Page 1 THE WORLD BANK Tajikistan: Public Expenditure and Financial Accountability Assessment Performance report June 2007 41121 Page 2 2 Table of Contents Abbreviations..................................................................................................................................3 Summary Assessment......................................................................................................................4 1. Introduction...............................................................................................................................12 2. Country Background .................................................................................................................14 2.1 Description of the Country Context .........................................................................................14 2.2 Description of budgetary outcomes..........................................................................................18 2.3 Legal and institutional issues and ongoing PFM reforms .........................................................19 3. PEFA Indicators........................................................................................................................23 A. PFM-OUT-TURNS: Credibility of the budget..........................................................................23 B. KEY CROSS-CUTTING ISSUES: Comprehensiveness and Transparency...............................28 C. BUDGET CYCLE....................................................................................................................39 D. Donor Practices........................................................................................................................73 4. Government Reform Process.....................................................................................................76 Annex 1 List of documents used....................................................................................................79 Annex 2 Tables for Indicator PI-2 .................................................................................................80 Annex 3: Organizational Chart, Ministry of Finance......................................................................82 Page 3 3 Abbreviations ACU Aid Coordination Unit AER Aid Effectiveness Review BO Budget Organization CS Civil Service CSIP Centralized State Investment Program EBF Extra Budgetary Funds GBAO Gorno-Badakhshan Autonomous oblast GOT Government of Tajikistan ICAD Internal Control and Audit Department IDF International Development Fund IGFR Intergovernmental Fiscal Relations INTOSAI International Organization of Supreme Audit Institutions KBO Key Budget Organization KRU Kontrolno-Revisionnoe Upravlenie (Office for Control and Revision) MDAs Ministries, Departments, Agencies MOF Ministry of Finance MOET Ministry of Economy and Trade MLSP Ministry of Labour and Social Protection MSRD Ministry of State Revenue and Duties MTEF Medium-Term Expenditure Framework NDS National Development Strategy PA Procurement Agency (MOET) PARS Public Administration Reform Strategy PEFA PMF PEFA Performance Measurement Framework PET(S) Public Expenditure Tracking (Survey) PFM Public Financial Management PFM-PR Public Financial Management Performance Report PIP Public Investment Program PITAPU Public Investment and Technical Assistance Program Unit (MOET) PRSP Poverty Reduction Strategy ROT Republic of Tajikistan RRS Rayons of Republican Subordination SCFCFC State Committee on Financial Control and Fighting Corruption SCIMSP State Committee on Investment and Management of State Property SFCC State Financial Control Committee SN Sub-national SPF Social Protection Fund TIN Taxpayer Identification Number UNCITRAL United Nations Commission on International Trade Law Page 4 4 Summary Assessment 1 This assessment is based on work undertaken between October 2006 and May 2007. The summary assessment covers the following three areas: 1. an integrated assessment of PFM performance, based on the 28+3 PEFA indicators, 2. an assessment of the impact of PFM weaknesses, and 3. the prospects for reform planning and implementation. The main report consists of four sections: 1. introduction, 2. country background (including a summary of fiscal performance), 3. a detailed discussion of Tajikistan’s performance against the PEFA indicators, and 4. a discussion of the government reform process. The PEFA assessment is based on an internationally agreed methodology to assess the public financial management systems of countries. It defines a set of indicators that guide an objective and standardized assessment. The PEFA methodology has been in use since 2005 and has thus far been applied to more than 60 countries. 2 Each indicator is rated from A to D, based on one or several dimensions and a precise definition of the criteria to be applied. Ratings for Tajikistan are based on four sources: a) primary fiscal data, laws and regulations of Tajikistan; b) information collected during missions in Tajikistan, involving two international and three local consultants; c) a self- assessment report by the GoT based on the PEFA methodology; and d) other existing reports. As this report reflects, the GoT has sought to improve Tajikistan’s public financial management; and this is resulting in good performance in a number of areas covered by the assessment. At the same time, the reform process is ongoing, and as discussed in sections 2 and 4, a range of further reforms are currently under discussion or at various stages of planning and implementation. The GoT is showing commendable commitment to a challenging PFM reform agenda. If reforms are well managed and pursued consistently, they will lead to improved PEFA scores in future assessments. It is also hoped that the current PEFA assessment report can offer some guidance – both to the GoT and to Tajikistan’s development partners – in further shaping the PFM reform agenda. 1 The Assessment has been carried out by the team of consultants (Ms. Verena Fritz, Team Leader, and Mr. Des Smallmann) from the Overseas Development Institute, United Kingdom, within the overall framework of the World Bank’s Tajikistan Programmatic Public Expenditure Review (PPER). Messrs. Hassan Aliev, Amirjon Ubaydulloev and Alijon Davlatov have provided local consultancy support. From SECO, which provided financing for the task, Mr. Martin Giesiger has been the coordinator. Overall oversight and task management was performed by the World Bank (Mr. Sudharshan Canagarajah – Team Leader for the PPER, Ms. Pascale De Lettenhove and Mr. Aziz Khaidarov managed the task from the WB Headquarters and Tajikistan Country Office respectively). 2 Further information, including the methodology in Russian, can be found at: www.pefa.org. Page 5 5 Integrated assessment of PFM performance The Public Expenditure and Financial Accountability Performance Assessment results are summarized in table 1. Table 1: Overview of Assessment Results Indicator Issue covered Rating A. PFM Out-Turns: Budget Credibility PI 1 Aggregate expenditure outturn B PI 2 Composition of expenditure out-turn C PI 3 Aggregate revenue out-turn A PI-4 Stock and monitoring of expenditure payment arrears C+ B. Key Cross-Cutting Issues: Comprehensiveness and Transparency PI-5 Classification of the budget D PI-6 Comprehensiveness of information included in budget documentation A PI-7 Extent of unreported government operations C+ PI-8 Transparency of inter-governmental fiscal relations B PI-9 Oversight of aggregate fiscal risk from other public sector entities C PI-10 Public access to fiscal information D C. Budget Cycle (i) Policy-Based Budgeting PI-11 Orderliness and participation in annual budget process B PI-12 Multi-year perspective in fiscal planning, expenditure policy, and budgeting D+ (ii) Predictability and Control in Budget Execution PI-13 Transparency of taxpayer obligations and liabilities C PI-14 Effectiveness of measures for taxpayer registration and tax assessment D+ PI-15 Effectiveness in collection of tax payments Not assessable (data not available for dim i) dim ii: B, dim iii: D PI-16 Predictability in the availability of funds for commitment of expenditures D+ PI-17 Recording and management of cash balances, debt and guarantees C+ PI-18 Effectiveness of payroll controls D+ PI-19 Competition, value for money and controls in procurement C PI-20 Effectiveness of internal controls for non-salary expenditure C+ PI-21 Effectiveness of internal audit D+ (iii) Accounting, Recording and Reporting PI-22 Timeliness and regularity of accounts reconciliation B PI-23 Availability of information on resources received by service delivery units C PI-24 Quality and timeliness of in-year budget reports C+ PI-25 Quality and timeliness of financial statements D+ (iv) External Scrutiny and Audit PI-26 Scope, nature, and follow-up of external audit D+ PI-27 Legislative scrutiny of the annual budget law C PI-28 Legislative scrutiny of external audit reports D D. Donor Practices D 1 Predictability of direct budget support D+ D 2 Financial information provided by donors for budget and reporting on project and program aid D+ D 3 Proportion of aid that is managed by use of national procedures D Page 6 6 C redibility of the budget (PI-1 to PI-4) In the period 2003 to 2006, the GOT has based its budget planning on conservative revenue forecasts. As reflected in indicator PI-3, this is generally positive, but it has led to a situation in which the budget outturn has differed considerably from budget plans. This discrepancy has declined but is still significant . Treatment of ‘own revenues’ of budget organizations has not been consistent, leading to discrepancies in the total amount of revenue reported. Furthermore, the actual composition of expenditure continues to differ from the composition in the budget plan. Given the tendency to exceed revenues it is important to increase the transparency of processes for allocating additional expenditures; including a more consistent involvement of the national parliament. This may also require more precise legal rules. All revenues and expenditures should be planned and discussed during the budget formulation process, restricting the use of supplementary budgets within-year to unexpected emergency requirements. Payment arrears have been declining and are currently at low levels; but the information that is recorded does not provide a clear distinction between payments due and arrears. Comprehensiveness and transparency (PI-5 to PI-10) The government has undertaken initial steps to improve the comprehensiveness and transparency of the budget; but further improvements are needed. The GOT has introduced functional and economic budget classifications that are generally in line with the 1986 GFS standard. However, an administrative classification is still missing; and until 2006, PIP resources have not been included in the functional breakdown. The Social Protection Fund has been integrated more closely with the budget in late 2006, but its funds are still managed through a separate bank account not integrated with the treasury. Sub-national governments are covered by the treasury system, and the central government monitors their fiscal situation. Reporting, monitoring and oversight of the fiscal risks from SOEs is weak; this is a major concern with regard to large enterprises and in the context of still sizable quasi-fiscal activities. The set of information included in the annual budget submitted for parliamentary scrutiny is relatively comprehensive, but lacks analytical material on new policy initiatives and their budgetary implications. Information on tax-expenditures (that is, the cost of tax exemptions) is not included in budget documents. There is some public access to fiscal information, but the information provided tends to be partial and incomplete. Accessibility of information should be improved further. Policy-based budgeting (PI-11 and -12) Some steps towards establishing more policy-based budgeting have been made and a renewed effort at establishing an MTEF has recently been initiated. The annual budget process is broadly orderly, in terms of following key dates set in the Law on State Finances; but there are still issues concerning the more detailed organization of the process and related