Report No. 34582-AF Afghanistan Managing Public Finances for Development (In Five Volumes) Volume II: Improving Public Financial Management and Procurement December 22, 2005 Poverty Reduction and Economic Management Sector Unit South Asia Region Document of the World Bank Table of Contents ACKNOWLEDGEMENTS ..................................................................................................................... i11 PART I THE PERFORMANCE OF AFGHANISTAN'S PUBLIC FINANCIALMANAGEMENT . SYSTEM ...................................................................................................................................................... 4 EXECUTIVESUMMARY .............................................................................................................................. 4 A.INTRODUCTION ................................................................................................................................... 10 Background......................................................................................................................................... 10 PFMPerformance Measurement......................................................................................................... 10 P F M Framework (Rules, Roles, and Information) .............................................................................. 11 B.Information Systems............................................................................................................................ 13 OVERALLPERFORMANCE OFPFMSYSTEM ....................................................................................... 14 PFMOut-turns:Credibility o fthe Budget........................................................................................... 14 Key Cross Cutting Features: Comprehensiveness and Transparency ................................................. 17 Budget Cycle ....................................................................................................................................... 21 b. a. Policy-Based Budgeting ............................................................................................................ 21 Predictability and Control inBudget Execution ........................................................................ 25 c. Accounting, Recording, and Reporting...................................................................................... 37 d. External Scrutiny and Audit....................................................................................................... 40 e. Indicatorso f Donor Practices..................................................................................................... 42 C.SUSTAININGAND STRENGTHENINGPFMPERFORMANCE:A ROADMAP............................................ 44 ANNEX................................................................................................................................................... 46 Annex 1 1:Roadmap to StrengthenPFMPerformance..................................................................... . 46 PART I1 PUBLICPROCUREMENT FORDEVELOPMENTEFFECTIVENESS . ........................ 52 EXECUTIVESUMMARY ............................................................................................................................ 52 A.INTRODUCTION ................................................................................................................................... 53 B.THEBUDGET c.PROCUREMENTTHROUGH AND PROCUREMENT ...................................................................................................... 53 THE PROCUREMENTAGENT ...................................................................... 55 Role ofthe Procurement Agent ........................................................................................................... 55 Procurement Services to Date ............................................................................................................. 56 Statistical Performance........................................................................................................................ 57 D.PROCUREMENTUNDERTHENATIONAL PROCUREMENTLAW a. ............................................................ 60 b. Pillar I- The Legislative and Regulatory Framework ............................................................... 61 Pillar I1- Central Institutional Framework and Capacity .......................................................... 65 c. Pillar I11- Procurement Operations and Market Performance ................................................... 68 d. Pillar IV - The Integrity o f the Public Procurement System...................................................... 72 E. PROCUREMENTUNDEREXTERNAL PROJECTASSISTANCE THROUGHTHE COREBUDGET ................75 F.ADDRESSING STRUCTURAL WEAKNESSES .......................................................................................... 75 Developing Capacity inthe Government ............................................................................................ 75 G.Legal Framework ................................................................................................................................ 76 H.PRIVATE SECTORISSUES .................................................................................................................... 76 CONCLUSIONSAND RECOMMENDEDACTIONS ................................................................................... 78 REFERENCE ............................................................................................................................................ 80 i FIGURES Part I Figure 1. 1: Share o fNon-Salary Expenditures Reimbursedby ARTF ...................................................... 36 Part I1 Figure2. 1:Procurement Implications o f Source o fFunds......................................................................... 54 Figure 2.2: Advertised Contracts................................................................................................................ 58 Figure 2. 3: Procedures Used ...................................................................................................................... 59 TABLES Part I Table 1. 1: Budget Out-turns....................................................................................................................... 15 Table 1 2: Expenditure Composition Variation . .......................................................................................... 16 Table 1. 3: Error Rate on Submissions to Ministry o f Finance................................................................... 35 Part I1 Table 2. 1:Fundingby Donor ..................................................................................................................... 56 Table 2.2: Summary of Procurement Performance Indicators................................................................... 57 Table 2 3: Summary o f Procurement Benchmark Indicators..................................................................... 61 Table 2 4: ABA Registrationand Classification ........................................................................................ .. 77 .. 11 ACKNOWLEDGEMENTS As part o f the overall Public Finance Management Review (see Acknowledgements for Volume I), this volume was prepared by the following team: Part 1(PFM Performance Assessment): Paul Sisk (SARFM), StCphane Guimbert (SASPR), The0 Thomas (IMF), William Byrd (SASPR), with advice from Nicola Smithers (PEFA). Part 2: (Procurement): Q u a m 1 Hasan (SARPS) and Peter Trepte (Consultant) This volume was processedbyJuliet Teodosio (SASPR). The process of carrying out the PFMReview involved extensive interactions with the Ministryo f Finance and other Government agencies, whose valuable cooperation and support was indispensable for completing the review and also greatly benefited this report. ... 111 PART I.THE PERFORMANCE OF AFGHANISTAN'S PUBLIC FINANCIAL MANAGEMENT SYSTEM Executive Summary i. Part Iof this volume presents an initial Public Financial Management (PFM) Performance Assessment for Afghanistan.' It i s based on a performance measurement framework developed by the Public Expenditure and Financial Accountability (PEFA) partnership program2 which identifies a set o f critical objectives o f a PFM system and a standard set o f high-level PFMindicators to assess performance against those objectives. The methodology provides a consistent framework for analysis and a monitoring tool for the Government to manage reform and demonstrate progress to domestic and international stakeholders. The ratings for Afghanistan are summarized in a table at the end o f this Executive Summary. After providing some background and a description o f Afghanistan's PFM system, the main body o f this assessment discusses each PFM indicator separately and outlines the reasoning behind the rating assigned to Afghanistan. The assessment ends with a roadmap for strengthening PFM performance. ii. The situation inAfghanistanhas some special features whichneedto be kept inmindwhen interpreting the ratings. First, some ratings reflect temporary factors (for instance the absence o f a Parliament inthe past). Second, some ratings are based on changes made recently, and it i s still uncertain whether these changes will be sustained. Third, the ratings assess the current situation in which significant external support (both advisory and operational) for the PFM system i s being provided to the Government; in several areas, this raises sustainability issues as the external support will have to be reduced over time, especially in relation to financial management operations. And finally, a new Public Finance and Expenditure Management (PFEM) Law and a new Procurement L a w were approved after this assessmentwas conducted; hence some ratings would be expected to change as a result of this major improvement inthe legal framework. iii. PublicfinancialmanagementinAfghanistanasdepictedthroughthePFMPerformanceIndicators as o f June 2005 i s summarized below under the six critical attributes o f a PFM system. The ratings against each o f the 28 individual PFM Performance Indicators and for the three indicators o f donor performance are shown inthe table below. Credibility of the budget (Performance Indicators 1-4) iv. Maintaining fiscal discipline all along has been a key Government priority. While the budget, before foreign grants, i s always in deficit given the low level o f domestic revenues, over the past three years the out-turn o f the deficit has been below plannedlevels. Three mechanisms have contributed to this outcome: (i) progress inrevenue mobilization, although as a share o f GDP it remains at well below strong half o f the level found inother low-income countries; (ii) a formal mid-year review process that monitors An earlier draft of this report was presented at the Afghanistan Development Forum in Kabul in April 2005. See www.adf.gov.af. This version incorporates additional analysis and a revision to the set o fperformance indicators. The members o f PEFA include the World Bank, International Monetary Fund, European Commission, UK Department for International Development, French Ministry o f Foreign Affairs, Royal Norwegian Ministry o f Foreign Affairs, Swiss State Secretariat for Economic Affairs, and Strategic Partnership with Africa (SPA). The analytical framework, indicator set, and detailed instructions for ratings are inPEFA (2005). Also see www,wfa.org. 4 execution and may adjust appropriations; and (iii) strictly applied no-overdraft policy that prohibits a central bank financing o f the budget. The methodology behind the ratings system is set out in the Public Financial Management Performance Measurement Framework, produced by the Public ExpenditureWorking Group (World Bank, IMF and Joint Donor Public Expenditure and FinancialAccountabilityUnit),available at http://www.pefa.org/Documents.htm. 5 v. Despite continuing improvements inbudget processes, the credibility o f the budget i s hampered by optimistic budget projections, particularly for development expenditures, mainly attributable to a lack o f realism at the budget formulation stage and limited capacity to implement the budget. Moreover, the External Budget (accounting for roughly three-quarters o f total public spending in the country) remains effectively outside the Government's control, further reducing the credibility o f the budget. There are two specific issues: (i)the External Budget is to a large extent outside the Government's control (and so to that extent cannot be used as an instrument o f policy); and (ii) gap between donor commitments the reported in the budget and actual expenditures i s large. A detailed analysis shows large deviations between budgeted amounts and actual out-turns (Table 1.1). Comprehensivenessand transparencyof budget information (Performance Indicators 5-10) vi. The Government i s committed to share budgetary information transparently. The annual budget for the national government i s impressive, with coverage o f most public spending o f the general Government sector, whether implemented by the Government or directly by donors. The budget is a useful public document relating expenditures on specific projects to National Programs. Revenue and expenditure records for the Core Budget are maintainedthrough the budget implementation system known as AFMIS (the Afghanistan Financial Management Information System). Comprehensive reports on budgetary aggregates are produced monthly and are now available on the Ministry o f Finance's (MoF) website. This information is generally reliable but is affected by lags inrecordingprovincial operations and weak reconciliation. There are also uncertainties about the quality o f reported information on the External Budget. In addition, fiscal risks which could arise fi-om the activities o f state-owned enterprises and municipalities are not effectively monitored because the financial information i s unverified and too limited for fiscal risk management inthese areas. Policy-based budgeting (Performance Indicators 11-12) vii. Some progress has been made toward the Government's key objective of making the national budget the central instrument o f policy and reform, but there are still major constraints. The full implications o f policy decisions, including for fiscal sustainability, are not taken into account due to lack o f capacity in Government to develop policy, agree on trade-offs among investments and recurrent costs, and match resources with policies. Multi-year planning i s also constrained by the complexity of aid coordination. Budgeting for recurrent and investment expenditures i s carried out separately, raising the fiscal sustainability risks fi-om recurrent cost implications o f investments. The budget process i s orderly and well understood within the Government, but it needs to be linked more closely with the Government's strategy; more widely owned politically, including by Cabinet and the Parliament; and more reflective of sector issues with more engagement by sector agencies, provinces, and civil society. Predictability and control in budget execution (Performance Indicators 13-21) viii. Much progress has been made in improving the implementation o f the budget (inparticular with respect to the allotment process and cash management), but there i s a need to enhance predictability o f funding for service delivery units. Line departments at the provincial level often cannot predict with accuracy the funds that will be available to them because o f delays incommunicating with the provincial offices o f M o F and the complexity o f the allotment process; this has led to cash rationing at the discretion o f the Treasury and the Mustoufiats. Many adjustments are made during the year, in particular through allotment transfers and use o f contingencies. N o information i s available at the level o f service delivery units(e.g. primary schools) on actual resources available. 6 ix. A particular weakness in the budget execution cycle is related to revenue collection. The overall system i s not very effective, as demonstrated by the large gap between revenues that could be collected and actual collections. In addition, despite some progress with taxpayer registration, the control framework on the revenue side remains weak. Nevertheless, clarification o f the legislation and strengthening o f financial management (banking arrangements to ensure a steady flow o f collected revenues to the Treasury Single Account, TSA) have generated some improvements in revenue performance. x. Much progress has been made in implementing the control framework, as reflectedby a significant increase in the eligibility rate for reimbursements o f expenditures by the Afghanistan Reconstruction Trust Fund(ARTF). There i s a division o fdutiesbetween the line ministries, which approve expenditure, and MoF, which keeps records and makes payments. These arrangements include external operational support in the Treasury and review by the ARTF Monitoring Agent. The controls also include effective cash management, allotment control on budget uses, and reconciliations o f records o f the Treasury and the line ministries. Significant weaknesses remain, however, compounded by the lack o f internal audit to provide feedback on performance and systems. Further improvements in the internal control framework are needed for both payroll and non-payroll expenditures and for revenues. The recent approval of a new PFEM Law and plannedfinancial regulations will help further strengthen the control framework. xi. Most procurement at present is carried out under standards agreed with donors. Procurement carried out under national rules, however, has been subject to an outdated regulatory framework which lacks institutional support adequate for effective enforcement. There i s insufficient competition as the basis for contract awards and no guarantee o f value for money. The new Procurement Law, approved in September 2005, provides a strong foundation for improving the regulatory and institutional framework. Accounting, recording, and reporting (Performance Indicators 22-25) xii. Considerable improvements have been made in this area. The recording o f expenditures and revenues has been strengthened by partial computerization o f AFMIS, notably in M o F (also piloted in two line ministries). Accountability o f line ministries and o f M o F to the Cabinet i s supported by frequent reporting on budget implementationby the Government, but improving accounting and reporting capacity inline ministries is required. Significant progress has been made toward consolidationo fbank accounts under a system o f TSA. However, much remains to be done to improve standards o f accounting, frequency o fbank reconciliation, and usefulness o freporting for line ministries' management. External scrutiny and audit (Performance Indicators 26-28) xiii. Government accountability generally is sought through the publication o f audited financial statements on the state budget and specific donor funds. The Auditor General i s working to international auditing standards with operational support from an audit advisor. However, the final audit o f the 2003/04 annual statements has not yet been released. Inaddition, the absence o f a Parliament (until the elections held in September 2005 and initial meeting o f ParliamentinDecember) has limited the extent o f external scrutiny of the budget and its performance. Donor practices (Donor Performance Indicators 1-3) xiv. Donors have mobilized a considerable amount o f external support for Afghanistan's budget. In particular, support to the Government's recurrent budget through the ARTF has been large and predictable. However, direct budget support still accounts for only a very small percentage o f total external assistance to Afghanistan, and the proportion o f aid that i s managed under national procedures remains low, with most aid executed by donors through their contractors. Financial information provided 7 by donors on donor-executed assistance activities, while reasonably comprehensive, i s subject to delays, usually relates to disbursements rather than to actual expenditures, and may be partial and suffer from inaccuracies and inconsistencies. Impact on development outcomes xv. The strengths and weaknesses o f Afghanistan's PFM system as summarized above have an important impact on the country's development. First, the strong achievements in terms o f fiscal discipline, cash control, and aggregate transparency have contributed to macroeconomic stability, as well as to sustained external assistance. There i s need for more progress, however, in broadening the comprehensiveness o f the Core Budget, in addition to efforts to improve the monitoring o f other fiscal activities, such as State-Owned Enterprises, municipalities, and external assistance outside the Core Budget. Second, weak sector strategies, inadequate prioritization, and lack o f information on results have made it difficult to allocate resources across and within sectors appropriately. Third, weaknesses in procurement, insufficient financial data given to line managers, and initial delays in m a l n g payments outside Kabul have been a constraint on the efficiency o f public expenditures. These problems, along with difficulties in getting specific data from donors on expenditures of externally-implemented projects and a lack o f realism in publicly available budget numbers, have exacerbated the disconnect between public expectations and actual delivery o f services. Concludingsummary and wayforward xvi. Afghanistan's ratings against the PFM performance indicators generally portray a public sector where financial resources are, by and large, being used for their intended purposes as authorized by a budget which i s processed with transparency and has contributed to aggregate fiscal discipline. The expenditure and financial position of the resources under the authority o f the Government are reported reliably in an understandable format, although there i s some uncertaintywith respect to revenuereporting. Performance regarding the allocative efficiency o f spending across programs and the efficiency o f operations i s not as good, however. xvii. Given that performance most likely would have been rated "1" on all dimensions in2001 or 2002, this assessment highlightsthe significant achievements o f the last four years. Inmost dimensions, the ratings are now comparable to other low-income developing countries (for instance African countries). Areas o f relative weakness include tax collections (indicators 13-15), as well as legislative oversight (27- 28) due to the absence of a Parliament hitherto. Payroll control (18) and predictability o f budget-support (Dl) are areas o f relative strength. The Government is m a l n g further progress in improving the PFM system by implementing a number o f reforms. A new Public Finance and Expenditure Management (PFEM) Law anda new Procurement Lawhave been approvedby the Cabinet; tax amendments are under implementation; customs and tax administrative reforms are moving ahead; and the M o F i s moving ahead with further structural improvements. All these developments are positive. xviii. Using the ratings o f this report as a baseline, hture progress in improving the performance o f the PFM system can be monitored. However, the nature o f the indicators and ratings means that measurable progress in terms of changes in ratings i s more likely to occur over periods o f a year or longer than on a short-term basis. Moreover, since the ratings assess the system's performance with significant external support (both advisory and operational), questions arise about the sustainability o f current levels o f P F M performance. This may mean that further improvements will be manifested only in maintenance o f current ratings based on sustainable national capacity and with declining levels o f external support. xix.The ratings against the PFM performance indicators also provide an assessment for the donor community o f the potential capacity of Government systems (currently supported by external assistance) to implement the operations which are now occurring outside Government systems. 8 xx. Annex 1.1 provides initial suggestions for a comprehensive, high-level roadmap o f policy and operational reforms required to sustain and further strengthen PFM performance. Concrete actions are recommended for the current Afghan fiscal year (2005/06), and medium-term priorities are identifiedthat should be developed into action plans later. The roadmap also highlightsa few cross-cutting issues. In particular, strengthening the core capacity o f the MoF i s vital, as MoF i s the chief custodian o f the PFM system and to a large extent responsible for its performance. 9 A. Introduction 1.1 This Public Financial Management (PFM) Performance Report is organized into three sections. The country background, objectives, and scope o f this study and the overview o f the PFM framework o f the national government are provided inthis introduction. The second section - Overall Performance o f the P F M System - provides the analysis and ratings for 28 PFM indicators and three donor activity indicators. The final section outlines a roadmap to improve PFM performance. Annex 1.1 provides a detailed matrix of concrete actions that would form part o f this roadmap. BACKGROUND 1.2 Political Context. As mandated by the Bonn process set in motion in December 2001 and reconfirmed at the Berlin Conference in March 2004, the Government has carried out Presidential elections in October 2004 and Parliamentary elections in September 2005. A new Cabinet was formed in December 2004. The Government i s committed to implementing a new legal and institutional PFM framework, and a new Public Finance and Expenditure Management (PFEM) Law and a new Procurement Law have recentlybeen approved. 1.3 Economic Context. Afghanistan i s a poor country devastated by a quarter-century o f conflict. The GDP was $6.0 billion in 2004105 (excluding income from opium roughly estimated at $2.3 billion). With a population of around 24 million, social indicators and income per capita o f around $250 per year rank among the lowest in the world. Afghanistan's economy has performed very strongly in the past three years with non-drug GDP increasing by more than 50% (29% in 2002, 16% in 2003, and 8% in 2004), albeit starting from a very low base. This mainly reflects the recovery o f agnculture from a protracted drought in the late 1990s, revival o f economic activity after major conflict ended, and the initiation o f reconstruction. Recent growth performance has been supported by sound macroeconomic policies - a highly successful currency reform in late 2002, a prudent "no-overdraft" policy prohibiting central bank financing of the budget deficit, and conservative monetary policy. PFMPERFORMANCE MEASUREMENT 1.4 Objectives and Scope. This assessment i s designed to measure Public Financial Management (PFM) performance in Afghanistan as of June 2005 using a methodology developed by the World BanMIMFDEFA Working Group! The methodology gauges the soundness o f the PFM framework by rating 31indicators usingspecific criteria provided inthe guidelines. The scope o f the assessment covers fiscal and debt management, budget formulation, budget execution, internal controls, procurement, accounting and reporting, auditing, transparency, and external scrutiny. The methodology draws on standard expenditure tracking benchmarks where appropriate, while also takmg a wider perspective including, for example, fiscal risk and predictability o f funding. This standardized methodology i s designed to allow meaningful inter-temporal and international comparisons. The ratings for all indicators are on a 1 to 4 numerical scale, with 4 being the highest possible rating and 1 the 10west.~ Specific criteria for ratings for each indicator are provided inthe methodological document referredto above. 1.5 It should be noted that this assessment focuses on the central or national level o f government; it does not cover the municipal level o f government (with the exception of Indicator 8, see below) or Public FinancialManagement Performance MeasurementFramework, by Public ExpenditureWorking Group (World Bank, IMF andJoint DonorPublicExpenditureandFinancialAccountability Unit), availableat http:i'/www.pefa.orgiDocunients.htm. The methodologicalpaper referred to earlier, preparedby the World BankAMFPEFA Working Group, uses arating scale of A to D, with A beingthe best rating and Dthe worst rating. Inthe PFM performanceassessment for Afghanistanpresentedhere, 4 correspondsto A inthe World BanWIMFPEFA scale, 3 to B, 2 to C, and 1to D. 10 arrangements for State-Owned Enterprises (with the exception o f Indicator 9, see below). This assessment also does not cover donor funding implemented outside the government's financial management framework (the External Budget, see below). Inaddition, it should be noted that the Government's PFM arrangements currently employ extraordinary measures through the use o f international firms for operational and advisory work in Treasury, procurement, and external audit. The assessment includes these extraordinary arrangements and puts forward issues for consideration as the Government transitions away from such support. PFMFRAMEWORK (RULES, ROLES, INFORMATION) AND 1.6 Afghanistan's public spending i s highly dependent on donor support for financing and implementation. In2004/05, total public expenditures amounted to $3.4 billion, o f which $0.9 billion was implementedby the Government. Domestic revenues amounted to $269 million. 1.7 Since 2002 the rules, roles, and information on public financial management have evolved and have largely satisfied donors' fiduciary and operational requirements. Significant steps were undertaken by the Government with donor support to implement emergency arrangements for public financial management, which at the same time aimed to build capacity in key areas o f Treasury and budget implementation, procurement, and external auditing. These reforms are still underway and will expand to include capacity buildingfor financial management inline ministries and internal audit. 1.8 Roles. Afghanistan i s a unitary state and includes a Presidency and Supreme Court as Constitutional bodies, 26 Ministries, and various other agencies and commissions. The National Assembly, re-established by the 2004 Constitution, has the authority to ratify, modify, or abrogate laws and legislative decrees, thus placing the laws that it passes at the top o f this policy hierarchy. The National Assembly has the exclusive power to approve economic plans and state budgets, and to grant and obtain loans. The Wolesi Jirga or lower house, once convened, will grant spending authority through the annual budget submitted by the executive. Ifno budget i s approved, the new financial year operates under the budget of the prior year. The Government i s required to present a budget proposal and a preliminary accounting of the current year in the fourth quarter o f each year. The Government must present its final accounts for each year in the first half o f the following year for approval by the lower house. 1.9 Central government ministries and institutions are primary budgetary units with specific budgets set in the annual budget law. The provincial departments o f the central government ministries are secondary budgetary units, and receive allocations at the discretion o f the primary budget unit. 1.10 Primary responsibility for implementingthe Budget Law lies with the Ministry o f Finance (MoF). The Ministry i s organized under three Deputy Ministers: Revenue, Finance (comprising Treasury and Budget), and Administration (comprising the oversight o f the Provincial offices o f the Ministry, Internal Audit, and General Services).6 Operations inKabul are executed by the Treasury Department through its financial management information system (AFMIS) on the basis o f requests from the budgetary units. Fiscal operations in the provinces (both revenue and expenditure) are processed by the provincial offices o f the M o F (Mustoufiats). This arrangement provides a basic segregation o f duties between authorization of uses and recording and custody of funds. 