownership and the capacity for reformimplementation.• Consistency & coherence with wider Donor policiesin country• Consistency with the Paris Declaration objectives asregards ownership, harmonisation and alignment.• Flexibility in design of support and existence ofadequate feed-back/ learning mechanisms in orderto adapt the design and implementation of support toJudgementneeds.Possible indicatorschangingcriteria/Findings
a inputs & context: the design of PfM reformb. outputs: the delivery of PfM reforms
A.1.What have been thehas been the na-B.1.ture andof the PFm re-outputs scale of PFmreform inputs providedform process and toby Government directwhat extent hasand Do-nors?donor support contrib-uted to these outputs?
• • • • •
•
• •
• •
A.2.What type of struc-tures have been usedfor the design andmanagement of thesereform inputs? Havethese structuresserved to provide a co-ordinated and harmo-nised delivery frame-work?
•
• •
Government fundsoutputs : reforms committed andDocumentation of for PFMactuallyfunction (according to PFM evaluation period.by PFM disbursed by year over the clusters inDonor funds for PFMAndrews 2010) and reforms committed andactually disbursed by year over the evaluation periodby type of output (People & Skills; Laws & Rules;Nature of& Business processes; OrganisationalSystems support provided to PFM reform efforts(equipment, training, TA, diagnostic work.)changes)Focal areas for reforms by reform (basedMatching of outputs to PFMfunctioninputs, on PFM“clusters” as inifAndrews 2010)differentiating possible Government & DonorFocal areas for reforms by organisational location/inputs.level of government (Ministry of Finance, SectorAnalysis of chronology of events to determine causalministries, Local Governments,reform and observedlinks between donor support to Parliament, CSOs,etc.)changes at output level.Consideration of counterfactual: would theseStructure of design team and related consultationoutputs have been generated in absence of donorprocess: Balance of inputs by Donors/ Govt/support to PFM reform?Consultants; Extent of input/ consultation with endusers of PFM system (Sector ministries, LGs andservice institutions).Management & co-ordination structure for PFMreforms (Ad hoc Project Units vs. Normalmanagement structures; Govt-controlled vs. SharedDonor-Govt management; Use of consultants formanagerial or purely advisory roles)Arrangements for monitoring & evaluation.Level of harmonisation and alignment of differentdonor contributions.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
A.3. What types of comple-mentary actions haveDonors taken to sup-port PFm reforms andwhat has been their
• Use of country systems: extent to which external aidis ‘on plan’, ‘on budget’ and ‘on treasury’?• Ranking of donor efforts to support national systemsas expressed in Paris Declaration monitoringreports and PEFA indicators D2 and D3.
• Inputs were primarily GoG-based, at approximately �12 mn annually.The main outputs during the study period include:• External of the Internal Audit an estimated €<0.5 mn was provided byPassage funding of less than Agency Act, 2003the EU, mainly forthe Internal Audit Agency (IAA) to co-ordinate,• Establishment of on-site training.• IAA provided capacityquality assurance forincluding preparation offacilitate and provide building to IAU staff, internal audit activitiesaudit manuals, training, on-site support programmes – going to thewithin MDAs and MMDAs. The operational framework of the internalfield to help internal auditors, do risk assessment, plan sector internalaudit units had been enhanced by the adoption of public and implementmanagement observations and response to the observations. manualaudit regulations and standards and the use of an internal auditand programmes.• Internal audit standards adopted from INTOSAI and AFROSAI E.• Establishment of Internal Audit Units (IAUs) in MDAs, MMDAs, asfollows:– Ministries: 21 established (out of 23);– Departments and agencies: 125 (out of 149);• The introduction of the internal audit function was designed and– MMDAs: 155 (out of 170).managed by the IAU with little external support.• The operational efficiency of the MDAs and MMDAs has been measuredand monitored by the IAA by reviewing the submissions receivedrelating to the: (i) internal audit charters; (ii) annual audit plans and (iii)quarterly audit reports. Data for 2010 show:– Annual audit plans: 216 (72%) had been prepared, comprising 14Ministries, 95 Departments and Agencies and 107 MMDAs.– Signed charters: 14 Ministries, 64 Departments and Agencies, 124MMDAs (total of 67%)– Around 27% of quarterly Internal Audit reports were received:– Ministries: 27 reports– Departments and agencies: 124 reports– MMDAs: 172.• There were little/no complementary actions takenfindings, includingResponsibility to enforce and follow up on auditing by DPs.for internal audit, is given to the Audit Reports ImplementationCommittees (ARICs) (80% of total).– Ministries: 22– Departments and agencies: 113– MMDAs: 124
113
significance? HaveB.2. they had any influenceHow efficiently werethese outputs gener-on the external con-ated? Was reform?straints to the pacingand sequencing of re-forms appropriate andcost-effective? Wasthe cost per output ac-ceptable?
• Scale and evolution over the evaluation period of aid• Consistency of outputs produced with plannedflows provided as GBS, SBS or debt relief and relativeprogramme (quantity andof treasury managementcontribution to reduction timing)• Extent of coordination between outputscosts and to discretionary resources available to• Quality of pacing and sequencing financingGovernment, (including effect on of outputproduction: for PFM reform outputs).constraints• internally coherent/ efficient? of dialogue on PFM• Nature and evolution over time• Consistent with capacity endowments?within GBS/PRSC, SBS structures or similar donor-• Appropriate to degree of ownership/ nature of
• The efficiency of the reforms has been hampered by the unclearreporting relationships of the IAUs in MDAs and MMDAs. At MDA andMMDA levels, IAUs have dual reporting responsibilities between theIAA and MDA and between IAA and MMDA. The IAA posts the InternalAuditors to MDAs and MMDAs and, although they report directly to theIAA, they are expected to be part of the MDA/MMDA. This dualallegiance creates tensions and suspicion among the IAs, MDAs andMMDAs, which has reduced the efficiency of reform.
114
Evaluation Question• The efficiency of the reforms has been hampered by the unclearreporting relationships of the IAUs in MDAs and MMDAs. At MDA and• Inputs were primarily GoG-based, at approximately �12 between theMMDA levels, IAUs have dual reporting responsibilities mn annually.External funding of less than an estimated €<0.5 mn was provided byIAA and MDA and between IAA and MMDA. The IAA posts the Internalthe EU, mainly for and MMDAs and, although they report directly to theAuditors to MDAs on-site training.• IAA provided expectedbuilding to of the MDA/MMDA. preparation ofIAA, they are capacity to be part IAU staff, including This dualaudit manuals, training, on-site support programmes – going to andallegiance creates tensions and suspicion among the IAs, MDAs thefield to help internal auditors, do risk assessment, plan and implementMMDAs, which has reduced the efficiency of reform.• management observations and response totogether, the efficiency ofAs internal and external audit need to work the observations.internal audit reforms depends on a good working relationshipbetween the IAA and Ghana Audit Service (GAS). To this end, attemptsto strengthen co-operation has been led by a formal dialogue betweenthe two institutions to examine areas of commonality. Formal dialoguehas been established with a Memorandum of Understanding that spellsout introduction of the internal work, dialogue processes and use• Thethe shared planning of auditaudit function was designed and ofoutcomes to enhance effectiveness and efficiency.managed by the IAU with little external support. GAS has tried to usethe work of IAA wherever possible in cases where the standardsadopted by the internal audits are sufficiently robust to permit theexternal auditor to place reliance on that work.• Main external constraint has been political commitment, with IAAreports being delayed at the senior Executive level (Office of thePresident) before being passed to Parliament. Slow responses frompolitical leadership, as well as weak administrative decision-making,affected negatively the pace of reform.• Effectiveness of internal audit constrained by inaction (insufficientfollow-up to audit reports) rather than funding constraints.• In addition, the IAA indicates that its main challenge in obtaining higherIAU compliance is capacity constraints amongst IAU staff to• There were little/no complementary actions taken by DPs. undertaketheir duties to the required standard.• Financial resources have not been a binding constraint.
Judgement criteria/ Possible indicators
– Ministries: 27 reports– Departments and agencies: 124 reports– MMDAs: 172.• Responsibility to enforce and follow up on auditing findings, includingfor internal audit, is given to the Audit Reports ImplementationCommittees (ARICs) (80% of total).– Ministries: 22– Departments and agencies: 113Findings– MMDAs: 124
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
a inputsefficiently wereofConsistencyB.2. How& context: the design• PfM reformof outputs produced with plannedprogramme (quantity and timing)these outputs gener-• Government funds for PFM reforms committed andA.1.What has beenpacing• Extent of coordination between outputsated? Was the the na-ture sequencing of re-• actually disbursed bysequencing ofevaluation period.Quality of pacing and year over the outputand and scale of PFmDonor fundsreform inputs provided• production: for PFM reforms committed andforms appropriate andby Government and Do-• actually disbursed by year over the evaluation periodinternally coherent/ efficient?cost-effective? Wasnors? per output ac-• Nature of support provided to PFM reform efforts• Consistent with capacity endowments?the cost• (equipment, training, TA, ownership/ nature ofAppropriate to degree of diagnostic work.)ceptable?• political & admin support for reform?Focal areas for reforms by function (based on PFM“clusters” as of outputs in relation to budgeted costs.• Relative cost in Andrews 2010)• Focal areas external support to PFM reform upon• Influence of for reforms by organisational location/level ofand sequencing.pacing government (Ministry of Finance, Sectorministries, Local Governments, Parliament,of• Counterfactual: would pacing & sequencing CSOs,etc.)outputs have been different in absence of externalPFM support?• Structure of design team and related consultationA.2.What type of struc-process: Balance of inputs by Donors/ Govt/tures have been usedConsultants; Extent of input/ consultation with endfor the design andusers of PFM system (Sector ministries, LGs andmanagement of theseservice institutions).reform inputs? Have• Management & co-ordination structure for PFMthesehave been thestructures• Degree of ownership of reforms at the administrativeB.3. Whatreforms (Ad hoc Project Units vs. Normal ofserved external con-and political levels (narrow vs. broad; depthbindingto provide a co-managementto reform and motivations for thisordinated and harmo-commitment structures; Govt-controlled vs. Sharedstraints on the deliveryDonor-Govt management; Use of consultants fornised delivery frame-commitment).of PFm reform: politi-work?Quality of interaction advisory roles)cal, financing or policy• managerial or purely between administrative and• Arrangements for monitoring & evaluation.political cadres.factors? How has this• Level of harmonisation and alignment of different• Extent and nature of political accountability (withinvaried across the dif-donor contributions. Legislature, vis-à-visruling party, vis-à-visferent PFm reform
A.3. components? comple-What types ofmentary actions haveDonors taken to sup-port PFm reforms andwhat has been their
significance? Havethey had any influenceon the external con-straints to reform?
• Electorate & Civil Society) and extent to which this isUse of country systems: extent to which external aidpatronage or performance-based.is ‘on plan’, ‘on budget’ and ‘on treasury’?• Economic contextefforts to supporton financing of• Ranking of donor and its influence national systemsPFM reform: per capita GDP and % growth inas expressed in Paris Declaration monitoringevaluation period, indicators D2 and D3.reports and PEFA significance of domestic revenue,significance of aid, level of total and discretionary• publicand evolution over the evaluationmacro crises.Scale spending; absence/ presence of period of aidflows provided as GBS, SBS or debt relief and relative• contribution to reduction of treasury managementHas timeliness of funding been an issue?• costs and “conventional wisdom” on PFM reforms:Nature of to discretionary resources available todoes prevailing thinking exclude financingGovernment, (including effect oncertain reformoptions?constraints for PFM reform outputs).• Has policy discussion on PFM reform been open in• Nature and evolution over time of dialogue on PFMterms GBS/PRSC, SBS structures or similar donor-within of range of participants and range of ideas?• govt fora, includingbeenof PFM reform conditions inOverall, what have use the binding constraints onthe PFM production possibility frontier?GBS/ SBS disbursement conditions.• How has this varied between different and “policy• Influence of dialogue on the “political” reformcomponents?space” constraints restricting choice of PFM reformoutputs.
straints on the deliveryof PFm reform: politi-cal, financing or policyfactors? How has thisvaried across the dif-ferent PFm reformcomponents?
reports being delayed at the senior Executive level (Office of thePresident) before being passed to Parliament. Slow responses frompolitical leadership, as well as weak administrative decision-making,affected negatively the pace of reform.• Effectiveness of internal audit constrained by inaction (insufficientfollow-up to audit reports) rather than funding constraints.• In addition, the IAA indicates that its main challenge in obtaining higherIAU compliance is capacity constraints amongst IAU staff to undertaketheir duties to the required standard.• Financial resources have not been a binding constraint.