internal deadlines. Furthermore, the institutional structures and systems for linking policies and budgets are still weak. Cabinet-level discussions of the budget and of key priorities and trade-offs to be reflected in fiscal planning are not well developed. A cabinet-level Budget Commission has been established in 2004, but its impact has remained marginal. An initial set of sector strategies has been elaborated; but thus far, only the education sector strategy is at a more advanced stage. Linking sector policy goals to realistic costing estimates will require further work for a number of years. Capital budgeting is fragmented, and needs to be further integrated with recurrent budgets. The high share of capital spending in total expenditures means that discretionary resources are a substantial share of the budget, and hence underlines the need for policy-based budgeting. Since a large part of capital spending is financed by development partners, it also underlines the need for a more effective approach to aid management on the side of the GoT, and to harmonization and alignment with national priorities on the side of donors. Some steps in this direction have been undertaken in 2007 in terms of budget presentation. However, institutionally, capital budget planning remains Page 7 7 s eparate and sub-divided into domestically financed and ext ernally financed ‘streams’, and aid policy and aid management structures continue to require greater attention. Predictability and control in budget execution (PI-13 to PI-21) A basic degree of predictability and control in aggregate budget execution has been achieved, but there are still systemic weaknesses. A core Treasury system has been established; and the government plans to extend automatization to local levels. Cash management has worked in a situation of continuously higher than expected revenues. Debt data recording and reporting has been improved; but debt sustainability analyses still need to include domestic debt. Important weaknesses in budget execution are the following: (a) effectiveness in tax collection is a concern, with significant levels of tax arrears especially from large enterprises, (b) public payroll controls are currently weak, relative to the standards demanded by the PEFA methodology, even though some improvements have been made in terms of clarifying pay structures and abolishing informal practices such as using funds from vacant posts to pay supplements, (c) procurement reform has begun, but while legal reform has been undertaken, the institutionalization of the new system is still missing, and (d) the internal audit function is very weak. Furthermore, as indicated above, the process of allocating additional resources during budget execution is not sufficiently clear and transparent. This means that that there is a situation of frequent ad hoc bidding for additional funds from various sources (presidential fund, excess revenue) and of generating own resources by budget organizations which have not been sufficiently taken into account at the budget planning stage. Expenditures budgeted and organized at short notice during the process of budget execution are unlikely to be employed most effectively. Furthermore, further improvements in cash- flow forecasting and management will be needed as the discrepancy between planned and actual revenue is reduced. According to the Public Administration Reform Strategy endorsed by the government and the president in 2006, payroll controls and internal audit are to be addressed in the short to medium term. It is currently not clear what further reforms are envisaged with regard to the revenue system and the cash-flow forecasting and management function (see also section 4). Accounting, Recording, and Reporting (PI-22 to PI-25) Basic systems for accounting, recording, and reporting are in place. Accounts reconciliation is generally carried out in a timely fashion. In-year reports are prepared on a quarterly basis by the Treasury. The information appears to be reliable. However, a modern reliable “double entry” Financial Management Information System is a vital priority for the future if quality accounting and reporting is to be achieved. A key weakness is the fact that the budget execution report is not being submitted for external audit on an annual basis. Furthermore, accounting standards do not comply with international standards, and the standards used are not systematically disclosed in fiscal documents. External Scrutiny and Audit (PI-26 to PI-28) External scrutiny and audit are weak. Legislative scrutiny of the annual budget law is satisfactory; but the absence of an administrative classification limits the ability of parliament to hold members of the executive to account. Tajikistan’s External Audit Institution continues to be accountable to the president and currently lacks independence as required by international (INTOSAI) standards. 3 Audit is largely focused on compliance with rules (regularity audits) which should be regarded as an adequate interim stage, given that the State Financial Control Committee (SFCC) was established in 2002 only (and has been dissolved and joined with the Anti-Corruption Department 3 Tajikistan’s external audit institution currently is not a member of INTOSAI. It holds membership in ECOSAI, the regional association for the countries of the Economic Cooperation Organization. This regional group is not a member of INTOSAI. Page 8 8 o f the Procurator General Office and the Tax Police into a new State Committee on Financial Control and Fighting Corruption (SCFCFC) since December 2006). Some training on new forms of audit (performance/value-for-money audits) has been received. Parliament is currently not reviewing audit findings. As noted above, the SFCC has not received the government’s financial statement for annual review (contrary to provisions stipulated by law), but rather has undertaken bi- annual audits of the budget (by conducting on-site audits of the MOF – in addition to audits of BOs). Audit findings are primarily submitted to the president, and to parliament only in summary format. The transformation of the SFCC in late 2006 is a problematic choice; as the roles of external audit and of combating corruption are distinct and should be kept separate according to established international practice. Donor assistance (D1 to D3) Donor assistance to Tajikistan still largely follows a traditional, project-based approach. Use of country systems has been very limited thus far as these systems were perceived to be weak. An intensified dialogue on aid effectiveness and the actions that the GOT and donors can take to improve aid practices would be desirable. Donor coordination should be a higher priority for the government, given Tajikistan’s substantial aid receipts (around 10 per cent of GDP) and the importance of these relative to domestically generated revenues. Donors need to make greater progress towards harmonization and alignment. Increasing the share of budget support may be one way of encouraging a more comprehensive approach to Tajikistan’s development needs; but it will require further progress with regard to the quality of the public financial management systems (and in particular internal and external audit functions). Assessment of the Impact of PFM weaknesses PFM systems have three objectives: macro-fiscal discipline, strategic allocation of resources, and operational or technical efficiency. These three objectives are linked. Fiscal discipline is the basis without which neither a strategic allocation of resources nor operational efficiency is possible. The findings of the strengths and weaknesses of the current PFM system in Tajikistan have the following implications in terms of these three objectives. Tajikistan has made important steps towards establishing macro-fiscal discipline. However, the situation remains fragile due to systemic risks and weaknesses, as shown by the substantially increased deficit in 2006. It will be essential not to loose sight of the fundamental issue of macro-fiscal discipline and the need to further strengthen relevant structures – including better revenue collection, sustainable debt management, and better oversight of fiscal risks arising from SOEs and quasi-fiscal activities. These issues should be addressed as a priority. Furthermore, significant progress still needs to be made with regard to the strategic allocation of resources and the technical efficiency in the utilization of public funds. The current situation reflects initial reform efforts and priorities since the re-building of the state and of the PFM system begun in 1997 (with an emphasis on re-establishing fiscal discipline). In the early 2000s, efforts were undertaken to improve the strategic allocation of resources (introduction of an MTBF), but they failed to have a systemic impact due to a combination of overambitious goals, the absence of sufficient political will in the context of wider political challenges, and capacity constraints. 4 A new round of ‘second-generation’ reforms is currently under preparation. To the degree that these reforms will be executed, they are likely to bring improvements with respect to the strategic allocation of resources (especially the implementation of an MTEF), and to technical efficiency. 4 See World Bank, Implementation Completion Report on IBTA2 (2006). Page 9 9 T he earlier experience implies that comprehensiveness and sequencing of these further reform efforts will require careful attention, especially with regard to: (i) avoiding overburdening the MOF and other key institutions during the reform process; (ii) ensuring that reforms cover implementation and do not stop at the adoption of new legislation; (iii) ensuring that initiatives are complementary and well linked (especially with regard to financial and sectoral management information systems; and complimentarily between information systems and underlying processes); (iv) ensuring that reforms are well adapted to the context of Tajikistan. Macro-fiscal discipline. The use of conservative budget estimates and the ability to limit demands on the budget have been important assets in the context of a very resource-constrained and uncertain economic environment in Tajikistan. However, there is some concern that the commitment to fiscal discipline is weakening. A core Treasury system has been established and equipped with the requisite tools for managing budget execution and reporting. While the situation has been stable in recent years, important risks remain. Tax revenue is still rather low, in part due to tax exemptions as well as to weaknesses in tax administration (and resulting high tax arrears). Also, there are substantial quasi- fiscal activities by state owned enterprises, and the practice of issuing state guarantees for debts incurred by SOEs may be resumed (see PI-17). Monitoring of fiscal risks arising from SOEs is weak. Furthermore, after a period of debt-write offs that helped to reduce the debt burden, Tajikistan is currently acquiring substantial new debt. The deficit and debt situation will require careful monitoring in order to ensure that the debt burden does not again become unsustainable. Finally, the current lack of a multi-year expenditure framework poses risks; particularly regarding the insufficient estimation of recurrent costs associated with large investments that are being undertaken or planned. Strategic allocation of resources. Open and policy- based competition for resources is not yet well developed in Tajikistan’s PFM system. The fragmentation of the government at the central level and the evolving system of intergovernmental relations pose challenges for the policy based allocation of resources. Transparency and public discussion of spending priorities in the legislature, civil society, and the media are still in their infancy. The GOT needs to decide whether to enhance the role of the cabinet- level Budget Commission or to consider a different mechanism for the discussion of the budget at this level. Given the high share of investments, the discretionary part of the budget is substantial, and hence requires strategic decision-making. In this regard, there is also an urgent need for greater government leadership and policy debate on the management and allocation of aid resources. 