1.11 Legislative Framework. As o f June 2005, there are four main pieces o f permanent legislation which govern Afghanistan's fiscal mechanisms. These include: the Budget Law (1983), the Accounting A recent decision has transferred the Accounting Department from a separate department under the Deputy Minister for Administrationto aparto fthe TreasuryPresidency. 11 Manual (circa 1964, date uncertain), the Control and Audit Regulation (issued under the Budget Law in 1985), and the Procurement Regulations (1987). The first o f these laws was replaced in late June 2005 by a new Public Finance and Expenditure Management (PFEM) Law, and the last of them by the new Procurement Law inSeptember 2005. 1.12 The I983 Budget Law provides appropriate checks and balances and assigns the key roles for fiscal management. It requires a pre-audit to be conducted by representatives o f M o F (the Financial Controllers, who currently report to the Accounting Department), who must pre-approve expenditures prior to distributingsalaries, procuring and purchasing supplies, and prior to signing any contracts. It also provides a requirement that the full cost o f donor projects be reflected in the Government's budget. It does not cover critical issues related to public monies, treasury operations, or the basis for accounting, among other dimensions. The law makes a separation between the "development" and "ordinary" budgets, rather than the more straightforward distinction between recurrent and capital budgets. The development and ordinary budget approach does not isolate recurrent costs from capital costs, which is a critical distinction that should be provided in the budget preparation and approval process. Neither does the L a w reflect the current budget or organizational structure o f Afghanistan; it divides spending into central and rural, development and social welfare budget, and a confidential budget. Frequent references are made to the Revolutionary Council (an institution from the Soviet occupation era) as the budget appropriationand amendment institution. N o provisions are made for budget financing. 1.13 The Accounting Manual i s considered to be the official accounting regulation, although the legal basis for it has yet to be clarified. It establishes the "legal requirements, organization, policies, procedures, and rules concerning the accounting system" of the Government o f Afghanistan. It can be said that the application o f the Accounting Manual was partially abandoned because the last full set o f financial statements drawn upby the Accounting Department inthe M o F date from 15 years ago. 1.14 The Control and Audit Regulation recognizes the General Control and Inspection Presidency o f the Council o f Ministers as the supreme audit authority for the whole o f government, under the auditor general, providing ex-post external audit. This authority audits the Final Account o f Revenues and Expenditures o f the MoF. The "Final Account" contains the appropriations o f the primary budgetary units, allocations issued duringthe fiscal year, revenues collected, and expenditures made by government entities. The Final Account i s prepared by the accounting, treasury, and revenue departments o f M o F and inthe past has beenthe primary source o f information for the preparation of the budget for the following fiscal year. It also provides for ministries to have their own internal audit departments which must coordinate with the General Control and InspectionPresidency. 1.15 Most public procurement in Afghanistan i s conducted through reliance on donor procurement procedures. The portion conducted under government rules i s governed by a series o f provisions: (i) a Law on procurement of 1987; (ii) a Purchase Regulation adopted as a Decree o f the Islamic Emirate o f Afghanistan in 2001, both o f which apply to the procurement of goods (and routine works and maintenance); and (iii) Rules and Regulations for the procurement of works adopted sometime after 1989 (since it replaces the rules adopted on that date). The Purchase Regulation effectively replaces the Law on Procurement. These regulations are reviewed indetail inPart I1o f this Volume. 1.16 Current Decrees and Regulations. In addition to the permanent laws described above which predate the current Constitution, the PFM legal framework also includes the annual budget decree which makes provisions on maximum levels o f new debt to be incurred, prohibits overdrafts, and authorizes M o F to issue instructions regarding finance and budget implementation. Furthermore, Presidential Decrees are used to provide specific PFM guidance and prohibitions. MoF has also, in the last three years, issued a number o f Budget and Treasury Circulars. Some o f the recent practices remain to be 12 coded (for example the decision for Treasury to prepare the Final Accounts or the decision to merge the Treasury and Accounting Departments; or the return o f the Financial Controllers to line ministries). 1.17 Conclusion. The current legal framework for PFM comprises multiple laws and regulations which are applied selectively and do not fix clear responsibility for PFMinthe line ministries. The partial application o f the Budget Law and regulations led to the imposition o f contravening provisions in the annual budget decree and undermined the demand for compliance with the emerging legal fi-amework supported by Presidential Decrees. 1.18 To address these issues, the Government has enacted a Public Finance and Expenditure Management (PFEM) Law in June 2005. The Law provides clear rules and roles for the exercise o f financial management, requires adequate reporting, provides for independent review, and holds public officials accountable. It covers the entire general (non-financial) government sector; as such, it does not apply to the SOEs which operate under a SOE (Tassady) Act. Responsibility for framing financial policy, regulating the use of public money, and applying the PFEM Law i s vested with MoF. The law has been gazetted and regulations to implement it are now being drafted. INFORMATION SYSTEMS 1.19 There are multiple information systems which record and process information on PFM in Afghanistan. Key systems include: AFMIS. The core financial management system i s the Afghanistan Financial Management Information System (AFMIS), which integrates budget control (appropriations and allotments), accounting, and treasury systems using FreeBalance software. The system operates under the control o f Treasury, records all budget uses, and i s the platform for reporting the financial position of the Government on a cash basis. AFMIS is not integrated with the budget database or the debt management database. Budget Database. The Budget Department of M o F keeps an unofficial budget appropriations and allotments database for the Core Budget. This i s a stand-alone system, and transfer o f information to AFMIS i s not automated. The official budget and allotment record i s considered to be in AFMIS. The current approved budget i s in the two systems and i s regularly reconciled; however, eliminating this redundant database would eliminate the need for such reconciliation. Donor Assistance Database. Although the implementation o f the External Budget i s not MoF's responsibility, the agreed plans and execution reported by the donors are recorded in the Donor Assistance Database inthe Budget Department. This i s a stand-alone system that i s focused on recording donor pledges, disbursements, and commitments. Annual donor commitments are contained inthe GLAD (Grant and Loan Database). Manual Accounting System. There i s a residual manual accounting system called for in the Accounting Manual that i s still partially maintained. The line ministries and the Accounting Department o f the M o F record allotments and expenditures in the system; it i s considered their base for chechng the balances reportedinAFMIS.' Debt System. The Debt Management Unit in M o F maintains a debt register which records terms o f all new obligations, disbursements received, and the debt service on each operation. This register i s a stand-alone applicationinExcel. 'MoF has decided to merge the Accounting Department with the Treasury Department, which will result in no longer keeping such manualaccountingrecords. This approachwill be supportedby thenew PFEMLaw andits regulations. 13 1.20 Systems are in place to cover the key minimum requirements for budget control and external reporting. Management control and more timely reporting depend on adding modules to the AFMIS and connecting both the line ministries and the provincial offices o f the MoF. For the Government o f Afghanistan to more effectively plan, manage, and track its resources, an overall systems architecture needs to be defined that portrays the long-term vision o f a fully integrated automated information system with the medium-term plan o f implementingspecific functional modules on a priority basis. Such a study was initiated in October 2005. B. OverallPerformanceofPFMSystem PFMOUT-TURNS: CREDIBILITYOFTHE BUDGET 1.21 The Government o f Afghanistan has developed a track record o f fiscal discipline over the last three years. However, its capacity to develop meaningful, policy-based budgets i s still weak, and as a result the budget remains an imperfect policy tool. Beyond the lack o f capacity in the Government (to develop policies, agree on trade-offs, and match resources with policies), this reflects the challenges o f aid coordination in a country where external assistance finances most public expenditures (see the indicatorson donor practices below). P F M out-turns are measured inIndicators 1-4. Indicator 1:Aggregate expenditureoutturn compared to original budget 1.22 The Core Budget i s divided into the Operating Budget and Development Budget. The Operating Budget includes expenditures funded by domestic revenues and external assistance (fast-disbursing assistance, reimbursements from ARTF and the Law and Order Trust Fundfor Afghanistan, LOTFA) and covers operating expenditures including the wage bill, most non-salary recurrent costs, and investments funded through the Operating Budget. The Development Budget includes investment operations (as well as recurrent costs) implemented by the Government. External Budget expenditures include technical assistance, most capital expenditures, and significant donor-financed recurrent expenditures. Because the development expenditures include both Core and External components, this has implications for the extent to which budget discipline i s applied to different expenditures. In essence, the core budget i s established around the need for aggregate fiscal discipline and allocative efficiency whereas the external budget i s not (see Indicator 12). 1.23 Public spending in Afghanistan i s very high, which i s the result o f donor-funded reconstruction activities. Operating expenditures inthe recurrent budget, excluding municipal expenditures, are slightly below 10% o f GDP, a rather modest level by international standards (however, substantial operating costs are funded by donors through the external budget - see below - e.g. in the security and health sectors). On the other hand, development expenditures reached 47% o f GDP in 2004/05. Revenues have been increasing over the past three years but are still below 5% o f GDP. The deficit i s covered by external assistance, as the Government has consistently refrained from overdraft financing. Most external assistance i s provided in the form o f grants. A small portion o f external assistance supporting development expenditures flows through the Government's accounts (inthe Core Budget), while the rest i s directly transferred from donors to non-Government implementing agencies (the External Budget). 1.24 The Government has laid out its medium-term fiscal strategy in the Securing Afghanistan's Future (SAF) report. Its objective, announced at the Berlin Conference in MarcWApril 2004, i s for domestic revenue to cover the wage bill by 2008/09 and all operating expenditures by 2013/14. Achieving this medium-term objective will require careful macroeconomic and expenditure management and substantial efforts to increase revenues. 14 Table 1.1: BudgetOut-turns 1 2002/03 I 2003/04 I 2004/05 Revenue expenditure) I I d a I -59.0 Expenditures(Total Core Budget) Budgeted (Af million) I 15,649 I 24,750 I 67,214 Actuals (Af million) 15,514 31,606 47,709 Difference (Afmillion) -135 6,856 -25,505 Difference ("hbudget) -0.9 27.7 -37.9 Total DevelopmentExpenditures(Core andExternal) Budgeted (US$ million) I 306 I 1,955 I 4,342 Actuals (US$ million) 503 I 1.419 I 2,806 Difference (US$ million) 197 -536 -1,536 Difference (% budget) 64.4 -27.4 -35.4 Source: MoF. IMF 1.25 Maintaining fiscal discipline has been a key Government priority. As documented in the World Bank publication Reforming Fiscal and Economic Management in Afghanistan, fiscal management today stands in stark contrast to the situation in 2001, when the public sector deficit was financed through currency printing and Central Bank borrowings in addition to covert direct external assistance to line ministries, all under a budget comprising mostly contingency funds. While the budget i s always in deficit given the low level o f revenues, M o F has over the past three years managed to keep the operating deficit below the planned level. The operating deficit i s defined as the difference between operating expenditures and domestic revenues, before external grants. Three important mechanisms have contributed to this positive outcome: (i) a very strong commitment to revenue mobilization; (ii)a mid- year review process and re-allocations; and (iii) strictly appliedno-overdraft policy. the 1.26 Rating 1: 2. The ability to implement the budgeted expenditure i s a key factor in supporting the Government's ability to deliver services as expressed inpolicy statements. Inthis assessment, the rating i s based only on the operating budget, as Core development expenditures was a concept introduced during 15 2004. Based on operating expenditures only, the rating i s "2" because operating expenditures have deviated from budget by a percentage higher than 10% but lower than 15%, inmore than one year Indicator 2: Composition of budget expenditureout-turn compared to original approved budget 1.27 The quality o f budget implementation, as measured by changes in expenditure composition, i s weak, A detailed analysis of deviations betweenbudgeted amounts and actual out-turns demonstrates that deviations by functional classification were very large in 2003/04, as shown in Table 1.2, but have been reduced in2004/05. Overall, this suggests that the Government has the ability to achieve fiscal discipline, but has a weaker capacity to achieve a satisfactory, policy-based allocation across sectors. The use o f large contingencies (much reduced inthe 2005/06 budget) also explains this variance. Table 1.2: ExpenditureCompositionVariation I I I I I a/ this differs from Table 1.1 since this i s based on the oDerating expenditures onlv. Source: MoF (annual budgets and annual financial statement), Staff calculation. 1.28 Rating 2: 2. Where the composition of expenditure regularly varies considerably from the original budget, that budget will not be a useful ex ante statement of intent. The rating criteria calls for a "2" where the variance between budgets and outturn by function does not exceed 10 percentage points above the variance o f total expenditures inmore than one year over the last three years. Indicator 3: Aggregate revenue out-turn compared to the original approvedbudget 1.29 Revenue collection has been somewhat lower than expected in 2004/05, but higher than anticipated in the previous two years (Table 1.1). This reflects the nature o f the projections, which were largely indicative targets given the weak capacity in the Government to develop more analytical projections (weak data, limited analytical capacity). The revenue outcomes also reflect the reality that actual collection depends on the degree o f compliance and remittances to the central Government, two factors which are still subject to considerable uncertainty. 1.30 Rating 3: 4. Strong revenue collection i s critical for Afghanistan to achieve its development objectives. Inaddition, accurate forecasting o f domestic revenue i s a critical factor in determining budget performance, because budgeted expenditure allocations are based on it. A comparison o f budgeted and actual revenue provides an overall indication o f the quality o f revenue forecasting. Although there i s significant variance betweenprojections and actuals, the rating i s "4" because actual revenues were below 97% o f budgeted revenue in only one year (83% in 2004/05), and this indicator i s not measured symmetrically. 16 Indicator 4: Stock and monitoring of expenditurepayment arrears 1.31 The strong centralized nature o f payment controls appears to have prevented any substantial build-up o f arrears. Even though it creates delays in payments (see below), there is no evidence o f a significant accumulation o f arrears. It i s known that there are no arrears on debt servicing for reconciled debt, and at the end o f 2003/04 the Treasury gathered information on payroll arrears and concluded that the level was very low.' This exercise was repeated recently for non-salary expenditures with a similar findingof a low rate o farrears. The recent sub-national administration analysis indicatedthe possibility o f payroll arrears in the education sector building up in some provinces (see Volume 111, Chapter 3), but these are not significant in the total expenditure picture. As pressure on domestic financing grows, there will be a need to improve the recording o f commitments and accounts payables, which will be addressed by improvements in financial management processes and systems. In addition, a number o f claims that pre-date this Government (domestic and external) will needto be reconciled. 1.32 Rating 4: 2. The rating on this indicator i s "2" because there may be a stock o f expenditure arrears but it i s likely to be below 10% o f total expenditure; the accumulation o f new arrears i s low and the net stock level declined inthe last year. However, there i s no track record yet o f an annual process to gather data on all arrears. KEY CROSS CUTTINGFEATURES: COMPREHENSIVENESSAND TRANSPARENCY Indicator 5: Classiflcation of the budget 1.33 Spending classification in the budget appropriations for the Operating Budget is formulated by major economic code for each budget entity. Until 2004/05, the Development Budget was structured around national programs, in many cases cutting across Ministerial mandates. As a consequence, it was not possible to present consolidated budget figures by economic classification or organizational unit for the total. This has been corrected for the 2005/06 Budget. MoF has introduced a modernized Chart o f Accounts in2005. This new system: (i) conforms with the structure o f the GFS classification system and (ii) combinedreportingontheDevelopmentandOperatingBudget. enables 1.34 In expenditure recording for the Core Budget, the classification is made for both operating expenditures and development expenditures at the level o f the organizational unit, source o f funding and specific contract, expenditure type (economic classification), and location (province). Bridging tables from the program classification to the functional (COFOG) classification are being used for 2005/06. The economic classification now follows the Government Finance Statistics (GFS) classification structure, produced by the International Monetary Fund (IMF). Difficulties arise in classifying Government functions, however, where activities o f a single organizational unit cover more than one function (e.g. Ministry o f Labor and Social Affairs, which encompasses the two functions o f economic affairs and social protection). Another significant departure (corrected in the 2005/06 budget) had been due to in- hndbenefits (food) givento employees inthe security sector andrecorded as goods insteado f wages and salaries. There remain issues with regard to revenue classification (existence o f redundant codes and issues with treatment o f financial assets and liabilities). 1.35 Rating 5: 2+. A robust classification system enables the trachng o f spending on the following dimensions: administrative unit, economic, functional, and program. Where standard international classification practices are applied, traclung poverty-reducing and other types o f spending can be facilitated. The rating i s "2+" because the existing operating budget classification is based on major Moreover,there is no incentiveinthe system to financeexpendituresthrough runningup arrearsbecausethe principal sourceof financingthe civilian recurrentbudgetis the ARTF, which makes reimbursementson the basisof settled invoices. 17 administrative and economic classifications, i.e. using GFS standards or a standard that can produce consistent documentation according to these standards. The new Chart o f Accounts (initiated for 2005/06) should result in a higher rating as it will allow consistent administrative, economic, and sub- functional classifications for the entire Core Budget. Indicator 6: Comprehensivenessof information included in budget documentation 1.36 Among other things, the annual budget documentation for 2005/06 included (i)clear definition a of the fiscal deficit (compatible with international standards); (ii) deficit financing and its composition (even though there are doubts as to whether the budget i s based on expected cash or commitments); (iii) estimates o f the current year's (2004/05) budget outturn; and (iv) summarized data for both revenue and expenditure according to the main heads o f the classification (see previous indicator). However, the budget documentation includes very little about macroeconomic assumptions, nothing about the debt stock, no information about financial assets or the previous year's (2003/04) budget out-turn, and little explanation o f major revenue policy changes and major changes inexpenditure program. 1.37 Another issue i s related to the Government accounting systems, including budget accounting, which should be organized and operated on a fund basis. A fund i s a fiscal entity which i s governed by specific regulation or restrictions. The major fund in Afghanistan i s the operating or general fund. This fund groups all the revenue, grants, and borrowings not restricted (earmarked) to individual project expenses. Expenditures funded from these sources would be appropriated under this fund. The other operations in the central government relate to administered funds which are held in trust while specific operations are carried out; these should be presented separately from the general fund since their resources are not fungible. The annual financial statements make this distinction and present a net result and the cash position o f the two separate entities: cash o f the Government which i s fungible and cash o f the administered funds which i s not fungible. The Budget Decree and the M o F Financial Report, on the other hand, mix general fund operations and administered fimds so the net borrowing requirement (or surplus) for the general fund is not presentedinthe budget documentation. 1.38 Rating 6: 2. Annual budget documentation (the annual budget and budget supporting documents) should allow a complete picture o f central Government fiscal forecasts, budget and out-turns. As well as revenues, expenditures, and financing, it should include debt level and composition, financial assets, and the fiscal impact o f contingent liabilities. It should also provide information comparable to the budget for the out-turns for the two preceding fiscal years, and forecasts o f the main budget aggregates for the two years following the budget. The rating i s "2" since the 2005/06 budget documentation includes some but not all the elements listed above. Indicator 7: Extent of unreported Government operations 1.39 The Government has demonstrated real commitment to present all public expenditure in the budget. The annual budgets comprise most sources and uses o f public funds o f the general Government sector, both those implemented by the Treasury through the Core Budget and those implemented by donors in the External Budget. There are no funds administered by the national Government for expenditure outside the budget. 1.40 Two issues arise, however. First, there are a few donor initiatives, mainly in the security sector, which are not in the budget. They provide no reports to the Government on these expenditures, and the budgetary impact i s not quantified. Second, reporting on the execution of the External Budget depends on reporting by donors, which i s somewhat incomplete and unverified. Related to this, the Government can only retrieve and report information on disbursements; there i s no information on actual expenditures. 18 1.41 Rating 7: 3. Budget reports (annual budget documentation, year-end financial statements, and other fiscal reports for the public) should cover all budgetary and extra-budgetary activities o f central government to enable a complete picture o f central government revenue, expenditures across all categories, and overall financing. The rating i s "3" because, while the level o f extra-budgetary activities of central government i s most likely insignificant, complete income/expenditure information on donors' grant-financed projects i s not included in the fiscal accounts. Moreover, for some operations in the External Budget executed by donors, the data made available are not considered highly reliable, are limitedto disbursement information, and are very aggregate (inparticular for the security sector). Indicator 8: Transparency of Inter-Governmental FiscalRelations 1.42 Afghanistan has a unitary state, with provinces and districts being de-concentrated units o f the central Government. The only exceptions are the municipalities, which have some fiscal autonomy and a potential for delivering urban services (such as water and ~anitation).~Revenues and expenditures o f municipalities are outside the national budget. Municipalities manage their own tax and fee base (although rates are agreed with the national Government) and must fund all o f their expenditures from this base. The Ministry of Interior (MoI) approves the budgets and staffing complements o f all municipalities. Oversight i s provided by M o I through the provincial governors; M o F approves the budget o f the municipality o f Kabul. However, there i s very little reporting from municipalities to M o F and no overall assessment o f their fiscal situation by MoF. There i s very little information on financial flows and fiscal positions o f the municipalities, but this should change with the implementation o f the new PFEM L a w which requires municipalities to comply with the financial reporting requirements established by MoF. Eventhe existence of transfers from the central Government (through MoI) to municipalities is unclear, and ifsuch transfers exist they would not appear to be very transparent. 1.43 Rating 8: 1. Information on the level o ftransfers to subnational governments i s important for the management o f both the central and subnational governments - and this should be communicated in a timely fashion. A clear allocation criteria, such as a formula, for the allocation i s needed to ensure transparency and predictability. Finally, sectoral information on expenditures at the subnational level i s also important to have a comprehensive view o f public spending and the quality o f its allocation across sectors. Keeping in mind that Afghanistan i s a unitary state and hence that this rating applies only to municipalities and does not have a wide scope, the rating i s "1" on the three dimensions o f the indicators: (i)transparency and objectivity in the horizontal allocation among subnational governments; (ii) timeliness o f reliable information to subnational governments on their allocation; and (iii) o f extent consolidation o f fiscal data for general Government according to sectoral categories. Indicator 9: Oversight of aggregatefiscal riskfrom otherpublic sector entities 1.44 The Government budget comprises all ministries and agencies o f the executive branch, the legislature (when formed), judiciary, decentralized operations o f the central government (provincial operations), and other agencies o f the central government (IARCSC, CAO). The budget does not include state-owned enterprises (although there may be some payroll o f SOEs mixed with line ministries' payrolls, and, in some cases, commercial-type activities are handled by government departments). The annual budgets from 2003/04 onward are exhaustive in that they include all sources and uses o f public funds - both those which are implemented by the Treasury through the Core Budget and those implemented by donors in the separate External Budget. There are, however, some donor-funded (and donor-executed) operations, particularly inthe security sector, outside the annual budget. See Evans et al. (2004) and Volume 111, Chapter 5. 19 1.45 SOEs.loRevenues and expenditures ofthe SOEs are consideredto be fiscal activities outside the national budget. However, there i s very limited information on these entities. The Government oversees 71 SOEs assigned to 15 line ministries. The significance o f the revenue, expenditure, and the financial position o f the SOEs i s not precisely known because o f their poor financial reporting. The operations and management o f SOEs are governed by the State Owned Enterprise Law (Tassady Law, 1991) which defines SOEs as legal entities with an independent balance operating with 100% capital o f the state." It does not cover companies partially owned by the state (e.g. Anana Airlines) which are governed by the Commercial Law (1955). For most SOEs, ownership rests with line ministries. 1.46 The SOE Department in M o F has an extensive role under law to analyze, monitor, and oversee financial operations o f SOEs. This has not taken place systematically due to weak financial control by the M o F and lack o f financial reporting by SOEs. There i s no consolidated information on the transfer o f public funds to and from SOEs partly due to lack o f information, but also because there are inconsistencies inthe way SOEs are treated across Government despite legislativeprovisions inthis area. Regarding financial reporting, the SOE Department o f M o F indicated that there are agreed fiscal plans with the majority o f SOEs and that most are submittingquarterly income statements. This information is manually recorded. Nevertheless, there are questions about the level and quality o f compliance. Fiscal transfers from and to SOEs are regulated by the SOE Law and are paid through transfers to the Treasury Single Account. Although SOEs are classified as independent legal entities, the SOE L a w has provisions that give line ministries and the M o F significant control over SOE finances. However, fiscal risks arise with SOEs because they implicate the state intheir financial dealings, as the sole or majority owner, and as a group they represent a significant financial operation. Given the poor reporting and the fact that the SOEs operate outside of the Government budget and internal control mechanism, the fiscal risk management inplace i s weak. 1.47 Rating 9: 1. Central government should monitor and manage fiscal risks arising from its own activities and fiscal risks that will impact nationally that arise from SOEs and municipalities. In good practice: (i) central government should require and receive quarterly financial statements and audited year-end statements from parastatals and SOEs, and monitor performance against financial targets and (ii) sub-national governments can generate fiscal liabilities for the central government, so their fiscal position should be monitored, at least on an annual basis. The rating i s a "1" because, while there i s oversight and monitoring by the national government o f parastatals and municipalities, the financial information i s too limited and unverifiedto serve to manage fiscal risks. Indicator 10: Public access to keyfiscal information 1.48 Fiscal. Government policy toward budget systems design has focused on accountability and transparency inboth policy makmg and information sharing. This has been one o f the great successes o f Afghanistan's budget process. To increase awareness among stakeholders, the Government has convened since the 2003/04 budget an annual Afghanistan Development Forum (ADF) to enable Government, the press, international donors, and NGOs to review the national budget. Transparency and accountability have also been enforced through the Consultative Groups (CG) for both the Operating and Development budget. 