Evaluation Question
commitment to reform and motivations for thiscommitment).• Quality of interaction between administrative andpolitical cadres.• Extent and nature of political accountability (withinruling party, vis-à-vis Legislature, vis-à-visElectorate & Civil Society) and extent to which this ispatronage or performance-based.• Economic context and its influence on financing ofPFM reform: per capita GDP and % growth inevaluation period, significance of domestic revenue,significance of aid, level of total and discretionarypublic spending; absence/ presence ofJudgement criteria/ Possible indicatorsmacro crises.Findings• Inputs were primarily GoG-based, at approximately �12 mn annually.External funding of less than an estimated €<0.5 mn was provided bythe EU, mainly for on-site training.• IAA provided capacity building to IAU staff, including preparation ofaudit manuals, training, on-site support programmes – going to thefield to help internal auditors, do risk assessment, plan and implementmanagement observations and response to the observations.
a inputs & context: the design• PfMtimeliness of funding been an issue?ofHasreform• Nature of “conventional wisdom” on committed and• Government funds for PFM reforms PFM reforms:A.1.What has been the na-does prevailing thinking exclude certain reformactually disbursed by year over the evaluation period.ture and scale of PFmDonor funds for PFM reforms committed andreform inputs provided• options?Has policy discussion on PFM reform been open inby Government and Do-• actually disbursed by year over the evaluation periodterms of range of provided to and reform ideas?• Nature of supportparticipantsPFMrange ofeffortsnors?• (equipment, training, TA,the binding work.)Overall, what have been diagnostic constraints on• the PFM production possibility frontier? on PFMFocal areas for reforms by function (based• “clusters” as varied between different reformHow has this in Andrews 2010)components?• Focal areas for reforms by organisational location/
• The 2006 PEFA did not score the internal audit indicator, as it was feltthat it was too early to assess the internal audit function• The introduction of the internal audit function was designed and• The 2009 PEFA assessed internal audit as follows: PI-21 (i) – coveragemanaged by the IAU with little external support.– frequency andand quality of internal audit function: C; PI-21(ii)distribution of reports: B; and PI-21(iii) – extent of managementresponse to IA findings: D.• These results follow a pattern common elsewhere of reports beingissued regularly, but with a lacklustre follow-up (implementation ofrecommendations more difficult than setting out what the changesshould be).• Not applicable. As internal audit is a management tool to ensureinternal control systems are operating effectively, it serves to facilitategreater efficiency of MDAs’ expenditures.• There were little/no complementary actions taken by DPs.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoNQuEStIoNS For CouNtry AND ComPoNENt CASE
115
level of government (Ministry of Finance,PfM reformc. outcomes & overall assessment of PfM reform and of donor support toSectorministries, Local Governments, Parliament, CSOs,Changes in performance of PFM system over periodC.1. What have been the in-• etc.)as measured by HIPC AAP indicators and PEFAtermediate outcomes• Structure of design team and related consultationA.2.What type of struc-indicators.of PFm reforms, intures of changes in the• process: Balance of inputs by Donors/ Govt/Nature of PFM system changes identified:termshave been usedfor the design sys-• Consultants; Extent of input/ consultation with endBy PFM function (according to PFM cluster)quality of PFm andmanagement of these• users of PFM system (Sector ministries, LGs andBy nature of change (Upstream, De jure,tems?service institutions).reform inputs? HaveConcentrated vs. Downstream, De facto,• Management & co-ordination structure for PFMthese structuresDeconcentrated).reforms (Ad hoc Project Units vs. Normalserved to provide a co-management structures; Govt-controlled vs. Sharedordinated and harmo-Donor-Govt management; Use of consultants fornised delivery frame-• managerialquality of PFM system directly relevant toChanges in or purely advisory roles)C.2. to what extent are thework?service delivery, especially for women & vulnerableoutcomes generated• Arrangements for monitoring & evaluation.groups:relevant to improve-• Level of harmonisation and alignment of different• donor contributions.budget (PI-5): does this permitClassification of thements in the quality ofprotection/ prioritisation of specific functions orservice delivery, par-A.3. ticularly forof comple-• Use of country systems: extent to which external aidWhat types womenprogrammes?mentary actionsInformation in budget’ and ‘on treasury’?• is ‘on plan’, ‘on Budget documentation (PI-6): doesand vulnerable have• Ranking of donor efforts to support national systemsDonors taken to sup-this include performance data or targets for prioritygroups?as expressed in Paris Declaration monitoringport PFm reforms andprogrammes?what has been their• reports and PEFA indicators D2 and D3.Does the quality of in-year reports and final accounts(PI-24 & 25) and the access of the public (PI-10)• Scale and evolution over the evaluation period of aidsignificance? Haveflows provided as GBS, SBS or debt relief and relativethey had any influencepermit analysis and discussion of spending &contribution to reduction programmes?on the external con-performance for priority of treasury managementcosts and to discretionary resources available tostraints to reform?• Is the predictability of funds for commitment ofGovernment, (including effect on financingexpenditure (PI-16) improving?• constraints for PFM reform on resources received byIs the quality of information outputs).• Nature and evolution over time of dialogue on PFMservice delivery units (PI-23) improving?• within GBS/PRSC, SBS structures or similar donor-Extent to which improvements may be attributed to
relevant to improve-ments in the quality ofservice delivery, par-ticularly for womenand vulnerablegroups?greater efficiency of MDAs’ expenditures.Findings• Inputs were primarily GoG-based, at approximately �12 mn annually.External funding of less than an estimated €<0.5 mn was provided bythe EU, mainly for on-site training.• IAA provided capacity building to IAU staff, including preparation ofaudit manuals, training, on-site support programmes – going to thefield to help internal auditors, do risk assessment, plan and implementmanagement observations and response to the observations.
116
Evaluation Question
groups:• Classification of the budget (PI-5): does this permitprotection/ prioritisation of specific functions orprogrammes?• Information in Budget documentation (PI-6): doesthis include performance data or targets for priorityprogrammes?• Does the quality of in-year reports and final accounts(PI-24 & 25) and the access of the publicJudgement criteria/ Possible indicators(PI-10)
• Effectiveness of reform is measured by whether or not reforms have• The introduction of the internal audit function was designed andbeen implemented as intended. In this case, it is arguably too early tomanaged by the IAU with little external support.assess the quality of the internal audit function, as opposed to itsorganisational establishment, which is still being implemented. As thePEFA assessment made clear, there are challenges with the follow-upto audit recommendations.
ANNEx A: SummAry mAtrICES oF rESPoNSES to EvALuAtIoN QuEStIoNSFor CouNtry AND ComPoNENt CASE
a inputs & context: the design of PfM reformpermit analysis and discussion of spending &performance for priority programmes?• Government funds forfundsreforms committed andA.1.What has been the na-• Is the predictability of PFM for commitment ofactually disbursed by year over the evaluation period.ture and scale of PFmexpenditure (PI-16) improving?reform inputs provided• Is the quality of information on committedreceived by• Donor funds for PFM reforms resources andactuallydelivery units year over the evaluation periodby Government and Do-service disbursed by (PI-23) improving?• Natureto which improvements mayreform effortsnors?• Extent of support provided to PFM be attributed to(equipment, training, TA, diagnostic work.)the PFM reform programme, and particularly to• external support?Focal areas for reforms by function (based on PFM“clusters” as in Andrews 2010) these changes have• Consider counterfactual: would• Focal areas for reforms by organisational location/happened in the absence of the reform programme/level of government (Ministry of Finance, Sectorthe externally supported component of the reformministries, Local Governments, Parliament, CSOs,programme?etc.)• Causes of changes, specifically contribution ofC.3. Have reform efforts• Structure of design team and related consultationA.2.What type of struc-identified PFM reform outputs.been effective? If not,process: Balance of inputs by Donors/ Govt/tures have been usedwhy not? If yes, to what• Consistency of actual changes with objectives andConsultants; Extent of input/ consultation with endfor the design andoutcome targets of PFM reform programme.extent PFm reformusers of PFM system (Sector ministries, LGs andmanagement of these• Extent to which external constraints (political,outputs been a causalservice institutions).reform inputs? Havefinancial, policy) have undermined effectiveness offactor in the changes• Management & co-ordination structure for PFMthese structuresPFM reform.identified in intermedi-reforms (Ad hoc Project Units vs. Normalserved to provide a co-• Consider counterfactual: what change could haveate outcomes?management structures; Govt-controlled vs. Sharedordinated and harmo-been expected at Intermediate Outcome level inDonor-Govt management; Use of consultants fornised delivery frame-absence of PFM reform as a whole, and in absence ofmanagerial or purely advisory roles)work?external support to PFM reform?• Arrangements for monitoring & evaluation.• Level oftrends in Outcomes:alignment of differentRecent harmonisation and do these suggest pastC.4. to what extent do thedonorwill be sustained?gains contributions.gains identified at theIntermediate outcome• Is there a commitment (at political andA.3. What types of comple-• Use of country systems: extent to which external aidadministrative levels) to continue PFM reforms?levels appear sustain-is ‘on plan’, ‘on budget’ and ‘on treasury’?mentary actions have• Are there organisational structures in place (onable? Is the process of• Ranking of donor efforts to support national systemsDonors taken to sup-supply & demand sides) to sustain PFM reforms?PFm reform sustain-as expressed in Paris Declaration monitoringport PFm reforms and• Is there the financial and technical capacity withinable?reports and PEFA indicators D2 and D3.what has been theirGovernment to sustain PFM reform in the absence ofexternal support?• Scale and evolution over the evaluation period of aidsignificance? Have• If not, is there a framework in place for continuingflows provided as GBS, SBS or debt relief and relativethey had any influenceexternal support while building local capacity?contribution to reduction of treasury managementon the external con-costs and to discretionary resources available tostraints to reform?Government, (including effect on financingconstraints for PFM reform outputs).• Nature and evolution over time of dialogue on PFMwithin GBS/PRSC, SBS structures or similar donor-govt fora, including use of PFM reform conditions inGBS/ SBS disbursement conditions.• Influence of dialogue on the “political” and “policyspace” constraints restricting choice of PFM reformoutputs.• As indicated, the internal audit function is in the process of beingimplemented.• Sustainability of the function will depend on political commitment,• There were little/no complementary actions taken by DPs.particularly in terms of demanding results and follow-up to actions.