5 Furthermore, weaknesses in revenue collection that result in an overall reduction of available resources should be part of the policy debate. Renewed efforts at establishing an MTEF are being initiated. The full implementation of an MTEF will be a medium term effort, in particular with regard to developing information on the cost of programs and realistically costed sector strategies. In the last three years, spending increased on average by more than 8 per cent during budget implementation. The uneven allocation of additional resources has resulted in substantial shifts in the sectoral allocation of funds, thus making allocative decisions taken during budget planning less binding. The main assets of the current system are the relative orderliness of the budget planning and execution processes, which are important prerequisite for the capacity to translate policy choices into actual fiscal allocations. Operational efficiency. 5 PIP funding has accounted for between 20 and 30 per cent of total budgetary resources per year. Page 10 10 O perational efficiency is important given that in Tajikistan public services have to be delivered with very limited resources. Some measure of predictability in resource flows to core activities has been achieved (while there still may be greater difficulties at local levels), and the internal control and reporting environment is in line with current technical capacities (including still limited automatization of the Treasury system). Key areas of concern with regard to operational efficiency are the following: firstly, procurement has yet to move from the stage of legal reform to full implementation in order to reap benefits in terms of efficiency. Secondly, institutions for internal and external audit are weak, do not cooperate, and are almost exclusively focusing on regularity in individual transactions, rather than on systemic issues and on the effectiveness with which public funds are used. A strategy for improving internal control and audit is under discussion, and discussions have recently intensified on how to improve the external audit function. Furthermore, the expectation of additional funds becoming available during budget execution distorts some of the incentive for budget organizations and their management to plan within a clear financial envelope. Currently, there is little transparency to citizens (as service users) regarding public funding allocations to front-line service providers such as primary schools or health facilities; and thus the potential for direct accountability between service users and service providers is not used. Prospects for Reform Planning and Implementation A new round of PFM reforms and capacity building covering a range of areas is under discussion between the GOT and donors. This follows previous reform and capacity building efforts which have been of mixed success (see WB April 2006); including the successful establishment of a core Treasury and budget management system since the late 1990s, and a range of legal reforms in the 2000s. Donor-supported efforts to reform the system of intergovernmental finances and to introduce an MTBF were not successful in the early 2000s. The overall environment is evolving (greater availability of resources, gradually improving capacity, step-by-step familiarization with reform ideas and options) and therefore conditions for further reform are improving. The key reforms and capacity building efforts currently under discussion include the following: · budget planning: establishment of an MTEF (including a strengthening of financial planning and management capacities in line ministries); revenue forecasting (already under way); introduction of an administrative budget classification; · budget execution: further development of the Treasury system, reform and expansion of internal control and audit; · reform of intergovernmental fiscal relations: per capita allocation of funds in social sectors; introduction of greater control of sector ministries over sector-related spending at local levels; · revenue administration; · reform of government structures and organization and of the public service; · further donor support for the development of an independent external audit function Page 11 11 · further donor support to develop the role of the parliament in scrutinizing all aspects of the budget process. As this list reflects, the reform agenda is broad and ambitious and will require good prioritization and sequencing. As a signatory to the Paris Declaration, Tajikistan has committed itself to undertake improvements of its PFM system (ongoing and planned reforms are discussed in greater detail in section 4 of this report). Exercising government ownership and active government management of this reform agenda will be crucial in maximizing the benefit from the financial and technical support that is being offered by donors. It will be important also to pay attention to links and interactions between the various components of the reform agenda that is under discussion. For example, the further development of the treasury system and changes in the role and capacities for public financial management in line ministries should be well coordinated; similarly, there are important linkages between reforms in sector planning and management and changes to the system of intergovernmental (fiscal) relations (around the division of responsibilities for service provision). Furthermore, while reforms are rightly focused on achieving improvements with regard to strategic allocation of resources, there are concerns about the apparent ambitiousness of these reforms, given the starting position. How exactly the introduction of the MTEF will be sequenced is a vital consideration. For establishing a meaningful medium-term planning system, basic building blocks of the fiscal system need to be strengthened further, including improvements in revenue administration, budget classification, payroll controls, procurement, and internal and external audit. The present tendency of substantial budget deviations at the sector level imply considerable levels of unpredictability even over a 12 months period. Until this unpredictability is reduced, attempts to program resources over the medium term will be futile. Furthermore, an important objective will be to find ways of integrating capital and recurrent budgeting (with implications also for aid policy and management). Even if not the entire reform agenda as outlined can be put on track in the short to medium term (2- 3 years), it is likely that PFM systems in Tajikistan will see further improvements over the coming period. This should lead to improvements in repeat rounds of the PEFA assessment at least in those areas where substantive reform and capacity building efforts can be set in motion. Page 12 12 1. Introduction Objective of the PFM-PR. The purpose of the PEFA assessment is to provide the Government of Tajikistan with an objective assessment of the country’s PFM systems. The assessment comes at a critical juncture, at which it provides an assessment of achievements made thus far, and as a new round of PFM reform is being launched. It aims to support a better understanding of the overall fiduciary environment of the budget and assist in identifying those parts of the PFM systems most in need of reform. It should contribute to a common understanding among the GoT and the donors wishing to support further PFM reforms. Process of preparing the PFM-PR. The PEFA study team has been composed of three national and two international consultants, led by a research fellow from the Overseas Development Institute (UK). The Swiss State Secretariat for Economic Affairs, SECO, is the main sponsoring partner of the assessment. The management of the process has rested with the World Bank. Other donors and external partners have been consulted; and are invited to provide comments on this draft performance report. The main GoT partner for the PEFA Assessment has been the Ministry of Finance (MOF). The MOF appointed a Working Group which coordinated the Government’s participation and input into the PFM Assessment. At various steps in the assessment process, meetings were held with all the various government institutions involved in the PEFA assessment (MOF, MOET, SFCC, MSRD, the parliamentary budget committee, the Aid Coordination Unit) 6 , and with key PEFA partners (IMF, DFID, EU) and others. The MOF also prepared a self-assessment report, which has been a key component of the assessment process. The PEFA assessment in Tajikistan has involved the following steps: (i) a kick-off mission providing an introduction and orientation on the methodology to the GoT and other stakeholders (October 8-19) (ii) a period of self-assessment, resulting in a draft self-assessment report by the MOF (October 18-November 17) (iii) a main mission resulting in a draft performance report (November 22-December 9) (iv) desk-based completion and revisions of the report (v) presentation and discussion of the report in country (April 23 to May 5, 2007). (vi) finalization of the report. The Performance Report sets out the findings from these combined efforts. Methodology. The PEFA methodology is set out in the Public Financial Management Performance Measurement Framework (available at www.pefa.org ). 7 It is based on 28 indicators covering a country’s PFM system, and 3 indicators addressing the interaction of donors with a country’s budget process and PFM system. PEFA Assessments provide cross-country comparable indications of the effectiveness of PFM systems, and of their improvements over time. They do not provide, however, for an analysis of the 6 Due to the reorganization of the government in late 2006, several of these institutions changed their designation during the PEFA process. 7 The PEFA assessment framework is also available in Russian through this website. Page 13 13 c auses of existing weaknesses, nor for an indication of the PFM system’s ability to deliver development objectives, e.g. on poverty reduction outcomes. Each indicator is scored on a scale from A to D. The basis for these ratings are the minimum requirements set out in the methodology. Many indicators include two or more dimensions, which are ‘added up’ using methods M1 or M2. For method M1 the weakest link is decisive, i.e. the overall rating of the indicator is based on the dimension with the lowest score. For M2, an average of the sub-ratings is used to arrive at the score for the overall indicator (see the PEFA Framework, ‘Scoring Methodology’). The main sources of information that have been used for this PEFA assessment are: (a) official GoT reports and data; (b) external evaluations and reports (WB, IMF, and others); and (c) numerous semi-structured interviews with key users and providers of PFM information and other stakeholders (government representatives; donor representatives; members of parliament; representatives of selected non-governmental organizations). To the extent possible, care has been taken to triangulate information. The self-assessment report has played an important role in terms of taking stock of current institutional features, practices, and laws concerning public financial management in Tajikistan; and has provided a key source on which the Draft Performance Report draws. Scope of the assessment. The PEFA assessment focuses primarily on the national level of a country’s PFM system. At the national level, it seeks to cover the entire PFM system, including cross-cutting and overall issues, the revenue side, the budget cycle from planning through execution to control and auditing; and the interaction of donors with the PFM system. In Tajikistan, more than 30 per cent of total spending is managed through local budgets; and in social service sectors, this share is even greater. A few of the indicators are specifically designed to probe into how the national level interacts with sub-national governments and with public sector service providers; and the study team has generally sought to include information on PFM practices at local levels. However, it has to be borne in mind that the scope for the verification of information related to sub-national levels has been weaker. Page 14 14 2. Country Background 2.1 Description of the Country Context Tajikistan is one of the five Central Asian states emerging from the former Soviet Union. Civil war from 1992 to 1997 severely hampered the development of a well functioning state and market. These efforts have been renewed since the late 1990s, and accelerated in the 2000s. Tajikistan’s structural environment is challenging. Tajikistan is a landlocked country; and large parts of mountainous terrain make transportation difficult and costly. Bordering Afghanistan, Uzbekistan, Kyrgyzstan, and China Tajikistan faces a regional neighborhood that is challenging but also offers potential. Of the 6.8m population, 40% are below the age of 15. Economic situation The macro-economic situation has significantly improved in recent years. After prolonged decline in the 1990s, GDP has grown by around 9 per cent per annum during 2000-2005. Inflation has been reduced to 7.1 in 2004 and 2005. External debt, which had risen to more than 100 per cent of GDP in 2000, has been brought down to below 40 per cent of GDP in 2005, as a result of strong growth, a prudent borrowing policy, and favorable debt restructuring. 