1.49 Fiscal information available to stakeholders and the public generally through the Government website includes all budget documents, plans and detail on budget revenue and expenditure for current, and estimates for future years. Detail on economic and monetary policy, assumptions, and forecasts are also available publicly on the IMF website in the form o f the periodic reports on the Staff Monitored lo This section draws on Volume 111, Chapter 4. There i s even less information on public enterprises with less than-100% state participation. 20 Program, but not directly in the budget document. Also, the budget document on the operating expenditures offers limited information on the role and objectives o f each ministry. 1S O InOctober 2004 the Government published a FinancialReportprepared by MoF to June 2004 on its website. This provides comprehensive information in an understandable format on budget execution and the financial position; the information i s appropriately summarized, and i s available in Dari and English. It provides accumulated revenue and expenditure figures for the Core and External Budgets (disbursements only) for 2001/02 to 2003/04 o f the transitional government and separately for 2004/05 to end o f the first quarter against budget. The report presents debt by creditor for debts prior to 2001/02 and by operation under the transitional government. It also presents complete information on the status o f donor pledges and performance against those pledges. However, this report lacks data by function and data suffers fiom the unreliability o f the figures on execution o f the External Budget, as recognized inthe report, and the lack o f coordinationwith Treasury to use final accounting data. 1.51 Since March 2005, monthly financial statements are also posted on the MoF's website (see also indicator 24). Annual financial accounts have not been made public. 1.52 Audits. Beginning in2004/05, annual audit reports on the results o fthe Auditor General's review o f the state budget and financial statements are to be distributed to Cabinet and divulged generally through press releases. Audit results on the ARTF are available to all donors and effectively to the public. Reports on work relating to compliance and special investigations are presented only to the Cabinet inthe semi-annual report o f the Auditor General (see indicator 26). 1.53 Procurement. Large-value contracts for goods are subject to competition (national and international) based on open tender following proper advertisement and subject to time limits. There i s public bid opening, recording, and measures in place to guide the process. For works as well, the regulations provide for open competition through advertisement in the newspaper. The regulation provides for the publication o f large contract awards. These awards are not made public by the central procurement unit, however, while decentralized procurement units have rarely the opportunity to procure large contracts (see also indicator 19). 1.54 Rating 10: 2. The general level o f transparency that exists regarding the fiscal plans, position, and performance o f the government depends on (i) annual budget documentation, (ii) within-year budget execution reports, (iii) year-end financial statements, (iv) external audit reports, (v) public procurement information on major contracts, and (vi) resources available to primary service units (such as schools). The rating is "2" because two o f these elements (budget documentation and within-year budget execution reports) are published without major delays and the format i s understandable. Improvement i s needed, however, notably with respect to external audit reports.12 BUDGET CYCLE a. Policy-Based Budgeting l3 Indicator 11:Orderlinessandparticipation in the annual budgetprocess 1.55 Legal Framework. The new Constitution adopted by the Grand Council (Loya Jirga) on January 3, 2004 provides the legal`authority for budget formulation, approval, and execution. Responsibilities are l2There is, however,the regularaudit reportonthe ARTF, which reportson financial control issues affectingthe statebudget. l3The discussion on planning and budget formulation i s drawn from Volume I,Chapter 6 and "Budget Formulation in Post Conflict Afghanistan(2004/05)", abackgroundpaper for the World BankPublic FinanceManagementReview. 21 clearly established between the Government (for formulation) and the bicameral legislature (for approval). Upon the establishment o f Parliament, the budget will be submitted through the upper house (Meshrano Jirga) with an advisory comment from the lower house (Wolesi Jirga), although final approval will rest with the lower house. A number o f articles in the Constitution (art. 137 and 139) would allow linkingbudget management responsibilities with sub-national administration in the years to come. The recently approved PFEM Law - and its regulations - will further specify the legal framework (importantly, the law also details the procedure for budgetary approval and amendment between the MoF, the Budget Committee, and the Parliament). However, as the Parliament has yet to be formally established, the Cabinet approves the nationalbudget through Presidential decree. 1.56 Institutional Framework. The Cabinet establishes the budget envelope, although in reality the M o F i s responsible on behalf o f Cabinet, based on domestic and external revenue forecasts, and with the assistance o f the IMF. M o F also sets the overall financial policy o f the Government, drafts the budget and appropriations law, and issues call circulars and budget guidance notes to the line ministries. Line ministries and agencies are responsible for formulating their expenditure plans for both the operating and development budgets, as appropriate. The line ministries and other first-level line agencies deliver their assigned public services mainly through their second- and third-level departments stationed at provincial and district administration, although these units are not responsible for their own budget formulation. In essence, the extent to which sub-national administration i s actively involved in budget formulation i s regulated by the central line ministriesand i s currently very low. 1.57 Participation. Donors are not directly involved in the formulation o f the operating budget, although it i s obvious that donor decisions on the development budget have implications for operating expenditures. Through the Consultative Groups (CGs), donors are involved in determining the formulation o f the development budget which has direct implications for each successive operating budget. Donors and NGOs are requested to provide information regarding the forthcoming year's expenditure plans, and integration with the various areas o f Government public investment. As noted, an annual Afghan Development Forum enables donors to communicate pledges and expected commitments by budget channel (direct support, Core Budget, or External Budget). Currently civil society has no formal role inbudget formulation or inthe development o fpublic policy. 1.58 Calendar. The annual budget cycle runs from March 21 -March20. Following the annual mid- term review, budget formulation takes place in the last quarter o f each year in a well-organized process for both the operating and the development budgets simultaneously, beginning with the issuance o f Budget Guidance documents. The operating budget has regularly been approved in a timely manner before the start o f the budget year, but there have been significant delays in the case o f the development budget, in large part related to difficulties in getting clarity on external financing, particularly for the External Budget. The 2004/05 development budget was approved four months into the budget year, whereas in 2005/06 an interim development budget was approved a few days after the start o f the budget year. 1.59 A Guidance Note provides an overview o f the complete formulation process for both budgets, although with a greater focus on the development budget. The note details activities related to: (i) formulating Ministerial Budget Worhng Groups (BWGs); (ii)setting policy results in terms of objectives, outcomes, and outputs; (iii) selecting sub-programs and clustering projects around coherent objectives; (iv) assessing donor pledges, commitments, and disbursements in a given budget year; (vi) setting existing expenditure requirements and new project proposals including the completion of Project Summary Sheets for years 1-3; (vii) ranking projects according to highest priority; (viii) detailing institutional arrangements for implementation; and (ix) mainstreamingcross-cutting issues on Gender, the Environment, Human Rights, HumanitarianAffairs, and Counter Narcotics. 22 1.60 Following Budget Committee endorsement o f individual expenditure plans, both the operating and development budgets are submitted for Cabinet (Council o f Ministers) approval on separate dates. The discussion focuses on whether the set o f proposals presents a coherent policy framework and provides for aggregate budget discipline. 1.61 The Presidential Decree formally approves the total budget with reference to the Council o f Ministers Approval. The Decree provides a breakdown o f the operating and development budget, by core and external category. It outlines sources o f domestic and external revenue, concessional loans, and appoints an ERC committee under the Budget Committee to oversee review and scrutiny o f programs. The decree also outlines the nationalpriority programs and project financing mechanisms. 1.62 Rating 11: 2. Effectiveparticipation inthe annual budget process impacts on the extent to which the budget reflects fiscal and sectoral policies. This requires an integrated top-down and bottom-up budgeting process, involving central agencies, spending agencies, and the political leadership in an orderly and timely manner, in accordance with a pre-determined budget formulation calendar. Clear guidance on the budget process should be provided inthe budget circular and budget formulation manual. Negotiations on allocations should be transparent and systematic. The indicator i s assessed along three dimensions. The rating i s "2" along the first dimension, "existence o f a fixed budget Calendar", because there i s a calendar, but delays occur and there i s little time for ministries and the Cabinet to make adjustments (this was in particular the case in 2004, as the mid-year review was postponedinto the tenth month o f the year). The rating i s also "2" along the second dimension, "guidance on the preparation o f budget submissions", as there i s a budget circular with ceilings for each ministry (for the operating budget), but these have not been approved by the Cabinet in advance. The rating i s "2" along the third dimension, "timely budget approval by the legislature", because the operating budget consistently has been approved by the President - who so far has had legislative powers - in a timely manner but there have been delays inthe approval o fthe development budget. The overall rating i s therefore "2". Indicator 12:Multiyearperspectiveinfiscal planning, expenditurepolicy, and budgeting 1.63 Financing Framework. Economic growth assumptions for the medium term have been presented in some budget doc~mentation.'~Nevertheless, projections o f domestic revenue are more targets than forecasts since they do not flow from proven forecasting models incorporating the economic assumptions and results o f specific policy actions. Donor contributions are projected for two outer years. 1.64 Operating Expenditures. There i s no projection o f operating expenses beyond the budget year. Overall, the operating budget i s not developed in any economic or fiscal multi-year framework (the only framework was developed for the Government's Securing Afghanistan 's Future, SAF, report). 1.65 Development Expenditures and Investment Program. Development expenditures, both in the Core and External budgets, are presented on a three-year basis. Linkingthe investment budget to policy i s challenging in Afghanistan because virtually all investments are donor funded. The National Development Framework (NDF) publishedby the Government in 2002, established three broad pillars o f concentration and within these 16national programs, i s meant to provide a strategic framework for donor- funded investment projects. The development budget arises from public investment proposals under each program. l4 Overview of the 2004/05 Afghanistan Operating Budget and Minister of Finance's address to Cabinet for the 2005106 OperatingBudget. The rating for this indicator does not take into account the Medium-Term Fiscal Framework developed in October 2005. 1.66 However, to date investment plans in most cases have involved little more than assembling a project shopping list. As such, the Development Budget tends to assemble donor financing around Government-determined investment priorities, but not as integral components o f Government programs. Projects financed by donors are largely aligned to national program areas such as education or health. However, in many cases high-priority Government projects do not receive financing ifthey fail to attract donor interest. Moreover, no formal mechanisms have been put in place to stop donors from programming projects outside o fthe budget framework. 1.67 To address the gap, in 2004 the Government prepared a seven-year investment program in the SAF report, which attempted to reconcile the availability o f resources with priorities for reconstruction and development infulfilling the objectives set by the strategy framework, through detailed sector studies for areas o f public investment under the budget. In addition, the PFEM Law requires the budget to be anchored ina multi-year framework. 1.68 Integration of Operating and Development Budgets. The operating and development budgets are prepared largely in parallel, even though the processes are coordinated by the Budget Department. Since the budget for 2004/05, there has been an attempt to assess the operating costs implications o f projects inthe development budget, but this has not been very successful so far given low capacity inline ministries. Because of the recurrent expenditure component in the Development Budget (both direct, such as the cost o f providing basic health services and grants to communities, and indirect, such as the cost o f operating and maintainingnew investments), better integration o f the Operating and Development Budgets can only occur after such spending is formally identified so that the operating costs embedded within the development budget can be progressively transferred to the core recurrent budget, assuming revenue availability. Another constraint to integration o f the budget i s that the Recurrent and Development Budgets are not presented under one view; the recurrent budget i s presented around the Government's organizational units whereas the Development Budget i s prepared based on programs which in many cases cut across organizational units. Finally, both the formulation and execution o f the External Budget are largely determined and implemented by the donors, outside o f constraints Government may suggest, and not necessarily in response to the specific investment proposals identified by the line ministries. 1.69 Rating 12: I+. Policy decisions have multi-year implications. Multi-year fiscal forecasts and estimates o f forward expenditures (including expenditures both o f a recurring nature and those involving multi-year investment commitments) are required to determine whether current and new policies are affordable within aggregate fiscal targets. At the same time, national and sectoral strategies are needed to guide the development o f forward estimates. The extent to which estimates are integratedinto the annual budget formulation process will complete the policy-budget link. Finally, decisions on investments and recurrent funding for a particular sector or ministry should be consistent and not be taken independently. This indicator i s assessedalong four dimensions. The rating i s "1" along the first dimension, "multi-year fiscal forecasts and functional allocations", because there are n o forward estimates in the budget (projections on the development budget mainly reflect multi-year commitments). The rating i s "2" along the second dimension, "scope and frequency o f debt sustainability analysis", as one such analysis was conducted recently (by the IMF). The rating i s also "2" along the third dimension, "existence of costed sector strategies" since such strategies exist for several sectors (often developed as part o f the S A F report) but they have not been reconciled with fiscal forecasts. The rating is "1" along the fourth dimension, "linkages between investment budgets and forward expenditure estimates", because no such linkage has been established yet (some progress has been made on road maintenance inthe revised 2005/06 budget). The overall rating i s therefore "I+". 24 b. Predictabilityand Controlin Budget Execution Indicator 13: Transparency of Taxpayer Obligationsand Liabilities 1.70 Clarity and comprehensiveness of tax liabilities. Concise and easily accessible tax laws are critical if tax policy i s to be legally enforceable. The legal framework in Afghanistan became progressively fragmented over the past decade by countless amendments and decrees by various administrations, and compliance rates diminished. Recent revenue reforms have tried to establish a clear and more modern legal framework that will enable the Government to gradually extend coverage in the provinces, enhance compliance rates, and also remove the plethora o f illicit charges and "taxes" that arose inthe administrativeandpolicyvacuum. 1.71 A new customs code and the consolidated Income Tax Law, which cover the majority of tax liabilities, were published on the official MoF website in March 2005, along with some new income tax measures. The new customs code provides the enabling legislation for implementing the revised tariff structure, introduced in 2004. Recent amendments also assert the supremacy o f revenue laws inrevenue- related matters, including the removal o f some tax concessions and exemptions which had been introducedthrough other statutes. 1.72 While the new legislation provides a more robust and transparent framework for revenue mobilization, it will take time for the Government to disseminate information about the new regime and to develop the requisite administrative capacity, procedures, and systems. For example, the enabling customs regulations and procedures, such as the compulsory use o f licensed customs brokers, are still being developed and rolled out to key provinces. Income tax provisions are also not applied evenly throughout the country as there i s currently no legal basis (or administrative capacity) for the tax administration- e.g. enforcement and collection procedures and powers, and penalty regimes have yet to be developed. The Afghan International Chamber o f Commerce (AICC) has reported instances of illicit taxes, double taxation, and areas where the law may be interpreted inconsistently or i s ambiguities. 1.73 Taxpayers' access to information on tax liabilities and administrative procedures. The new customs code and the consolidated Income Tax Law, which cover the majority o f tax liabilities, were published on the official M o F website in March 2005. M o F has also started to publish advice and interpretative rulings for taxpayers (see Article 96 o f the Income Tax Law and Article 17 o f the customs code). Inaddition, MoF has launched a number of initiatives aimed at taxpayer education, including holding consultations with the private sector and the publication o f various "tax guides" on the official website. However, coverage of the tax guides, public rulings, education campaigns, and administrative procedures i s still far from complete, partly because new taxes are still being implemented and the procedures are still being developed andor rolled-out to the major provinces. 1.74 Tax appeals mechanism. The tax appeals system i s inthe process o fbeing established as part o f the customs and tax administration reform programs, but it i s unlikely to be functioning until after the proposed administrative restructuring o f the Afghan Customs Department (ACD) and the General Presidency o f Revenue (GPR) within MoF. The new customs code sets out the procedures for the review o f customs decisions (Article 18) and arbitration (Article 19), although the administrative and procedural arrangements to support the law have yet to be fully developed. The Income Tax L a w does not include such provisions. Currently, if taxpayers wish to dispute the amount o f their income tax liability, their formal rights allow only that the matter be taken to court, which can be costly and time consuming. Consistent with a modern tax administration, the authorities are proposing to amend the Income Tax law to allow persons who wish to dispute their tax liability to give notice inwriting to M o F explaining their grounds for dispute. M o F will be required to review the matter. Should a dispute not be resolved to the 25 person's satisfaction, the dispute may then be taken to a court for resolution. The amendments to the Income Tax Law have subsequently been approved. 1.75 Rating 13: I+.This indicator i s composed o f three components. On the first dimension, "clarity and comprehensiveness o f tax liabilities", the rating i s "2" since the recent amendments to the income tax law and the adoption o f a new customs code have made the legislative framework relatively comprehensive and clear for some major taxes. However, not all aspects o f the legislative framework are being implemented in practice, and the supporting administrative procedures, human capacity, and physical infrastructure need to be developed. The fairness o f the tax system i s also questioned by many taxpayers due to the reported existence o f many illicit taxes, the lack o f comprehensive coverage, and discretionary powers o f the government entities involved. On the second dimension, "information on tax liabilities and administrative procedures", the rating i s "2" since taxpayers have access to some information on tax liabilities and administrative procedures, but the usefulness o f the information i s limited mainly due to the coverage o f selected taxes only and limited geographical coverage, and the ongoing development o f the administrative procedures and tax rulings. On the third dimension, "existence and functioning o f a tax appeals mechanism", the rating i s "1" as there i s n o such system. The overall rating i s therefore "1+." Indicator 14: Effectiveness of measuresfor taxpayer registration and tax assessment 1.76 Controls in the taxpayer registration system. Inorder to make tax administration more efficient and to ensure proper crediting o f taxes paid to taxpayer accounts, M o F introduced a unique taxpayer identification number (TIN) program in September 2003. All individual persons, companies, and organizations which are, according to the Income Tax Law and Customs Law, required to pay taxes or customs duties, including social, non-profit, and welfare organizations that are withholding taxes from the salaries o f employees or from rentalpayments are requiredto obtain a TIN. 1.77 As of March 22, 2005 19,341 TINShad been issued in Kabul with a further 5,377 inprovinces (total 24,7 18), resulting inpartial coverage. The Nationwide roll-out o f the TINSto provinces commenced in April 2004 with the inclusion of the four largest provinces: Herat, Kandahar, Mazar-e-Sharif, and Jalalabad (plus Kunduz). 1.78 Incoordination with the Afghan Investment Support Agency (AISA), all new investment license applicants receive a TIN as part of AISA's concept o f a one-stop shop. TIN registration will also be required as part o f the Money Changers and Customs Brokers licensing processes. These agreements highlight the cooperation fostered with AISA (investors), D a Afghanistan Bank (Money Changers), and the Customs Department (Brokers). 1.79 Effectiveness of penaltiesfor non-compliance with registration and tax declaration. Currently, M o F has no powers to enforce compliance with respect to the Income Tax Law. Inparticular, revenue officers do not have power to access taxpayers' premises to collect information regarding income tax matters or to collect unpaidtaxes. These powers may be exercised by the Ministryo f Interior but there are no formal arrangements with the tax authority. The tax system therefore relies on taxpayers voluntarily meeting their tax obligations, and as a result tax compliance i s very low. 1.80 Enforcement powers were introduced in the new customs code (Chapter 14), including a provision for the establishment o f a Customs Police Administration and new provisions for penalties. The Government i s talung a gradual, risk management approach in establishing mobile customs teams to restrict smugglingand tax evasion. The mobile customs teams are being trained and deployed first to the major customs offices (e.g. Herat, Mazar-e-Sharif, Jalalabad, Kandahar, etc). It will take some time before the customs police and processing and control procedures are established. 26 1.81 For both tax and customs administrations, enforcement capacity needs to be developed in conjunction with the implementation o f improved processing and control procedures to respond to the current situation o f widespread non-compliance. 1.82 Planning and monitoring of tax audit programs. The new revenue systems being introduced will increasingly rely on self-assessment and the use o f risk targeted auditing o f taxpayers as a key activity to improve compliance and deter tax evasion. However, the Government is at a very early stage o f developing the capacity and processes that will enable the collection and analysis o f information on non-compliance and other risks that are necessary for focusing tax audit activities and resources toward specific sectors, and types o f taxpayers with the highest risk o f revenue leakage. It i s hoped that the establishment o f a dedicated Large Taxpayer Office (LTO), which began in 2004, will help to pilot many o f the reforms associated with monitoring, auditing, and enforcing collection for taxpayers. However, in the short-term capacity constraints and the absence o f a well-functioning legal system will severely limit the ability o f either the ACD or the GPR to investigate and successfully prosecute major evasion and fraud cases on a regular basis to ensure that taxpayers comply with their obligations. 1.83 Rating 14: I+. This indicator has three dimensions. On the first dimension, controls in the taxpayer registration system", the rating i s "2" because, while taxpayers are registered with a unique TIN in a database system, the coverage is currently partial (in addition, the linkages to other registratiodlicensing functions - AISA, DAB, Brokers - are being strengthened and may be supplemented by occasional surveys of potential taxpayers). On the second dimension, on penalties, the rating i s "1" since penalties for non-compliance are generally ineffective and rarely imposed due to the lack o f enforcement powers and the need to develop capacity and processes for both the tax and customs administrations. On the third dimension, audit programs, the rating i s "1" as tax audits and fraud investigations are undertaken on an ad-hoc basis ifat all. The overall rating i s therefore "1+". Indicator 15:Effectiveness in collection of taxpayments 1.84 InAfghanistan, the tax gap is extremely large -broadly, this is the difference between the current tax level, around 4.5% o f GDP in 2004/05, and the potential, which given the existing laws and statutes, i s assumed to be closer to 11% o f GDP.'' As there are few effective enforcement mechanisms, compliance i s largely voluntary, and the compliance rates are assumed to be extremely low. This results inthe accumulation of significant tax arrears, althoughbecause there is no comprehensive system o f tax assessment precise figures are unknown. The main emphasis o f the tax and customs administration reforms i s to develop the capacity and processes to collect current tax liabilities and there i s little effort made to collect tax arrears from previousyears. 1.85 A major fiscal reform was the establishment o f the Treasury Single Account (TSA) in 2004. The TSA helps to integrate Government accounts into an account or set o f linked accounts through which the Government collects revenues and transacts all its payments. The apex o f the system i s a single account held by the M o F (Treasury) at the central bank. The TSA i s designed to ensure that all Government revenues, wherever they are collected, are transferred to the Treasury. Once funds have been deposited in provincial revenue accounts they are automatically swept to the Treasury, at least on a monthly basis. Fundscollected inthe major urban centers are swept into the TSA at least on a weekly basis. 1.86 Reporting on tax collections i s done monthly by the revenue department and Treasury in MoF. The revenue authority and the Treasury generally report on revenue collections, and a system has recently See Volume 111, Chapter 1 27 been introduced to reconcile the cash balances with collectionreports on a monthly basis. However, there i s no system inplace to reconcile tax assessments, collections, arrears, and transfers to the Treasury. 1.87 Rating 15: I+.The effectiveness o f tax payments i s a critical factor in a country that seeks to increase its revenues from a very low base. Key factors toward this goal consist o f avoiding the accumulation o f tax arrears, promptly transferring revenue payments to the Treasury, and ensuring adequate reconciliation between payments, tax assessments, arrears, and bank account statements. The effectiveness of tax payments i s very low, with very little reconciliations and a low level o f collection. However, considerable progress has been made over the last few months in consolidating the TSA arrangements and ensuring periodic transfers o f hnds to the Treasury. The rating is therefore "1+". Indicator 16: Predictability in the availability of funds for commitment of expenditures 1.88 The existing PFM laws and recent legislation through the annual budget decrees and budget circulars are clear that all budget uses are subject to annual appropriations and quarterly allotments approved by MoF. The objective o f these mechanisms i s to support budget execution within the cash resources o f the Government and to provide a predictable flow o f cash to executing units. The PFEM Law and the annual budget decree establish the budget's role as the single authority for expenditure, through appropriations and ceilings to public employment which are provided with the appropriations. The annual budget decree also provides the Treasury with full control over public banlung and cash and calls for all receipts to be turned over to the Treasury. The Treasury and provincial Mustoufiats are the sole authority to control the flow o f Core Budget expenditure payments fiom official Government accounts. Hence the instruments for expenditure control are in place. The execution o f the External Budget, the part o f the budget that i s not implemented by the Government, i s more difficult to assess because reporting on these activities i s limited to disbursements. 