Annex b: list of PersonsConsultedNameAdam HusseinAngela PeasahB.G. GodjoeBaptiste MandouzeBrigidCharlotte AfudegoD.B. TenkorangDan BoakyeDaniel Atwere NuerDaniel ChachuDavid PedleyDennis NchorEdmundBarimah-SarpongEdward FiawoyifeEline OkudzetoEmmanuelAkrofi-TiboEnoch HemansEric LampleyEva MendsF. G. DakpallahGabriel TandohGrace AdzroeH. A. MustaphaHarald Küppersorganisation & PositionDFO, Gomoa West District AssemblyHead, Research & Development, CAGDAsst. A-G, GASProgramme Officer, Macro-Economic and TradeSection, EU DelegationKfWPolicy Analyst, IFPDPO, Gomoa West District AssemblySr. Economist and Former PFM Country Lead,World BankPrincipal Research Analyst, Ghana RevenueAuthorityIFPI/ISODEC Policy AnalystGovernance Adviser, DFIDPolicy Analyst, ISODECHead, GIFMIS, MoHDep. Director, Audit, MoHMacroeconomist, AfDBAsst Clerk to the Committee on Finance,ParliamentChief Director, MoFEPGAS Board Member and Advisor to AuditorGeneralDirector of Budget Development Unit, BudgetDivision, MoFEPDirector, PPME, MoHDep. Director, PPME, Min of TourismDeputy CAG (F&A), CAGDAg. Dep. Director, FinanceProgramme Manager, Good FinancialGovernance, BMZ117
ANNEx B: LISt oF PErSoNS CoNSuLtED
Helmut SchönHemis UssifHerman DusuHon. Albert KanDapaahHon. James KlutseAvedziHumphrey A. KumaIrene NordjoIsaac AgyareIsmaila CeesayJames N. AmponsahK. K. KufeK.E. GhansahKafui Ken-SeneyaKirsten RichterKwabena AdjeiMensahKwabena BoaduOku-AfariKwabena GyanKwakyeKwame GyesawKwame QuandahorKwasi OwusuLeonardShong-QuarteyLinda LarosonMarie-LaureAkin-OlugbadeMohummed MuseahNana Juaben-BoatenSiriboeOsa AhinakwahPatrick Nomo118
KfWMacroeconomist, Swiss Co-operationDOF, MoHChairman, Public Accounts Committee,ParliamentChairman, Public Finance Committee,ParliamentDirector, PPME, Min of TourismDanidaHead of PFM Reform Unit, Budget Division,MoFEPWorld Bank, FM AdvisorDeputy CAG (T), CAGDAg. DCAG (FMS), CAGDAg. DAG, GASHead, PBU, MoHBMZFormer Budget Director (MoFEP) and FormerDeputy Controller, CAGDDirector – ERFD, MoFEPAsst Economist, ERFD, MoFEPAsst Economist, ERFD, MoFEPPPME, GHS, Head of BudgetDirector, Adm, CAGDPolicy Asst II, ISODECDBA, Gomoa West District AssemblyRes Rep AfDBA/D, Gomoa West District AssemblyFormer Chief Director, MoFEPIMFHead of Internal Audit Agency(currently on leave)Clerk to the Committee on Finance, Parliament
Peace Fiawoyife
ANNEx B: LISt oF PErSoNS CoNSuLtED
Peter AidooPhilomena JohnsonRamatu Ude UmantaRanford AgyeiRapheal TufourRichard AsamoahAsieduRobert QuarshieSally LakeSam K. BoatengSammy ArkhurstSamuel BoatengSeidu KotomahSmart ChegabatiaSolomon OtooSteve GurSulemana B. BeningTheophilus Aidoo-MensahThomas PeasahYakubu ZakariaYaw SifahYukiyo Oda
Economic Planning Officer, MoFEPCo-ordinator, Institute of Fiscal Policy (IFP)Ag. Director, Finance, GHSAg. Director General, Internal Audit AgencyController and Accountant General (CAG)Deputy Auditor General, Audit ServiceAg. DAG/GAD, GASSenior Adviser, MoHHead, Fin Reporting and Monitoring, MoHHead of Public Expenditure Monitoring Unit,Budget Division, MoFEPDirector, Procurement & Supply, MoHDirector, GIFMIS, CAGDDep. Director, HRD, Min of TourismHead, FMS, MoHSenior Advisor, IMFPrincipal Planner, MoHDCE, Gomoa West District AssemblyBudget Division, MoFEPISODECAg. DAG/PSAD, GASProject Formulation Adviser, JICA
119
Annex C: list of ReferencesAfrican Economic Outlook (2011),Ghanadownloaded on 18thMay 2011 from:http://www.africaneconomicoutlook.org/en/countries/west-africa/gha-na/Allsop, T., Attah, R., Cammack, T., Woods, E (2009),Mid-Term Evaluation ofthe EFA Fast-Track Initiative: Ghana Case Study.Cambridge Education,Mokoro Ltd and Oxford Policy Management, 10thSeptember 2009.Andrews, M. (2011) Strengthening PFM – How are we doing? What’s next onthe agenda? Presentation delivered for CIPFA Conference February 2011.Betley, M. (2008),Putting Aid on Budget – Ghana Case Study.Mokoro April 2008a study for the Collaborative Africa Budget Reform Initiative (CABRI)and the Strategic Partnership with Africa (SPA)Betley, M. (2008,)Assessing the Impact of the PEFA Framework,Ghana CaseStudy, Final Report.Betley, M. and Burton, J (2011), Draft –Ghana: Assessing the Benefits of Multi-Donor Budget Support.Draft Report: March 2011.Betley, M., Ganguli R., Ochieng M., (August 2010), Public Expenditure andFinancial Accountability, Volume II: Sub-National Government, FinalReport, Ecorys Netherland BV.Brinkman, H.-J., Bauer, J.-M., and Mahama, A. 2009, Assessingthe Impacts ofthe Global Economic and Financial Crisis on Vulnerable Households in Ghana.WFP.Cox, M. and MacCarthy, M. (2009), GhanaJoint Assessment Strateg y (G-JAS) –Mid-Term Review.African Centre for Economic Transformation and Agul-has Applied Knowledge: 30 November 2009.De Renzio, P., Andrews, M. and Mills, Z. (2011),Does donor support to publicfinancial management reforms in developing countries work? An analytical study ofquantitative cross-country evidence.ODI: Working Paper 329.ECORYS (2010),Republic of Ghana Public Expenditure and Financial Accountability2009: Public Financial ManagementPerformance Assessment Report, VolumeI: Central Government. Final Report. Financed by the European Union.Economic Commission for Africa (2006),African Governance Report.Addis Aba-ba November 2005.European Union (2008), Ghana-European Community Country StrategyPaper and National Indicative Programme for the period 2008-2013.120
ANNEx C: LISt oF rEFErENCES
European Union (2008), Analysis of the EC Public Finance Managementmonitoring in budget support countries.Gordon, A. and M. Betley, ( June, 2011),Joint Evaluation of Public Financial Man-agement Reform in Burkina Faso, Ghana & Malawi – Ghana Desk Report.FiscusPublic Finance Consultants Ltd & Mokoro Ltd, Oxford.Government of Ghana (2006),Short and medium term plan for PFM reform.Government of Ghana (1999-2011),The Budget Statement and Economic Policy.Ghana Audit Service (2003-2009),Audited Annual Financial Statements.Ghana Audit Service (December 2010), IT Audit Report of the Auditor-Gen-eral on the Integrated Personnel and Database 2 of the Accountant Gen-eral’s Department, IT Audit Report Not 007/10.Government of Ghana and UNDP (2010),2008 Ghana: Millennium DevelopmentGoals Report.April 2010.IMF (2009),Ghana: Staff Report for the 2009 Article IV Consultation and Request for aThree-Year Arrangement Under the Poverty Reduction and Growth Facility.Pre-pared by the African Department: June 30, 2009Kavalsky, B. and Hartmann, A. (2010),Poverty Reduction Support Credits: GhanaCountry Study.IEG Working Paper – World Bank.Lawson, A. ( June 2011),Joint Evaluation of Public Financial Management Reform inBurkina Faso, Ghana & Malawi – Inception Report.Fiscus Public Finance Con-sultants Ltd, Oxford.Lawson, A. and De Renzio, P. (2009),Approach and Methodolog y for the Evaluationof Donor Support to Public Financial Management (PFM) Reform in DevelopingCountries. Part B: Indicative Evaluation Framework.Fiscus Public Finance Con-sultants Ltd. September 2009.Lawson A., Boadi G., Ghartey A., Ghartey A., Killick T., Kizilibash Agha, Z.and Williamson T., ( June 2007),Joint Evaluation of Multi-Donor Budget Sup-port to Ghana, Final Report, Volume One: Evaluation Results and Recommendationson Future Design and Management of Ghana MDBS,Ghana Centre for Demo-cratic Development & Overseas development Institute, London.MDBS (2010),Ghana: Multi-Donor Budget Support – Aide Memoire of the 2010 Annu-al Review.Accra May 2010.MoFEP (2011),Key Objectives of the Ministry.Downloaded on 18thMay 2011from: http://www.mofep.gov.gh/about_objective.htmODI (2007),Budget Support to Ghana A Risk Worth Taking?ODI Briefing Paper 24July 2007.OECD-DAC (2006), 2006Survey on Monitoring the Paris Declaration: CountryChapters.Downloaded on 6thApril 2011 from: http://www.oecd.org/datao-ecd/27/48/38901668.pdf121
ANNEx C: LISt oF rEFErENCES
OECD-DAC (2008),Survey on Monitoring the Paris Declaration: Country Chapters –Ghana.Downloaded on 6thApril 2011 from: http://www.oecd.org/datao-ecd/26/2/41951381.pdfOECD-DAC (2010),Statistical Database – OECD Stat Extracts.DownloadedMay 2011 from: http://stats.oecd.org/Index.aspx?DataSetCode=CRSNEWOPM (2006),Developing Capacity? An Evaluation of DFID-Funded Technical Coop-eration for Economic Management in Sub-Saharan Africa – Volume 2: Case Studies.DFID Reports-EV667, June 2006.Quartey, P., Ackah, C., Dufe G., Agyare-Boakye, E. (2010), Evaluation of theImplementation of the Paris Declaration on Aid Effectiveness: Phase II –Ghana Country Report.Quist R. (December 2009), Technical Review of the GIFMIS ProjectShort, J. (May 2003), Country Case Study 4: Assessment of the MTEF inGhana, Overseas Development Institute, London.Roberts, J. and Andrews, M, (2005), Something Funny Happened on the Wayto Reform Success: The Case of Budget Reform Implementation in Gha-na.International Journal of Public Administration,28: 291–311, 2005.van der Helm, A., Le Mounier, X., Nehmeyer-Srocke, I., Ostler, G., Owusu-Ayim, J.(May 2008), Report on BPEMS and IPPD2 Field Visits.UNDP (2010),The Real Wealth of Nations: Pathways to Human Development: TheHuman Development Report.Downloaded on 17th May 2011 from: http://hdr.undp.org/en/reports/global/hdr2010/World Bank (2003),International Development Association Program Document for aProposed Credit in the Amount of SDR 63.9 million (US$ 88 million equivalent) andProposed Grant in the Amount of SDR 26.9 million (US$37 million equivalent) to theRepublic of Ghana for a Poverty Reduction Support Credit and Grant.Report No.25995-GH: May 29 2003.World Bank (May 2004), Implementation Completion Report on a Credit tothe Republic of Ghana for a Public Finance Management Technical Assis‑tance Project, World Bank Report No. 28089‑GH.World Bank ( June 2004), Republic of Ghana, Country Financial Accountabil-ity Assessment, World Bank Report No. 32291-GH.World Bank (2005a),Implementation Completion Report (IDA-39570) on a credit inthe amount of SDR 58.6 million (us$85 million equivalent) and a grant in the amountof SDR 27.6 million (us$40 million equivalent) to the Republic of Ghana for a secondpoverty reduction support credit.Report No: 34591-GH: 14 December 2005.World Bank (2005b),Status of Projects in Execution FY-2005-SOPE. Africa Region,Country – Ghana.Operations Policy and Country Services, 23 September2005.122
ANNEx C: LISt oF rEFErENCES
World Bank (2006a),Public Finance Management Performance Report and Perfor-mance Indicators. 2006 External Review of Public Financial Management – VolumeII.PREM 4 Africa Region.World Bank (2006b),Status of Projects in Execution FY-2006-SOPE. Africa Region,Country – Ghana.Operations Policy and Country Services, 19 September2006.World Bank (2007),Status of Projects in Execution FY-2007-SOPE. Africa Region,Country – Ghana.Operations Policy and Country Services, 10 October 2007.World Bank (2008),Status of Projects in Execution FY-2008-SOPE. Africa Region,Country – Ghana.Operations Policy and Country Services, 11 October2008.World Bank (2009a),Ghana 2009 External Review of Public Financial ManagementVolume I – Main Report.Draft March 2009 PREM 4 Africa Region.World Bank (2009b),Status of Projects in Execution FY-2009-SOPE. Africa Region,Country – Ghana.Operations Policy and Country Services, October 02, 2009.World Bank (2010a),International Development Association Program Document for theSeventh Poverty Reduction Support Credit in the Amount of SDR 141 million(US$215 million equivalent) to the Republic of Ghana.REPORT No. 51920-GH:December 17th 2010.World Bank (2010b),Status of Projects in Execution FY-2010-SOPE. Africa Region,Country – Ghana.Operations Policy and Country ServicesWorld Bank (2010c), Joint Development Partners Preparation and Pre-Appraisal Mission, Ghana Integrated Financial Information Systems Pro-ject Component of the eGhana Project. Aide MemoireWorld Bank (2010d), Project Paper on an Additional Financing Credit to theRepublic of Ghana for the eGhana Project, World Bank Report No.54603-GHWorld Bank (2011a),World Development Indicators.Downloaded on 16thMay from:http://data.worldbank.org/data-catalog/world-development-indicators.World Bank (2011b),Ghana Country Brief.Downloaded on 18thMay 2011 from:http://web.worldbank.org/WBSITE/EXTERNAL/COUNTRIES/AFRICAEXT/GHANAEXTN/0,,menuPK:351962~pagePK:141132~piPK:141107~theSitePK:351952,00.htmlWorld Bank (2011c),Tackling Poverty in Northern Ghana.Report No. 53991-GH,1 March 2011.World Food Programme (2010),Summary Evaluation Report Country ProgrammeGhana: 2006-2010.Executive Board Annual Session June 2010.Wynne A, (December 2005), Public Financial Management Reforms inDeveloping Countries: Lessons of Experience from Ghana, Tanzania andUganda, African Capacity Building Foundation, Working Paper.
123
Annex d: Summary of PFM reforminputs, outputs, andintermediate outcomes
124
PFm focal areaPartial (e.g. excludessalary-related costs)activity-based MDAbudgetsPilot phase has last-ed nearly 10 years.Differences in classifi-cation systems betweencentral, local not re-solved prior to pilotingIndirect (policy dia-logue)Weaknesses in, & ambi-tiousness of, MTEF de-sign. Pressure to roll outearly.Direct (design, im-plementation)
Inputs1
Specific reform
outputs2
Causal factors inpositive/negativeprogressrole of DPContributions3,4Intermediateoutcomes
Strategicbudgeting
MTEF
Deterioration incredibility of multi-year framework
Support to MTEF: TA,design of MTEF, soft-ware. Funding of$4.6 mn (WB) (andGoG of $0.2 mn). Co-financing from DFIDand CIDA.More comprehen-sive budget info,more timely budgetcalendar: solely GoGinputsFinance and Admin-istration ActFinancial Adminis-tration RegulationsMore comprehen-sive budget informa-tionApprop. Act passedbefore beginning offiscal yearCitizens’ GuidesPre-budget consul-tations w/ publicBudget informationon websiteTIN introducedVAT re-introducedGRA ActTax Policy Unit es-tablished in MoFEPLarge TaxpayersCommitment by GoGCommitment by GoGN/A/indirect (policydialogue)
Composite budgets
Budgetpreparation
GoG inputs
Legislative basis forPFM
Budget contents
N/A/indirect (policydialog)N/A/indirect (policydialogue)
Improvement inbudget calendar,More transparencyin budget informa-tion (greater ac-countability to thepublic)
Timing of budgetsubmission to Par-liament
Commitment by GoG
More communica-tion with public onbudget
Commitment by GoG
N/A
ANNEx D: SummAry oF PFm rEForm INPutS, outPutSAND INtErmEDIAtE outComES
Resourcemanagement –revenues73
Direct taxes
Commitment by GoG,with DP supportCommitment by GoG,with DP supportCommitment by GoG,with DP support
Direct (design, im-plementation)Direct (design, im-plementation)Direct (design, im-plementation)
TA, training, soft-ware, hardwareFunding of €6.8 mn(BMZ) + €2 mn(SECO). Prior fund-ing from DFID.