8 However, Tajikistan has subsequently taken on a major Chinese loan for infrastructure development. 9 The economy has a rather narrow base with aluminum and cotton as the main export products (the aluminum smelter TADAZ contributes more than 50 per cent of total export earnings). An important driver of the economy in recent years have been remittances from Tajik workers in Russia and elsewhere. Major investments in the country’s potential hydropower generation capacity have been resumed since the early 2000s. 10 Aggregate fiscal discipline has been maintained in 2003 to 2005, but appears to be currently weakening. Compared to a budget deficit of 6 per cent in the mid-1990s, Tajikistan has been running a primary surplus during 2003 to 2005. However, a primary deficit and a substantial overall deficit have emerged in 2006 and are planned for 2007 (see section 2.2, table 3). Growth has slowed somewhat since 2005, but is still expected to be substantial over the next five years (around 6.5 per cent per annum). Table 2: Key Macroeconomic and Social Indicators 2001-2006 2001 2002 2003 2004 2005 2006 (estimate) GDP growth 10.2 9.1 10.2 10.6 6.7 7 Inflation (CPI) 36.5 10.2 17.1 6.8 7.8 12.0 GDP per cap USD 171 190 237 310 336 400 Poverty rate, %* .. .. 67 58 .. .. Source : World Bank, IMF; MOF * at <$2 per day 8 Status Report on External Debt , Department of Public Debt and State Financial Assets, Ministry of Finance, Dushanbe, 2006. 9 The total loan amounts to approximately USD600m to be provided for construction projects over a period of 3 years. In the 2007 budget, a loan of USD200m is forecast. 10 Projects had stalled in the 1980s when the Soviet Union began to disintegrate. Page 15 15 Overall system of government and government reform program Recent key political developments The last parliamentary elections took place in February 2005. As a result, the pro-government People’s Democratic Party currently holds 57 of 63 seats in parliament. Presidential elections took place on November 6, 2006. The incumbent, Emomali Rahmonov, was re-elected for another 7-year term. The presidential elections were followed by extensive re-structuring of the government. On November 30, 2006, the President issued a Decree on Reforming the Structure of Central Organs of Executive Power of the Republic of Tajikistan. 11 According to the Decree, the reform is aimed at “further deepening of the public administration reform, forming of transparent system of functioning of state organs, more effective and efficient management of public funds, and providing for sustainable economic, social and cultural development of Tajikistan”. One of the major objectives of the reform was to eliminate duplication of functions as well as provide for more rational redistribution of functions among government agencies. The number of ministries and state committees was reduced from 25 to 17. The Ministries of Industry and of Energy were merged into the Ministry of Industry and Energy. The Social Protection Fund was dissolved and its functions were passed to the Ministry of Labor and Social Protection (a Social Protection Agency was set up under the MLSP). Over the last years, the Social Protection Fund and the Ministry of Labor and Social Protection have had a shared responsibility for social protection policy. However, the coordination of activities as well as division of responsibilities between those two agencies was poor, and the integration is hoped to bring gains in terms of policy formulation and implementation in the sector. Other changes include the transfer of the anti-monopoly function to the Ministry of Economic Development and Trade, while support of entrepreneurship is managed by the newly created State Committee on Investment and Management of State Property (SCIMSP). SCIMSP is also responsible for foreign investment promotion, the function that was managed by the Ministry of Economy and Trade; and it has assumed the functions of the Aid Coordination Unit which previously was a separate unit in the presidential administration. As to budget preparation and execution process, all main agencies involved in the process retained their functions (no changes were made with regard to the Ministry of Finance). However, changes affect the wider set of institutions related to public financial management. The Ministry of State Revenues and Duties was split into the Tax Committee and the Customs Committee under the Government. Furthermore, the former State Financial Control Committee (SFCC) was dissolved and now is a part of the State Committee on Financial Control and Fighting Corruption (SCFCFC). This new Committee combines the previous functions of the Anti-Corruption Department of the Procurator General Office, the Tax Police and the SFCC. While the reduction in the number of ministries and state committees is generally positive, some aspect of the restructuring are problematic. This concerns especially the combination of the external audit function with the anti-corruption function in one state committee, a combination of functions which is not in line with international good practices; and the creation of the Tax Committee and the Customs Committee “under the government” without clear lines of accountability for these important functions. As discussed below (PI-26), in early 2007 development partners and the GoT held intensive discussions about the establishment of an external audit office which would meet international standards (also with regard to the independence of such an office), and further change with regard to these functions is likely to take place in 2007 and beyond. 11 Presidential Decree #9, November 30, 2006. Page 16 16 The levels of government The Constitution defines Tajikistan as a unitary state. Below the Republican level, there are (a) the city of Dushanbe, (b) three regions ( oblasts ) and (c) 13 districts ( rayons ) of “republican subordination”. 12 Oblasts are further subdivided into rayons , and then further into cities and rural communities. The lowest level of public administration is the jamoat . The legal status of sub-national governments is defined by the Constitutional Law on Local Public Administration and in the Law on Self-government in Towns and Villages . A Public Administration Reform Strategy for the RoT was adopted in March 2006 which envisages further changes to the system of government, and the distribution of power, responsibilities, and resources between the different levels of government. 13 While oblasts and rayons have elected councils, the dominant power rests with local executives, the heads of which are appointed and dismissed by the president (and approved by respective local councils). Thus, while local governments have some autonomy vis-à-vis central government structures, they are politically dependent on the central executive. Local budgets are formed down to the level of cities and rayons . The Ministry of Finance interacts directly with 17 sub-national entities: the 3 oblasts, Dushanbe, and the 13 rayons of republican subordination. In principle, sub-national governments have substantial control over revenue and expenditures. The Constitution grants local governments the right to develop and implement their own budgets and to establish local fees, taxes and duties “in accordance with the law”. 14 Substantial revenues flow directly to local governments through revenue sharing arrangements which are defined in the annual budget law. For example, for 2005, it was envisaged that the Republican budget would receive around 80 and sub-national governments around 20 per cent of total revenues (excluding the revenues to the SPF); and in addition, around 10 per cent of non-SPF revenue were to be transferred from the Republican to the local level in the form of subventions and subsidies (see also table 7 in section 2.2). Especially in social service sectors such as health and education, a large share of total spending takes place at sub-national levels. 15 Overall government reform program A National Development Strategy (NDS) seeks to bring all medium to long term strategies together in one single plan, covering the period 2006-2015. The PRSP2 is to serve as the action plan to cover the initial 3-year implementation period of the NDS. Both the PRSP2 and the NDS were approved by the government in March 2007. Key priorities set out in the NDS are reforms of the system of government; the promotion of economic growth (through developing the agro-industrial complex, infrastructure, communications, energy, and industry; and including an emphasis on private-sector led growth); and social development and achieving the MDGs (with an emphasis on developing access to basic social services). 12 This section draws on the LGI report on Local Government, 2001. 13 Presidential decree #1713, March 15, 2006. 14 Constitution of the Republic of Tajikistan, version as amended in 2003, article 77. 15 Up to 90 per cent of the education budget are spent at local levels (WB education sector fiduciary assessment, August 2006). Page 17 17 T he PRSP2 concretizes th ese priority activities for the period 2007 to 2009. Pillar 1 are ‘functional reforms’, including: 1) reform of the state administration, 2) reform of macro-economic policies, 3) improvements of the investment climate, 4) development of regional cooperation and integration into the world economy. The sub-section on administrative reform is not very detailed; but refers to the Public Administration Reform Strategy (discussed below) that was adopted in 2006. Budget and tax reforms are set out as the top priority for the set of reforms addressed in sub-section 2. Specifically mentioned are: the introduction of an MTEF, and the reform of budget processes in the first instance in the health and education sectors, the development of proposals on the optimization of state debt, and on the system of internal controls, and finally reforms of the tax and customs administration. The PRSP2 contains some costings by broad functional categories. The total estimated cost of the PRSP2 is 5bn USD over the three year period. The largest share of this – 3.6bn USD – are envisaged for investments in the transport, communications, industry, and energy sectors. The cost for improvements in the health and education sectors are estimated at 328mn USD. Reforms to the public administration are estimated to cost 10mn USD, and reforms of macro-economic policies 82.9mn USD. The Public Administration Reform Strategy of March 2006 sets out the envisaged reforms to the system of government; and has important implications for the development of the PFM system. Like the NDS, the PARS covers the period up until 2015. A number of donors are envisaged to support the implementation of the PARS (WB, UNDP, ADB, EU, USAID, JICA, SDC, AKF, CIDA). The PARS defines the following areas for reform: 1) increase effectiveness of national development management, 2) improve public administration in line with market economy principles, 3) increase effectiveness of the public finance management, 4) form a modern professional civil service, 5) develop administrative and territorial management, 6) form local self-governance capable to provide qualitative services to the population. PFM reforms are a crucial component of the overall PARS, as strategic allocation of resources and the quality of financial management will have important implications for the overall implementation. Components of PFM reform as set out in the strategy are the following: (i) greater coordination between the budget and economic policy (including closer links between recurrent and capital budgets), (ii) greater coherence in responsibilities for PFM (with a view to integrating investment planning and the budget process more closely; providing the MOF with overall responsibility for linking strategies and programs with the budget process); (iii) establishment of better procedures for financial decision-making in ministries; (iv) improvement of audit systems (strengthening internal and external audit; introducing new audit methods), (v) greater effectiveness of public procurement, and (vi) rationalization of revenue allocation mechanisms among levels of government. 