1.89 Appropriations and Revisions. For the operating budget, appropriations expressed inthe budget law are made to primary budgetary units (the ministries). Many adjustments are made during the year including the assignment of contingency funds to budgetary units. Contingency funds are available inthe appropriations to deal with unforeseen demands, but these must be re-appropriated to specific budgetary units and provided allotments before use. Annex 1.1 o f the Budget Decree specifies the uses permitted and who has authority over each fund. These funds are usually allocated in numerous tiny amounts, causing fiequent - though small - changes to appropriations. In addition, since mid-2003/04 the Government has prepared a major mid-year review which seeks to reallocate budgets when appropriate. Otherwise there are no revisions to the appropriations. 1.90 The External Budget i s contained inthe Budget Decree but does not constitute spending authority since these funds are under the control o f the donors; spending and reporting are governed by the agreements between the respective donors and the Government. 1.91 Allotments. Allotments assign the spending authority over the four quarters o f the fiscal year. These quarterly allotments are inturn assigned by each ministryto locations (which include the provincial offices o f the line ministries), all o f which are recorded in AFMIS. The last step in the expenditure process i s the availing o f funds under the allotments so payments may be made. Allotments, however, do not reflect the quarterly pattern o f expenditures as they are usually set at 25% o f the annual appropriation regardless o f expenditure patterns. Allotments can therefore constrain expenditures where the expenditure pattern does not follow the 25% per quarter distribution allowed by allotments. Second, allotments intotal do not take into account cash availability, so there i s uncertainty as to whether they will be funded. 28 1.92 For provincial operations, evidence from the Mustoufiats indicates that there are two sets o f allotments approved at their level: the allotments advised from Treasury and in AFMIS and the allotments duly approved and supported by B-20 advices but which may not all be in AFMIS and which have been advised to the provincial line ministries' offices (see Volume 111, Chapter 3). Mustoufiats are empowered by the B-20s to spend, and these are the amounts expected by the provincial spending units. However, funding from the Treasury is inline with the AFMIS recorded allotments which are necessarily o f a lower amount in total. The causes for the discrepancies could not be precisely determined. These differences implythat allotments advised to the provincial offices cannot all be funded, and that the Mustoufiat must apply his own criteria in rationing the available funding. Given the different cycles o f allotment approvals (seven advices per year plus those due to adjustments in appropriations) and cash transfers (monthly), and given the disconnect between allotments advices to provincial spending units and cash transfers at the Mustoufiat level, the provincial departments cannot ascertain the true fbnding status o f their allotments. While cash constraints have not been significant for the Government as a whole, there are many allotted expenditures which are delayed or never funded because o f perceived cash constraints. The 2005/06 Budget Decree has established AFMIS as the sole source o f allotment information to address this issue. 1.93 Development expenditures in the Core Budget, on the other hand, are allotted annually and managed through the Special Disbursement Unit (SDU) on the basis o f availability o f cash inthe Special Accounts concerned. 1.94 Cash Management. Treasury uses the information available in AFMIS to prepare daily calculations and consolidation o f all the Government's cash balances supporting the operatingbudget and matches this against the requests for payments and transfers (see next indicator). Since the end o f 2004, forecasts o f all cash inflows and outflows under the operating budget are also prepared for the` fiscal year and updated monthly. Revenue information i s input from the actual to date and the difference required to meet the target agreed with the IMF. Donor contributions to the recurrent budget are identified for the balance o f expenditures based on existing commitments. Expenditure forecasts are only based on approved allotments; these, however, do not arise from procurement plans which may result inthe current level o f idle balances inthe provincial accounts.16 Difficulties are also faced inestablishing the pattern o f expenditures given that the prior year's recording by period was affected by delays in recording on the one hand and the use o f advances to suppliers on the other hand. The poor communications infrastructure with the provinces results in both delays in recording o f expenditure and limited effectiveness o f the single treasury account since it cannot support expenditures ordered by the provincial offices. 1.95 Rating 16: I+. Predictability in the availability o f funds is critical for line management to operate effectively. The rating i s "1+" because some progress has been made with regard to cash flow forecasts (now prepared by the Treasury, but at an aggregate level and with limited analysis on actual cash inflows and outflows), but more progress i s requiredwith regard to the allotment process and within- year adjustments. The current allotment procedure in Afghanistan does not support effective budget execution because o f the difference between the executing unit's records and the controlling unit's records o f authorized amounts. Also, rationing o f cash in line with transparent rules has yet to be achieved. Similarly, the use o f a year-end decree to regularize expenditures over appropriations indicates that there are uncontrolledwithin-year deviations from allotments. Indicator 17: Recording and management of cash balances, debt, and guarantees 1.96 Cash management. Cash management i s the responsibility o f the Treasury, although much o f its execution depends on the provincial offices o f M o F (Mustoufiats). The main tools for cash management Idlebalances intheprovinces stood at two months of expendituresin January 2005. 29 are the Treasury Single Account (TSA) in the Central Bank o f Afghanistan (DAB) and the integrated financial management system (AFMIS) under Treasury. These operate under a policy of no overdrafts (i.e. no financing by the Central Bank) and the use o f external budget support and ARTF to finance the shortfall between planned operating expenditure and domestic revenue. 1.97 According to the Constitution, all public revenue must be transferred to the TSA, including cash collected by the Treasury and by other agencies, and deposits made to DAB. Expenditures for the operating budget all originate in the TSA, either as payments to suppliers or as transfers to provincial Mustoufiats, which in turn operate single treasury accounts for expenditures and a separate account for revenues. The provincial revenue accounts are swept weekly or monthly (depending on the revenue generation capacity o f the province) and transferred to the central TSA. Cash related to donor earmarked funds (investment operations financed by donors) are deposited inSpecial Accounts outside the TSA (278 o f these accounts are operated by the SDU inTreasury). 1.98 Regulations for cash management are set by Treasury and communicated through Treasury circulars. Within the operating budget, the flow of funds for expenditures follows two routes: 0 For headquarters line ministries, expenditures are made from the TSA inresponse to payment requests from line ministries. There i s no specific authorization or cash allotment because budget allotments are expected to be funded. Shortfalls in cash availability result in postponing payment but not canceling expenditures; the decisions on what i s postponed are made on an ad hoc basis by the Director o f Treasury (paying salaries invariably i s assigned high priority). There is no data available on the amount or frequency of such delays (see above on arrears). Advances to line ministries for purchases from suppliers are common; these are charged to the related budget sub-allotment and followed up by Treasury for timely settlement. Payments made by Treasury to suppliers are made by direct deposit to supplier bank accounts, i.e. a check i s generated in AFMIS which serves as payment instructions to DAB for effecting a deposit inthe commercialbank account o f the supplier. As o fFebruary 2005, approximately $11million o f advances to line ministries were outstanding (1% o f Core Budget expenditures). 0 For provincial operations, expenditures are paid through provincial Mustoufiats. These offices are provided an advance at the beginning o f the year equal to one or two months o f expenditures based on allotments;" then each month during the year funds are requested to match the expenditures forecast on the basis o fthe approved allotments. Funds are transferred inline withthe request and cash availability, but these transfers do not identify specific uses. They are assigned to payment requests by the Mustoufiats in line with their priorities and within the rules onbudgetary uses. 1.99 Debt and Guarantees. Authorization and negotiation o f debt i s the responsibility o f MoF, with new debt specifically authorized through provisions in the annual budget law. The PFEM L a w specifies that the annual budget documentation should include how the annual appropriations will be financed and designate a maximum amount o f borrowing and guarantees that can be entered into. The Debt Management Unit (DMU) inthe Treasury i s responsible for maintaining a debt register and for budgeting debt service. The Unit is also the repository for the debt legal agreements. The Financial Report published by MoF, available on the Government's website, contains information on credit terms and amount o f debt by creditor for pre-2003/04 debt, and by specific loan for post-2003/04 borrowing. Given the lack o f records, there remain however a few unconfirmed claims. Similarly, while the Government is "There areuncertaintiesaboutthecalculation ofthisinitialamountandtheextenttowhich ithasconstrainedexpenditures in provinces. See Volume 111, Chapter 3. 30 not granting any guarantees, the poor state o f financial control over SOEs does not permit conclusions to be drawn on the existence or amounts o f these. 1.100 The debt register is an Excel spreadsheet application used to record loan terms, movement on principal and debt service, and the runningbalance payable. The DMU established the debt position in 2003/04 from the records in Treasury and from a circular and other contacts with donors. The DMU participates in the negotiation o f all new lending and receives a copy o f all loan agreements which comprise the basis for establishing a new entry in the debt register. The register i s updated for new disbursements from the advices received from creditors and from consulting the websites o f the creditors where all disbursements on specific loans can be viewed. This i s complemented by confirmation with DAB on receipt o f the disbursement. Recording in the accounting o f Treasury, however, is not initiated by the DMU and may or may not coincide with the movement reported in the DMU.Reconciliation is complicated because Treasury accounting i s on a cash basis. Debt service payments are initiated in the DMU;these are preparedmanually outside the register and arenot recordedinthe register untilDABand the creditor confirms payment. Again, these are recorded separately in Treasury accounting and not integrated with the debt register. 1.101 There i s no automation under the debt register for budgeting o f debt service for existing loans, forecasting for new loans and disbursements, analysis under different scenarios, or query reporting and costing. Similarly, the system does not offer the requisite data integnty or audit trails to support integrationwith AFMIS or reliance on its reports without manual verification. 1.102 There i s no formal debt strategy, but developing such a strategy i s a structural benchmark in the IMF's SMP. MoF has never conducted any debt sustainability analysis itself, but such analysis has been conducted by the IMF (which suggests that the level o f debt will remain sustainable if there i s strong economic growth, complete forgiveness o f all old debt and claims, robust increases in revenues, and external assistance overwhelmingly inthe form o f grants). 1.103 Rating 17: 2+. Good cash balance, debt, and guarantee management i s required to make most effective use o f public financial resources, avoid unnecessary costs, and guarantee predictability and sustainability. This indicator i s rated along three dimensions. The rating i s "1" on the first dimension, "quality o f debt data and reporting", since the unreconciled amount o f debt remains substantial (even though spreadsheet records are now maintained on reconciled debt, and reports on debt stock and service are prepared periodically). The rating i s "2" on the second dimension, "extent o f consolidation o f the Government's cash balances", because, although there i s daily calculation and consolidation o f government cashbalances, a number o f provincial accounts are reconciled and consolidated only monthly and development funds for the most part remain outside the cash planning and consolidation arrangements since they are not fungible. The rating i s "4" on the fourth dimension, "systems o f contractingloans and issuance o f guarantees", since the system for loans i s transparent and effective (with authority given to the M o F and policies set inthe Budget Decree). This rating assumes that n o guarantee has been issued recently inthe Government. Overall, the rating i s therefore "2+". Indicator 18:Effectiveness of Payroll Controls 1.104 Wages are the most significant cost element inthe operating budget, representing more than 50% o f the actual operating budget. Controls over payroll are based first on the staffing establishments (Tushkeels) which are proposed by the Ministries and approved by the Office o f Administrative Affairs ( O M ) during the annual budget process. The tushkeel provides the organizational chart for the ministry, detailing staffing positions and levels. In the absence o f a nominal roll, actual headcount i s estimated through a database kept by the ARTF Monitoring Agent on the basis o f summary reports of monthly payrolls in the provinces and detail provided by the Kabul ministries. The other component needed by 31 ministries and departments for processing payrolls i s the takhsis (the salary allotment for the budgetary units) 1.105 StafJig, personnel records, and pay scales. The Independent Administrative Reform and Civil Service Commission i s responsible for oversight o f appointments and promotions (for positions at grade 2 and above), civil service management, and administrative reform. Historically, the line ministries held personnel files for all staff, but these are no longer maintained, and nor do the ministries have information on the position and grade for all hiring.Hence the payroll arises from the staffing and attendance records maintainedby the individual budgetaryunits. 1.106 Every public employee has a personnel grade. Civil servant wage rates and pay policy are established centrally for all public employees in Afghanistan. Two base pay scales - one for contract staff and another for permanent staff - apply equally everywhere inAfghanistan. There are separate pay scales, however, for teachers, the army, and police. In addition, several new programs allow higher pay rates, based on a clear vision o f core responsibilities, structural reorganization, and merit-based appointments. 1.107 Payroll processing (see also Volume 111, Chapter 2). Payrolls are usually prepared manually by line departments. They may also be prepared from stand-alone automated payroll systems, for instance in the M o F (MS-Access database). Central authorities do not impose any standard for this processing or establish any minimumrequirement to link to personnel records. The request for funds to pay payrolls follows the procedure used for any payment, with the payroll sheet used as support. After the requisite authorizations in the budgetary unit, including MoF's controllers within the budgetary unit or in the Mustoufiat, the requests are reviewed by M o F inKabul or by the Mustoufiats." This review i s limited to matching the payroll sheet to the request for funds, viewing the allotment ceiling for salary and for positions (Tashkeel), but does not extend to underlying records. Except for those covered by the Individualized Salary PrograminKabul (see below), payment o fpayroll depends on a bondedtrustee who receives a check for the payroll amount for a budgetary unit, withdraws cash from the bank, and makes the settlement with each staff member who evidences receipt by signing the payroll sheet. 1.108 M o F i s leading an initiative to verify the national payroll. The program aims to cover all civil servants inKabul ministries and agencies (200,000) and comprises three stages: issuance individual salary payments monitored by Treasury staff, salary payment from the list inDAB (requiring an ID), and deposit o f salary to the bank accounts o f Government employees. To date, 13 ministries have been processed through the first stage, and individual salary payments are being processed for 26,000 employees in Kabul, but the rate o f transition to more advanced stages i s limited by the low capacity for issuing identification cards. Individual salary payment implies witnessing the payroll distribution by a team comprising the budget unitrepresentative, a representative o f the DAB,bank staff, and a representative o f the MoF. Significant savings will accrue to the Government from the full implementation o f the program, as duplicate and unauthorized payments will be eliminated. 1.109 Monitoring. The payroll cost for the recurrent budget (except for Ministry o f Defense, the security forces, and the uniformed staff o f the Ministry o f the Interior) i s subject to review by the ARTF Monitoring Agent. This review relates the payroll sheet entry back to the accounting entries and to personnel and attendance records. It also includes witnessing individuals who are in the payroll records. The record o f compliance i s very high (95%), but it i s not determined if the individual receives his full payroll or ifhis attendance records represent time at work. Inany case, the sample tested out o f the three In 2002103 and 2003104, this control was centralized in Kabul, which created delays in payroll processing, as provincial representatives had to travel to Kabul every month, without improving the controls, as the central Treasury had no capacity to review the paperwork indetail. 32 million annual payroll transactions i s too small to provide full assurance o f the non-existence of "ghosts" on the payroll, or that actual work inbeingdone by those who are paid. 1.110 Effectiveness of controls. There is evidence that controls on establishment are effective, as the observed headcount inthe Monitoring Agent database has remained flat over the last three years (with the exception o f teachers). Although controls are manual and decentralized, this i s compensated for by the centralized controls on positions, paper documentation o fpayroll expenditures, and the concurrent review by the ARTF Monitoring Agent. The accuracy o f the provincial data on headcount, however, is affected by difficulties inreporting head count where payroll delays or back pay occur, so the payroll for a month may not coincide with the head count. Significant risks inpayroll remain, related to: (i) weak oversight o f payroll support in areas not covered by the Monitoring Agent (most notably in the security sector); (ii) decentralized personnel records; and (iii) reviews by internal audit and central authorities interms limited o fmatching payroll to personnel records and evidencing o f final settlement with the employees. 1.111 Rating 18: 2. Controls exist over payroll expenses in aggregate through the tashkeel and the allotments (takhsis). Inaddition, audits have been undertaken, including by the ARTF Monitoring Agent, showing good results. The link between the payroll and the nominal roll i s a key control which i s decentralized to the operating unitsand i s not closely supervised, but as explained above, payroll integnty i s not undermined due to the existence o f other controls. Hence although manual personnel records may be incomplete, and there may be delays inprocessing o fpayroll changes, the rating i s "2". Indicator 19:Proc~rernent'~ 1.112 Currently, most procurement o f any significant scale i s conducted through reliance on donor procurement procedures. Very little procurement i s conducted by line Ministries themselves; where it is, the tendency i s to recruit outside consultants, generally NGOs, to carry out the work on their behalf. Only a very small proportion o f total procurement (by value) appears to be conducted on the basis o f a series o f national provisions regulatingprocurement: (1) a Law on procurement o f 1987; (ii) a Purchase Regulation adopted as a Decree o f the Islamic Emirate o f Afghanistan adopted in 2001, both o f which apply to the procurement o f goods (and routine works and maintenance); and (iii) Rules and Regulations for the procurement o f works adopted sometime after 1989 (since this replaces the rules adopted on that date). The Purchase Regulation effectively replaces the Law on Procurement. 1.113 Goods. The national provisions relating to goods are clearly inneed o f significant amendment to provide further detail and clarification. Nevertheless, the direction o f these provisions i s acceptable, if a little confused, suggesting that the basic concepts o f a good procurement system are understood. Large- value contracts are subject to competition (national and international) based on open tender following proper advertisement and subject to time limits. There i s public bid opening, recording, and measures in place to guide the process. Contracts o f smaller value are subject to direct purchasing methods under which one to three officials will purchase the required goods depending on their value. Smaller value contracts for international procurement are to be procured by way o f a Request for Quotation (WQ) procedure. The main problems with the provisions are lack o f detail and clarity in the qualification (as opposed to eligibility) requirements, setting o f specifications, drafting o f bidding documents, and award criteria, particularly with the apparent predilection for negotiation and the tendency to re-open biddingto attract better prices. 1.114 Works. In the case o f works, the Government i s currently operating on the basis o f regulations adopted in 1989. The main difficulty with these regulations i s that they resemble much more a set o f bid and contract preparation guidelines than they do a regulation to ensure competitive bidding. Hence they l9 This section draws on Part I1of this Volume. The rating does not take into account the approval of the new Procurement Law. 33 require significant updating. The regulations provide for open competition through advertisement in the newspaper. The regulations are inadequate in terms o f setting out qualification criteria, requirements as to specifications, award criteria, processes, and minimumtime limits. 1.115 Controls. There are, in effect, no controls over the public procurement system. There i s no formal complaints or appeals mechanism inthe existingrules, although there appears to exist a commonly understood system o f sequential complaints to the contracting authority, the Government, and the Attorney General. There are, however, no documented cases o f complaints, successful or otherwise. Annual audits of procurement are possible intheory but few have been conducted. Auditing staff lack the capacity to conduct appropriate procurement audits. 1.116 Rating 19: 312. Public procurement i s a major component o f the P F M system which directly impacts on the efficiency and economy o f expenditures and also contributes to the budget formulation and expenditure management process. This indicator has three dimensions. The rating along the first dimension, "use o f open competition for award o f contracts that exceed the nationally established monetary threshold for small purchases", i s "1" for decentralized procurement under national procedures as there i s insufficient data to assess this; it i s "4" for procurement done by A R D S as there i s accurate data showing that more than 75% o f contracts above the threshold are awarded on the basis o f open competition. The rating along the second dimension, "justification for use o f less competitive procurement methods", i s "4" for procurement through A R D S and "3" for national systems (both justify the use o f less competitive methods in accordance with regulatory requirements, but these requirements are not hlly clear for national systems). The rating along the third dimension, "existence and operation o f a procurement complaints mechanism", i s "1" as no such mechanism exists. The overall rating i s therefore "3" for A R D S procurement and "2" for procurement carried out directly by Government. Indicator 20: Effectiveness of internal controlfor non-salary expenditures 1.117 To the extent that public spending i s implemented by the Government, the public financial management control system comprises the budget, allotment and reporting procedures, and release o f funds executed by the central authorities and their delegates inthe provinces. Inaddition, controls inline ministries that relate to authorization o f expenditure and approval o f services rendered and internal audit are important. 1.118 Internal Control Framework. Under the 1983 Organic Budget Law, MoF had the responsibility to pre-audit expenditure; this i s carried out by their representatives in the line ministries, where all payment requests are reviewed for allotment availability, adequate supporting documentation, and compliance with authorization procedures. At the same time, the Control and Audit Regulation provides line ministries with "Domestic Auditors" who conduct ex post audit under a reporting relationship to the line ministries and to the Auditor General. M o F delegates in line ministries do not have access to the allotment ledgers, so review and approval o f these i s perfunctory. In the case o f the Mustoufiats, allotment information from AFMIS distributed by the Treasury frequently does not reconcile with information inthe line ministries, because o f different timing in communicating changes approved in the center from two different sources (see above). 1.119 In2002, an integrated public financial management system (AFMIS) was set up inMoF. AFMIS i s a cash-based financial system using FreeBalance software, which was initially configured to accept expenditure information and to produce checks. Treasury currently handles all data entry and check runs for M o F and line ministries in Kabul. It records in summary the operations o f the Mustoufiats based on their monthly reports on expenditure and revenue. The structure o f AFMIS was modified at the beginning o f 2003/04 to enable recording o f the Approved Budget, Primary Budget Unit (PBU) allotments, and location allotments. The system provides facilities to ensure that for each economic major group item: (i) 34 spendingo fthe budget agencies is within the limitso frelevant unspentlocation allotments; (ii) allocation o f location allotments i s within the limits o f the relevant unused PBU allotments, and (iii) allocation o f PBU allotments is within the limits o f the relevant unallocated approved budget. Similarly, under the reforms in Treasury all processes are well documented, there are job descriptions for all positions, and staff are trained in their respective duties. AFMIS records revenue in the accounts o f the Chart o f Accounts and enables budget revenues to be recorded and reported under appropriate classifications. New procedures have been developed for posting andrecording revenues and reconciling revenue receipts with bank statements. AFMIS can now produce regular and ad hoc reports on revenue performance o f the budget for each revenue type and class, for eachrevenue collection center, for eachprovince, and for each accounting period. 1.120 Internal controls in the central government are currently strengthened by the presence o f an independent Monitoring Agent paid for by ARTF that pre-approves requests for reimbursement o f recurrent expenditures to be funded by the ARTF. The ARTF finances a large part o f the Government's recurrent budget including civil servant salaries, but not including defense and security expenditures. The Monitoring Agent's review helps to verify compliance with procedures and witnesses evidence o f services or goods received. Defense and security payrolls are reviewedby M o F under a special project on payroll verification (see below on payroll controls). 1.121 Effectiveness. The effectiveness o f controls - including pre-audit and internal audit - can be inferred from the results o f the M o F review o f payment requests, form M-16, and the Monitoring Agent's report on eligibility for recurrent expenditures. The error rate incurredby ministies and agencies inKabul on the submissions o f M-16 (payment requests) to the Treasury has been improving over time (Table 1.3), but non-compliance remains high. The ex-post review by ARTF shows drastic improvements from last year to the current year (Figure 1.1). However, ARTF's Monitoring Agent finds that capacity and understanding o f the control framework remain weak inline ministries. 1.122 Rating 20: 2. Internal control has several dimensions. An effective internal control systemi s one that is: (i) based on an assessment o f risks and the controls required to manage the risks; (ii)incorporates a comprehensive and generally cost-effective set o f controls which address compliance with rules in procurement and other expenditure processes which are widely understood and complied with; (iii) prevents and detects mistakes and fi-aud; (iv) safeguards information and assets; (v) ensures quality and timeliness o f accounting and reporting; (vi) i s circumvented only for genuine emergency reasons; and (vii) for which top management takes full responsibility. Evidence o f the effectiveness o f the internal control system should come from regular audits. Internal control goes beyond formal rules and procedures to include the attitude, awareness, and actions regardinginternal control. 2003104 2004/05 M 16 37.0% 24.0% ARTF MA 69.4% 34.5% Source: MoF. 35 Figure 1.1: Share of Non-Salary ExpendituresReimbursed by ARTF r I 2004105- 2004105- 2004105- 2004105- 2005l06- 2005106 2003104 1 2 3 4 1 2 Source: ARTF MonitoringAgent. Note: Ratio of non-salaryexpenditureapprovedfor reimbursement onnon-salaryexpenditure submitted to ARTF. 1.123 The problems with the control environment in Afghanistan can be inferred from the problems withrevenue remission to the center bythe provinces, confusion with and access to authorized allotments, and general variance from formal rules and regulations. The rating for internal control i s "2" because the system consists o f a basic set of rules for the processing and recording o f transactions which are rather well understood by those directly involved in their application, and which audits indicate are observed in a significant majority of transactions. Inparticular, effective controls are in place in Treasury with the computerized financial management system. Commitment controls are ad-hoc, manual, and not integrated in the budget, but there i s evidence o f improving and most recently adequate effectiveness o f controls. However, capacity in line ministries remains weak. Finally, the work o f the Monitoring Agent forms a key element o f the internal controls today, but it should be rendered redundant over time with advances in systems, training, and clarification o f the regulatory framework. Indicator 21: Effectiveness of Internal Audit 1.124 Legal and Institutional Framework. The Control and Audit Regulation provides for the internal audit function as part o f line ministries' organization and calls for coordination with the external auditors. There is no regulation, however, to support the work o f the auditors. Internal Audit Departments exist in 12 major Ministries, including Finance, Defense, Interior, Education, Higher Education, Health, Public Works, Telecommunications, and Commerce. Each Internal Audit Department i s organized along similar lines and is under the control o f an Audit Director, with separate audit units for relations - to co-ordinate and supervise audit work; planning -to plan audit work; analyses - to carry out analysis o fthe work done bythe audit groups; and audit groups -to carry out the financial andaccounting audits. 