Indirect taxes
125
Significant increasein revenues, notclear to what extentthis was due to im-proved efficiency
Greater administra-tive efficiency
sive budget informa-tionpublic)Approp. Act passedbefore beginning offiscal yearCommitment by GoGCausal factors inpositive/negativeprogressrole of DPContributions3,4IntermediateoutcomesDirect (design, im-plementation)Direct (design, im-plementation)N/ACommitment by GoGN/A/indirect (policydialogue)
dialog)
Timing of budgetsubmission to Par-liament
126
More communica-tion with public onbudget
PFm focal area
Inputs1
Specific reform
Citizens’ GuidesPre-budget consul-tations w/ publicBudget informationoutputs2on website
StrategicResourcebudgetingmanagement –revenues73
MTEFDirect taxes
Indirect taxes
Partial (e.g. excludesTIN introducedsalary-related costs)activity-based MDAVAT re-introducedbudgets
Weaknesses by GoG,Commitmentin, & ambi-tiousness of, MTEFwith DP support de-sign. Pressure to roll outCommitment by GoG,early.with DP support
Deterioration inSignificant increasecredibility of multi-in revenues, notyear to what extentclearframeworkthis was due to im-proved efficiency
ANNEx D: SummAry oF PFm rEForm INPutS, outPutSAND INtErmEDIAtE outComES
Commitment classifi-Direct (design, dia-GRA phaseIndirect (policy im-Pilot Act has last- Differences inby GoG,with DP supportplementation)Tax Policy Unit es-ed nearly 10 years.cation systems between logue)tablished in MoFEPcentral, local not re-Large Taxpayerssolved prior to pilotingUnit established inMoFEPBeginning the pro-cess of integratingN/A/indirect (policyImprovement inBudgetGoG inputsLegislative basis for Finance and Admin- Commitment by GoGrevenue Actdialogue)budget calendar,preparationPFMistration servicesMore transparencyFinancial Adminis-Slippage in fiscalCommitment by GoG un- Direct (design, soft-Aid, debt manage-Regular reports onAdvisory support,Resourcein budget informa-tration Regulationsware)discipline (weak-der PUFMARP reformsmentamt, composition ofsoftwaremanagement –tion (greater ac-nesses in commit-public debt, guaran- External technical sup-Funding of $0.2 mnTreasury, aid &Commitment by GoGN/A/indirect (policyBudget contentsMore comprehen-countability toment controls)theteesbudgetloans toof ext. informa- port (software)(WB) + $0.03 mndebtdialog)sivepublic)parastatals(GoG) (debt manage-tionment), $0.2 mn (WB)Direct (limited, notGoG-DP policy dialogueCommitment con-New commitmentCommitment by GoGN/A/indirect (policyApprop. Act passedTiming of budget+ $0.03 mn (GoG)sustained)trolscontrolbeginning ofsystem intro- (to initiate)dialogue)before(cash mngmt). Other submission to Par-Commitment/Institu-ducedyear sus-fiscal (notliamentfunding from UStional factors/ inter-tained)Treasury, WB, andCommitment by GoGN/ACitizens’ GuidesMore communica-agency co-operation (toIMFPre-budget consul-tion with public onsustain)tations w/ publicbudgetN/A/indirect (policyGoG-DP policy dialogueTreasury realign-Establishment ofBudget informationdialogue)mentseparate Treasuries Commitment/Institu-on websitetional factors/ inter-in some MDAs, re-Resourceagency co-operationgions (improve-Direct taxesTIN introducedCommitment byDirect (design, im-Significant increaseTA, training, soft-73 The Petroleum Revenue Management Act is not included in this analysis, since its promulgation (AprilGoG,occurred after the period being assessed and2011)management –ments in cash flowswith DP supportplementation)in revenues, notware, hardwarethus itstax efficiency was beyond the scope of the study.comprehensiverevenues73effect onFunding of €6.8 mnnotclear to what extentIndirect taxesVAT re-introducedCommitment by GoG,Direct (design, im-or sustained)this was due to im-(BMZ) + €2 mnwith DP supportplementation)proved efficiency(SECO). Prior fund-Direct (but not fullyGoG-DP policy dialogueSingle Treasury ac-Closing of a numbering from DFID.Commitment by GoG,Direct (design, im-GreaterGRA Acteffective), indirectCommitment/Institu-count administra- of redundant bankwith DP supportplementation)tive efficiencyTax Policy(thoughUnit es-(policy dialogue)tional factors/ inter-accountstablished in MoFEPthis still leaves many agency co-operationLarge Taxpayersin place)Unit established inSlippage in fiscalCash managementAttempts to improve GoG-DP policy dialogue, Direct (but not fullyAdvisory supportResourceMoFEPdisciplineeffective), indirectDP triggerinformation onpro-Funding from CIDA,managementBeginning the cash(policy dialogue)flowsof integratingWB, IMF, DFID, DKK,-expenditurescess to/from MDAs Commitment/Institu-tional factors/ inter-not comprehensiveand UNICEF.revenue services
Support to MTEF:TA, training, soft- TA,design of MTEF,ware, hardware soft-ware. Funding ofFunding of €6.8 mn$4.6 mn (WB) (and(BMZ) + €2 mnGoG of $0.2 mn). Co-(SECO). Prior fund-financing froming from DFID. DFIDand CIDA.More comprehen-sive budget info,more timely budgetcalendar: solely GoGinputs
Greater administra-Composite budgetstive efficiency
on websiteTIN introducedVAT re-introducedCommitment by GoG,with DP supportDirect (design, im-plementation)Commitment by GoG,with DP supportDirect (design, im-plementation)Commitment by GoG,with DP supportDirect (design, im-plementation)
Resourcemanagement –revenues73Significant increasein revenues, notclear to what extentthis was due to im-proved efficiency
Direct taxes
TA, training, soft-ware, hardwareFunding of €6.8 mn(BMZ) + €2 mn(SECO). Prior fund-ing from DFID.
Indirect taxes
Greater administra-tive efficiency
PFm focal areaRegular reports onPartial (e.g. excludesamt, composition ofsalary-related costs)public debt, guaran-activity-based MDAtees of ext. loans tobudgetsparastatalsPilot phase has last-New commitmented nearly 10 years.control system intro-duced (not sus-tained)Differences in dialogueGoG-DP policyclassifi-cation systems between(to initiate)central, local not re-Commitment/Institu-solved prior to pilotingtional factors/ inter-agency co-operation (tosustain)GoG-DP policy dialogueCommitment by GoGCommitment/Institu-tional factors/ inter-agency co-operationCommitment by GoGGoG-DP policy dialogueCommitment/Institu-Commitment by GoGtional factors/ inter-agency co-operationIndirect (policy notDirect (limited, dia-logue)sustained)Commitmentin, & ambi-Weaknesses by GoG un-der PUFMARP reformstiousness of, MTEF de-External technical sup-sign. Pressure to roll outport (software)early.Direct (design, soft-im-ware)plementation)
Inputs1
Specific reform
GRA ActTax Policy Unit es-tablished in MoFEPLarge TaxpayersUnit established inMoFEPBeginning the pro-cess of integratingoutputs2servicesrevenueCausal factors inpositive/negativeprogressrole of DPContributions3,4IntermediateoutcomesSlippage in fiscalDeterioration indiscipline (weak-credibility of multi-nesses in commit-year frameworkment controls)
ResourceStrategicmanagement –budgetingTreasury, aid &debt
Aid, debt manage-MTEFment
Advisoryto MTEF: TA,Support support,software MTEF, soft-design ofFunding of $0.2 mnware. Funding of(WB)mn (WB) (and$4.6 + $0.03 mn(GoG) (debt manage-GoG of $0.2 mn). Co-ment), $0.2 mnDFIDfinancing from (WB)+ $0.03 mn (GoG)and CIDA.(cash comprehen-More mngmt). Otherfunding from USsive budget info,Treasury, WB, andmore timely budgetIMFcalendar: solely GoGinputsN/A/indirect (policyN/A/indirect (policydialogue)dialogue)
Composite budgetsCommitment con-trols
Budgetpreparation
GoG inputs
Treasury realign-Legislative basis formentPFM
Budget contents
N/A/indirect (policydialog)Direct (but not fullyeffective), indirectN/A/indirect (policy(policy dialogue)dialogue)
Improvement inbudget calendar,More transparencyin budget informa-tion (greater ac-countability to thepublic)
Single Treasury ac-count of budgetTimingsubmission to Par-liament
Resourcemanagement-expenditures
Advisory supportFunding from CIDA,WB, IMF, DFID, DKK,and UNICEF.
More communica-tion with public onCash managementbudget
N/ADirect (but not fullyeffective), indirect(policy dialogue)
Slippage in fiscaldiscipline
ANNEx D: SummAry oF PFm rEForm INPutS, outPutSAND INtErmEDIAtE outComES
Resourcemanagement –revenues73
Direct taxesProcurement
Indirect taxes
Establishment ofFinance and Admin-separateActistration Treasuriesin some MDAs, re-Financial Adminis-gions (improve-tration Regulationsments in cash flowsnot comprehensiveMore comprehen-or sustained)sive budget informa-tionClosing of a numberof redundant bankApprop. Act passedaccounts (thoughofbefore beginningthis still leaves manyfiscal yearin place)Citizens’ GuidesPre-budget improveAttempts to consul-tations w/ publicinformation on cashBudget informationflows to/from MDAson comprehensivenotwebsiteor sustainedTIN introducedProcurement Actpassed; challengesVAT re-introducedwith implementationGRA ActIPPD2 being used forTax Policy Unit es-ministries, depts.tablished in MoFEPbut only some agen-Largecies TaxpayersCommitment by GoGGoG-DP policy dialogue,DP triggerCommitment/Institu-tional factors/ inter-agency co-operationCommitment by GoG,GoG-DP policy dialogue,with DP supportDP triggerCommitment by GoG,Institutional factors (im-with DP supportplementation)Commitment by GoG,Commitment by GoG,with DP policy dialogue,GoG-DPsupportDP trigger
Direct (design, im-Direct (design),plementation) indi-rect (implementa-Directtion) (design, im-plementation)Direct (design, dia-Indirect (policy im-plementation)logue)
127
Significant increasein revenues, notclear to what extentthis was due to im-proved efficiencyInternal audit func-tion initiated
Internal controls,audit, &monitoring
TA, training, soft-ware, hardwareFunding of €6.8 mn(BMZ) + €2 mn(SECO). Prior fund-ing from DFID.TrainingFunding of €<0.5 mn(EU). Other fundingfrom DFID and
Greater administra-Payroll controlstive efficiency
ment
separate Treasuriesin some MDAs, re-gions (improve-ments in cash flowsnot comprehensiveor sustained)GoG-DP policy dialogueCommitment/Institu-Causal factors intional factors/ inter-positive/negativeagency co-operationprogressIntermediateoutcomesDeterioration inSlippage in fiscalcredibilitydiscipline of multi-year frameworkWeaknesses in, & ambi-GoG-DP policy dialogue,tiousnessDP triggerof, MTEF de-sign. Pressure to rollCommitment/Institu-outearly.tional factors/ inter-agency co-operationDifferences in classifi-GoG-DP policy dialogue,cation systems betweenDP triggercentral, local not re-(im-Institutional factorssolved prior to pilotingplementation)Commitment by GoG,GoG-DP policy dialogue,DP trigger by GoGCommitmentUnder PUFMARP re-formsCommitment by GoGIndirect (policy dia-Direct (design),logue) (implemen-indirecttation)(design, im-Direct (but not fullyplementation)effective), indirect(policy dialogue)Direct (but not fullyeffective), indirect(policy dialogue)role of DPContributions3,4
Commitment/Institu-tional factors/ inter-agency co-operation
dialogue)
128
Single Treasury ac-count
PFm focal areaPartial (e.g. improveAttempts to excludessalary-related cashinformation on costs)activity-based MDAflows to/from MDAsbudgetsnot comprehensiveor sustainedPilot phase has last-Procurement Acted nearly 10 years.passed; challengeswith implementation
Inputs1
Specific reform
Closing of a numberof redundant bankaccounts (thoughthis still leaves manyoutputs2in place)
StrategicResourcebudgetingmanagement-expenditures
MTEFCash management
Composite budgetsProcurement
Internal controls,audit,& monitoringBudgetpreparation
Advisoryto MTEF: TA,Support supportFunding from CIDA,design of MTEF, soft-WB, IMF, DFID,ofware. Funding DKK,and UNICEF. (and$4.6 mn (WB)GoG of $0.2 mn). Co-financing from DFIDand CIDA.More comprehen-sive budget info,more timely budgetcalendar: solely GoGTraininginputs of €<0.5 mnFunding(EU).inputs fundingGoG Otherfrom DFID andFrance.Indirect (policy dia-logue)N/A/indirect (policydialogue)N/A, indirect(policy dialogue)N/A/indirect (policydialog)N/A/indirect (policydialogue)Direct (underPUFMARP), thenN/Aindirect (policy dia-logue)Slippage in fiscaldiscipline
Payroll controls
Legislative basis forPFMInternal audit
Budget contents
Internal audit func-tion initiatedImprovement inbudget calendar,More transparencyin budget informa-tion (greater ac-countability to thepublic)
ANNEx D: SummAry oF PFm rEForm INPutS, outPutSAND INtErmEDIAtE outComES
Commitment by GoG
Accounting andreporting
Funding of $15.3 mn(WB)+ $4.6 mn (GoG)(BPEMS)
Timing of budgetsubmission to Par-liamentFMIS/BPEMSMore communica-tion with public onbudget
IPPD2 being usedfor ministries, depts.but only someFinance and Admin-agenciesActistrationFinancial Adminis-Internal Audittration RegulationsAgency Act passedInternal AuditMore comprehen-Agency establishedsive budget informa-Standards, manualtionproducedIAUs establishedApprop. Act passedARICs establishedbefore beginning offiscal yearImplementation in 6Citizens’ Guidespilot MDAs not sus-Pre-budget consul-tained. Systemtations w/ publicabandoned.Budget informationon websiteTIN introducedMore timelyVAT re-introducedgazettingGRA ActExternal Audit Agen-Tax Policy Unit es-cy Acttablished in MoFEPStandards, manualLarge TaxpayersproducedUnit established inPerformance, VfM MoFEPaudits begunBeginning of auditClearance the pro-cess of integratingbacklogrevenue servicesAmbitiousness of de-Commitment by GoGsign, insuff. high-levelcommitment in practice,seen primarily as tech-nol. reform, manage-ment problems (agencyownership), technicalCommitment by GoG,problemswith DP supportGoG-DP policy dialogue,CommitmentDP trigger by GoG,with DP supportCommitment by GoG,Commitment by GoG un-with DP supportder PUFMARP reforms,with DP supportGoG-DP policy dialogue,DP trigger