16 The way in which priorities are set out in the PRSP2 and the PARS differs somewhat; in particular with regard to a greater emphasis on the introduction of an MTEF in the PRSP2 (which appears related to ongoing discussions with development partners). The PARS is a helpful reference point for the combined effort of the government and its development partners. The extensive re-structuring of the government subsequent to the November 2006 elections signal a political interest in undertaking these reforms. However, it should be borne in mind that the political economy challenges as well as the technical and coordination challenges for implementing such a broad reform strategy are substantial. 16 The introduction of an MTEF is not explicitly mentioned but is implicit in points (i) and (ii). Page 18 18 T he PARS envisages changes in the short (end of 2006), medium (2007-08, and long-term (2009- 15). To the degree that these changes materialize, they have the potential to bring substantial improvements of the system of government, including those areas covered by the PEFA PMF. 2.2 Description of budgetary outcomes Fiscal performance Table 3: Key fiscal indicators 2003-2006, in % of GDP 2003 2004 2005 2006 (exp) 2007 (planned) Total revenue including grants 17.3 17.9 19.6 19.4 19.3 - own revenue 17.0 17.2 19.2 19.2 19.0 - grants 0.3 0.7 0.4 0.2 0.3 Total expenditure 19.1 20.3 23.1 27.8 34.5 - non-interest expenditure 17.8 19.6 22.6 27.2 33.7 - interest expenditure 1.2 0.7 0.5 0.6 0.7 Overall deficit including grants and PIP -1.8 -2.4 -3.5 -8.7 -14.3 Primary deficit 0.9 0.7 0.2 -0.2 -1.0 Net financing 1.8 2.4 3.5 8.7 14.3 - external 2.8 4.0 3.4 8.2 13.3 - domestic -1.1 -1.7 0.1 0.5 1.0 Source: MOF; includes state budget (central and local) and SPF In recent years, fiscal performance has been sound with gradual increases in revenue, and deficits of < 1 per cent of GDP. The ratio of revenues to GDP is still low by regional comparison (CIS-7), although it is in line with other low income countries. An issue of some concern is the fact that fiscal deficits have rapidly increased again in 2005 and 2006. These are covered by external financing from a range of development partners. The debt stock has been reduced from a peak at more than 100 per cent of GDP in 2000 to less than 40 per cent of GDP by 2005, due to the combination of debt cancellation and strong GDP growth. The debt to GDP ratio will grow again from 2007 as Tajikistan is assuming new debt, including a USD 600mn loan from China over a three year period. Table 4: External debt (USD mn) 2000 2001 2002 2003 2004 2005 2006 (prelim) 2007 (expected) External debt Debt/GDP (%) 1027 108.2 1018 97.0 984 82.5 1030 66.3 871 42.2 895 38.9 860 30.6 1265 40.3 Source : MOF Allocation of resources Table 5: Actual expenditures as % of total expenditures* by sectors Actual budgetary allocations by sectors (as a % of total expenditures) 2003 2004 2005 2006 1. Public administration 19.7 16.3 16.0 13.8 Page 19 19 2. Defence 7.0 8.0 7.5 6.4 3. Law enforcement 6.9 7.2 7.5 7.3 4. Education 16.7 16.7 19.8 19.6 5. Health 6.4 5.9 6.4 6.5 6. Social protection 16.3 16.1 18.2 18.1 7. Communal services (incl. utilities) 6.6 7.5 6.3 6.4 8. Culture and sport 3.4 3.4 3.3 3.8 9. Heating and energy complex 1.3 2.1 1.4 2.3 10. Agriculture 3.4 3.5 3.0 2.7 11. Industry 1.2 2.6 1.8 2.9 12. Transport and communications 8.5 8.5 6.2 5.4 13. Other economic activities and services 0.6 0.4 0.7 0.2 14. Other services 1.8 1.8 1.9 4.4 Source: MOF; note: no functional breakdown is available for the PIP; so the functional breakdown only reflects primary expenditure, excluding interest payments and the PIP. Including the PIP would significantly affect the allocation across sectors. * including central and sub-national levels of government Table 6: Actual expenditures as % of total expenditures* by economic classification Actual budgetary allocations by economic classification (as % of total expenditures) 2003 2004 2005 2006 Current expenditure 65.7 59.8 63.1 54.4 - wages and salaries 14.6 13.4 16.8 16.4 - goods and services 28.7 27.7 26.4 19.6 - interest payment 6.5 3.4 2.2 2.0 - transfers 16.0 15.1 17.8 16.4 Capital expenditure 34.1 40.0 36.9 45.6 - externally financed 14.1 15.1 15.9 30.7 - domestically financed 20.0 24.9 21.0 14.9 * including central and sub-national levels of government Table 7: Distribution across budgets (% of total expenditures) 17 2003 2004 2005 Republican budget 58.2 59.3 56.5 Local budgets 31.8 31.3 32.4 Social Protection Fund 10.0 9.4 11.0 Total 100.0 100.0 100.0 Source : MOF/Treasury 2.3 Legal and institutional issues and ongoing PFM reforms A number of reform steps around public financial management (including also tax policy and tax administration), have been undertaken in recent years, of which many are still in the process of being (fully) implemented and/or are still being deepened; while new rounds of reform are already on the horizon (see section 4). The legal framework for PFM 17 Including recurrent expenditure and domestically financed capital expenditure. Page 20 20 T he basic legal document on PFM is the ‘Law on State Finances of the Republic of Tajikistan’. 18 The law defines three main components of the budget: “The system of state budget of the Republic of Tajikistan consists of the budgets of following levels: - the first level - the republican budget and budgets of state target funds; - the second level - local budgets. Local budgets consist of budgets of GBAO and its cities and rayons, of oblasts, cities and rayons of oblast subordination, city of Dushanbe and its rayons, cities and rayons of Republican subordination and jamoats .” 19 The budget classification was amended in 2004, but an administrative (or ‘organizational’) classification is still missing (see PI-5). A new Tax Code and a new Customs Code came into force in early 2005 (Law no. #61 and Law no. #62, December 3, 2004). A new Law on Public Procurement was adopted in 2006 (Law #168, March 3, 2006). Further relevant laws with regard to public financial management include: the Law on Treasury (Law #19, May 12, 2001), the Law on the State Financial Control in the Republic of Tajikistan (Law #66, December 2, 2002), and the Law on State and State Guaranteed Borrowing and Debt (Law # 886, December 11, 1999). The institutional framework for PFM The main responsibility for PFM rests with the MOF and its two main departments (budget department, treasury). An organizational chart of the MOF is provided in appendix 3. As the chart reflects, there are issues regarding the internal organization of the MOF, such as the separation of budget departments for public administration and for defense and law enforcement from the main budget department, and the division of responsibilities for sub-national budgets. In 2004, a (cabinet level) budget commission was set up to facilitate better linkages between government policy and budget planning. 20 The Treasury system is under development (see also section 4). Regarding internal control and audit systems, the Office for Control and Revision (KRU – an institution dedicated to control created during the Soviet period) was dissolved in 2001 and replaced by an internal audit unit within the Treasury department. An External Audit Institution, the State Financial Control Committee (SFCC), was established in 2001/2002. The SFCC was dissolved in the government restructuring in November 2006 and re-established as a State Committee on Financial Control and Fighting Corruption. The MOF interacts directly with a large number of budget organizations; as well as with the 17 main sub-national governments and their budgeting processes. The role of Line Ministries in the budgeting of sectors has been relatively limited; and financial departments of Line Ministries primarily deal with the financing of the central ministry apparatus only. MOF sector kurators (e.g. for the social sphere, infrastructure, energy) in the Budget Department and in the Treasury are responsible for assisting BOs in their budget preparation process and for monitoring budget execution. Further legal and institutional reforms that are currently being prepared or planned are discussed in section 4 of this assessment. The key features of the PFM system 18 Law #77, December 2, 2002. 19 Article 5, 2002 Law on State Finances. 20 Chairman: Prime Minister; deputy chair: Minister of Finance; members: key ministers, representatives of presidential office. The commission does not appear to have a real impact on budgeting thus far (see section C (i) below). Page 21 21 Capital and recurrent budgeting are still rather disjointed, but the government is taking some steps towards closer integration. Achieving closer integration would also require stronger management of aid, particularly at the planning stage of externally financed projects, and further efforts at harmonization and alignment by donors. Capital budgeting is subdivided into the domestically financed CSIP, and the externally financed PIP. Key steps in preparing the PIP and the CSIP rest with relevant departments in the Ministry of Economy and Trade. 21 The list of proposed projects with costings is then handed over to the MoF during the process of the budget preparation (in June/July). 22 Since 2004, the PIP has been prepared on a rolling basis and is updated annually (previously in three year cycles). In the 2007 budget, PIP funding is included into the main budget plan for the first time. Capital expenditures account for more than 35 per cent of total non-interest expenditures, with a further substantial rise in 2007 due to the commencement of Chinese loans (see table 6 above). 23 The PRSP department in the Presidential Administration is very small (3 staff members). As discussed in section 2.1, a second generation PRSP (to cover 2007-2009) was adopted in March 2007, i.e. substantially after the conclusion of the 2007 budget plan. Furthermore, Tajikistan has had a fragmented system of government with a large number of KBOs ( rasporoditel’ pervogo urovniya – 1st tier spending units) of which there were around 80 until the government restructuring in late 2006. KBOs, and even 2nd tier spending units (may) interact directly with the MOF. Line ministries play a relatively minor role in sectoral budget planning; and as a consequence have small financial departments with limited capacity. 