1.125 Audit Activity. Audits are carried out for all entities on an annual basis. Fieldwork inprovinces also reveals significant audit activity (see Volume 111, Chapter 3). Audit plans are submitted to the Control and Audit Office (CAO) to coordinate effort across departments and prevent duplication o f work. A Director in the CAO is tasked with coordinating work with the Internal Audit Departments. Internal audit work i s focused on the identification o f irregularities and the investigation o f potential corruption and fraud. When a potential crime i s identified, cases are passed to the Attorney General for legal action. Summaries of audit results are submitted to CAO for information. However, audit plans are not based on risk-related factors. Reviews o f internal control processes are not carried out, so the external audits carried out by CAO place no reliance on the work done by the Internal Audit Departments. 36 1.126 Despite this lack o f activity, the ARTF Monitoring Agent provides reports to M o F with data on performance o f line ministries with regard to fiduciary standards. Eventhough this mechanism should be seen as temporary, it provides feedback to management on performance and progress in meeting these fiduciary standards, and shows that actions have been taken as a result o f that. There i s Government response and follow-up on the observations o f the Monitoring Agent to address specific uses o f budget funds which were disallowed, and also onthe trends inquality o f submissions which are trackedbyunit. 1.127 Audit Capacity. All Internal Audit Departments recruit staff from within their own Ministries. It i s common for Internal Audit staff to have in excess o f 20 years' experience within the ministry concerned. Ministries carry out training themselves mainly though the use o f internal seminars to discuss problems and resolve issues. No formal academic or professional training for internal audit i s available in Afghanistan. Internal auditingmanuals do exist, but these are 30 years old and only refer to work methods and not to modem internal audit practices. The level o f information technology knowledge and experience inthe Internal Audit Departments i s very low or non-existent, and there i s no attempt to carry out internal audits usingcomputed aided audit techniques. 1.128 The Internal Audit Departments cannot be regarded as independent because they are heavily influenced by the ministries within which they function. They do not carry out work that would be consideredto comply with recognized international internal auditing standards. 1.129 Under the new PFEM Law, M o F will establish an internal audit administration and appoint internal auditors in the line ministries; this provision will have to be coordinated with the emerging amendments to the Control and Audit Law. A reform i s underway inM o F to reorganize its own Internal Audit Department and those o f the line ministries, develop an internal audit regulation, and, at the same time, bringtheir work into line with international internal auditing practices. 1.130 Rating 21: 2. Regular and adequate feedback to management i s required on the performance o f the internal control systems through an internal audit function (or other systems monitoring function) that i s appropriately structured; has adequate independence, breadth o f mandate, and power to report; utilizes appropriate professional standards; and reports on systemic issues. Specific evidence of an effective internal audit function would also include assessment and monitoring o f error rates in procurement and expenditure transactions, a focus on high-riskareas, reporting on correction rates, reliance by the external auditor on the internal audit reports, and actions taken by management on internal audit findings. 1.131 Monitoring o f the internal control system i s in place through the extraordinary arrangements relying on the FM Advisor, Procurement Advisor, and Monitoring Agent (even though these operations examine only transactions which are reported through AFMIS and do not formally examine compliance or evaluate the internal control system in the central government). The rating i s therefore "2" as some monitoring o fthe internal control framework i s provided by these arrangements. c. Accounting,Recording,and Reporting Indicator 22: Timelinessand regularity of accounts reconciliation 1.132 Reliable reporting of financial information requires constant checking and verification o f the recording practices o f accountants; this i s an important part o f internal control and a foundation for good quality information for management and for external reports.20 2o Inaddition to accounts reconciliation, two other types o f reconciliation are important. First, reconciliation with line ministries budget figures i s critical for application o f the segregation o f duties between authorization o f use of funds, which rests with the line ministry, and recording and custody o f funds, which is concentrated in Treasury. Line ministries keep records o f both their 37 1.133 With regard to the reconciliation o f bank accounts, the official budget control and accounting records of the Core Budget are those o f AFMIS operating under Treasury, which apply the cash basis o f accounting. The movement o f the TSA inthe DAB i s now compared daily to the accounting records inan Excel application; however, full automation in AFMIS i s yet to be developed. All differences are identified although not all o f the corresponding adjustments in the accounting records or DAB records have been effected. Also, the opening balance o f this account for the current fiscal year was not reconciled with the DAB balance, so there i s no external corroboration for the financial position reported in the financial statements. The unreconciled difference carried forward from prior years is significant, and the required adjustments (or recovery from DAB) are pending. This issue has recently been addressed for the preparation o f the 2004/05 annual statements. Procedures have also been put in place for monthly reconciliation o f the bank accounts o f the Mustoufiats, but reconciliations o f provincial accounts remain weak. 1.134 Related to the line ministries and the Mustoufiats records i s the issue o f cash advances. At any given time, there i s $11million of outstanding advances to the line ministries for payment to suppliers. Treasury has introduceda system to control these through the use o f a holding account and reconciliation o f eachholding account i s done periodically. 1.135 Rating 22: 2+. Timely and frequent reconciliation o f data from different sources i s fundamental for data reliability. This indicator has two dimensions. Along the first dimension, "regularity o f bank reconciliations", the rating i s "2" because bank reconciliation takes place at least quarterly, usually within eight weeks after the end o f each quarter. Along the second dimension, "regularity o f reconciliation and clearance o f suspense accounts and advances", the rating i s "3" as the Treasury has established and implementednew procedures by which these accounts are cleared at least annually. The overall rating i s therefore "2+". Indicator 23: Availability of information on resources receivedby service delivery units 1.136 All public spending in the central Government is subject to the budget, which is expressed in appropriations and authorized through the allotment process under the responsibility o f MoF. For the Core Budget, there i s good information on what resources are received by provincial departments. This does indicate that funds are being used for the intended purposes and that by and large the budget i s made available to the authorities under the appropriations. 1.137 While the reporting on actual uses o f funds i s available and reliable, the supply o f funds to front- line service units i s undermined by problems with the allotment procedure (see above). As a result, the distribution of resources to provincial offices for non-salary items i s delayed and often affected by discretion applied at the Mustoufiat level. Information on the receipt o f resources by service units (at the facility or district level) i s often laclung. Front-line service units (e.g. primary schools) have little or no information on resources allocated to them (beyond employees). A small public expenditure traclung was piloted in the education sector under the World Bank Public Finance Management Review (see direct revenue, which i s deposited into the TSA, and expenditure payments authorized via forms M-16. The accounting of these in AFMIS, however, originates in Treasury for Kabul operations and with the Mustoufiats for provincial operations. Treasury circulates a report of revenue and expenditure from AFMIS to all line ministries monthly and follows up on the replies for reconciliations. Average compliance by the Ministries i s 70% for expenditures and 80% for revenues. Second, it is also important to reconcile revenue and debt. Revenue collected through the Mustoufiats, largely deposited directly in DAB, is recorded in revenue suspense accounts in AFMIS when transferred to the TSA revenue account, until the corresponding accounting by exact revenue code is received from the Mustoufiat. Loan transactions reported under the DMU and those recorded inAFMIS are reconciled monthly. 38 Volume I,Chapter 6 and Volume IV, Chapter 2). Also, communities do not have access to information on budget implementation outside o f the aggregated reporting available inGovernment publications. 1.138 With respect to the External Budget, no verified information is provided by the Government on the extent to which reported disbursements are reaching the beneficiaries. There i s no record o f these expenditures in the Government accounting, and donor reporting on expenditures i s not regulated, monitored, or aggregated for management or external reporting. The Auditor General does not access the external audit reports on the performance o f the service delivery contractors financed by the investment operations o f the External Budget. 1.139 Rating 23: 1. This indicator rates information on the provision o f goods and services to front-line service delivery units which were intended for their use. Information on the receipt o f resources by service units i s critical; it can be provided through the accounting system, if sufficiently extensive, reliable, and timely. Overall, despite good information on Core Budget execution, the rating i s "1" because there i s no data collection on resources to service delivery units (except the pilot expenditure traclung survey ineducation). Indicator 24: Quality and timelinessof in-year budget reports 1.140 Information for M o F on budget execution i s on-line for Kabul operations because it i s processed by the Treasury inAFMIS. Such information i s available monthly for provincial operations through the summary report provided by the Mustoufiats, although 30% o f this i s rolled up with a month's delay. Revenue from the provinces i s recorded inthe cash position and to a revenue holdingaccount on a weekly basis. Funds are swept from the revenue accounts, but specific accounts are only assigned after summary reports from the Mustoufiats are received (see above). The Treasury has been posting monthly fiscal reports on its website since March 2005, but they focus on the aggregate fiscal position. 1.141 As noted, line ministries keep their own records o frevenue and expenditure based on the requests made for payments by Kabul and fi-om information received from their provincial offices. Reporting to line ministries from AFMIS i s done monthly through a report o f revenue and expenditure which i s followed-up for replies on reconciliations between the line ministries' own records and AFMIS records. Line ministries are responsible for disseminating this information to their secondary budgetary units. 1.142 Effective reporting i s hampered, however, by the delay in submission o f information on transactions effected through the Mustoufiats, because 30% o f these are only recorded in the month following the transactions. Revenue i s not fully coded until a month after collection. Moreover, for both Kabul and provincial operations, the heavy use o f supplier advances distorts the exact timing o f the reporting since they are reported as expenditures inthe period o f the advance and not inthe period o f the service. Because there i s no month-end cut-off, information for a given month may change with enties made long after month-end. 1.143 Rating 24: 2. The ability to monitor the budget requires timely and regular information on actual budget performance to be available to M o F to monitor performance and, if necessary, identify actions to get the budget back on track. Line ministriesneedthis information to manage the activities for which they are accountable. The role o f M o F i s both to effect payment and maintain the primary budget execution records. The line ministry i s responsible for reconciling M o F data with their own records. Efforts are being made to achieve effective two-way flow of information between M o F and line ministries. 1.144 The rating i s "2" because budget reports, with classification that allows comparison to the budget at some levels and which incorporate expenditure and revenue data, are disseminated inthe Government within four weeks after the end o f each quarter. Although there are some problems o f accuracy (e.g. 39 when data from a large province i s missing), these do not fundamentally undermine the usefulness o f the reports, since the sources o f inaccuracies (e.g. omissions) are known and can be taken into account. Indicator 25: Quality and timelinessoffinancial statements 1.145 The 2004 Constitutionrequires the Executive to report on the execution o f the budget inthe first semester following the end o f the fiscal year. Under the Organic Budget Law, the M o F i s responsible for reporting. The system for budget execution reporting i s AFMIS under Treasury (since the Accounting Departmenthas been amalgamated with the Treasury). 1.146 Timeliness. For 2003/04, audited copies o f the state budget financial statements were requiredto be submitted to the Afghan Government's council o f Ministers by 20 September 2004. The statements were submitted to and approved by the Council on 27 September 2004, subject to evaluation and approval by the Auditor General. These statements were subsequently revised on 7 November 2004. While the timeliness does not meet the deadlines o f the Constitution, improvements were made as compared with the prior year, and the results were in time to be considered in the development o f the 2004/05 budget, comparing well to international practice. 1.147 Quality. Externalreporting for the Government-implemented Core Budget to the Cabinet (which has final approving authority until the legislature i s formed), the donor community, creditors, and the public arises from two separate processes, which are not reconciled. Requisite reports on donor-financed projects, annual audited financial statements, and annual financial reports o f donor-financed investment operations are prepared from bank accounts and related supporting expenditure documents and from donor advices on disbursements, which are also duly reconciled to the related bank statements. These accounts are prepared on a modified cash basis in that they reflect operations pertaining to 2003/04 even ifcashmovementtookplace after the endofthe fiscalyear. 1.148 On the other hand, reports on the state budget and financial position are prepared fi-om the accounts in AFMIS, on a cash basis. These are presented in line with international public sector accounting standards, except that they do not reflect donor contributions under direct implementation which, under the standard for cash accounting, are payments by third parties on behalf o f the central government. These operations should be presented in a separate column in the financial statements as sources and uses. The difficulty arises in that the Government does not have verified information on the exact uses, although it does have information on the sources. 1.149 Rating 25: 2. Consolidated year-end financial statements are a critical element o f transparency in the system. In addition, the ability to prepare year-end financial statements in a timely fashion i s a key indicator o f how well the accounting system i s operating, and o f the quality o f the records maintained. A complete set o f financial statements (but without information on financial assets and liabilities) was presentedto the fiscal authorities within 12months o fyear-end. However, given the significance o f donor in-kmd contributions which are not reflected in the cash statements, the financial statements are not presented fully in accordance with international accounting standards. With the delay in reporting on 2003/04, the quality o f the financial statements presented to the highest fiscal authority mustbe rated "2". d. ExternalScrutiny andAudit Indicator 26: Scope, nature, andfollow-up of external audit 1.150 Legal Framework. The 1985 Control and Audit Law recognizes the General Control and Inspection Presidency o f the Council o f Ministers (now the Control and Audit Office, CAO) as the supreme audit institution which i s responsible for attesting to the veracity o f the annual accounts, 40 evaluating the performance o f the "domestic auditors", and reviewing the State Owned Enterprises. A Presidential Decree o f Oct 17, 2002 established the Control and Audit Office (CAO) to implement the 1985 law and provides that Ministries, offices, banks, enterprises, governmental organizations, and related social organizations will assist the Audit Groups. The Presidential Decree o f November 30, 2002 named Dr. Sharif Sharifi as the Auditor General, reportingto the President. Inline with the audit law, the C A O audits the "Final Account" o f M o F and also undertakes external audits inthe provinces. 1.151 Under the 2004 Constitution, the Executive must present to the National Assembly a final account o f the prior year's budget transactions, but there i s no provision for an external review o f this rendering. The PFEM Law supports the external audit function under the CAO, and a draft external audit law i s under preparation. 1.152 CAO Activities. To support the fiduciary framework for donor funding to Afghanistan in the immediate post-conflict period, the first World Bank-supported Emergency Public Administration Project funded technical assistance to the CAO through an audit advisor to assist in the conduct o f the external audits o f the ARTF, LOTFA, investment operations administered by the Government, and state budget financial statements. This has enabled the C A O to upgrade the capabilities o f its staff and to complete the audit o f these operations in 2003/04 to international standards, except for the state budget review o f 2003/04 where the opinion does not follow international standards o f auditing. The Auditor General offered an opinion with qualifications on the statements o f the ARTF relating to issues on bank reconciliations in 2003/04 and 2004/05. The CAO i s currently undergoing reorganization as part o f the Priority Reform and Restructuring (PRR) program within the framework o f a special salary re-structuring already granted to the CAO by the President. Since 2002, most senior staff inthe organization have been replaced, more qualified persons have been recruited at lower levels, and there are plans to increase staff fiom 160 to 300 through hrther recruitment o f graduate junior staff in accordance with new selection criteria. Because local graduates with modem accounting knowledge are generally unavailable and also because many o f the recruits will have law or other degrees, there i s a need for additional professional audit training on entry. There i s still inadequate capacity to conduct financial statement audits without the support o f the audit advisor. 1.153 Follow-up on findings. M o F has been closely involved in responding to CAO audit reports, including the observations inthe management letters on World Bank grants and trust funds, especially the multi-donor ARTF. For example, a matrix o f corrective measures itemized the actions to be taken following the 2002/2003 audit. Other Ministries, however, are less responsive, and overall only 2% o f Government bodies respond to audit reports on their activities. The Auditor General does prepare reports on departments and special reviews which are presented to the Cabinet semi-annually. These reports are not public, and no response or action i s offered by the reviewed administrative units. 1.154 Rating 26: 2. For the Government to fulfill its fiduciary responsibility, external audit has to be carried out to acceptable standards, with the participation o f the audit advisors to plan, supervise, and execute core components o f the audit. There should be direct follow-up on audit findings by the Executive, which may include follow-up by M o F and follow-up by the individual audited entity. Evidence o f effective follow-up o f the audit findings includes the timely reduction inuncleared findings, and the provision by the Executive or audited entity o f a formal written response to the audit findings, indicating how these have been and are being addressed. The rating is "2" because over 50% o f government expenditures are audited annually, reports are issued within 12 months o f year-end, and significant issues are identified. All the financial statement audits include an assessment of internal control systems and reports on systemic issues as well as irregular transactions. The response by M o F to the observations on internal control made in the annual audit o f financial statements i s thorough and supported by an action plan, whereas responses in other areas on compliance audits are practically non- existent. 41 Indicator 27: Legislative scrutiny of the annual budget law 1.155 The 2004 Constitution provides the legal authority for budget formulation, approval, and execution. As was stated earlier, responsibilities are clearly established between the Government (for formulation) and the bicameral legislature (for approval). However, as the Parliament has only just been formally established, the Cabinet currently approves the national budget through Presidential decree; the decrees are issued by the President, after agreement of the Council o f Ministers. Upon the establishment of Parliament, the budget will be submitted through the upper house (Meshrano Jirga) with an advisory comment from the lower house (Wolesi Jirga), although final approval will rest with the lower house. The Constitution (art. 137 and 139) allows linkingbudget management responsibilities with sub-national administration in the years to come. Importantly, the law also details the procedure for budgetary approval and amendment between MoF, the Budget Committee, and Parliament. 1.156 Rating 27: 1. The power to give the government authority to spend will rest with the legislature after this body i s convened. Ifthe legislature does not rigorously examine and debate the law, that power i s not being effectively exercised and will undermine the accountability o f the government to the electorate. A rating of "1" i s given since so far legislative scrutiny has not occurred. This rating is obviously temporary as it reflects the absence o f a Parliament as o f June 2005. Indicator 28: Legislative Scrutiny of External Audit Reports 1.157 The state budget financial statements are formally approved by the Cabinet, but these may be considered before the audit i s finalized as was the case in 2003/04, when financial statements were approved by Cabinet but subject to subsequent presentation o f the auditor's report. This does indicate that the auditor's report on the financial statements i s critical to the review by the Cabinet. A formal evaluation o f options for developing an overall framework for the scrutiny and follow-up o f audit reports based on models in other comparable countries will be made and a system for Afghanistan designed. The Government i s planning to form an interim ad hoc scrutiny committee to examine responses o f heads o f agencies to recent audit reports and subsequently to help the National Assembly in forming a Public Accounts Committee. 1.158 Rating 28: 1. The legislature plays a key role in exercising scrutiny over the execution of the budget that it approves. A common way in which this i s done i s through a legislative committee/commission that examines external audit reports and questions responsible parties about the findings of such reports. The operation o f the committee will depend on adequate financial and technical resources, and on adequate time being allocated to keep up to date on reviewing audit reports. The committee may also recommend actions and sanctions by the Executive. This indicator i s rated "1" since the National Assembly has not yet reviewed the public accounts (as inthe case o f indicator 27, this rating i s temporary). e. Indicators of Donor Practices 1.159 The PFM performance measurement framework includes three indicators reflecting donor practices, measuring: (i) the predictability of direct budget support; (ii) the predictability o f donor flows (based on how closely actual aid flows correspond to donor forecasts and the comprehensiveness and timeliness of information); and (iii)the portion o f aid managed under national procedures. 42 Indicator One:predictability of direct budget support 1.160 Direct budget support has consisted o f two multi-donor trust funds (ARTF and LOTFA) and program loanshudget support operations (from the Asian Development Bank and the World Bank). It should be noted that this rating does not pertain to the amount o f budget support provided, its share o f total external assistance, or its adequacy in relation to national external funding requirements; instead it rates whether the budget support that i s provided occurs ina predictable manner. 1.161 Rating One: 4. The rating i s "4" as the Government has been able to fully rely on these sources, ARTF in particular, to meet its recurrent budget financing requirements in a flexible and predictable manner. There have, however, been some delays in disbursements from LOTFA due to donor delays in replenishing that trust fund. Moreover, external budget support (from program loanshudget support operations) that the Government has designated for use in financing development projects in the Core Budget also has been predictable on a year-to-year basis. Indicator Two: Financial information provided by donorsfor budgeting and reporting on project and program aid 1.162 Donors provide significant information to M o F on external assistance. Also, the coverage i s estimated to be relatively high,with most donors giving some information to M o F so that the Government can monitor donor activity. However, the main issues are (i) the absence o f data on expenditures (most donors only focus on disbursements from their own account to the implementing agency), (ii) concerns often raised about the accuracy and consistency o f the data, and (iii) delays in providing information to the Government. 1.163 Rating Two: I+.Donor practices can support or hinder PFM inpartner countries. This indicator captures two key aspects o f donor practices: (i) the extent to which donors provide in a timely manner adequate information to the Government about the funds that are to be provided and on the funds or other forms o f development assistance that have been provided, and (ii) the extent to which donors provide quarterly reports. The rating i s "1+" because, while there i s a good record o f providing financial information, for the 2005/06 budget not all major donors provided budget estimates at least three months before the start o fthe fiscal year. Indicator Three: Proportion of aid that is managed by use of nationalprocedures 1.164 Out o f the total public spending o f $3.4 billion in 2004/05, all but $269 million was donor financed. And o f the $3.4 billion, only $0.9 billion was Government implemented (Core Budget). These resources flow through accounts controlled by M o F and thus at least partially follow national procedures. The oversight by the Government o f the operations implemented under donor responsibility i s limited, which M o F recognized publicly (MoF Financial Report in June 2004). From the perspective o f budget implementation, some recurrent spending and most public investment occurs outside national budgetary channels and procedures. The Government's control over planning o f total public spending, implementation o f public spending, and reporting on its execution i s rendered more difficult by the large portion which occurs outside o f Government control. Donors are not subject, through their donor agreements, to plan, report, and operate inline with Government procedures. 1.165 Rating Three: I. The requirement that national authorities use different procedures for the management o f aid funds diverts capacity away from managing the national systems. This i s compounded when different donors have different requirements. Conversely, the use o f national systems by donors can help to focus effort on strengthening national procedures. The use o fnational procedures would not mean that donor funds cannot be kept separate from government funds, but that the banlung, authorization, 43 procurement, accounting, disbursement, and reporting arrangements are the same as those used for government funds. The rating is "1" because less than 50% o f aid funds to central government are managed through national procedures. C.SustainingandStrengtheningPFMPerformance:a Roadmap 1.I Afghanistan's ratings against the PFM performance indicators generally portray a public sector 66 where financial resources are, by and large, being used for their intended purposes as authorized by a budget which i s processed with transparency and has contributed to aggregate fiscal discipline. The expenditure and financial position o f the resources under the authority o f the Government are reported reliably in an understandable format, although there i s some uncertainty about revenue reporting. Performance regarding the allocative efficiency across programs and agencies as well as the efficiency o f operations i s not as good, however. 1.167 Reforms which have contributedto the progress to date have been externally supported, including the commitments under the IMF Staff Monitored Program, the World Bank's Programmatic Support for Institution Building operations, and externally-supported operational functions and capacity building (in Treasury, Government procurement, and external audit, with IDA'SEmergency Public Administration Projects). The objectives o f M o F under these reforms include providing a strong legal foundation for the reforms, extending current systems to line ministries and provincial offices (Mustoufiats), and most important replacing the external operational advisors with core Government capacity over time. Further technical assistance operations are intended to address some o f these issues. 