Resourcemanagement –revenues73
Direct taxes
Externalaccountability
TA, training, soft-ware, hardwareFunding of €6.8 mn(BMZ) + €2 mn(SECO). Prior fund-ing from DFID.TA/capacity building(GAS), training, au-dits of selectedflows. Funding of €8mn (EU) (Externalaudit), €0.5 mn/yr(BMZ) (PAC), DFID(GAS, PAC) £0.4 mn,KfW, CIDA, DANIDA
Timeliness of in-yearIndirectexecution re-budget taxesportsGreater administra-External audittive efficiency
Direct (design, im-plementation)N/A/ indirectDirect dialogue)(policy(design, im-plementation)Direct (design, im-Direct (design, TAplementation)support)
Significant increasein revenues, notclear to what extentthis was due to im-proved efficiencyImproved account-ability to the publicfor public finances
producedIAUs establishedARICs establishedImplementation in 6pilot MDAs not sus-tained. Systemabandoned.Ambitiousness of de-sign, insuff. high-levelcommitment in practice,seen primarily as tech-nol. reform, manage-ment problems (agencyownership), technicalproblemsCausal factorsdialogue,GoG-DP policyinpositive/negativeDP triggerprogressN/A/ indirectrole ofdialogue)(policyDPContributions3,4IntermediateoutcomesImproved account-Deterioration inability to the publiccredibility of multi-for public financesyear frameworkDirect (design, im-TA support)plementation)Commitmentin, & ambi-Weaknesses by GoG un-der PUFMARP reforms,tiousness of, MTEF de-with DP support roll outsign. Pressure toGoG-DP policy dialogue,early.DP triggerDifferences in classifi-cation systems betweencentral, local not re-solved prior to pilotingCommitment by GoG,public pressureIndirect (policy dia-logue)Direct (underPUFMARP), thenindirect (policy dia-logue)Slippage in fiscaldiscipline
Accounting andreporting
Funding of $15.3 mn(WB)+ $4.6 mn (GoG)(BPEMS)
FMIS/BPEMS
PFm focal area
Inputs1
Timeliness of in-yearbudget execution re-Specific reformports
More timelygazettingoutputs2
ExternalStrategicaccountabilitybudgeting
External auditMTEF
Composite budgets
External AuditPartial (e.g. excludesAgency Actsalary-related costs)Standards, manualactivity-based MDAproducedbudgetsPerformance, VfM audits begunPilot phase has last-Clearance of audited nearly 10 years.backlog
TA/capacity buildingSupport to MTEF: TA,(GAS), training, au-design of MTEF, soft-dits ofFunding ofware. selectedflows. Funding of €8$4.6 mn (WB) (andmn (EU) (ExternalGoG of $0.2 mn). Co-audit), €0.5 mn/yrfinancing from DFID(BMZ) (PAC), DFIDand CIDA.(GAS, comprehen-More PAC) £0.4 mn,KfW, CIDA, DANIDAsive budget info,(DKK 0.5 mn/yr)more timely budgetcalendar: solely GoGinputs
Parliamentary scru-tiny of public finance
N/A/indirect (policyMore timely PAC re-dialogue)view of audit reportsPublic access to PACdebatesN/A/indirect (policyBudgetGoG inputsLegislative basis for Finance and Admin- Commitment by GoG1DP fundingto match this breakdown.dialogue)preparation only. GoG funding not available PFMistration Act2Main outputs shown. Does not include staff training, considered an input. Where outputs have not been sustained, this has been noted.Financial Adminis-3Analysis from Box 7 above.tration Regulations4DP support shown may not be comprehensive due to non-availability of data.Commitment by GoGN/A/indirect (policyBudget contentsMore comprehen-dialog)sive budget informa-tionApprop. Act passedbefore beginning offiscal yearCitizens’ GuidesPre-budget consul-tations w/ publicBudget informationon websiteTIN introducedVAT re-introducedGRA ActTax Policy Unit es-tablished in MoFEPLarge TaxpayersUnit established inCommitment by GoGN/A/indirect (policydialogue)N/A
Improvement inbudget calendar,More transparencyin budget informa-tion (greater ac-countability to thepublic)
Timing of budgetsubmission to Par-liament
More communica-tion with public onbudget
Commitment by GoG
ANNEx D: SummAry oF PFm rEForm INPutS, outPutSAND INtErmEDIAtE outComES
Resourcemanagement –revenues73
Direct taxes
Commitment by GoG,with DP supportCommitment by GoG,with DP supportCommitment by GoG,with DP support
Direct (design, im-plementation)Direct (design, im-plementation)Direct (design, im-plementation)
Indirect taxes
TA, training, soft-ware, hardwareFunding of €6.8 mn(BMZ) + €2 mn(SECO). Prior fund-ing from DFID.
Significant increasein revenues, notclear to what extentthis was due to im-proved efficiency
129
Greater administra-tive efficiency
Annex E: Consultant Termsof Reference1. introductionThese Terms of Reference are for case studies to be carried out in BurkinaFaso, Ghana and Malawi on the reform of public financial management (PFM)systems, the results achieved, the role played by donors and other institutionaland contextual factors that may contribute to or hinder PFM reform outcomes.The case studies will build on empirical analysis that investigates where andwhy PFM reform has delivered results and, conversely, where and why it hasnot. The main focus of the case studies will be to investigate whether and howdonor behaviour and the design and implementation of PFM reform makesa difference to the achievement of results, or whether other domestic contextualfactors carry more weight.
2. backgroundThe Paris Declaration on Aid Effectiveness and the associated emphasis on theuse of country systems, budget support, and governance and anti-corruptionhave triggered increased attention on the reform of PFM. Strong PFM systemsare a key element of the institutional and governance framework needed forbuilding peaceful and stable societies and successful economic and social devel-opment, essential to improved service delivery and to the achievement of theMillennium Development Goals.Nevertheless, PFM systems in many developing countries remain weakand there is lack of certainty or consensus on the role of donors and the con-text under which external support can best assist the process of PFM reform.To address this, the evaluation departments of DANIDA (Denmark), Sida(Sweden) and the AfDB (African Development Bank) have agreed, in consul-tation with the OECD-DAC Evaluation Network, to manage a joint evalua-tion of PFM reforms in developing countries. This and other planned jointevaluations, including the joint evaluation of the impact of budget support,public sector governance reform, support to anti-corruption programmes, andthe implementation of the Paris Declaration,will feed into discussions prior tothe next High Level Forum on Aid Effectiveness (HLF-4) in Busan (29November – 1 December 2011).The PFM evaluation is interested in finding answers to two related ques-tions:(a) Where and why do PFM reforms deliver results (i.e. improvement inthe quality of budget systems); and(b) Where and how does donor support to PFM reform efforts contributemost effectively to results?
130
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
To answer these questions the evaluation design is made up of several compo-nents outlined in summary below.First,analytical background workhas been undertaken both to:– define what is meant by PFM reform results; and– consider how results can be measured across countries and overtime to assess the degree to which change in the quality of PFMsystems has occurred (see Lawson/De Renzio Approach and Meth-odology for the Evaluation of Donor Support to PFM in Develop-ing Countries Part A July 2009 and Part B September 2009).• S econd a literaturereviewhas been completed looking at the rangeof approaches to PFM reform, donor support and existing evidence onsuccess/failure of PFM reform approaches74.• hird, a quantitativeanalysishas been undertaken to identifyTcountries where PFM reform has delivered results in the qualityof PFM systems; where it hasn’t, and the contextual factors thatmight explain these differences as well as the correlation with donorsupport75.• ourth, countrycase studieswill follow up the findings from theFquantitative analysis and explore why, in some cases, donor supportappears correlated with results, and why in others it does not. The casestudies will explore whether and how donor behaviour and theapproach to PFM reform design and implementation makes a differ-ence to results. Five case studies are planned in Sub-Saharan Africastarting with Burkina Faso, Ghana and Malawi.• inally, aregional Africa synthesis reportwill be compiledFthat will bring together findings from each of the evaluation productsoutlined above.Country case studies have been selected on the basis of data availability (seebelow) and because they provide examples where budget institutions improvedwith: (i) high donor effort; (ii) low donor effort; and where regardless of donoreffort, budget institutions did not improve.
3. PurposeThe purpose of the evaluation is to identify what factors – institutional andcontextual – contribute to successful PFM reform and how donors can bestsupport PFM reform processes given the influence of contextual factors on theprocess of change. Conversely, the case studies will also identify where PFMreform has not worked, and whether the application of aid effectiveness princi-ples to PFM reform is important to results. The evaluation findings areintended for Governments, donors and PFM practitioners. The intention is toimprove the design of external support for country led PFM reform efforts.
74 Pretorius, C and Pretorius, N. (2008) Review of the Public Financial ManagementLiterature. London: DFID.75 de Renzio, P., M. Andrews and Z. Mills (2010) Evaluation of Donor Support to PublicFinancial Management (PFM) Reform in Developing Countries. Analytical study ofquantitative cross-country evidence.London: Overseas Development Institute.
131
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
4. Scope and limitationsCountry case studies cover the period from 2001–2009.76The period repre-sents a time in which donors became increasingly interested in PFM andagreed to increase the effectiveness of aid expenditure, including by usingcountry systems to channel and deliver aid finance.As PFM performance information is only widely available for central gov-ernment, the scope of the case studies is restricted to central governmentorganisations. Nevertheless, country visits may provide an opportunity togather information (in addition to that contained in PEFA) on the extent towhich PFM reforms are beginning to extend beyond central government insti-tutions to local government and to service providers. The extent to which thegovernment is taking a lead in this may indicate ownership and reform sus-tainability.The quantitative analysis found a positive and significant, albeit weak, cor-relation between donor support to PFM reforms and improvements in PFMsystems. It also found some positive correlations between the way aid is pro-vided and the strength of PFM systems. However, these average effects cannotbe taken as causal and universal, and need to be further investigated. There-fore the main purpose of the country case studies is to unpack the nature ofPFM reform in different cases where there is found to be: (i) a positive correla-tion with donor support; (ii) a negative correlation with donor support; and (iii)no correlation between PFM results and donor support. As such, case studieswill focus on the history of PFM reform inputs; what has been provided, forwhat purpose, in what sequence, for how long and at what cost that might helpto explain the correlation (positive or negative) with PFM results or lack ofthem. The case studies will therefore not investigate the impact of PFMreform (particularly on service delivery) but will instead focus on inputs in theevaluation framework; how they have been identified, designed and deliveredand the significance of this for the delivery of intermediate outcomes(explained in detail below). In countries where there has been high donor sup-port to PFM, a key line of enquiry is the extent to which the application of aideffectiveness principles is found to make a difference to results.However, given the range of factors that contribute to PFM results, it maybe difficult to directly attribute results to donor support.77PFM reform inter-ventions are treated as inputs in the evaluation framework and the case studymethodology is centred on the ability to assess the institutional and contextualfactors that helped to support success and/or failure of these inputs at eachstage of the evaluation framework. It may also be possible to link intermediateoutcomes to outputs and donor inputs. For example, one dimension of PFMreform – linking policy to planning and budget (an intermediate outcome inthe evaluation framework) – may receive substantial donor support becausewhile it is difficult from a technical perspective it may be relatively easier froma political perspective. Donor support for the achievement of the other inter-76 Or the date of the second PEFA report, which might be earlier than 2009.77 In fact, the quantitative analysis highlights how economic factors in particular explaina large part of variation in the successful implementation of PFM reforms.