2 4 These three features, the disjoint between capital and recurrent budgets, the relatively weak role of the PRSP, and the fragmentation of the system of government contribute to rather weak linkages between policies and budgeting in Tajikistan. Such linkages would be tackled in the context of establishing an MTEF which is currently being debated by the GoT and with development partners. Revenue collection agencies are separate from the MOF. Until late 2006, tax and customs authorities were joint in a Ministry of State Revenue and Duties; since then they are established as two separate agencies. The revenue administration has weaknesses in particular with regard to ensuring payments by large taxpayers. Establishing good databases will be one essential element in improving tax compliance; as will be data sharing between the tax and the customs authorities in particular since these have been separated in late 2006. A core Treasury system has been developed in Tajikistan and – within its current paramenters and limitations (being a “single entry” based system, and lack of computerization at local levels) – it operates effectively . An IMF report issued in July 2005 drew attention to many potential improvements in treasury procedures. The Treasury has made significant progress in the practical implementation of those proposals. They have introduced an “Electronic Payment System” for the 21 The Public Investment and Technical Assistance Program Unit (PITAPU) responsible for the PIP and the Capital Investment and Municipal Housing Department (CIMHD) responsible for the CSIP. 22 The Investment Policy Division under the Budget Department of MOF. 23 The overall loan has a size of about USD 600m (see also EIU 06/06). It is part of a USD900m credit provided by China for member states of the Shanghai Cooperation Organization. The terms of the loan to Tajikistan are concessional (2% interest rate; 2 year grace period). The biggest project (USD 269mn; under implementation) is the reconstruction of the road Dushanbe – Khujand (to the border of Uzbekistan), including the construction of a tunnel which will make the road passable year-round. A loan of USD146mn USD is provided for the construction of a South- North electricity transmission line with the capacity of 500 kV. The assistance is tied (i.e. implemented by Chinese companies). 24 Their role is also reduced by the relative importance of budget planning at local levels for spending in service sectors. See PI-8; 23. Page 22 22 d irect payment to creditors’ bank accounts. Significant further improvements are dependent upon the development of modern computer, communication and networking facilities, in particular covering also sub-national levels of government. Ideally such a major exercise should incorporate the provision of quality financial and accessible information for sector budget managers (as these are to be established as part of the MTEF reforms). Internal and external audit are currently underdeveloped. Both are new functions in post-Soviet environments, and hence take some time to develop. Given that Tajikistan is a latecomer in this process, ample experience in developing these functions is available from other transition countries. The GoT has committed itself to develop internal control and audit functions and has enlisted external support for this from the EU. Further efforts at developing the external audit function still need to be defined. Page 23 23 3. PEFA Indicators This secti on provides a detailed analysis of the performance of Tajikistan’s PFM system against the criteria set out in the PEFA framework. The assessment is based on the status quo, covering the system’s performance over a three-year time horizon (2003 to 2005; where possible also taking data for 2006 into account). A. PFM-OUT-TURNS: Credibility of the budget PI-1 Aggregate expenditure out-turn compared to original approved budget It is essential that governments can translate a budget plan into actual funding allocations in the course of the fiscal year. This makes the original approval of the budget plan meaningful, and reflects that the central government is reliable from the perspective of spending units. In Tajikistan the problem in recent years has been an over- rather than an under-execution of the planned budget on the revenue side (see PI-3), which has fed through to over-execution of expenditures, which therefore has taken place without undermining fiscal discipline. This is in line with the experience in other CIS countries in recent years. It has been additionally fuelled by conservative growth estimates in 2003 and 2004; and by incentives built into the Tajik revenue system to underestimate revenues. It appears that the degree of over-execution on the revenue side is leveling off as these exceptional circumstances are no longer present. Table 8: Comparison of Original Budgeted and Actual Expenditures, 2003-2006 2003 2004 2005 2006 Budgeted primary expenditure (mn. Somoni) 571.4 859 1234 1501 Actual primary expenditure (mn. Somoni) 668.7 984 1276 1619 Difference between actual & budgeted primary expenditure (mn. Somoni) 97.3 125 42 118 Difference as % of budgeted primary expenditure (%) 17 14.6 3.4 7.8 Note : Primary expenditures exclude debt servicing payments and foreign and domestic funding of the PIP; funds shown include the CSIP, and the Social Protection Fund. Source : MOF/Treasury. Contrary to provisions specified in the Law on the State Budget (art. 46), supplementary budgets in 2003 and 2004 were not approved by parliament, even though this should happen when the difference between planned and actual revenues (and subsequently expenditures) exceeded 10 per cent. 25 Supplementary budgets were adopted by government decree. 25 The Law on State Finance, article 46 states: “In case of decrease or increase of budget revenues, that can lead to the change of financing in comparison with approved budget for more or less then 10% of yearly prescription, the Government of the Republic of Tajikistan in this case submits to the Majlis Namoyandagon Majlisi Oli of the Republic of Tajikistan the draft Law of the Republic of Tajikistan about introduction of amendments and additions to the Law of Page 24 24 The initial underestimation of the budget both on the revenue and on the expenditure side does reflect a strong top down control in the aggregate, i.e. unrealistic spending demands are not being included in the budget; and budget preparation has been conservative. However, the expansion of the deficit in 2006 (see table 3 in section 2) indicates that this conservative stance may be weakening. Indicator Brief explanation Rating (M1) PI-1 Aggregate expenditure out-turn compared to original approved budget In 2004 the deviation was just below 15%. In 2005, the deviation has declined markedly. It re-increased again somewhat in 2006, but remained below the threshold of 10% for a ‘B’ rating. B PI-2 Composition of expenditure out-turn compared to original approved budget Where the composition of expenditure varies considerably from the original budget, the budget will not be a useful statement of policy intent. Also, there may be questions about both the rationale and the procedures for re-allocation of expenditures during budget execution. There is currently no administrative classification used in the budget preparation process (see PI-5). Therefore, the classification which can be used to assess this indicator is the functional classification. Table 9 - Results Matrix for PI-1 for PI-2 year total exp. deviation total exp. variance variance in excess of total deviation 2003 17.0% 27.5% 10.5% 2004 14.6% 23.9% 9.3% 2005 3.4% 13.3% 9.9% 2006 7.8% 15.5% 7.7% Source : MOF/Treasury; see table 1. Expenditure variance has been high and while it declined significantly in 2005 compared to the previous two years, it still exceeded total expenditure variation by 10 per cent. Variance of expenditure composition compared to overall variance has declined in 2006; however, expenditure composition variance has slightly re-increased again from 2005 to 2006. The main contributing factor to high variance is the way in which additional resources are allocated during budget execution. There have been several sources of additional funds the use of which leads to shifts in sectoral allocations: higher than expected tax revenues, allocations from the presidential reserve fund (which is equivalent to 2 per cent of the total budget), and the use of own resources generated by BOs themselves (see PI-3 and PI-7). As reflected in the tables in appendix 2, the functional categories which have consistently seen the greatest additional allocations during budget execution are “public administration” and “communal services”, while “other services” have seen the greatest reductions during budget execution. In 2006, under-execution also affected law enforcement (-16.3%), education (-5.6%), and agriculture the Republic of Tajikistan on State budget of the Republic of Tajikistan for the next fiscal year together with necessary documents.” The GoT has expressed the view that approval of a supplementary budget by the national parliament is necessary only if the Republican level budget has experienced a change in excess of 10%; but not if the over- or under- execution of revenues in excess of 10% is attributable to local level budgets. Page 25 25 ( -11.3%). It should be borne in mind that decision making about the allocation of (additional) resources takes place both at the central and at sub-national levels of government. Indicator Brief explanation Rating (M1) PI-2 Composition of expenditure out-turn compared to original approved budget Variance in expenditure composition has exceeded overall deviation in primary expenditure by around the benchmark of 10 per cent in two of the past three years, and by 7.7 per cent in the most recent year. C PI-3 Aggregate revenue out-turn compared to original approved budget Accurate forecasting of domestic revenue is a critical for determining budget performance, since expenditure allocations are based on that forecast. The main thrust of PFM reforms is generally to strengthen forecasting capacity and to promote the development of institutional forecasting systems that allow objective forecasts based on the best available data and methodology. As mentioned under PI 1, actual revenue has exceeded planned revenue in Tajikistan. Similar trends have appeared in other post-Soviet countries during the recent period of rapid economic growth after a decade of decline. According to the GoT, IMF demands for conservative revenue estimates were also a contributing factor to the degree of underestimation particularly in 2003 and 2004. The degree to which actual revenues have exceeded planned revenues has declined between 2003 and 2006. Overall macro-economic and macro-fiscal forecasting responsibilities are split between the MOET, the MOF, and the NBT; and there are concerns about the coordination of forecasting activities between these various institutions. The main responsibility for revenue forecasts rests with the Tax Policy and State Charges Unit in the Ministry of Finance. Revenue execution responsibility rests with the Ministry of State Revenues and Duties (MSRD) (since December 2006: with the Tax Committee and the Customs Committee). Efforts to improve forecasting capacity are supported by the EC. There is uncertainty about revenue figures. This uncertainty has two main sources: (i) figures for own revenues of budget organizations, which make up a substantial share of total revenues (around 7.5 per cent, with some year-to-year variation) were not fully included in budget plans before 2007; (ii) there are – usually small – divergences the records on ‘actual’ revenues among different departments. Despite these discrepancies, the overall trends that these figures indicate are similar, showing a decline in the excess of planned to actual revenues from around 120 per cent in 2003 to around 108 per cent by 2006. Table 10: total domestic revenue – planned and actual 2003-2006 MinFin budget department MinFin tax policy department Tax revenue Plan Actual % Plan Actual % 2003 574.9 712.6 123.9 574.9 707.3 123.0 2004 824.3 933.3 113.2 824.3 934.2 113.3 2005 1148 1192.5 103.9 1148 1195.9 104.2 2006 1424.3 1543 108.3 1424.3 1543.1 108.3 Non-tax revenue 2003 68.9 74 107.4 35.2 48.2 136.9 2004 98.7 100.8 102.1 50.6 53.6 105.9 2005 150.3 168 111.8 70 77.1 110.1 Page 26 26 2006 186.2 193.9 104.1 93.5 91.1 98.3 Total revenue 2003 643.8 786 122.1 610.1 755.5 123.8 2004 923 1034.1 112.1 874.9 987.8 112.9 2005 1298.3 1360.5 104.8 1218 1273 104.5 2006 1610.2 1736.9 107.9 1517.8 1635 107.7 Source : Ministry of Finance, respective departments. Non-tax revenue includes: state duties, sanctions and fines, sale of goods, income of the National Bank and income from communal services, as well as own revenues generated by budget organizations. The latter category is shown as a separate category for the first time in the 2007 budget plan. As has been noted in previous reports, there are incentives for local governments to underestimate revenues during the budget preparation process. 