1.168 Given that performance most likely would have been rated "1" on all dimensions in 2001 or 2002, this assessment highlights the significant achievements o f the last four years. Inmost dimensions, the ratings are now comparable to other low-income developing countries (for instance Afncan countries). Areas o f relative weakness include tax collections (indicators 13-15), as well as legislative oversight (27-28) due to the absence o f a Parliament hitherto. Payroll control (18) and predictability o f budget-support (Dl) are areas o f relative strength. The Government i s making further progress in improving the P F M system by implementing a number o f reforms. A new Public Finance and Expenditure Management (PFEM) Law and a new Procurement Law have been approved by the Cabinet; tax amendments are under implementation; customs and tax administrative reforms are moving ahead; and the M o F i s moving ahead with further structural improvements. All these developments are positive. 1.169 M o F has developed a vision to guide it in identifying and implementing a wide range o f actions required for strengthening its current functions and capacities and moving forward from the emergency efforts o f the last two years toward a modern and effective public financial management organization.21 Usingthe ratings of this assessment as a baseline, future progress in improving the performance o f the PFM system as a result o f continuing reforms can be monitored. The indicators also reveal to the donor community the potential capacity o f Government systems to implement the operations which are now occurring outside Government systems. However, the nature o f the indicators and the scale o f the deficiencies highlighted by the ratings means that measurable progress, in terms o f improvements in ratings, i s more likely to occur over periods o f at least a year and more likely several years rather than on a short-term basis. Moreover, it should be kept in mind that the ratings assess the current situation in which significant external support (both advisory and operational) for the PFM system i s being provided to the Government; hence in some areas, future improvements may be manifested not in higher ratings but in maintaining adequate ratings while phasing out external capacity and replacing it with core Government capacity. ''"Afghanistan MinistryofFinance-The Way Forward",backgroundpaper for the WorldBankPFMReview. 44 1.170 Annex 1.1 seeks to provide a comprehensive, high-level roadmap of policy reforms required. It i s comprehensive in the sense that all areas o f PFM are covered. It i s high-level in the sense that some concrete next steps for the coming few months are identified, but many actions are less specific and will need to be converted into detailed action plans by the responsible government departments. Finally, the roadmap focuses on policy actions, which means that for the most part it does not cover issues o f institutional development, such as training, recruitment practices, pay structures, organizational structures, detailed business processes, specific IT systems, etc. These issues also will need to be addressed and will be closely associated with the overall process o f strengthening M o F and other key institutions. 1.171 This roadmap highlightsa few critical cross-cutting issues: 0 Enhancing the capacity of MoF is vital to improve PFMpe$ormance. Since M o F i s chief custodian o f the PFM system, its capacity to develop and implement policies i s at the core o f this action plan. Progresstoward a Five-Year Strategic Plan-initiated inMay 2005 -will be key to developingMoF's capacity and performance. Certain high-level policy and institutional parameters require careful analysis and Cabinet ownership. The Public Finance and Expenditure Management Law brings some clarification with regard to line ministries' responsibilities, but more work is required to flesh out these responsibilities (for budget formulation, internal control and audit, extent o f delegation, payment authority, commitment and fixed asset controls). With regard to the center / province relationship, the Constitution has established a unitary state: with the exception o f municipalities, subnational administration only has the powers that the center delegates to it. Also, in terms o f financial management, the system i s at present "MoF-centric" in the sense that most budget implementation issues are dealt with directly between provincial departments and Mustoufiats (as opposed to within line ministries). The Cabinet needs to consider whether (and for how long) this i s most appropriate, and to fully own the approach chosen. 0 Thesuccess of the reform program will also depend on the sustained support of donors. The content, mode, and capacity-building elements o f technical assistance will be very important. Itis hopedthat discussions around this action plan can help align donor support and identify gaps in assistance. 1.172 The next steps o f an effective action plan would include the following: (i) prioritize the actions around a set o f mutually supportive measures that are feasible, realistic, and sustainable, and that can generate a step change in PFM performance during each year; (ii) develop support for these reforms (among staff o f MoF, the Cabinet and other Government agencies, with the Afghan people, and with the international community); (iii) reinforce the policy agenda with a solid organizational and institutional development plan; (iv) prepare detailed action plans with a clear assessment o f technical assistance requirements; and (v) develop a process to monitor actions and performance, review progress, and adjust the programas required. 45 ! SC i ; iC ; S aE c e iY E "II IC i e 1 7 I cd 4 W . . '1 & IY Y 3 0 r- hl 8 6 > a ! N n -4 -4" -4 3h' 0 v, WI N W N' N N B .r .r c ._ c > L R b K .r F .r E: i .r 2F r r t PART11. PUBLICPROCUREMENTFORDEVELOPMENT EFFECTIVENESS Executive Summary 1. The overriding objective o f a national public procurement system i s to deliver efficiency and value for money in the use o f public funds, while adhering to fundamental principles o f non- discrimination, equal treatment, and transparency. Procurement i s therefore at the core o f the Public Finance Management (PFM) system and contributes greatly to several o f its objectives, including efficiency, transparency, and accountability. The state o f procurement in Afghanistan suggests that the regulatory framework has been in need o f improvement. While fiduciary standards are reasonable (though in need o f further improvement), procurement focuses excessively on prices at the expense o f quality. In addition, there is a lack of implementation regulations, such as handbooks, operational procedures, and standard contracts. And the absence o f monitoring, complaint, and appeals mechanisms weakens the overall performance o f the procurement system. The recent passage o f the new Procurement Law, and its implementation through specific regulations (currently beingprepared), will address many o f these issues. ii. Besidesthelegalframework,thefirstkeyissueislackofownershipandaprocurementchampion inthe Government. This is a serious impedimentto reform and to inter-Ministerial dialogue. There are a number of obstacles to such ownership, for example: (i) o f adequate training (both inthe public and lack private sectors); (ii)misunderstanding o f the scope o f the procurement function (i.e. the belief that only major works or purchases are "procurement"); (iii)inappropriate regulations; (iv) lack of common internal structures to deal with procurement in the line Ministries; (v) lack o f communication between trained procurement staff and senior administrators; and (vi) no clear incentives for improvement and ownership. iii. A second, related issue is the lackofcapacity inthe lineMinistries, as evidencedbytheir inability to define and communicate effectively their desired technical specifications in procurements conducted with the assistance o f the procurement consultant. Lack o f capacity i s also evident inthe local private sector: while the number o f bids i s reasonably high, there i s limited understanding o f and capacity to implement the public procurement rules. iv. Finally, a number o f mechanisms, such as monitoring o f performance, complaints, appeals, procurement audits, are missing, therefore reducing the effectiveness o f the procurement system. v. Priority actions should include the following: 0 Create a Procurement Policy Unit; 0 Adopt regulations to implement the new Procurement L a w 0 Develop and implement a large-scale capacity buildingprogram for procurement officers in the Government; and 0 Create a training program for the local private sector to enhance its capacity to participate to bids. 52 A. Introduction 2.1 As part o f the broader review o f Afghanistan's Public Finance Management (PFM) system, this assessment o f procurement procedures and performance focuses on the legal framework applicable to the conduct o f procurement in Afghanistan as well as on the performance of those responsible for applying the procurement procedures. Procurement inAfghanistan i s conducted under a multiplicity o f procedures of greater or lesser sophistication and by various public and private sector entities. It comes as no surprise, therefore, that performance has been uneven and dependent on the precise procedures which are applied and by whom they are applied. 2.2 The assessment seeks to identify the applicable rules and to assess their appropriateness and suitability for the task o f malung the most efficient use o f the resources available. It further seeks to assess the performance of those responsible for applying the procedures, either within the Government or external agents acting on behalf o f the Government. In doing so, this assessment also seeks to take a holistic approach to the procurement environment by investigating both sides o f the procurement equation: not only the quality and competence o f the purchaser (Government) but also the ability and preparedness o f suppliers and contractors to tender for and provide the Government with the goods, works, and services it needs at competitive prices and at satisfactory levels o f quality. 2.3 The ultimate objective o f this assessment i s to guide and assist M o F and more generally the Government in identifying and implementing actions required for strengthening the functions and capacities o f those involved in procurement and the procurement function itself. It will also help M o F to move forward from the "emergency" needs and approaches o f the last three years o f political and economic transition, toward a modem and effective national public procurement system with the emphasis on sustainability and nationalpublic and private sector development. 2.4 Section B below outlines the structure o f the procurement system as it relates to the budget. Section C reviews the performance o f the system, focusing on the portion for which performance data are available (funds channeled through the procurement agent). Section D assesses the national rules and institutions for public procurement. Section E reviews procurement rules for implementation o f donor- assisted projects in the Core Development Budget. Section F outlines key issues related to the legal framework and the lack o f capacity in the Government. Section G analyzes issues facing the private sector with respect to participating in public procurement. The final section concludes and summarizes recommendations. B. TheBudgetandProcurement 2.5 Procurement for expenditures using funds from the national budget will usually be subject to procurement rules set by the Government. Procurement for expenditures using donor funds will usually be subject to the procurement rules o f the donor organizations (for example, the World Bank's or ADB's procurement guidelines). This may apply to both loans and grants. A distinction needs to be drawn, however, between donor funds which are used directly and those which are provided indirectly through various budget support mechanisms. In the former case, the funds remain identifiable with the donor organization. Inthe latter, they become "fungible" and are considered to be an integral part o f the national budget. Procurement expenditure using such fungible hnds will, absent any agreement to the contrary, be spent according to the procurement rules applicable to the national budget. 2.6 The emergency situation inAfghanistan inthe immediate post-conflict period led to a plethora o f donor funds and mechanisms for providing and using those funds. This generated confusion, since the 53 multiplicity o f fundingsources ledto a multiplicity o f applicable procurement rules, sometimes inconflict with each other. This resulted in inconsistencies, reduced capacity in the sense that knowledge was dispersed unevenly, and prevented the emergence o f a common process capable o f being applied consistently throughout the country. 2.7 Afghanistan's national budget i s divided in two parts (see Part Io f this Volume). The Core Budget includes all funds flowing through the Government's accounts. Within the Core Budget, expenditures are divided into two broad categories. Operating expenditures are primarily recurrent and continuing, and include all civil servants' wages and pensions, plus goods and services for operations and maintenance (O&M); however, this component also includes some small investment expenditures. These expenditures are financed by both Government revenues and external aid, notably grants through the Afghanistan Reconstruction Trust Fund (ARTF) and the Law and Order Trust Fund for Afghanistan (LOTFA). Development expenditures include projects that finance mainly capital goods (e.g. roads), but also some recurrent expenditures such as technical assistance, training, health services, and grants. These expenditures are financed through direct budget support such as the World Bank Programmatic Support for Institution Building (PSIB) and ADB Program Loans and through external project assistance channeled through the Core Budget, mainly IDA or ADB funded projects as well as investment projects financed by ARTF. The External Budget includes all external assistance funds not flowing through the Government's accounts, i.e. those directly disbursed by donors to implementing agencies (Non- Governmental Organizations, private firms). 2.8 The External Budget inevitably i s subject to the procurement rules applied by the donors, although donors may chose to apply the National Procurement Law provided it i s deemed acceptable. Core Budget procurement i s a more complicated matter, but despite a certain lack o f clarity in the practical application o f the procurement rules (to be discussed in the various sections below), the procurement implications o f the various budget components can, based on the broad categorization given above, be represented schematically by Figure 2.1: Figure 2.1: ProcurementImplicationsof Source of Funds CORE BUDGET v v OperatingExpenditures DevelopmentExpenditures v v v v Revenues Assistance 3. & & 3. Special National Procurement Law, including use of ARDS a/ arrangements, including use of ARDS 2.9 This assessment i s concerned with procurement under the Core Budget. As such, it needs to address the various procurement systems which may apply. Because o f its crucial role, the assessment will begin with considering the way in which procurement i s conducted under the Procurement Agent 54 ( A R D S mechanism). This will be followed by consideration o f the national procurement system. Finally, the assessment will consider the procurement rules applicable to external project assistance,other than that which uses the Procurement Agent. 2.10 This assessment thus concentrates on the nationalbudget. Some expenditures, however, are made at a lower administrative level. The municipalitieshave some measure o f fiscal autonomy inthat revenues collected remain at the municipal level in order to fund municipal expenditures. For non salary expenditures under the national budget, provinces have the authority to make purchases according to the national procurement procedures. That i s the same for municipal procurement expenditure. Inpractice, very little in the way of non salary budget has been sent to the provinces or districts. The issue o f provinces and municipalities i s reviewed inVolume 111, Chapters 3 and 5 o fthis PFMReview.' C.Procurementthroughthe ProcurementAgent ROLEOFTHE PROCUREMENTAGENT 2.1 1 Following various pledges o f donor support for the massive reconstruction and development program required in Afghanistan, mechanisms were put inplace to coordinate donor funds and to ensure capacity to procure, disburse, and account for their expenditure. One such mechanism is the Afghanistan Reconstruction Trust Fund (ARTF), which provides financing for priority expenditures including recurrent costs. The ARTF i s administeredby the World Bank. 2.12 At the same time and to cope with the urgent task o freconstruction, the Transitional Islamic State of Afghanistan (TISA), with the help o f the donor community, established the Afghan Assistance Coordination Authority (AACA) in May 2002. Responsibility for the AACA lay with the Ministries o f Finance, Planning, Reconstruction, and Foreign Affairs. AACA has since been dissolved (in September 2003), and the part responsible for procurement was renamed Afghanistan Reconstruction and Development Services (ARDS) and made responsible to the Ministers o f Planning and Reconstruction, now the Minister of Economy. As the capacity to plan and implement reconstruction projects in the Government ministries was weak, arrangements were quickly put in place to provide `centralized' technical assistance in such critical areas as financial management, procurement, and auditing. * Given the constraints in Afghanistan at the time of the missions, it was difficult to collect extensive information regarding procurement at sub-central level. Nevertheless, some visits were possible, which largely provide confirmation o f the general impression that the existing regulations are followed. It would appear that the same processes are respected inthe provinces, with similar difficulties. Decentralization o f the procurement function i s not advanced. While many municipalities and local offices (presidencies) o f the various Ministries which operate in the field do conduct their own procurement, that procurement i s o f very low value. Responsibility for procurement is given only for contracts with a value o f a maximum o f about Afs 10,000, although in some cases that can be extended to Afs 30,000. In all other cases, officers must seek the approval o f the provincial governor. Even then, expenditure will only be authorized by the Ministry concerned in Kabul or, inthe case o f the municipalities, by the Ministry o f the Interior. National shopping i s the most common procurement method at this level, although competitive bidding i s not unknown. There is no experience o f international competitive bidding given the low value contracts involved. Any large contracts will be awarded centrally. The information collected (by way o f interview) suggests that the local officers follow the existing regulations and carry out their duties in the same way as the officers in the purchasing departments of the central Ministries. This means that there i s compliance (even ifthe regulations themselves are inadequate), and it would also appear that local officers understand the importance o f the integrity o f the process, with actions being taken in the event o f conflicts o f interest and instances o f corruption. However, this work suffers from the absence of any standard bidding documents and are not provided with any systematic or formalized training. What training there is involves on-the-job training, which i s likely to perpetuate any inefficiencies that exist. It i s clear that, with the new Procurement Law there will be a significant need o f re-training at this level. 55 2.13 Under the "Procurement Strengthening Component" o f the IDA-financed Emergency Public Administration Project (EPAP), the Crown Agents was contracted as the country's central Procurement Agent (PA) to put in place emergency procurement capacity to facilitate rapid, transparent utilization o f donor resources for reconstruction and development. The PA i s not, however, a true agent in the sense that it does not actually purchase on behalf o f the Government but rather facilitates purchasing by the Government. The Crown Agents contract became effective on August 15, 2002. It came to an end, following an extension, toward the end of 2004 when a new PA was appointed (RITES India Ltd). The role o f the PA is to handle procurement for all goods, works, and services under operations financed directly by IDA, as well as activities financed under the ARTF. Inaddition, the P A facility can be used (i) for budgetary expenditures from domestic resources and (ii) the Government-executed expenditures for funded by other donors. Where these options are chosen, the implementing agencies would provide technical inputs for the process, but they would not be responsible for any procurement actions duringthe course o f the project. Annex I11o f EPAP-Iprovides that procurement under IDA grants will be camed out in accordance with the World Bank's procurement Guidelines. Similarly, all procurement done under the ARTF Investment Window should follow the Bank`s procurement guidelines. 2.14 The scope o f the services contracted with the PA originally was limited to the provision o f advice and assistance inprocurement procedures, even though this would cover nearly 100% o f the procurement work o f the Government, which would simply be signed o f f by the Government. Interms o f training, the PA was required to carry out a training needs analysis and to develop a framework for training. This all proved to be significantly less than what was required during contract implementation. A s a result, the contract was subsequently renegotiated and amended inDecember 2002 to take into account the increased resources required to cope with the rapidly rising workload and to accommodate a greater focus on capacity building. The contract was further amended to include hiring ten local procurement staff and a National Counterpart Manager. In addition to handling procurement work for IDA and other donor funded items, the updated terms o f reference included staff traininghapacity building in A R D S and line Ministries and a review/recommendations on the legal and institutional framework. These activities have a significant bearing on the national procurement system discussedinthe following section. PROCUREMENT SERVICESTO DATE 2.15 The PA has worked mainly on procurement for IDA-financed projects (Table 2.1). It has started work on some ADB and EU projects, and a few applications on projects financed directly by the Government (own resources or budget-support proceedings). As o f November 9, 2005, the PA has facilitated Government procurement with the placement o f 243 contracts valued at $496 million. In addition, the PA i s workmg on 151new contracts with a total value o f $477 million. Table 2.1: Fundingby Donor Note: * includes1contractunder signature;** includes55 demands under ARTF funding (around $16 million) receivedin November2005. Source: ProcurementAgent. 56 STATISTICAL PERFORMANCE' 2.16 To assess the performance o f these procurement activities, this Volume uses a methodology under development by a group o f institutions ledby the World Bank and O E C D D A C (Table 2.2). However, not all o f the performance indicators developed by this group are usedhere, because they are designed to be applied to national procurement systems rather than procurement activities for specific donor-funded projects. For example, under a national procurement system it would be expected that a complaints review mechanism would be in place, which allows tenderers to bring complaints against alleged breaches o f the national system. N o such mechanism applies to donor-funded projects, which instead use prior and post review mechanisms. There is, additionally, the possibility to complain, but this i s a rarer occurrence in the case of donor projects as a result o f the close monitoring. Nevertheless, one complaint was lodged against the activities o f the original PA and was dealt with appropriately (with a finding that the complaint was unfounded). Table 2.2: Summary of Procurement Performance Indicators ~ Indicator name Indicates Measured by Satisfactory threshold3 Advertisement of bids Transparency and Number of bids (in %) for which 95% ormore andpublicationof opennessof system invitation to bid and contract award awards results are publicly advertised Time for preparationof Real opportunityfor Number of days between invitation to 21 days or more for open bids biddersto submit bids bid andbid opening bidding, 10 days or more for restricted bidding, and 3 days or more for shopping Time for bid evaluation Efficiency of bidding Number of days between bid opening 90 days or less process andpublicationof award Biddersparticipation Level ofconfidence of Average number of bidders submitting 5 bids or more (calculatedas private sector inthe bid ineachbidprocess the average for the sample of process transactions) Methodofprocurement Level of competition Number of bidding processes using a 1%or less method less competitive than the process recommended for the estimatedcontract amount. Direct contracting Transparency and level Percent of contracts (by number and 10% or less of number o f of competition value) awardedon a sole-sourcebasis contracts and 5% or less o f total value of contracts Processescancelled Quality ofbidding Number (in %) of bid processes 5%or less process declarednullbefore contract signature Latepayments Quality and consistency Number (in %) of payments made 10%orless ofpaymentprocess more than 45 days late Contract amount Quality ofbidding and Percentage increase o f final contract 15% or less (calculated as increase contract management amount due to changes and the average for the sample of amendments transactions) Restrictedcompetition Quality of advice Number (in %) of processes for the 5% or less for consultants selection of consultants using open competition instead of a restricted competitionor shortlist methodology Selectionmethodfor Weight of quality to Number (in %) of processes for the 15%or less consultants price ratio usedin selection of consultants having price selection weighted more than 20% of the total scoring points It shouldbe pointedout that this current assessment applies only to the performanceof the original PA, since the replacement PA has notbeenworking on the task sufficiently long for any datato be meaningful. These thresholds values are currently only tentative and not yet approved by the Bank's Board. Even then, they may be amendedfollowing a series ofpilot projectsbasedonthese performanceindicators. 57 Advertisementof bids andpublication of awards 2.17 From the available figures, it would appear that, in the case o f goods and works at least, all contracts have been advertised either nationally or internationally. With respect to services, around 60% have been advertised internationally, although none appear to have been advertised nationally. It may be that national consultancy contracts had a value which was below the threshold for advertising and didnot, therefore, require advertising. 2.18 The 95% threshold has been achieved in the case o f both goods and works for advertisement o f contracts. However, no contract award has been advertised (this has recently be initiated). Figure 2.2: Advertised Contracts 100% 80% ' O N o A d 1 60% Int'a1Ad 40% 0 NationalAd 20% Goods works Services Timefor preparation of bids 2.19 In the case of consultancy services, there is an average of 44 days for the preparation and submission o f proposals, which i s satisfactory. No data was available for goods and works, however. Timefor bid evaluation 2.20 The threshold i s met inmost cases: 76 days for large works (above $100,000 equivalent), 74 days for large goods contracts (above $50,000 equivalent), and 30 days for small goods contracts (below $50,000 equivalent). Only in the case o f consultancy services can the time be judged excessive (101 days). Bidders'participation 2.21 The threshold i s achieved only in the case of works, and then only barely (5.3 bidders on average). This i s a rather surprising result given the information otherwise collected on the availability o f local contractors. The reasons are not clear, but certainly point to a lack o f contractor capacity (see below). On large goods contracts (above $50,000 equivalent), the number i s only three, and only 2.5 on small goods contracts. For consultancy services contracts let by a competitive method, the outcome i s only three bidders on average. 58 Method of procurement used 2.22 With respect to the methods used, the open procedure has been used in 100% o f cases for goods and in 94% of cases for works. Instances o f methods other than open procedure inthe case o f works arose only in contracts whose value was above the threshold o f $100,000 equivalent which i s the trigger for international competitive bidding using Bank guidelines in Afghanistan. Of the eighth above this threshold, two were let by single source method. 2.23 Inthe case of goods, the threshold was met. Inthe case ofworks, it shouldbe recalledthat the designed to cover mature systems, and the level o f procurement activity in this instance may not be sufficient to enable firm conclusions to be drawn. Inthe circumstances, this assessment considers that the threshold i s also met inthe case o f works. 2.24 The threshold has not been met in the case o f consultancy services contracts, as competitive procedures were used inonly 62% o f cases. Figure 2.3: Procedures Used HOther E! Open I Goods Works Services Direct contracting 2.25 This was used in 6% o f cases in respect o f international works contracts (the value o f these i s unknown, however) and appears to meet the threshold. Other data 2.26 No data was available on: Processes cancelled; Late payments; Contract amount increase; Restricted competition for consultants; Selection method for consultants. Overall Performance 2.27 Given the post-conflict emergency situation in Afghanistan, the PA has managed to perform its contracted tasks well. However, the overall performance o f this arrangement i s more nuanced given a number o foperating constraints that the PA has faced: 59 Lack of appreciation o f donor requirements by the counterpart ministries, which have delayed procurement processing. Lack o f competence o f counterpart line ministry staff in providing inputs on the technical aspects/specifications o fprocurement. Lack o f procurement knowledge and capacity o f counterpart staff, malung their participation inany procurement process less effective. Non-coherentrequirements and procedures o f different donors. Lack o f capacity o f the local private sector to follow donor procedures and international best practices inbidding. Lack of overall Government ownership inthe procurement process. Agreements made with technical staff in line ministries being changed by heads o f departments. 2.28 At the same time, implementationraised anumber ofissues with the PA, including that: 0 They were understaffed at the outset, leadingto delay. 0 They should have includedmore experienced procurement specialists inits team. 0 The processing o f procurement has sometimes been slow. 0 The training courses could have been better designed. 0 The training emphasized donor procurement procedures more and current country procurement regulations less. 0 They are overly bureaucratic. 0 They do not appear to have established a good dialogue with the big procuring ministries, which has not facilitated implementationand has worked against capacity building. 2.29 The constraints and weaknesses identified above indicate that the Government should become both more involved in the procurement process currently conducted by the PA and more educated in procurement matters generally to enable a transfer o f functions and capacity. D,Procurementunderthe NationalProcurementLaw 2.30 There i s much less data available on procurement processed under national procurement rules. The main indicator comes from the ARTF's Monitoring Agent, as one o f ARTF's fiduciary standards i s the availability o f adequate procurement documentation. This indicates very strong improvements (see Part I). However, while the amount o f ineligible expenditures i s decreasing, approximately a third remains ineligible because o fprocurement issues. 