132
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
mediary outcomes – greater transparency and comprehensiveness and con-trol, oversight and accountability – may not be as strong because the politicalcosts of these reforms may be higher despite their relative technical ease. Itwill therefore be important for the evaluation case studies to explore the widercontext of reform intervention and whether certain reforms are pursedbecause they are politically more palatable than others. In other words, thecase studies would examine the extent to which donor support is concentratedat particular phases of the budget cycle.While the evaluation framework identifies final outcomes including theoperational efficiency of public spending, it may be too soon to draw conclu-sions about the impact and sustainability of results. This is largely due to thefact that the evaluation period is relatively short and while quality PFM is nec-essary for the quantity and quality of service delivery, it is not sufficient. Nev-ertheless, it may be useful to consider what factors help to support on goingreform and what factors risk sustainability e.g. is PFM reform supported bycivil society, the Parliament and the business community; implemented ina stable and growing economy; building and developing the capacity of theMinistry of Finance as a key central government body; and spreading furtherthan the centre to include sector Ministries, local government and servicedelivery units; and supported by on going donor support and technicalassistance?
5. Methodology for the Selection of case StudiesCase study countries have been selected on the basis of data availability. Allcase studies have at least two Public Expenditure and Financial Accountabil-ity (PEFA) assessments available (which covers a period of at least three years)plus World Bank and IMF HIPC assessments which extends the evaluationperiod by at least another four years (or more depending on the date of theHIPC assessments) which may be sufficient to observe changes in the qualityof PFM systems.The table below separates the 14 countries in Sub-Saharan Africa thatmeet the data requirements, into countries where reforms have deliveredresults (i.e. an increase in HIPC/PEFA scores between 2001 and 2007 – or thedate of the second PEFA assessment – and countries where reforms did notbring about any improvement or where the quality of PFM systemsdeteriorated.
PEfa Performance in 14 african countries 2001–2007Countries where budgetinstitutions improvedBurkina Faso, Ethiopia, Ghana,Mali, Tanzania, ZambiaCountries where budget institutionsdid not improve or deterioratedBenin, Guinea, Madagascar, Malawi,Mozambique, Rwanda, Sao Tomeand Principe, Uganda
Source: PFM Evaluation Approach Paper Part A: Assessing Budget Institutionsand Budget Reforms in Developing Countries
133
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
Of these countries, case studies will be selected so that three types of situationcan be examined:• ne in which donor support appears to be positively correlated with OPFM improvement;• ne in which donor support appears to be negatively correlated with OPFM improvement; and• ne in which significant PFM improvements appear to have occurred Odespite relatively low levels of donor support.This suggests that countries are selected from across the following table.
Relative impact of Donor Support to PfM reforms in SSa (1998–2007)PFm reformHigh donorsupportLow donorsupportSource: As above
Countries where BudgetInstitutions ImprovedBurkina Faso, Tanzania,ZambiaEthiopia, Ghana, Mali
Budget Institutionsdidnot improveBenin, Malawi, Mozam-bique, Rwanda, Sao Tomeand Principe, Uganda
However, as stated above, donor support is not the only factor influencing thedesign and implementation of PFM reform measures and their results. Other fac-tors, notably a range of domestic economic, political and institutional factors, arelikely to determine the dynamic of the reform process as well as the resultsachieved. Hence, country case studies have been specifically selected to includecases where budget performance has improved with little or no donor support forPFM reform to highlight what specific factors contributed to reform outcomesand to provide a relevant counter factual.
6. analytical approachThe first step in answering the related questions of where and why do PFMreforms deliver results and where and how does donor support contribute mosteffectively to results, is to build a common definition of what is meant byresults. The second step is to identify empirical information that might help tomeasure results and compare them across countries and over time (see Assess-ing Budget Institutions and Budget Reforms in Developing Countries: Over-view of theoretical approaches and empirical evidence. Paolo de RenzioJuly 2009).For the purpose of this evaluation, the following three dimensions of budg-et institutions provide a basis for assessing their overall quality:• Transparency and comprehensiveness:looks at issues related tothe quality of budget information, from the classification system to thecoverage and clarity of budget documents; accessibility to budget infor-mation by the Legislature, the general public, media and civil society134
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
• inking budgeting, planning and policy:assesses the extent toLwhich the budget is effective in converting policy objectives into rele-vant taxation and spending actions; budgets are derived from accuratemedium term forecasts and contain a policy perspective• ontrol, oversight and accountability:considers whether ade-Cquate mechanisms are in place to promote overall accountability forthe use of public resourcesQuality budget institutions are defined as those that exhibit higher degrees oftransparency; policy orientation and control/accountability (see Table 1below). At the opposite end, weak budget institutions are identified by theiropacity, their lack of linkages with planning and policy, and the absence orweakness of mechanisms for monitoring and accounting for the use of publicfunds.These dimensions are consistent with parts of the PEFA assessment as wellas with indicators developed for HIPC assessments (Table 1).
table 1 PEfa indicators of budget PerformanceIntermediateoutcomeTransparencyand Compre-hensivenessDefinitionPEFAIndicators
The quality of budget information,HIPC 1, 2, 4, 5from the classification system to the PEFA 5, 6, 7, 8, 9,coverage and clarity of budget docu- 10, 13, 14, 15ments; accessibility to budget infor-mation by the Legislature, the gen-eral public, media and civil societyBudget is effective in converting pol- HIPC 7, 10icy objectives into relevant taxationPEFA 11, 12, 16,and spending actions; budgets are23derived from accurate medium termforecasts and contain apolicy per-spectiveAdequate mechanisms are in place HIPC 8, 9, 11, 15to promote overall accountability for PEFA 17, 18, 19,the use of public resources20, 21, 22, 24,25, 26, 27, 28
Links betweenplanning, policyand budget
Control over-sight and ac-countability
Unsurprisingly, there are limited sources of information and cross-country datawhich can be relied on to assess and compare the quality of budget institutions.The most comprehensive attempt at constructing a framework to assess thequality of budget institutions is the PEFA PFM Performance MeasurementFramework (PEFA 2005) based on 31 indicators, which cover institutionalarrangements at all phases of the budget cycle. Moreover, the framework con-tains all the information needed to measure the quality of budget institutionslong the three dimensions identified above.135
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
The country case studies will explore the extent to which PFM reform is morelikely to produce results when there is an enabling environment for reform,when donor behaviour follows the principles of aid effectiveness, and whenPFM reform interventions follow certain principles.78This suggests that PFMreform is more likely to produce results in the following circumstances:• Economic Growth and Political Stability:PFM reforms takeplace in a stable environment that allows for the time, policy space andflexibility needed to implement complex governance reforms, and theadditional public funds generated by growth.• Reform Planning and Design:PFM reform inputs consider thelocal context taking into account the strength of existing institutions.Reform plans have been prioritized and sequenced to implement basicsfirst and do not overwhelm existing administrative capacity.• Strengthened Approach:reforms are country owned and managedthrough existing processes; with donor support harmonized andaligned behind country led reform programmes and aid is channeledthrough country PFM systems79.• Political Economy:PFM reforms have sustained high level politicalsupport for governance reforms in general (including civil servicereform) and reflect political priorities and feasibility; political economyfactors (such as patronage networks) are less powerful.• Demand side governance:PFM reforms build on existing publicdemand for improved PFM through strengthening transparency ofdecision making and financial information, and there is greateraccountability to the public and users of public services. Countrieswhere these processes exist are more likely to deliver results in PFMreform, but this could also be a necessary pre-condition to more diffi-cult or politically sensitive PFM reforms.The relevance of these factors to the PFM change process has been incorpo-rated into an evaluation framework. The purpose of the pilot country casestudies is to test the evaluation framework and to elaborate further on whereand why PFM reforms deliver results and how donor support can more effec-tively support the PFM change processes.
7. Evaluation QuestionsDetailed questions for each country case study regarding PFM reforms are list-ed below against the OECD DAC evaluation criteria of relevance, efficiencyand effectiveness of PFM reform. The questions have been structured to reflectthe theory of change set out in the evaluation framework in Annex 1. At each78 See Paolo de Renzio ( July 2009): PFM Evaluation Approach Paper, Part A chapter 6:“Explaining success in budget reforms: lessons from the political economy of govern-ment reforms.” The theory of change largely draws on the lessons learnt from firstgeneration structural adjustment reforms, rather than second generation governancereforms. However, the theory closely suggests that the principles of aid effectivenessalso apply to PFM reform.79 While this may appear tautological, there can often be several PFM reform projectinterventions and TA initiatives (World Bank, IMF, ADB and so on) operating within oneinstitution, usually the Ministry of Finance with no single agreed strategy for PFM reform,
136
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
level of the evaluation framework (inputs, outputs, intermediate and final out-comes) the evaluation will consider the institutional and contextual factors thatinfluenced the design and implementation of PFM reform and their significancein delivering reform success.80It is important to note that these questions refer toall PFM reform inputs whether or not they are financed by external donors.• How relevant is PFM reform to local context and existingsystems?– Is there a government led PFM reform programme that has highlevel political support?– Does PFM reform respond to domestic priorities, e.g. politicallydriven public sector reform agendas, macroeconomic and fiscalneeds, political priorioties for improved service delivery?– Is donor support designed and structured to support governmentled and government managed initiatives?– Do PFM reform programmes include a component aimed atstrengthening budget reporting e.g. to the public. Or do PFMreform programmes include components to include the public inresource allocation decisions?– Is external support to PFM reform designed to fit with the nature ofpolitical support for reform, to the institutional strengths and weak-nesses of the existing PFM system, and to the organisational capa-bility of the lead agencies (e.g. finance ministry) in PFM reform?Are international models of PFM reform transplanted on a “one-size fits all” basis or is PFM reform developed incrementally to fitwith existing administrative capacity?81– Are PFM reforms consistent with on-going public administrationreforms?– Is donor support based on building existing PFM systems ratherthan creating new ones?– Is there evidence that donor supported reforms have overwhelmedexisting institutional capacity?– What is the role of technical assistance in PFM reform design?– Are PFM reform and management processes supported by orinclude active consultation and communication with a wide rangeof stakeholders involved in the reforms, as well as active measuresto broaden support for reform?• How efficient and cost effective is PFM reform?– Efficiency should look at the ratio between costs and output or out-comes. It will be important to estimate what PFM reform costs?How much has been spent by Government initiatives and bydonors e.g. on personnel and equipment, to achieve particularPFM objectives?80 These questions are a summary of a much longer list of questions taken from Lawson/de Renzio81 Assessing political support is not straightforward but evidence could be gathered throughinterviews (TA, donor, government officials, and civil society) and through the ability ofthe Ministry of Finance (and the Minister of Finance) to lead and implement PFM reformsacross government with strong support from Cabinet and Parliament.
137
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
– Is donor support for PFM reform coordinated around a single PFMreform plan or strategy or is support fragmented across several ini-tiatives?– Have donor efforts been slow at getting started or taken longer thanexpected, requiring on-going TA support?– Is donor support reliant on specifically designed PFM reform man-agement units (project implementation units)?• How effective is PFM reform?– Is there any additional evidence of PFM reform that is not capturedin the PEFA framework e.g. reforms extending beyond central gov-ernment institutions?– Have PFM reforms (including donor support to reforms) movedbeyond de jure reform aspects, such as approving laws and regula-tions, to de facto aspects, such as changes in actual budget practic-es, and have these elements of PFM systems improved?– Have PFM reforms extended beyond the centre (e.g. CentralFinance Agencies) to include, for example, sector Ministries, localgovernment and service delivery units and what explains thisspread? Have reforms been effective in improving PFM perfor-mance beyond central finance agencies?– In aid dependent countries, to what extent has the use of generalbudget support, PFM-related conditionality, and efforts to reduceaid fragmentation contributed to strengthening PFM performance?Have these efforts impacted across all aspects of PFM, or on specif-ic areas such as de jure and concentrated PFM processes?– Are country systems for financial reporting and accountability uti-lised by donors?– To what extent is aid expenditure included in different stages of thebudget process82?