26 Local governments can use additional own tax revenues at their discretion; and additionally collected own revenues do not result in any reduction of subventions from the republican budget. However, there are two counter-balancing factors to this incentive structure: 1) tax estimation for specific localities is carried out top down by the Tax Policy Unit in the MOF which issues revenue targets for each sub-national government; 2) local tax administrations are under Republican, not local subordination. Furthermore, there are incentives from the expenditure side as well for underestimation: in a situation where cash management is basic, estimating expected revenues conservatively is an important ingredient for ensuring the constant availability of funds during budget execution. Table 11 provides a detailed breakdown of non- tax revenue. The main category are ‘own revenues’ ( spetsial’nye sredvsta ) generated by BOs, for example, from renting out premises. The latter were not included in estimations of non-tax revenue before 2006 (information on such funds was included as an annex, but not included in the budget law voted on by parliament; information was included in budget execution reports). In 2007, such own revenues of budget organizations for the first time were shown as a special and separate category in the budget plan (see also PI-7 with regard to the treatment of these revenues). Table 11: non-tax revenue, planned and actual 2003-2006 (MOF budget department) 27 2003 2004 2005 2006 Plan* act. % Plan* act % Plan* act % Plan* act % Total non-tax revenue 69 74 107.4 99 101 102.1 150 168 111.8 186 194 104.1 of which own revenues of BOs ( spetsial’nye sredstva ) 48.5 49 101.0 79.2 75.5 95.3 115.7 115.7 100.0 129.4 129.4 100.0 State duties 5.5 7.3 132.7 7 8 114.3 10.2 24.6 241.2 23.1 26 112.5 Sanctions and fines 3.8 3.2 84.2 4.4 3.5 79.5 8.8 4.7 53.4 9.5 6.1 64.2 Other non-tax revenue 11.1 14.5 130.6 8.1 13.8 170.4 15.6 23 147.4 24.2 32.4 133.6 Source: MOF; budget department. *Information on own revenues of BOs included as an annex to the budget plan only. Indicator Brief explanation Rating (M1) 3. Aggregate revenue out-turn compared to original approved budget Actual revenue has exceeded planned revenue levels in recent years (the definition of the PEFA indicator and scoring criteria is such that excess revenue or uncertainty over A 26 See IMF, FAD (2005). 27 There are significant discrepancies between the data on ‘own revenues of BOs’ ( ‘spetsial’nye sredstva’ ) originally submitted to the PEFA team, and the figures later provided by the MOF. Page 27 27 revenue figures does not affect the rating). PI-4 Stock and monitoring of expenditure payment arrears Expenditure payment arrears are expenditure obligations that have been incurred by government, and for which payment is overdue; and can indicate problems such as inadequate commitment controls, poor cash rationing, inadequate budgeting for contracts, or under-budgeting of specific items. Given the situation that in recent years revenues have been higher than expected, there seem to be no current problems with payment arrears (this is also the conclusion reached by the IMF/FAD 2005 report). Ensuring that only obligations are included in the budget that can be serviced has been a major policy objective of the GoT. The way in which records are kept does not distinguish between outstanding obligations and overdue obligations (there is no definition of arrears in use in the public accounting system); and there are concerns about underdeveloped commitment control. Table 12 reflects outstanding payments as registered by the Treasury at year-end, showing that their levels are low. Expenditure arrears could re-emerge in a situation in which a conservative fiscal stance is relaxed (see PI-1); and in which unexpected revenue short-falls may be difficult to manage cost-effectively through access to short-term borrowing. Dimension (i): Stock of expenditure payment arrears According to available figures, the current stock of arrears is low (see table 12; and IMF Art IV 2007, p. 14). The figures include all types of arrears (i.e. including wages, utilities, capital expenditures). Table 12: Outstanding payments, end of period 2003 2004 2005 Expenditure arrears, end of period % of total expenditures Expenditure arrears, end of period % of total expenditures Expenditure arrears, end of period % of total expenditures 42,2 6,3 23,5 2,4 10,9 0,9 *of the total sum of expenditures, excluding interest payments on external debt and the externally funded part of the PIP If local governments incur any wage arrears, they have to inform the MOF; and will receive a short- term credit from the MOF to cover the arrears. In 2005 this happened in several rayons in the context of wage increases (which were not matched by expected decreases in staff levels).28 Nonetheless, there are some doubts whether all outstanding payments are reflected in full in the central accounts. Firstly, not all arrears may be fully reported by local level governments to the central government; and this may be difficult to trace in the context of manual systems. Secondly, currently no system of commitment controls exists, so that arrears for the purchase of goods and services may not be fully reflected. Thirdly, the definition of arrears (in contrast to accounts payable) is not clear in the accounting system. Payment arrears to utilities have still been somewhat of a problem in recent years; however, the MOF appears to have brought this under control in 2005 and 2006. If there are any delays in 28 The PETS for the health and education sectors will provide more detailed information in this regard, e.g. whether any delays in wage payments at local levels occur. However, this information will only become available after the conclusion of this assessment. Page 28 28 p ayments by a BO to a utility company, the company can turn to the MOF directly to request the payment at year end. Neither the World Bank nor the IMF currently monitor the situation of payment arrears, so no verification of the situation could be provided from their side; but the decision to drop arrears monitoring reflects confidence by the IFIs that these are in fact low and currently do not constitute a problem in Tajikistan. Dimension (i): A (note: the rating is based on the last available year only) Dimension (ii): Availability of data for monitoring the stock of expenditure payment arrears The treasury quarterly collects data on outstanding obligations of budget organizations. As evidenced from the data reflected in table 12, the monitoring that has been in place has contributed to a reduction in arrears. However, there is no age profile of arrears available from the existing databases. A commitment control system is not yet operational in Tajikistan; and a system which would monitor in more detail creditors and an age profile would require the implementation of Commitment Control System (and possibly an FMIS). Dimension (ii): C Indicator Brief explanation Rating (M1) 4. Stock and monitoring of expenditure payment arrears The stock of arrears was below 2 per cent in 2005. The monitoring system is continuous, but it fails to adequately distinguish arrears from outstanding payments. C+ B. KEY CROSS-CUTTING ISSUES: Comprehensiveness and Transparency PI-5 Classification of the budget A comprehensive government classification system and chart of accounts allows the preparation and execution of budget to be based on the following heads: administrative unit, economic item, functional and program. Since 2002, the Tajikistan classification of fiscal accounts follows the 1986 GFS manual. Since 2002 the budget has been prepared following these classifications, as have been budget execution reports. In 2004, a somewhat modified version of the classification came into effect. The 14-sector GFS 1986 functional classifications did not anticipate the use of modern financial management systems or of organization-based budgets. Only a functional and an economic classification are used for budget preparation. De facto, the funding of a number of main administrat ive units is ‘inserted’ into the functional classification (see figure 1 below), mostly under the residual category of ‘other activities’ in the respective sector. Reform projects are also included under this sub-heading. Figure 1: Example of the actual application of the classification system (2005 budget execution report) 5. Health (1) Hospitals Page 29 29 ( 2) Policlinics (3) Protection of the health of the population (4) Other activities in the area of health care - Ministry of Health of the Republic of Tajikistan (central apparat ) - Projects of reforms in the health sector The use of the classification system in the adopted budget and in the annual budget execution report is broadly, but not fully consistent. While sector headings are consistent, at the sub-sector level and below there are inconsistencies between the two sets of reports. Furthermore, until 2006, the PIP has not been integrated into the functional breakdown (nor has a separate functional breakdown following the same classification as for recurrent expenditures been presented). This hampers opportunities for effective scrutiny of spending priorities. In 2007, the PIP has for the first time been integrated into the functional breakdown. The GoT prepares an internal organizational breakdown of the budget at the start of the fiscal year, based on which funds are actually allocated. However, this breakdown is not part of the budget documentation that serves budget approval, and budget execution reporting and monitoring. The absence of an administrative breakdown reduces the potential for holding budget managers to account. 29 There are plans to introduce an administrative classification system; and the WB is planning to provide technical support for this reform. If this reform is implemented, the rating of this indicator would improve, possibly significantly, in the next assessment. Indicator Brief explanation Rating (M1) 5. Classification of the budget There is currently no administrative classification. The functional classification is broadly in accordance with GFS. D PI-6 Comprehensiveness of information included in budget documentation Annual budget documentation as submitted to the legislature for scrutiny and approval should allow a complete picture of central government fiscal forecasts, budget proposals and out-turn of previous years. The PEFA PMF sets out 9 pieces of information/documentation which governments should be submitting to the legislature. Budget documentation submitted to parliament includes the following: (i) a breakdown of revenues and expenditures (ii) macro-economic forecasts and data on the financing of the budget deficit (iii) expected turn- out of the current year’s budget (iv) summary tables on revenues and expenditures (including data for the current and previous year) (v) the budget of the Social Protection Fund (vi) a breakdown of capital expenditures (vii) estimates by each tax (viii) estimates of payments of external debt, including estimate of debt stock (ix) the government’s program on external debt (x) the budgets of local governments (xi) the budget execution report for the previous year. 29 See also IMF ROSC (2007), fn.16. Page 30 30 A ccording to discussions with the parliament’s budgetary committee, budget documentation submitted to parliament had improved significantly in recent years. Comparing the list of documents received by Tajikistan’s parliament with that set out under the PEFA indicator, it currently does not include the following items: - information on financial assets - explanation of budget implications of new policy initiatives and estimates of the budgetary impact of all major revenue policy changes and/or changes to expenditure programs. Furthermore, as noted under PI-5 (budget classification) and PI-27 (legislative scrutiny), while a relatively complete set of documents is received by parliament, the informational value with regard to the budget has been limited due to the absence of an administrative classification, and the separation between recurrent and investment funding in the presentation of the sectoral allocations. The latter has been improved in the 2007 budget plan; but the former still remains to be addressed. It would be important in particular to include more explanatory materials in the budget in future; as this helps in the understanding and policy-oriented discussion of budgetary issues in parliament. Indicator Brief explanation Rating (M1) 6. Comprehensiveness of information included in budget documentation Recent budget documentation fulfils at least 7 out of the 9 information benchmarks. A PI-7 Extent of unreported government operations Annual budget estimates, in-year execution reports, year-end financial statements and other fiscal reports for the public, should cover all budgetary and extra-budgetary activities of central government to allow a complete picture of central government revenue, expenditures across all categories, and financing. The indicator covers extra-budgetary activities of the government on the one hand, and expenditure information on donor funded projects on the other hand as the main categories of potentially unreported government operations. (i) The level of extra-budgetary expenditure (other than donor funded projects) which is unreported i.e. not included in fiscal reports As of early 2007, Tajikistan has no extra-budgetary funds (e.g. off-budget road funds or pension funds). However, there is an issue around the collection of so-called ‘special resources’, i.e. revenue generating activities by budget organizations at the central and local level, ranging from the issuing of passports by the Foreign Ministry to the renting out of school-rooms by local schools. Until 2007, such special revenues and related expenditures have not been systematically included in budget plans, although they have been included in budget execution reports. 