2.31 In addition to this crude measure of performance, procurement under the existing procurement rules and regulations in Afghanistan may be assessed using the benchmark indicators developed by the group o f institutions led by the World Bank and OECD / DAC, the same forum which developed the performance indicators used in the previous section. The purpose o f using benchmark indicators is, however, different. These are used to assess the acceptability and appropriateness o f the national system. The methodology i s based not on a comparison o f the national system with any other given set of procurement rules such as the Bank's or other donors' own guidelines, but rather on an assessment o f the quality o f the system against benchmarks that are designed to gauge whether the objectives o f comparable systems and the principles that underlie them are met and whether the methods employed, the organization o f procurement, and market conditions are able to meet those objectives and principles. The benchmark indicators provide greater flexibility for countries to achieve an internationally acceptable standard intheir own way and tahng account o f their own particular legal or constitutional environment. 60 2.32 The benchmark indicators used are set forth inTable 2.3, and the assessment i s based on two key assumptions: 0 First, the activities of the PA are not taken into account inthis exercise because they are not based on the national procurement system but on the Bank's and other donors' procurement guidelines. 0 Second, the assessment i s based on the national rules and regulations in place at the time o f the assessment (June 2005, before the adoption o f the new Procurement Law). This new Law will bring the legal framework in line with international standards and the indicators for Pillars Iand I1will be fully achieved. At the same time, it will imply a significant change in the procurement environment and therefore will require a substantial capacity buildingeffort. Pillars and Indicators Assessment4 I.LegislativeandRegulationFramework Indicator 1: Public Procurement legislative andregulatory framework complies with 2 P A applicable obligations with regard to national and international requirements Indicator 2: Availability of Implementing Regulations andDocumentation 1 NA 11.CentralInstitutionalFrameworkandCapacity Indicator 3: Mainstreaming Proceduresinto Public Financial Management 2 PA Indicator 4: FunctionalNormative Body at the Centre 1 NA Indicator 5: Existence o f InstitutionalDevelopment Capacity 1 NA 111.Procurement OperationsandMarket Performance Indicator 6: Efficient Procurement Operations and Practice 1 NA Indicator 7: Functionality o f the Public Procurement Market 2 PA Indicator 8: Existence of Contract Administration and Dispute Resolution Provisions 2 P A IV. Integrity of the Public Procurement System Indicator 9: Effective Control andAudit System 1 NA Indicator 10: Efficiency o f Appeals Mechanism 1 NA Indicator 11:Degree o f Access to Information 2 PA Indicator 12: Ethics and Anti-Corruption Measures 1 NA a. Pillar I - The Legislative and Regulatory Framework 2.33 Afghanistan currently has a series o f provisions regulating procurement: (i)a L a w on procurement of 1987; (ii)a Purchase Regulation adopted as a Decree o f the Islamic Emirate of Afghanistan in 2000, both o f which apply to the procurement o f goods (and routine works and maintenance), and (iii) Rules and Regulations for the procurement o f works adopted sometime after 1989 (since this replaces the rules adopted on that date). The Purchase Regulation effectively replaces the Law on Procurement. However, it i s very much a copy o f the Law with very few amendments, mostly o f a religious nature. 4 The assessmenti s based on the following measures: BaselineAchievement Score Assessment Key FullyAchieved (FA) 4 Almost all benchmarksmet SubstantiallyAchieved (SA) 3 Some proceduraldeficiencies,but mostbenchmarksmet PartiallyAchieved (PA) 2 Proceduraldeficienciesandless than 50% of the benchmarksmet NotAchieved(NA) 1 Seriousproceduraldeficienciesandonly afew benchmarksmet. 61 2.34 In general terms, the provisions relating to goods are in need of significant improvement to provide further detail and clarification. Nevertheless, the direction o f these provisions i s acceptable, if a little confused, suggesting that the concepts o f a good procurement system, at least, will not be new. Large-value contracts are subject to competition (national and international) based on open tender following proper advertisement and subject to time limits. Smaller value contracts with international procurement are to be procured via an RFQ("Request For Qualification") procedure. The main problems with the provisions are the lack o f detail and clarity in the qualification (as opposed to eligibility) requirements, the setting o f specifications, drafting o f bidding documents and award criteria, particularly with the apparent predilection for negotiation, andthe tendency to re-openbiddingto attract lower prices. 2.35 Inthe case o fworks, the Government is currentlyoperating on the basis o fregulationsadoptedin 1989. The main difficulty with these regulations i s that they resemble much more a set o f bid and contract preparation guidelines than they do a regulation to ensure competitive bidding. As such, they clearly require significant updating. The regulations provide for open competition through advertisement in the newspaper. But they are inadequate in terms o f setting out qualification criteria, requirements as to specifications, award criteria, processes, and minimumtime limits. Indicator 1:Public Procurement legislativeand regulatoryframework complieswith applicable obligations with regard to national and international requirements 2.36 Overall, the current legal system offers a reasonable, though confused, framework. The key issues are: (i)regulations are not comprehensive or user-friendly and are in need o f considerable modernization; (ii) they over-emphasize prices to the detriment o f quality and transparency; (iii) they create the possibility o f opening bids before the end o f the process; and (iv) they include no complaint review process a) Scope of application and coverage of the legislative and regulatoryframework 2.37 Though not stated explicitly in the Purchase Regulation, a number o f provisions make reference to the full range o f Government entities including Government authorities, municipalities, regional authorities, and state-owned enterprise. There i s a clear presumptionthat the rules apply to these entities. No mention i s made o f the entities covered by the works regulations. The rules apply to all contracts o f whatever value, subject to different procurement methods. 2.38 Procurement o f works, goods, and services are all included, although n o specific reference to consultancy services i s made in the Purchase Regulation which also applies to services more generally. These services are, however, explicitly includedwithin the scope o f the works regulations. 2.39 Insummary, the basic principles o fan acceptable procurement system may be discerned from the applicable legislative and regulatory framework. However, this framework i s not well structured, consistent, user friendly, or easily accessible to users and all interested stakeholders. b) Procurement methods 2.40 Overall, acceptable procurement methods are recognizable but not properly defined. Furthermore, the overemphasis on price has a negative impact on quality and possibly technical standards. 2.41 There i s a stated preference, in the case o f both goods and works, for open, competitive procurement. There appears to be no alternative in the case o f works contracts other than in cases o f emergency or when the contract i s declared secret (by the contracting authority). Inthe case o f goods and services, open bidding i s mandatory above certain thresholds and, where these are not met, alternative 62 procedures such as requests for quotations and direct contracting may only be used in very few, well- defined cases: emergency situations, where the contract has been declared secret by the Council o f Ministers, and where spare parts may only be purchasedfrom the original manufacturer. 2.42 The regulations foresee the use o f competitive biddingfor contract values above Afs 1million. 2.43 Inthe case o fgoods and services, two openproceduresare envisaged, althoughboth appear to fall short o f international best practice. What i s referred to as "open bidding" i s effectively a live reverse auction which i s held at a determined time and place where the bidders will deposit a sample o f their products. "Closed bidding" or "sealed-envelope'' biddingapproximates more closely to the internationally accepted open biddingprocedure and i s used (implicitly) inthe case o f more complex goods and services where the services of an ad-hoc evaluation committee are required. Very little i s said, however, about the evaluation process. Where the prices offered by two or more bidders (offering acceptable quality) are the same, they are all asked to submit a further bid, within five days if not all are present. Where bid prices are again the same, the decision i s made by lottery. 2.44 Inboth cases, bidders and non-bidders are given the opportunity ofbettering the prices achieved at the live auction or following the sealed bidding process. Whenever another party offers to provide the goods or services at a reduction of at least 15% o f the winning price and where the winner does not agree to match that price, the other party will be awarded the contract. 2.45 There i s clearly an overemphasis on price to the detriment o f quality, which i s largely neglected in all of the regulations. The poor quality specifications are frequently criticized by both the public and private sectors. c) Advertising rules and time limits 2.46 Where competitive bidding applies, there i s mandatory publication o f opportunities for competitive procurement, and the relevant regulations set out acceptable information requirements for publication. Publication takes place more than once in national newspapers and on national radio, and invitations are delivered to Embassies o f foreign missions in Afghanistan. There is, however, no central organ for publication. 2.47 The regulations set out minimumtime limits inthe case o f goods and services. These are 10 days inthe case ofnationalprocurement (which appears a little low) and 2 months inthe case of international competition. Where internationalbiddingi s used, notices may also be sent directly to bidders. 2.48 Although not explicit, it appears that contract awards are also published but at the expense o f the winning tenderer. d) Rules onparticipation and qualitative selection 2.49 There i s a serious lack o f rules on qualification criteria. Even if foreign firms are not too heavily penalized, there i s a stated preference for State Owned Enterprises (SOEs) that i s detrimental to the efficiency o f the outcome. 2.50 There are no rules which discriminate directly against foreign bidders in terms o f participation or qualification, though there are domestic preferences. Domestic preferences o f 15% are applied where the goods offered are o f the same quality. Furthermore, if foreign bidders do not have local representatives who are taxed inAfghanistan, their bids are increased, for evaluationpurposes, by an additional 8%. 63 2.51 Even more fundamentally, however, at least in the case o f goods and services, there are n o indications o f what qualification criteria may be used or how qualifications are to be evaluated. The works regulations are marginally better but still do not set out a clear and objective set o f qualification criteria. 2.52 Where SOEs operate, procuring entities are obliged to purchase goods and services (including routine maintenance and other works inthe case o f the works regulations) from such entities even where their prices are higher. The rules state that they may be as highas 10% more, but inpractice the price preference for SOEs i s often as much as 20% more. SOEs are also exempted from the requirement to provide bid and performance securities inthe case of works procurement. e) Tender documentation and technical specijcations 2.53 The regulations contain some basic references to the content o f tender documentation, but this i s not supplemented by any additional guidelines or forms, except in the case o f works where these are contained inthe regulationsthemselves. 2.54 Neither set o fregulations provides any adequate guidance on technical specifications or the use o f standards, international or otherwise. This i s also related to broader issues such as lack o f capacity and absence o f standards: frequent complaints have been made on the level o f technical competence o f procurement officials and on the general absence in Afghanistan o f accepted standards and technical specifications. The result i s that evaluations take place in respect of very different products with the resulting absence o f a level playing field. 8 Tender evaluation and award criteria 2.55 Both sets o f regulations state that award should be based on quality and price, althoughthere i s no mention o f what quality elements may be taken into account. But for several reasons there i s an over- emphasis on prices. First, no criteria are mentioned and no mechanism introduced to enable the evaluation of non-price criteria. N o procedures are provided to the evaluation committee on the conduct o f the valuation, althoughthe works regulations do attempt a rudimentary set o f tables which might assist. 2.56 Second, despite the stated combination o f quality and price, the remaining provisions o f the regulations place emphasis on price, notably in the provisions relating to the live auction and the continuing price bidsinvited following the sealed biddingprocedure. Given the absence o f any significant qualification criteria and o f consistency intechnical specifications, it i s difficult to see how awards would be made on anything but lowest price. g) Submission, receipt, and opening of tenders 2.57 Bids are opened publicly (and the auction takes place) at a prescribed time and place. However, the procuring entity has the option to re-conduct bid opening if it i s not satisfied with the bids or, inthe case of the works regulations, not all bids have been received, which i s not acceptable. There are few provisions relating to the conduct o f bid opening, and other than in cases where the oral submission o f illiterates in the auction procedure i s reduced to writing, no requirement to record and keep information disclosed at the bidopening. There i s no requirement to keep information confidential. h) Complaint reviewprocedures 2.58 No complaint andremedy procedures are included inthe current legislation. 64 Indicator 2: Availability of Implementing Regulations and Documentation 2.59 The availability o f a series o f implementing procurement regulations, such as operational procedures, handbooks, model tender documentation, and standard conditions o f contract, is important for correct and consistent application o f the legislative and regulatory framework as well as for effective handling o f procurement operations. 2.60 None o f the following benchmarks have been achieved, inpart because o f the deficiencies in the legislation itself, and also because there is no organization (policy unit) which is able to provide the necessary assistance inthis area (see Pillar I1below): 0 Implementing regulation that provides defined processes and procedures not included in higher level legislation. 0 Model documents for goods, works, and services. 0 Procedures for evaluation o f tenders and award o f contracts. 0 Procedures for pre-qualification. 0 Procedures for contracting for consultant services, ifnot defined inhigh-level legislation. 0 User's Guide or manual for contracting entities. 0 General Conditions o f Contracts for public sector contracts covering goods, works and services consistent with national requirements and international requirements when applicable. 2.61 The State structures are subject to the national provisions, although in some cases such as mining and hydrocarbons there may be specific measures ineffect for SOEs. b. PillarI1 CentralInstitutionalFrameworkandCapacity - 2.62 Modernizing and maintaining a country's public procurement system i s an on-going and complex process. It benefits strongly from the existence o f a focal point within the Government administration with sufficient capacity and qualifications to manage the procurement system and the monitoring o f public procurement implementation. All bodies with legitimate interests in public procurement, such as the contracting entities, private sector entities, the Government, and the judicial system, will require central support o f varying nature and degree. 2.63 At present, a critical issue for the Government is the lack o f a central body to lead the effort on public procurement reform (i.e. oversee, manage, and support efficient implementation as well as provide leadership in modernizing and maintaining the public procurement system). The issue i s the lack o f any comprehensive initiative inrespect o f long-termreform and capacity development (see below). 2.64 While ARDS has been assigned the role o f managing the PA contract, it i s not designed as a central procurement policy unit. Furthermore, its functions and location are, to a large extent, incompatible with such a role, and it has insufficient capacity to take on such a role. Under the IDA- financed EPAP project, the PA has been starting to carry out activities (such as training, documentation preparation, drafting o f a new law, etc.) which would normally fall within the remit o f a central policy unit.Ithas beendoing so, however, only inthe absence o fany Government leadinthis sphere, andthis is not its proper function. 65 Indicator 3: Mainstreaming Procedures into Public Financia1Management 2.65 At a basic level, procurement is well integrated in the overall public finance management framework (for instance, adequate procurement documentation i s one o f the requirements o f financial controls). However, due to weaknesses both in the procurement system and in the overall PFM system, procurement processes are not fully integrated inthe PFM system. a) Procurementplanning and data on costing arepart of the budgetpreparation process 2.66 Budgetingand planning are functions which are not yet fully developed, with budget formulation remaining largely incremental for operating expenditures. Procurement officials state that requisitions are made to the accounting departments as and when necessary and that, if there is sufficient budget available, the requisitions will be approved. b) Budget law supports efective procurement, contract execution, andpayment 2.67 It appears that from a formal point o f view at least, procurement expenditures are subject to scrutiny ensuring that payments are made only when the correct documentation has been presented. There i s an external monitoring agent inplace for expenditure from the donor trust fund (ARTF) which monitors payroll and O&M expenditure before reimbursement by the ARTF. However, the monitoring i s confined largely to checking the quality and existence o f the appropriate documentation used for requesting money, advances, and payment. There i s no effort to monitor the effectiveness o f procurement, contract execution, and payment. Nevertheless, although control i s mechanical, there i s a control. 2.68 There has been significant improvement in the eligibility o f expenditures for ARTF reimbursement in the last few years. In 2003/04, some 75% o f expenditures were rejected by the Monitoring Agent. By the end o f 2004/05, that figure was closer to 10% (see above). Of the ineligibles, an estimated 30% were due to the procurement procedures. The MonitoringAgent has conducted some 12 workshops each year dealing with the relevant procedures but considers that the improvement has less to do with increased understanding and more to do with strong disincentives to manipulationbrought about bythe sanction o floss o ffbture budgets tiedto ineligible expenditures. 2.69 Similarly, the new Public Finance and Expenditure Management Law makes no mention o f procurement activities. c) Initiation of procurement operations is dependent on sufficient budget appropriations for the duration of the contract 2.70 Procurement expenditures will be sanctioned once accounting departments have verified the existence o f sufficient funds. There are, however, two limits to this rather strong control. First, there i s no commitment recording or control. Therefore, authorization to proceed with a purchase might create arrears if it i s later found that there i s not enough cash to settle the payment. Second, long-term or multi- annual projects are not taken into account properly in the system. However, most such projects are conducted usingdonor funds and are not subject to local laws and regulations. d) Contract execution is linked to budgetary controls 2.71 Payment (as opposed to contract execution) i s made based on the presentation o f the correct documentation and availability o f financial allotments. As such contract execution i s linked to budgetary controls. 66 Indicator 4: Functional Normative Body at the Center 2.72 The absence o f a functional normative body i s one o f the major issues raised by this analysis. N o such body exists at the central level. While the absence o f such a body does not necessarily mean that there i s no guidance for procurement reform, the clear lack o f ownership o f the Government for procurement reform means that the reform process has no direction and little substance. 2.73 Inaproperly functioningsystem, such abodywouldhavethe following features: 0 The status and basis for the functional normative body at the center (the Body) would be covered inthe legislative and regulatory framework. 0 The responsibilities o f the Body would address a well-defined set o f functions which would include, but would not be limited to: the provision o f advice to contracting entities; drafting amendments to the legislative and regulatory framework and implementing regulations; providing oversight and monitoring o f public procurement; providing procurement information; reporting on procurement to other parts o f Government; developing and supporting implementation o f initiatives for improvements in the public procurement system; and providing implementing tools and documents to support capacity development o f implementing staff. 0 Organization and staffing o f the Body and its level o f independence and authority (formal power) to exercise its duties would need to be determined in a trustful and credible manner and ina way that i s sufficient to implement the responsibilities. 0 Responsibilities o f the Body would provide for sufficient separation and clarity so as to avoid conflict o f interest. Indicator 5: Existence of Institutional Development Capacity 2.74 Overall, institutional development capacity i s weak inthe following four dimensions. a) A system exists for dissemination of procurement information, including tender invitations and tender result information 2.75 Tender invitations are published in national newspapers, broadcast on national radio, and sent directly to foreign missions in Kabul. However, there i s no central system for the collection, publication, and dissemination of such information. The PA i s responsible for disseminating tender invitations for the procurements handled by them, mostly donor funded. For new projects, subject to the Bank's latest Procurement Guidelines, the P A i s also publishingthe tender results as per the requirement. It is unclear whether information on other procurements, not handled by the PA, i s available to the general public, as there i s no central body or institution which has the responsibility for dissemination o f such information for all procurements by the Government. There i s no information website. However, the new National Procurement Law requires creation o f a procurement website by the Procurement Policy Unit which will ensure dissemination o f procurement information, including biddingopportunities, bidresults, etc., for all procurement by the Government. b) Training, advice and assistance is provided to develop the capacity of Government and private sectorparticipants to understand the rules and regulations and the how they should be implemented 67 in order for the system to provide the required service within the public expenditure management system 2.76 There i s no systematic plan to train procurement officials throughout the Government. The PA provided some limited training programs to procurement officials at the national level. In addition, it provided intensive formal and on-the-job training to 10 national staff - the Procurement Liaison Officers (PLOs) - who are intended to form the core procurement capacity within the Government. Currently, however, these PLOs are used for the purposes o f the Procurement Agent and are not integrated within the Government. The national training program reached up to 200 people, but this i s inadequate as a national training program. It i s unclear what has happened to this training program with the replacement o f the procurement agent. Giventhe lack o f central responsibility for the procurement reformprocess (see above), these initiatives are now well behindschedule. 2.77 One o f the tasks o f the PA i s to provide procurement assistance to line Ministries on request. This process implies, however, that it i s the procurement agent which conducts the procurement. This is, indeed, the way in which this facility i s being used inpractice. This process does not, therefore, develop the capacity o fthe Government (see also sectionbelow on capacity). c) Systems andprocedures existfor collection and monitoring of national procurement statistics 2.78 No such systemexists. d) Quality control standards are disseminated and used to evaluate per$ormance of staff and address capacity development issues 2.79 No such standards exist. However, there i s a planto develop performance standards as a basis for gradually transferring responsibility for procurement to the line Ministries against milestones achieved. At the outset, these would be based on completion o f training programs, but thereafter the performance standards would depend on performance, based on an attestation given by independent attesters (assisted by auditors) and usinga benchmarlungsystem. c. PillarI11 ProcurementOperationsandMarket Performance - 2.80 There are some encouraging signs that procurement operations are functioning and that the performance o f the market could be good enough. However, lack o f data makes the evaluation o f performance very sketchy and, by itself, suggests underdevelopment o fprocurementinAfghanistan. Indicator 6: EfJicient Procurement Operations and Practice 2.81 The regulations and processes could support effective procurement practices. However, capacity (both o f people and o f systems) severely limitsthe actual efficiency o f procurement operations. a) The level of procurement competence among procurement oflcials is consistent with their responsibilities 2.82 It is clear that most procurement officials are aware o f the existing rules and regulations applicable to their functions. There i s also an apparent willingness to apply those rules and regulations. The absence o f any clear central guidance on the interpretation o f these rules and regulations means, however, that procurement officials in different branches o f the Government tend to provide their own interpretations, leading to inconsistency and some instances o f questionable practices. The difficulties lie less with the competence and willingness o f procurement officials to apply procurement laws than with the inadequacy o f the rules themselves. Hierarchical approvals are just that, approvals. Decisions are often taken at high levels, with an overemphasis on price. b) The training and information programs implemented for procurement officials, as well as for private sector participants, are consistent with demand 2.83 There is n o evidence to suggest that, other than in the case o f the limited training programs initiated by the PA (see indicator 5b), there i s any coordinated approach to the dissemination o f information in relation to procurement or to the conduct o f any training programs whatsoever. This i s a neglected area o f activity. It i s also likely that this inaction i s due, at least in part, to the lack o f recognition o f public procurement as a core activity o f Government. This i s reflected in the failure o f the Government (line Ministries) to benefit from the improved abilities o f procurement specialists who have obtained some training through the A R D S program and in the more general the lack o f ownership o f the procurement reformprocess. 2.84 A further problem raised by the procurement officials themselves is the severe lack o f language slulls, which means not only that they are ill-equipped to conduct international competition but also that they are unable, in setting specifications, to take account o f standards and specifications which are adopted and usedinternationally. 2.85 There has been no training o f the private sector, resulting inits inability to respond to public calls for competition. Bothprocurement officials and representatives o f the private sector raisedthis as an issue which needs to be addressed urgently. 2.86 Unusually, procurement training has been taking place in the Ministry o f Defense, with the assistance o f the U S military. With support from independent consultants, it i s expected that up to 500 procurement officers will be trained before the end o f 2006. This is also being done in cooperation with the US'S Defense Acquisition University (DAU), which will ensure recognition o f the qualifications achieved (DAUlevel 4 equivalent). e) The existence of administrative systems for public procurement operations, and information databases to support monitoring of peq4ormance and reporting to and responding to the information needs of other related Government systems 2.87 There i s no formal system inplace to report procurement activities and, as a result, no system for monitoring performance in respect o f public procurement activities. On the other hand, requests and authorizations are prepared inwriting and kept in files, suggesting that there i s at least a paper trail which, with adequate provisions and implementation, could be turned into a record-keepingand reporting system capable o f permitting monitoring functions to take place. d) The existence and implementation of internal control mechanisms for the undertaking of procurement operations at the contracting level, including a code of conduct, separation of responsibilities as a checWbalance mechanism and oversight/control of signature/approval authority 2.88 There are few internal control mechanisms in place. The only consistent reference i s to ex ante budgetary control through the accounting departments o f the line Ministries where there i s usually a 69 representative o f the MoF.' There is no suggestion, however, that this control considers the procurement operations conducted. It appears to be restricted to financial control. 