8. tasksThe consultants will conduct country case studies in Burkina Faso, Ghanaand Malawi.Task 1: Inception Phase• Organise consultation workshop (in Tunis) as part of the inceptionphase involving key users and stakeholders of the evaluation.Prepare an Inception Report that would:• Further develop the evaluation questions, evaluation framework, ana-lytical tools and overall work plan for the country case studies thatbuilds on the original approach papers and the literature review toensure complementarities and the best possible synthesis report83.82 For instance see CABRI – dimensions of aid on budget including procurement83 Lawson/De Renzio Approach and Methodology for the Evaluation of Donor Supportto Public Financial Management (PFM) in Development Countries Part A July 2009and Part B September 2009
138
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
• Incorporate an approach to test the findings and hypotheses emergingfrom the quantitative study.Task 2: Desk Review• Review existing PEFA/HIPC assessments and performance data (forthe three case study countries) – what does the evidence tell us aboutPFM reform progress over the evaluation period? What specificreforms have been followed and is it possible to track them over time?• Gather information on donor support to PFM reform over the evalua-tion period (including project documents where these are available).While 2001 is the date of the first HIPC assessment, it might also makesense to extend that period backwards to capture earlier donor PFMsupport, and initial reforms that took place in the mid to late 1990s.• Identify the level and character of government initiatives, and (changesin) the level of political will to undertake PFM reforms during the eval-uation period, classifying it according to the following categories:– Stage of the budget cycle (e.g. preparation, approval, execution,audit)– Budgeting time horizon (e.g. annual budget, MTEF)– Involved stakeholders (e.g. Parliament, CSOs, DPs)– Type of input (e.g. legislative, human capital, infrastructure)– Cost (direct and indirect) and time• Identify donor support (inputs in the evaluation framework) to PFM reformduring the evaluation period in each country case study and classify itaccording to the three dimensions of quality PFM (described above) andthe following:– its phase in the budget cycle e.g. preparation, approval, execution,audit– type e.g. technical assistance, training, capacity building, softwareand computer installation, budget support, dialogue on PFMreform;– process of delivery e.g. project management unit, or through Gov-ernment systems;– donor providing the support e.g. World Bank, IMF, AFDB, andbilaterals and whether it is joined up or implemented through sepa-rate project agreements;– cost and time;• Using project documents, identify the outputs that support interven-tions are intended to deliver e.g. people, skills and organizationalcapacity; changes in laws, rules and procedures; improved informationsystems and business processes; and changes in incentives and controls(see evaluation framework);• Compile a timeline of support showing the sequencing of donor sup-ported reform activities (plus investments). Can results in PEFA assess-ments be linked to specific reforms that have been supported bydonors?139
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
Task 3: Undertake country visitsThe purpose of the country visits is to assess the relevance, efficiency, effec-tiveness and sustainability aspects of the theory of change framework and howdonor support has been designed and implemented.The evaluation team would be expected to carry out structured and semistructured interviews with Ministry of Finance officials, other officials in gov-ernment, sector ministries, local government, politicians, civil society anda range of donors both those involved in PFM reform and those that are not.The consultants could consider sending a limited number of questions inadvance of the country visit to both donors and government officials. If a jointdonor/government budget support or PFM group exists detailed discussionshould be held with these groups to review experience over time.Task 4: Report draftingThe consultants will be expected to produce stand alone reports for each coun-try case study. Reports should be no longer than 30 pages with additional infor-mation included as annexes as necessary. Reports should be succinct as it isimportant to produce written information which is accessible to a wide audienceand to readers whose first language may not be English (or conversely French).As far as possible, the initial findings of the evaluation should be discussed withthe participating government, with donor partners and other stakeholders incountry for comment and feedback before the evaluation team departs. Draftreports would be presented to the Management Group and Evaluation Refer-ence group for comments and feedback before the final report is produced.
9. budgetThe total cost (fees and reimbursables) for the evaluation must not exceed SEK3,500,000.
10. Deliverables84and timetableThe consultant will undertake the following tasks within the timeline set outbelow:DeliverableInception phase consultation workshop in TunisPrepare an Inception Report setting out the ap-proach to the case studies. Undertake in countryconsultations during the inception stageSubmit byMarch 201131stMarch 2011
Undertake adesk based review of PEFA reports30thApril 2011and other available evidence about PFM reforms inthe case study countries. The objective is to developapreliminary overview of and hypotheses abouthow the reforms have been undertaken, partnergovernment initiatives and the level and characterof the external support.140
84 All deliverables will be submitted electronically in English and French
ANNEx E: CoNSuLtANt tErmS oF rEFErENCE
Deliverable
Submit by
Visit case study countries and undertake arange of May–June 2011interviews, workshops, and ade-briefing work-shops at the end of the field visit (following acom-mon evaluation approach)Briefing report with preliminary findings fromcountry visitsCountry case study reports submitted (followingacommon format)Final Synthesis Submitted30thJune 201131stAugust 201130thSeptember2011
11. consultant qualifications and skillsThe work will require a small team of consultants who have experience inPFM reform, with part of that experience being in the evaluation of develop-ment policy, programs or project operations. Support may be required togather information on donor support in both case studies including projectdocuments, PEFA reports and HIPC data and to provide support with thedesk study. Two consultants will be required for each case study and it will beimportant that the team leader is fluent in both English and French (in read-ing and writing) and involved in all three case studies. Consultants would beexpected to be familiar with reform approaches, partner government initia-tives and interventions of donor agencies in African countries particularly inthe area of PFM. Familiarity with PEFA assessment systems will beimportant.Compulsory requirements for personnel are specified in section 4.2.1 forteam leader and 4.2.2 for other personnel. Evaluation criteria for qualificationand competence are specified in 7.2.1 for the team leader and 7.2.2 for otherteam members.Final reports should be submitted in both English and French. All reportssubmitted should be professionally edited.
12. Management and administrationThe consultant will report to the evaluation task manager for methodologicalguidance, preparation and drafting of the report. Draft reports will be sub-mitted to the Management Committee made up of Sida, DANIDA andAFDB evaluation departments. Drafts may be circulated to PFM professionalstaff in these institutions including country specialists for comments. Finaldrafts will also be circulated to a wider group of stakeholders mainly includ-ing bilateral and multilateral donors and PFM experts who may submit fur-ther comments. In country support would be provided by donor countryoffices.
141
Joint Evaluations
1996:1
The international response to conflict andgenocide: lessons from the Rwanda experience:Synthesis ReportJohn Eriksson, Howard Adelman, John Borton, Krishna Kumar,Hanne Christensen, Astri Suhrke, David Tardif-Douglin, SteinVillumstad, Lennart WohlgemuthSteering Committee of the Joint Evaluation ofEmergency Assistance to Rwanda,1996.
1997:1
Searching for Impact and Methods: NGOEvaluation Synthesis Study.Stein-Erik Kruse, Timo Kyllönen, Satu Ojanperä, Roger C. Rid-dell, Jean-Louis VielajusMin of Foreign Affairs Finland, OECD-DAC, Sida, 1997.
1997:2
Measuring and Managing Results: Lessons forDevelopment Cooperation: Performance ManagementDerek PoateUNDP/OESP Sida, 1997.
2003:1
Local Solutions to Global Challenges: Towards Effec-tive Partnership in Basic Education. Final Report.Joint Evaluation of External Support to Basic Educa-tion in Developing Countries.Ted Freeman, Sheila Dohoo FaureNetherlands Ministry of Foreign Affairs, CIDA, DFID, Depart-ment for Foreign Affairs Ireland, EU, BMZ, JICA, Ministry ofBasic Education and Literacy Burkina Faso, Danida, Norad,Sida, UNESCO, UNICEF, World Bank. 2003.
2003:2
Toward Country-led Development : aMulti-PartnerEvaluation of the Comprehensive Development Frame-work : Synthesis reportCarol Lancaster, Alison Scott, Laura Kullenberg, Paul Collier,Charles Soludo, Mirafe Marcos, John Eriksson, Alison Scott;Ibrahim Elbadawi;John Randa,World Bank, OED, CIDA, Danida, Norad, ODI, JICA, Sida, 2003.
142
JoINt EvALuAtIoNS
2005:1
Support to Internally Displaced Persons: Learningfrom Evaluation. Synthesis Report of aJoint Evalua-tion Programme.John Borton, Margie Buchanan Smith, Ralf OttoSida, 2005.
2005:2
Support to Internally Displaced Persons: Learningfrom Evaluation. Synthesis Report of aJoint Evalua-tion Programme: Summary VersionJohn Borton, Margie Buchanan Smith, Ralf OttoSida, 2005.
2005:3
Humanitarian and Reconstruction Assistance toAfghanistan 2001–2005: From Denmark,Ireland, the Netherlands, Sweden and the UnitedKingdom; A Joint Evaluation. Main reportDanida, Sida, Chr. Michelsen Institute, Copenhagen, DFID,Development Cooperation Ireland, BMZ, 2005.
2005:4
Humanitarian and Reconstruction Assistance toAfghanistan 2001–2005: From Denmark,Ireland, the Netherlands, Sweden and theUnited Kingdom; A Joint Evaluation. SummaryDanida, Sida, Chr. Michelsen Institute, Copenhagen, DFID,Development Cooperation Ireland, BMZ, 2005.
2005:5
An Independent External Evaluation of the Interna-tional Fund or Agricultural DevelopmentDerek Poate, team leader, Charles Parker, Margaret Slettevold …IFAD, Sida, CIDA, 2005.
2006:1
Joint Evaluation of the International response to theIndian Ocean tsunami: Synthesis ReportJohn Telford, John Cosgrave, contribution Rachel HoughtonTsunami Evaluation Coalition (TEC) Action aid, AusAID, BMZCIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO, IFRD,Federal Min for Economic Cooperation and Development Germa-ny, JICA, Min des Affaires Étrangères France, Min des AffairesÉtrangères Luxembourg, Norad, NZAID, DEZA, Sida, UN,UNDP, UNFPA, Unicef, Usaid, WFP, WHO, World Vision, 2006.
2006:2
Impact of the tsunami response on local and nationalcapacitiesElisabeth Scheper, Arjuna Parakrama, Smruti Patel, contribu-tion Tony Vaux143
JoINt EvALuAtIoNS
Tsunami Evaluation Coalition (TEC) Actionaid, AusAID, BMZ,CIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO, IFRD,Federal Min for Economic Cooperation and Development Germa-ny, JICA, Min des Affaires Étrangères France, Min des AffairesÉtrangères Luxembourg, Norad, NZAID, DEZA, Sida, UN,UNDP, UNFPA, Unicef, Usaid, WFP, WHO, World Vision, 2006.2006:3Coordination of International Humanitarian Assis-tance in Tsunami-affected countriesJon Bennett, William Bertrand, Clare Harkin, Stanley Samaras-inghe, Hemantha WickramatillakeTsunami Evaluation Coalition (TEC) Actionaid, AusAID, BMZ,CIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO, IFRD,Federal Min for Economic Cooperation and Development Ger-many, JICA, Min des Affaires Étrangères France, Min desAffaires Étrangères Luxembourg, Norad, NZAID, DEZA, Sida,UN, UNDP, UNFPA, Unicef, Usaid, WFP, WHO, WorldVision, 2006.2006:4Funding the Tsunami Response: A synthesis of findingsMichael Flint, Hugh GoyderTsunami Evaluation Coalition (TEC) Actionaid, AusAID,BMZm CIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO,IFRD, Federal Min for Economic Cooperation and Develop-ment Germany, JICA, Min des Affaires Étrangères France, Mindes Affaires Étrangères Luxembourg, Norad, NZAID, DEZA,Sida, UN, UNDP, UNFPA, Unicef, Usaid, WFP, WHO, WorldVision, 2006.2006:5Links between relief, rehabilitation and developmentin the Tsunami response: A synthesis of initial findingsIan ChristoplosTsunami Evaluation Coalition (TEC) Actionaid, AusAID, BMZmCIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO, IFRD,Federal Min for Economic Cooperation and Development Ger-many, JICA, Min des Affaires Étrangères France, Min des AffairesÉtrangères Luxembourg, Norad, NZAID, DEZA, Sida, UN,UNDP, UNFPA, Unicef, Usaid, WFP, WHO, World Vision, 2006.2006:6The role of needs assessment in the Tsunami response– Executive summaryClaude de Ville de Goyet, Lezlie C MorinièreTsunami Evaluation Coalition (TEC) Actionaid, AusAID,BMZm CIDA, Cordaid, Danida, Dara, Irish Aid, DFID, FAO,IFRD, Federal Min for Economic Cooperation and Develop-
144
JoINt EvALuAtIoNS
ment Germany, JICA, Min des Affaires Étrangères France, Mindes Affaires ÉtrangèresLuxembourg, Norad, NZAID, DEZA, Sida, UN, UNDP, UNF-PA, Unicef, Usaid, WFP, WHO, World Vision, 2006.2006:7Evaluation of Coordination and Complementarity ofEuropean Assistance to Local Development: with Ref-erence to the 3C Principles of the Maastricht TreatyRobert N. LeBlanc and Paul BeaulieuSida, Ministry for Foreign Affairs, Austria, Ministry for ForeignAffairs, Department for International Development Coopera-tion. Belgium, Min. des Affairs étrangères/Direction General dela Cooperation International, France, Department of ForeignAffairs Development Co-operation Division, Ireland and Minis-try of Foreign Affairs/Directorate-General for InternationalCooperation, the Netherlands, 2006.2007:1Evaluation of General Budget Support – Note onApproach and Methods. Joint Evaluation ofGeneral Budget Support 1994–2004AFD, DFID, MOFA, NZAID, USAID, AusAID, BMZ, JBIC,NORAD, Danida, SECO, CIDA, JICA, Min of Foreign AffairsSpain, Portuguese Development Cooperation, Sida, 2007.2007:2Evaluating Co-ordination, Complementarity andCoherence in EU development policy:a synthesisEvaluation Services of the European Union, Sida, Ministry forForeign Affairs, Austria,Ministry for Foreign Affairs, Department for InternationalDevelopment Cooperation. Belgium, Min. des Affairsétrangères/Direction General de la Cooperation International,France, Department of Foreign Affairs Development Co-opera-tion Division, Ireland and Ministry of Foreign Affairs/Directo-rate-General for International Cooperation, Netherlands, 2007.2007:3Evaluating Democracy Support: Methods and Experi-ences.Sida, Department for Evaluation and Internal Audit and Internation-al Institute for Democracy and Electoral Assistance (IDEA), 2007.2007:4Peer Review Evaluation Function at the World FoodProgramme (WFP). Peer Panel Members: Jock Baker, Stefan Dahl-gren, Susanne Frueh, Ted Kliest, Zenda Ofir.Advisors to the Panel:Ian Christoplos, Peta Sandison Sida, BMZ, UNEG, WFP, 2007.145
JoINt EvALuAtIoNS
2008:1
Managing Aid Exit and Transformation:Lessons from Botswana, Eritrea, India, Malawi andSouth Africa: Synthesis ReportAnneke Slob, Alf Morten JerveSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:1
Managing Aid Exit and Transformation:Summary of aJoint Donor EvaluationJesper HeldgaarSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:2
Managing Aid Exit and Transformation: India Coun-try Case StudyAlbert de Groot, CK Ramachandran, Anneke Slob, Anja Wil-lemsen, Alf Morten JerveSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:3
Managing Aid Exit and Transformation: South AfricaCountry Case StudyElling N Tjønneland, Pundy Pillay, Anneke Slob, Anje Willem-sen, Alf Morten JerveSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:4
Managing Aid Exit and Transformation: Eritrea Coun-try Case StudyTeferi Michael, Rudy Ooijen, Anneke Slob, Alf Morten JerveSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:5
Managing Aid Exit and Transformation:Malawi Country Case StudyEsther van der Meer, Arne Tostensen, Anneke Slob, Alf MortenJerveSida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
2008:1:6
Managing Aid Exit and Transformation:Botswana Country Case StudySida, Netherland’s Ministry of Foreign Affairs, Danida andNorad, 2008.
146
JoINt EvALuAtIoNS
Charity Kerapeletswe, Jan Isaksen, Anneke Slob, Alf MortenJerve2008:2Evaluation of the Implementation of the Paris Declara-tion: Phase One Synthesis ReportBernard Wood, Dorte Kabell, Nansozi Muwanda,Francisco SagastiInternational Reference Group comprising members of the DACNetwork on Development Evaluation, 2008.2008:3Joint Evaluation of Citizen’s Voice andAccountability: Synthesis ReportAlina Rocha Menocal, Bhavna SharmaCommissioned by Directorate-General for Development Coop-eration (Belgium) – DGCD, Danish International DevelopmentAssistance – Danida, Federal Ministry for Economic Coopera-tion and Developmen (Germany) – BMZ, Norwegian Agency forDevelopment Cooperation – Norad, Swedish InternationalDevelopment Cooperation Agency – Sida, Swiss Agency forDevelopment and Cooperation – SDC, Department for Interna-tional Development – DFID, 2008.2009:1Anti-Corruption Approaches: A Literature ReviewArne Disch, Endre Vigeland, Geir SundetCommissioned by Asian Development Bank – ADB, DanishInternational Development Assistance – Danida, Department forInternational Development – DFID, Norwegian Agency forDevelopment Cooperation – Norad, Swedish Agency for Devel-opment Evaluation – SADEV, Swedish International Develop-ment Cooperation Agency – Sida, 2009.2009:2Public Financial Management ReformLiterature ReviewCarole Pretorius, Nico Pretorius (Evaluation Report EV698)Commissioned by Department for International Development –DFID, Dutch Ministry of Foreign Affairs, Swedish InternationalDevelopment Cooperation Agency – Sida, Canadian Internation-al Development Agency – CIDA, African Development Bank –AfDB, 2009.2009:3A ripple in development? Long term perspectives onthe response to the Indian Ocean Tsunami: A joint fol-low-up evaluation of the links between relief, rehabili-tation and development (LRRD)Emery Brusset (team leader), Mihir Bhatt, Karen Bjornestad,John Cosgrave, Anne Davies, Adrian Ferf, Yashwant Deshmukh,147
JoINt EvALuAtIoNS
Joohi Haleem, Silvia Hidalgo, Yulia Immajati, Ramani Jayasun-dere, Annina Mattsson, Naushan Muhaimin, Adam Pain, Ric-cardo Polastro, Treena Wu.Commissioned by LRRD2 Joint Steering Committee, Sida,Norad, Danida, the Netherlands Ministry for Foreign Affairs,CIDA, BAPPENAS, Indonesia; BRR, Indonesia;Ministry for Plan Implementation, Sri Lanka, Ministry forNational Building, Sri Lanka; ISDR, Bangkok; IFRC, Bangkok;CARE International;OCHA; UNICEF, 2009.2009:3:1A ripple in development? Document review: Annotatedbibliography prepared for the joint follow-up evalua-tion of the links between relief, rehabilitation anddevelopment (LRRD) in responses to the Indian OceantsunamiJohn Cosgrave, with the assistance of: Emery Brusset, MihirBhatt, Yashwant Deshmukh, Lucia Fernandez, Yulia Immajati,Ramani Jayasundere, Annina Mattsson, Naushan Muhaimin,Riccardo PolastroCommissioned by LRRD2 Joint Steering Committee, Sida;Norad; Danida; the Netherlands Ministry for Foreign Affairs;CIDA; BAPPENAS, Indonesia; BRR, Indonesia; Ministry forPlan Implementation, Sri Lanka; Ministry for National Building,Sri Lanka; ISDR, Bangkok; IFRC, Bangkok; CARE Interna-tional; OCHA; UNICEF, 2009.2009:3:2A ripple in development? Long term perspectives onthe response to the Indian Ocean Tsunami: A joint fol-low-up evaluation of the links between relief, rehabili-tation and development (LRRD) – Summary ReportEmery Brusset (team leader), Mihir Bhatt, Karen Bjornestad,John Cosgrave, Anne Davies, Adrian Ferf, Yashwant Deshmukh,Joohi Haleem, Silvia Hidalgo, Yulia Immajati, Ramani Jayasun-dere, Annina Mattsson, Naushan Muhaimin, Adam Pain, Ric-cardo Polastro, Treena Wu.Commissioned by LRRD2 Joint Steering Committee, Sida;Norad; Danida; the Netherlands Ministry for Foreign Affairs;CIDA; BAPPENAS, Indonesia; BRR, Indonesia; Ministry forPlan Implementation, Sri Lanka; Ministry for National Building,Sri Lanka; ISDR, Bangkok; IFRC, Bangkok; CARE Interna-tional; OCHA; UNICEF, 2009.2010:1Evaluation of the Joint Assistance Strategy for Zambia(JASZ) 2007–2010.Anne Thomson, Dennis Chiwele, Oliver Saasa, Sam Gibson
148
JoINt EvALuAtIoNS
Commissioned by Ministry of Foreign Affairs ofDenmark – Danida, Swedish International Development Coop-eration Agency – Sida, Irish Aid, 2010.2011:1Supporting Child Rights – Synthesis of LessonsLearned in Four Countries: Final ReportArne Tostesen, Hugo Stokke, Sven Trygged, KateHalvorsenCommisioned by Swedish International Development Agency –Sida and Norwegian Agency for Development Cooperation –Norad, 2011.2011:2Aiding the Peace. A Multi-Donor Evaluation of Supportto Conflict Prevention and Peacebuilding in SouthernSudan 2005–2010. Final ReportJon Bennett, Emery Brusset, Chris Barnett, Sara Pantuliano,Wendy Fenton, Anthony VauxCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation –Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairsof Denmark – Danida, 2011.2012:1Joint Evaluation of Support to Anti-Corruption EffortsSynthesis 2002–2009Derek Poate, Charlotte VaillantCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairs ofDenmark – Danida, 2012.2012:2Joint Evaluation of Support to Anti-Corruption EffortsBangladesh Country ReportDerek Poate, Charlotte Vaillant , Imran Ahmed, Deborah Mans-field, Mozammel Hoque, Zarina Rahman KhanCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development
149
JoINt EvALuAtIoNS
Cooperation Agency – Sida and the Ministry of Foreign Affairsof Denmark – Danida, 2012.2012:3Joint Evaluation of Support to Anti-Corruption EffortsNicaragua Country ReportDerek Poate, Paul Harnett, Imran Ahmed, Mignone Vega, JoseLuis VelasquezCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairs ofDenmark – Danida, 2012.2012:4Joint Evaluation of Support to Anti-Corruption EffortsTanzania Country ReportCharlotte Vaillant, Imran Ahmed, Deborah Mansfield, AnneBartholomew, Isaac KiwangoCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairs ofDenmark – Danida, 2012.2012:5Joint Evaluation of Support to Anti-Corruption EffortsVietnam Country ReportDerek Poate, Edmund Attridge, Tim McGrath, Dang NgocDung, Nguyen Thi Minh HaiCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairs ofDenmark – Danida, 2012.2012:6Joint Evaluation of Support to Anti-Corruption EffortsZambia Country ReportCharlotte Vaillant, Imran Ahmed, Paul Harnett, DeborahMansfield, Gilbert Mudenda, Stephen TemboCommissioned by the Evaluation Departments of NorwegianAgency for Development Cooperation – NORAD, United King-
150
JoINt EvALuAtIoNS
dom Department for International Development – DFID, AsianDevelopment Bank – ADB, Swedish Agency for DevelopmentEvaluation – Sadev, Swedish International Development Coop-eration Agency – Sida and the Ministry of Foreign Affairs ofDenmark – Danida, 2012.2012:7Evaluation of Public Financial Management Reform inBurkina Faso, Ghana and Malawi 2001–2010 Final Syn-thesis ReportAndrew LawsonCommissioned by the African Development Bank – AfDB, theSwedish International Development Cooperation Agency – Sida,and the Danish International Development Assistance – DANI-DA, 2012.2012:8Evaluation of Public Financial Management Reform2001–2010 Ghana Country ReportMary Betley, Andrew Bird, Adom GharteyCommissioned by the African Development Bank – AfDB, theSwedish International Development Cooperation Agency – Sida,and the Danish International Development Assistance –DANIDA, 2012.2012:9Evaluation of Public Financial Management Reform inMalawi 2001–2010 Final Country Case Study ReportAlta Fölscher, Alex Mkandawire, Ruth FaragherCommissioned by the African Development Bank – AfDB, theSwedish International Development Cooperation Agency – Sida,and the Danish International Development Assistance –DANIDA, 2012.2012:10Evaluation of Public Financial Management Reform inBurkina Faso 2001–2010 Final Country Case StudyReportAndrew Lawson, Mailan Chiche, Idrissa OuedraogoCommissioned by the African Development Bank – AfDB, theSwedish International Development Cooperation Agency – Sida,and the Danish International Development Assistance –DANIDA, 2012.
151
Evaluation of Public financial Management Reforminghana, 2001–2010final country case Study ReportWhere and why do Public Financial Management (PFM) reforms succeed? Where and how does donor support toPFM reform contribute most effectively to results? To answer these questions, an evaluation of PFM reforms hasbeen carried out, primarily based on country studies of Burkina Faso, Ghana and Malawi. An internationalquantitative study and a literature review were also undertaken. This report presents the findings of the study inBurkina FasoThe findings from the three country studies are summarised in a separate synthesis report, concluding thatresults tend to be good when there is a strong commitment at both political and technical levels, when reformdesigns and implementation models are well tailored to the context and when strong, government-led coordina-tion arrangements are in place to monitor and guide reforms.Donor funding for PFM reform has been effective in those countries where the context and mechanisms wereright for success, and where external funding was focused on the Government’s own reform programme. Thewillingness of some Governments to fund PFM reforms directly shows that external funding may not be thedeciding factor, however. Donor pressure to develop comprehensive PFM reform plans has been a positiveinfluence in countries receiving Budget Support, but attempts to overtly influence either the pace or the content ofPFM reforms were found to be ineffective and often counter-productive. Key lessons for donor agencies are thusto focus on where the right preconditions exist, to align to government programmes and, under all circumstances,to ensure that aid works in favour of the PFM system and not against it.The evaluation has been commissioned jointly by the African Development Bank (AfDB), the Swedish InternationalDevelopment Cooperation Agency (Sida) and the Danish International Development Assistance (DANIDA).
SwEdiSh inTERnATiOnAl dE vElOPMEnT COOPER ATiOn AGEnCyAddress: SE-105 25 Stockholm, Sweden.visiting address: valhallavägen 199.Phone: +46 (0)8-698 50 00. Fax: +46 (0)8-20 88 64.www.sida.se [email protected]