2.89 Procurement contracts are signed by the Minister or Deputy Minister depending on thresholds. Ministers have the authority to approve and sign contracts up to a value o f Afs 20 million inthe case o f the development budget and Afs 7 million inthe case o f the operational budget.6 Above these thresholds, contracts are approvedby the President. There i s no suggestion, however, that this provides any technical control over the procedures used. There i s also an approval committee in MoF. The danger with such a committee, at the present time, i s that its role and functions appear to cut across the attempt to impose accountability on line Ministries and may confuse the roles o fpurchaser and regulator. Indicator 7: Functionality of thePublic Procurement Market 2.90 Despite widespread concerns about insufficient numbers o f bidders, there are a number o f producers and suppliers in the private sector who could help make the public procurement market work. However, their capacity to follow regulations and submit quality bids i s weak. In addition, the private sector has very limited confidence inthe system. a) The public procurement market in the countly has sufficient numbers of producershppliers to providefor competition and comparative pricingfor goods, works, and services when compared to prices paidfor similar requirements in local, regional, or international markets as applicable 2.91 There i s very little data on which to base the assessment o f comparative price, although there i s little to suggest that the prices achieved by public purchase are radically different from those obtained p r i ~ a t e l yHowever, there i s one significant exception (see. indicator Id). It would appear that, where . ~ there i s an SOE which offers specific products for sale that have the same quality as those offered by private suppliers, Government purchasers are obliged to purchase from the SOE. This appears to happen even where the prices offered are higher than those obtained inthe otherwise private market place (up to 20% higher). 2.92 On the other hand, it appears that there i s a sufficient number o f domestic companies to provide adequate competition for a number o f products (see also Section G). Purchasing departments report consistently high numbers of bidders for most large-value contracts, with numbers reaching up to 50 in the case o f works maintenance contracts. The tendering base i s made up not only o f domestic tenderers but also returning Afghans, predominantly from Pakistan. These bring with them capital for investment and some know-how. There i s significant technical expertise in some areas, and large returning Afghan companies report no lack o f qualified sub-contractors. However, local tenderers face a series o f structural difficulties. They are largely unable to offer volume, standardized products as local manufacture to the extent it exists (such as sash or furniture factories) is fragmented. Inputs are largely imported. It was reported that there i s some local pipe manufacturing and some quarrying, but little else. This goes for most basic materials (cement, steel, etc) as well as virtually all equipment. The situation i s exacerbated by a lack o f product standardization and incomplete or non-existent technical specifications. There i s little formal structure to domestic contracting or materials supply markets. Industry associations are limited, althoughthere i s a chamber o f commerce. This line o f reportingi s expectedto be changed, see Part I. 'Thesecaseof thresholdshaverecentlybeenincreased. The the highwaysector is discussedinVolume IV. Chapter 5. 70 b) The quality and consistency of submissions by the private sector to the Government enables the application of theprocess leading to the award and completion of contracts 2.93 The problems with the market would appear to be qualitative rather than quantitative: local tenderers face difficulties inmeeting the formal (procedural) requirements o f the biddingprocess. There are a number o f reasons. First, there has been insufficient training o f the private sector in terms o f operating within a system o f regulated procurement. Training i s certainly required inthis respect interms o f completing bid forms and preparingtenders. Second, it i s clear that procurement officials are unable to adequately set or evaluate technical specifications, qualification criteria, or award criteria. A third issue is the lack o fproduct standardization as noted above. 2.94 While it would appear, therefore, that contracts are awarded according to the (outdated) process currently applied and are completed without too muchdifficulty, the quality o f the requirements set by the public purchaser i s fundamentally deficient, leadingto unsatisfactory results overall. e) Theparticipation rate of tenderers in response to competitive tender invitations is consistent with the capacity of the marketplace 2.95 There i s no lack o f bidders responding to calls for competition. The difficulties lie elsewhere. The private sector's lack o f confidence in the legal and regulatory fkamework governing the procurement process i s evidenced by their submission o f complaints or protests, through an informal mechanism described consistently by many procurement officials (although none could identify any case which had been filed, let alone any case which had been successful - see below on appeals and complaints mechanisms). Indicator 8: Existence of Contract Administration and Dispute Resolution Provisions 2.96 While there are contract administration provisions, they need to be made more explicit, notably by developing more detailed implementation regulations (which should be done once the new law is approved). There i s a lack dispute resolutionmechanisms. a) Procedures are defined for undertaking contract administration responsibilities to include inspection and acceptanceprocedures, quality controlprocedures, and methods to review and issue contract amendments in a timely manner 2.97 Inthe caseofworks, these procedures are contained inan annex to the regulationswhich apply to works contracts. While they are not complete, they are nonetheless based on acceptable provisions found elsewhere. In the case o f goods, inspection and acceptance provisions are contained in Purchase Regulation. b) Dispute resolution procedures are included in the contract documentprovidingfor an efficient and fair process to resolve disputes arising during thepe$ormance of the contract 2.98 Given the absence o f any standard form contracts in the case o f goods, there are n o such provisions. In the case o f works, the standard form contract contained in the regulations themselves includes a very brief dispute resolution clause which calls for amicable settlement in the first place followed by recourse to the courts. No attempt i s made to provide a timely and efficient alternative to the court systemwhich i s considered to be slow and unreliable. e) Procedures exist to enforce the outcome of the dispute resolutionprocess 71 2.99 There i s n o alternative system to the court procedure and, consequently, no other outcome to enforce. The degree o f enforcement o f court decisions i s unknown, but presumed very low. d. Pillar IV The Integrityof the PublicProcurementSystem - 2.100 The overall framework o f control and audit i s largely centered on financial issues, with little on the performance of procurement activities. The latter should be developed over time, but without developing such controls and audits as a way to second-guess systematically all decisions or to penalize weak capacity. The problems o f the anti-corruption commission highlight the risk o f an excessive focus on prosecution. This issue i s compounded by lack o f publicly available information on public procurement and the absence o f an appeal mechanism. Indicator 9: Effective Control and Audit System 2.101 There i s an effective control and audit system, but it focuses on financial matters. a) A legalframework, organization, policy, andprocedures for internal and external control and audit of public procurement operations exists and operates toprovide afunctioning controlframework 2.102 At present, there is a legalrequirement that any procurement be audited by the Auditor-General's Office (AGO). The Auditor General's Office has the theoretical power and duty to conduct external audit o f the procurement function but has not yet done so other than, ina limited sense, inrespect o f externally funded contracts. These are currently ex-post controls, with the reports communicated to the President (see Part I). 2.103 There are a number o f audit activities, including at the provincial level,' but they mainly focus on financial controls. No procurement-specific audits have been conducted at the national level. The AGO i s given the power o f auditing the regularity o f procurement and intends to do so by way o f sampling (proposed at 10% o f contracts, although it i s not yet clear whether they intend to do this by value or by number). During interviews conducted in 2004 with the AGO, it was apparent that there is a lack o f capacity within the AGO to conduct audits o f procurement. Current training initiatives are concentrated on core audit functions and do not contain any training targeted specifically at procurement compliance audits. During interviews, it was also clear that there was no understanding o f the difference between compliance andpe$ormance audits and that, were audits to be conducted, it i s likely that attempts would be made to second-guess the decisions o f procurement officials. This would imply failing to carry out a relatively straightforwardcompliance audit and replacing it with a corrupted performance audit. b) Enforcement andfollow-up onfindings and recommendations of the controlframework provide an environment thatfosters compliance 2.104 There has been some follow-up to financial audits.' Since there are n o procurement-specific audits, the issue o f follow-up i s moot.. e) The internal control system provides timely information on compliance to enable management action See Volume 111, Chapter 3. See Part Iof this Volume. 72 2.105 The exceptional mechanisms set in place by the ARTF (Monitoring Agent) provide timely information on compliance. However, this i s not part o f the Government's own system, and the internal control system itselfdoes not generate such information. d) The internal control systems are sufficiently defined to enableperformance audits to be conducted 2.106 Currently, it i s unlikely that even compliance audits are capable o fbeing conducted. e) Auditors are sufficiently informed about procurement requirements and control systems to conduct quality audits that contribute to compliance 2.107 There i s a significant lack o f capacity inthis respect. Indicator 10: EfJicienq of Appeals Mechanism 2.108 There i s no formal complaints mechanism provided for in the existing rules and regulations. It i s reported consistently by those interviewed that there i s an informal mechanism which consists o f a first- level complaint to the procuring entity which may set up an investigation committee to hear the complaint, followed by a second- level complaint to the "Government". Insome cases, it was said to be a complaint to the auditor general whose office had set up a specific committee. This could be followed by a third-level complaint to the attorney-general, although it was not clear whether this could be independent o f the decision o f the auditor general or whether it was in the form o f an appeal from that decision. In any event, this informal mechanism would appear to be hypothetical: none o f those interviewedcould identify any instance o fa complainthaving been filed or progressed. 2.109 A proper appeals mechanism would have the following properties: 0 The existence and operation o f a complaint review system that gives participants inthe public procurement system a right to file a complaint within the framework o f an administrative and judicial review procedure. 0 The decisions are deliberated on the basis o f available information and the final ruling can be made by a body withjudicial capacity under the law. 0 The complaint review system has the capacity to handle lodged complaints efficiently and a means to enforce the remedy imposed. 0 The system operates in a fair manner, with outcomes o f decisions balanced and justified on the basis o f available information. 0 Decisions are published and made available to the public. 0 The administrative review body i s separate fiom the regulatory body. Indicator 11:Degree of Access to Information a) Access to information is supported by publication and distribution of information through available media with supportfrom information technology whenfeasible 2.1 10 There i s an official Government gazette which contains information regarding laws, including procurement, but its availability and distribution i s unknown. In the case o f procurement, there i s no information website or other means o f accessing information related to procurement. Procurement officials have copies of the rules and regulations but they are otherwise difficult to come by. It i s not known whether they are available to the general public. 73 b) Systems exist to collect key data related to performance of the procurement system and to report regularly 2.111 No such systems exist. e) Records are maintained to validate data 2.112 Procurement officials keep a file o f the paper requisitions, etc. There appears to be no formal record-keeping requirement, however. Indicator 12:Ethics andAnti-corruption Measures 2.113 The procurement system should be perceived to operate with integrity, providing for clear definitions o f unacceptable practices and stating the consequences for participants in the procurement system who engage in fraudulent, corrupt, or unethical behavior. a) The legal and regulatory framework for procurement, including tender and contract documents, includes provisions addressing the issue of corruption, fraud, conflict of interest and unethical behavior and states actions which can be taken with regard to such behavior (either directly or by reference to other laws). 2.114 There are no such provisions in the existing legal and regulatory ffamework. However, the new Procurement Law will provide a strong basis inthis area. b) Special measures by the Government toprevent and detectpotentialfraud and corruption in public procurement is addressed in an anti-corruption program 2.1 15 Although not specifically referred to, corrupt practices in the context o f public procurement would fall within the ambit o f recent anti-corruption initiatives. These, however, appear to be inadequate. A new anti-corruption law was introduced in 2004 and i s implemented by a senior official in the President's office rather than by a commission or task force. The implementing office was set up only toward the middle o f 2004 and suffers from lack o f staff and facilities. For the purposes o f procurement, there i s a serious issue with the law in that it does not distinguishbetween corrupt procurement practices and mistakeshegligence in applying the procurement law. Any alleged breach o f procurement and/or other laws would be perfunctorily investigated and, regardless o f hard evidence o f corruption (a breach gives rise to a barely rebuttable presumption o f corruption), would be sent to the attorney-general for prosecution and the courts would be expected (without guidance) to make the appropriate decision. Despite the stated number o f corruption cases before it (60), none were specifically related to procurement. When cases were cited, it was clear even ffom a first impression that these were cases where a breach o f the rules had taken place (at least as a result of incompetence) but where the charge o f corruption lackedevidence. e) Stakeholders (private sector and civil society) take active measures to support the creation of a procurement market knownfor its integrity and ethical behaviors 2.1 16 There i s no evidence that this i s the case. 74 E. ProcurementunderExternalProjectAssistancethroughthe CoreBudget 2.1 17 IDA and ARTF use the services o f the PA (see above). Other donors are increasingly using the same approach, but with variations. For instance, ADB does not apply its normal procurement guidelines to program loans and requires only that procedures generally adhere to principles o f economy and efficiency. The only specific requirements are compliance to country eligibility (procurement must originate in a member country) and compliance with negative lists. When usingthe PA, the arrangements are project-specific. So far, for the procurement related to the roads sector, the PA at ARDS advertises the SPN, issues bidding documents, undertakes bid opening, and evaluates the proposals along with a team from ministry, prepares the evaluation report, and submits to ADB on behalf o f the ministry. The same process i s followed for the Ministry o f Civil Aviation. In the case o f energy projects (power transmission), the plan i s for the P A to do everything except for evaluation o f proposals, which will be handledby staff from Ministry o f Energy and Water. F.AddressingStructuralWeaknesses DEVELOPING CAPACITY INTHE GOVERNMENT 2.118 Duringits mandate, the original PA, through ARDS, trained more than 100 procurement staff from various ministries. The training was conducted in a three-stage sequential process. Stage 1 represents the most basic level. Stage 2 begins to look at specific procurement rules for goods and works. Stage 3 focused on the process o f selection o f consultants. Participants received a certificate o f completion, but this certificate i s not a h n to a qualification. The training appears to have been based on World Bank rules rather than on the national system. 2.119 In addition to the more general procurement training program for line ministry staff, 10 individuals were selected for intensive procurement training which included on-the-job training and mentoring with the PA's staff. These 10 individuals are intended to provide core Government procurement capacity which will ultimately assist the line Ministries inbuildingtheir own capacity. They are known as the Procurement Liaison Officers (PLOs). Currently, the plan is to retain them within A R D S inthe short term in order to provide centralized procurement and then to place a number o f them inthe procurement unitsto be set upinthe key line Ministriespursuant to the new Procurement Law (see below). Their current salaries are significantly higher than those available inthe line Ministries. Training of the PLOsbegan inJanuary 2004 and has resulted ina number o f well-trained and competent PLOs. 2.120 Besides the availability o f central procurement facilitation services under the ARDS, some of the line ministries have established Program Implementation Units (PIUS), which include international procurement consultants, and all procurements under these line ministries are carried out by its internal experts, involving the Ministries' own procurement staff. This i s a possible short-term solution; however, sustainable procurement capacity building will be addressed through the Public Administration Capacity Building Project. This project aims to develop and implement (i)a sustainable nationwide capacity buildingstrategy to take into account the requirements not only o f central government but also provincial and municipal government as well as the private sector; (ii)a program to assist and facilitate the achievement o f PRR status for a number o f key Ministries to be identified; (iii)a plan to transfer procurement capacity and responsibility to line Ministries according to a set o f milestones; and (iv) a system o f attestation whereby the performance o f Ministries can be assessed against identified benchmarks as part of the milestone mechanism. 75 LEGAL FRAMEWORK 2.121 A new National Procurement Law has recently been approved. The purpose o f this law is to establish an open, transparent, competitive procurement system, based on effective budgetary and expenditure controls and reporting requirements designed to achieve efficiency, economy, the prevention o f abuses, and a fair opportunity for participation by all potential contractors, including small enterprises and individuals. In terms o f policy, the law has three main features: (i) the creation o f a Procurement Policy Unit and a review mechanism; (ii) the introduction o f a common organization o f the procurement function in all Ministries and Government departments in Afghanistan; and (iii) the introduction o f procurement methods and procedures in line with the international standards. The law introduces a number o fnew approaches and processes; hence its implementationwill require significant training. G.Private Sector Issues 2.122 Surprisingly, a number o f contractors and suppliers o f varying capability, including some large companies, can be found implementing works in Afghanistan. Although not well organized and facing constraints related to availability o f materials, transportation limitations, power outages, and security concerns, they provide a good base for the implementationo fpublic works and reconstruction. 2.123 Private sector firms include foreign contractors capable o f implementing large contracts; some are implementingcontracts exceeding $50 million annually. Many o f these companies are owned or managed by Afghans returning from Pakistan, the US, and other countries. Others include Turlush and Palustani companies seeking opportunities in the Afghan reconstruction boom. Collectively, these firms bring capital, technology, equipment, and skdls. 2.124 Inaddition, there are a significant number ofdomestic companies offering arange o fconstruction services and products. This i s shown in the large number o f companies competing for smaller contracts, reaching up to 50 inthe case o f a recent biddingo f works maintenance contracts. These companies offer technical expertise in a number o f areas, and the larger firms (returning Afghan companies and others) are able to draw upon them as a source o f qualified sub-contractors. 2.125 Three organizations support the structuring o f the local contracting industry. Afghan Builders' Association (ABA) i s an industry association with a current membership o f 160 companies working inall aspects o fthe construction sector. Membership fees are $200 per year. It has established links to other international organizations (such as the Turlush Builders' Association) and intends to expand these. Although it i s an association o f Afghan companies only, it recognizes the business importance o fjoint venturing and subcontracting with foreign firms. ABA acts primarily as an industry liaison and lobby group but i s also involved in capacity building o f its members." It maintains a database o f its members and classifies them according to capability (see Table 2.4). ABA's planned future activities include establishment o f an equipment pool and materials laboratory. lo Capacity building focuses on training for members employees. These include short courses in engineering and management done in cooperation with Kabul University Engineering school and a trades training and certification program consisting of six months of training (classroomtheory plus on thejob training) incarpentry, plumbing,etc. 76 Table 2.4: ABA Registration and Classification Category Number of Firms Annual CategoryA 20 Large (more than $50 million) CategoryB 32 Medium CategoryC 108 Small Afghan Investment Support Agency (AISA), an investment department under the auspices o f the Ministry o f Commerce, has taken over the company registration system. Approximately 8000 companies were registered by the Ministryo f Commerce, and it i s an objective o f AISA to reduce such registrations to a realistic number. Highregistration fees are expected to help; they have been set at $700 for small, $1,500 for midsized, and $2,000 for large companies. The Afghanistan Chamber of Commerce (ACC) has been operating for eight years. It has 18,000 members, about 25% o f which are foreign. ACC i s a member o f the International Chamber o f Commerce, Islamic Chamber o f Commerce, and ECO (a regional economic development organization that works in neighboring countries such as Iran and Palustan as well as Central Asia). ACC fulfills an industry development role, assisting its members through seminars and information dissemination. It also represents its members vis-a-vis the Government on issues o f taxation, importation o f goods, and development o f commercial regulations. 2.126 Contractors and suppliers face a series of constraints and structural limitations. Suppliers are unable to offer volume, standardized product as there i s no established manufacturing capacity in areas relatedto the construction industry(plumbingfixtures, electrical materials, sash or furniture fabrications). This applies also to basic materials such as cement and steel, and to construction equipment. Contractors must therefore rely largely on imported materials, goods, and equipment. Add to this the difficult circumstances o f Afghanistan related to security, transportation, and inadequate infrastructure, and the construction environment becomes quite difficult. The following summarizes the concerns reportedby the private sector: The reliance on imported goods introduces uncertainty o f supply related to unreliable importers, transportation constraints, and importation requirements. Even when imported, there i s a general problem in obtaining quality materials. Cement from Pakistan and Iranwere cited as examples. There i s a need for local testing facilities. For locally produced goods and materials there are concerns about inconsistent quality. The situation i s exacerbated by lack o f product standardization and incomplete or non-existent technical specifications. Trades skills are available locally, but that there i s a shortage o f management capability and inconsistency o f technical skills. There i s a need for certification o f trades. Government procedures are seen as outdated and in need of modernization, especially with respect to payment procedures. Some believe the Government i s too focused on price, which i s resulting inbelow-cost bids and the attendant results o f low quality. There is no suitable bid challenge procedure. Bribery exists, and there is no accepted formal process for lodging complaints. 2.127 In addition to the external constraints identified by the private sector, it also faces internal limitations, especially the smaller locally owned and operated companies. Although there are usually sufficient numbers o f bidders involved in procurement, their ability to follow regulations and submit 77 quality bids i s weak. The difficulties relate mainly to compliance with procedural requirements. This i s exacerbated by lack o f confidence inthe public procurement system. 2.128 Not much has happened so far in terms o f training the private sector. In January 2004, the Ministry of Commerce facilitated a meeting with the Afghan business community to identify ways in which Afghan participation in donor fundedprocurement could be enhanced. Little has been done since. 2.129 The following recommendations are put forward to address the constraints facing the private sector and its internal capacity limitations (see the previous section on the public sector side). These recommendations could feed into a curriculum to be developed by the new Procurement Policy Unit and contracted out to firms or institutions specialized indelivering training. 0 Development o f general business capabilities, to help smaller local contractors with the basics o f managing a company. 0 Training private sector firms to function within a system o f regulated procurement - preparing bids, pricing works, implementing contracts, managing variations, payment procedures, etc. This would include training in Government procurement practices as well as ininternationalbestpractices andthose ofthe mainmultilateralandbilateraldonors. 0 Assistance to develop a system o f trades training and accreditation to promote both availability and consistency o f shlls inthe private sector. 0 Assistance to develop capability within Afghanistan to secure a reliable supply o f inputs (primarily construction materials) and to assess the quality o f those inputs.This could include review and possible revision to importation procedures, establishment o f a national testing laboratory, and information dissemination on availability, price, and quality o f products. H.ConclusionsandRecommendedActions 2.130 The overriding objective o f a national public procurement system i s to deliver efficiency and value for money in the use o f public funds, while adhering to fundamental principles o f non- discrimination, equal treatment, and transparency. Procurement i s therefore at the core o f the Public Finance Management (PFM) system and contributes greatly to several o f its objectives, including efficiency, transparency, and accountability. 2.13 1 This assessment suggests that the rules before the new Procurement L a w were not providing an acceptable framework for public procurement in Afghanistan. While fiduciary standards were reasonable (though inneed o f improvement), procurement focused excessively on prices - at the expense o f quality - and absence o f a number o f monitoring, complaint, and appeal mechanisms weakened overall performance. This means that the donor community, despite its commitment to rely increasingly on national systems, was not be able to rely on the national system for procurement o f goods, works, and services funded by them. Regardless o f the position o f the donors, the failure o f the current rules to provide a kamework which will ensure efficiency and value for money was o f serious concern to the Government. With scarce resources, it i s imperative to ensure that resources available are spent to the greatest advantage. The new Procurement Law i s a significant step toward a modem procurement system. 2.132 Moving forward, the key issues can be summarized as follows: 0 There i s currently n o procurement champion at any level in Afghanistan. This i s a serious impediment to reform and to inter-Ministerial dialogue. Ways must be found to create 78 interest inthis reform at the highest political level as well as at the operational level. The creation o f a Procurement Policy Unit (PPU) is therefore a very urgent priority. 0 There i s also clearly a lack o f ownership inprocurement. Ina sense, this was inevitable in the early stages o f the reconstruction process. The early stages are now over, however, and thought needs to be given to ways o f creating local ownership. A way must be found o f moving the procurement function from ARDS and external agents to the Government itself. This i s a question o f capacity transfer but also a question o f creating an appropriate environment for such a transfer inthe line Ministries. 0 There are a number o f obstacles to such ownership, for example: (i) lack o f adequate training (inboth the public and private sectors); (ii) misunderstanding o f the scope o f the procurement function (i.e. the belief that only major works or purchases are "procurement"); (iii) inappropriate regulations; (iv) lack o f common internal structures to deal with procurement in the line Ministries; (v) lack o f communication between trained procurement staff and senior administrators; and (vi) no clear incentives for improvement and ownership. 0 One critical aspect o f the lack o f capacity inthe line Ministries i s their inability to define and communicate effectively their desired technical specifications in procurement conducted with the assistance of the PA. This needs to be addressed as a matter o f urgency, whether coordination continues at the level o f the procurement agent or whether responsibility for procurement i s transferred to the line Ministries. 0 The new law sets out a system o f review by a Review Committee chosen from a Standing Review Panel to be established and maintained by the director of the PPU. Again, the desirability o f a review mechanisms i s not in question. Care needs to be taken, however, that the solution chosen i s both appropriate and workable. For any such system to succeed, it i s critical to identify a procurement champion in the Government capable o f making it work. 2.133 The priority actions should thus be as follows: 0 Create a Procurement Policy Unit; 0 Finalizeregulations to implement the Procurement Law; 0 Develop and implement a large-scale capacity buildingprogram for procurement officers inthe Government; and 0 Create a training program for the local private sector to enhance its capacity to participate to bids. 79 REFERENCES Evans, Anne, Nick Manning, Yasin Osmani, Anne Tully, andAndrew Wilder (2004). A Guide to Government in Afghanistan. Washington, DC andKabulAfghanistan:The WorldBankandAfghanistan ResearchandEvaluationUnit. PEFA Secretariat (2005). "Public Financial Management Pe$ormance Measurement Framework." Washington, DC: WorldBank (June). 80