30 As a result, there has been a substantial underestimation of non-tax revenue and expenditures in recent years. Special revenues are managed by local governments and budget organizations, and hence translate into additional expenditure by those organizations able to generate them. Furthermore, there is widespread anecdotal evidence of user fees in the health and education sectors. Since user fees are not legalized in many cases (i.e. except in cases of private service provision), these funds are also not included in fiscal reports. Social Protection Fund. Prior to December 2006 the SPF operated with a degree of independence, having its own bank account and being responsible for the payment of the majority of social 30 See also IMF ROSC (2007), p. 5. Page 31 31 r ecipients. In reporting terms, however, the State Budget included all expenditure and revenues of the Fund, as does the Budget Execution report. The IMF ROSC report does, however, make an important observation that “no systematic assessment of the SPF’s long-term financial sustainability had yet been undertaken”. The Fund was dissolved by presidential decree #9 of November 30, 2006, and the new arrangements established the Social Protection Agency (under the MLSP). The initial intention was to integrate the social fund payments into the single treasury at the end of 2006; however, this was not implemented and the Social Protection Agency still maintains a separate account. The collection of social security contributions is now allocated to the Tax Committee. So-called ‘special resources’ (‘ spetsial’nye sredstva ’ or own resources of budget organizations). Non-tax revenues in total include sale of goods by government, income of the National Bank, and income from communal services, as well as own resources (‘special resources’) generated by budget organizations. The latter were not included in estimations of non-tax revenue before 2006 (information on such funds was included as an annex, but not included in the budget law voted on by parliament; and in budget execution reports; and there is considerable inconsistency in the figures for this revenue category – see PI-3). At the suggestion of the IMF, own revenues of budget organizations are now included and shown throughout in the 2007 budget plan; as a separate category (i.e. separate from other non-tax revenue). Table 13: Tax revenue to BO ‘own resource’ revenue 2003-2005 Year/Item Total Somoni 000 Tax Revenue Somoni 000 Non Tax Revenue Somoni 000 Of which ‘special resources’ 000 SF % of total 2003 – Plan 643.8 574.9 68.9 48.5 7.5 2003 – Actual 786.6 712.6 74 49 6.2 2004 – Plan 923 824.3 98.7 79.2 8.5 2004 – Actual 1034.1 933.3 100.8 75.5 7.3 2005 – Plan 1298.3 1,148.0 150.3 115.7 8.9 2005 – Actual 1,360.5 1,192.5 160 115.7 8.5 2006 -- Plan 1610.2 1424 186.2 129.4 8 2006 -- Actual 1736.9 1543 193.9 129.4 7.5 Source: MOF; note: non-tax revenue includes: state duties, sanctions and fines, and so-called ‘extra-budgetary funds’. Special resources are only available to Budget Organizations (BO) and can be spent with the approval of oblast or rayon heads for local budget organizations, and of the MOF for republican budget organizations. BOs have considerable freedom in spending such resources and are also able to carry forward any unspent balances of own revenues as at 31 December. It is not possible for the study team to give any view on whether these resources are prudently spent on priority areas, although procedures do allow Treasury Staff to check all payment requests made. Unofficial Fees. Concerns have been raised in previous donor reports concerning unofficial charges levied in the Education and Health sectors by BOs. No direct evidence of this practice has been uncovered; but there is anecdotal evidence from various sources about such fees. Constitutional amendments undertaken in 2002 permitted the establishment of fee-based services in education and health; but unt il now the state has guaranteed access to free basic education and to “free medical assistance in state medical establishments” (art. 38 and 41). Overall, the level of unreported extra-budgetary expenditure appears to be low; but there is a substantial degree of uncertainty. While own revenues of BOs have been underestimated and the practices around these may need to be further regularized, they are reported on. We do assume that unofficial fees exist in the main service sectors (health and education) but that these amount to less than 5% of total expenditure (taking into account that government spending on health and education combined accounts for 27% of total primary expenditure, and that the amounts of these informal Page 32 32 c harges suggested by anecdotal evidence is relatively low). However, as these are truly unrecorded there is a great deal of uncertainty. The forthcoming results from the PETS may provide a more precise picture on this issue. Dimension (i): Score B (ii) Income/expenditure information on donor-funded projects which is included in fiscal reports The table below summarized the Overseas Development Aid (ODA) disbursed in Tajikistan in the period 2003 to 2005. Table 14: Development Assistance Total Disbursements (PIP and Non-PIP) 2003-2005 (in USD mn) 2003 2004 2005 GDP in USD 1,555 2,076 2,298 Total DA disbursed 182.6 192.2 179.6 PIP 47.2 82.9 70.2 Non-PIP 135.5 109.4 109.4 DA in % of GDP 11.7% 9.2% 7.8% Source : Aid Coordination Unit of the Administration of the President of RT. See also section 3.D regarding divergence in data on aid disbursements. The Aid Co-ordination Unit (ACU) has maintained a database of all aid projects and this appears to present a reasonably comprehensive position, although the database is not directly linked to fiscal reporting. Budget organizations are required to report any aid received to the MOF; but the Project Team could not verify that these procedures were complied with. Reporting procedures for official development assistance to date has been as follows: the government circulates a form among donors providing investment funding. Donors submit the completed form once they have approved a project. TA, such as provided by many bilateral donors, is reported twice annually to the ACU, but apparently not to the MOF. Direct budget support is negotiated with and reported to the MOF. The main budget only contains references to aid received as direct budget support. Grant aid that is not budget support is not currently included in fiscal documents. Aggregate budget tables reflect as grants received: Somoni 14mn for 2003 (~4.7mn USD), 41mn for 2004 (~ 13.5mn USD) and 31mn for 2005 (~10mn USD). Thus, in each year, less than 50 per cent of disbursed grants was included in budget reports. The PIP accounts for aid received as investment funding; it includes all loans from development partners (including concessional/grant elements). In sum, complete income and expenditure information for all loan financed projects is included in fiscal reports; but grants are included in the budget to less than 50 per cent. The Aid-Coordination Unit was dissolved in December 2006; and the aid coordination function was given to the newly created State Committee on Investment and Management of State Property. The brief in respect of donor project monitoring indicates that major improvements will be made, including reporting procedures. It is, however, far too early to say how the practical operations will work; as few concrete steps had been completed by April 2007. Page 33 33 Dimension (ii): C Indicator Brief explanation Rating (M1) 7. Extent of unreported government operations (i) The level of unreported extra-budgetary expenditure (other than donor funded projects) is less than 5% of total expenditure. Score B (ii) Complete income/expenditure information for all loan financed projects is included in fiscal reports. Score C C+ PI-8 Transparency of Inter-Governmental Fiscal Relations This indicator reflects the fact that sub-national governments play an important role in service delivery, and that they are important recipients and users of budgetary funds. The indicator is assessed based on three dimensions. The MOF directly interacts with 17 sub-national budgets (Dushanbe, three oblasts, 13 rayons of republican subordination). Governments of these SN entities in turn interact in the budget preparation and execution of the SN entities subordinated to them. Relevant legal acts are the 2002 Law on State Finance, and the 2004 Organic Law on Local Bodies of State Power, which contains articles regarding local public finance procedures and rules. (i) Transparent and rules based systems in the horizontal allocation among SN governments of unconditional and conditional transfers from central government Tajikistan does not currently have an overall system of formula or similarly rule-based resource allocations between levels of government, and as a consequence with regard to the horizontal allocation among local governments. However, there is a set of rules and practices. SN governments have three sources of income: own revenues – based on a revenue-sharing arrangements specified in the annual budget law; subventions from the republican budget to meet minimum funding levels in regions without sufficient own revenue; and subsidies which are provided to finance specific spheres (housing complex; culture). 31 In aggregate terms, more than two-thirds of local revenues are own revenues retained at local levels (see table 15). 32 The level of subventions is calculated based on a funding gap between expenditure needs and own revenue capacity. Expenditure needs are currently calculated on an input basis (e.g. classes/teacher workloads) (see also WB PEIR 2005, chap. 5) Table 15: Local budget revenue 2005 and 2006 (mn Somoni) 2005 2006 Plan actual % p to a % of loc rev Plan Actual % p to a % of loc rev Own revenues 285.1 318.5 111.7 68.8 404.9 439.7 109 73.9 Remaining funds from previous year 0 14.1 - 19.9 19.9 100 3.3 Targeted funds 95 95 100 20.5 122.6 122.6 100 20.6 31 In addition, the following functions at local levels are financed directly from the republican budget: law enforcement; land use committees; financial departments of local government; and courts. 32 The description here includes the allocation of revenue shares, as this is part of the resource transfer package between central and local governments. Page 34 34 (subvention) Additional funds from republican budget 0 31.9 - 6.9 7.2 7.2 100 1.2 Loans 0 0 0 0 0.4 - 0.1 Maintaining capital city function of Dushanbe 0 3.7 - 0.8 0 5 - 0.8 Total local revenues 380.1 463.2 110.3 554.7 594.9 107.2 Source: MOF Responsibility for service provision (central, local, and joint) is set out in the Law on State Finance, article 18 (with substantial responsibilities allocated to local levels). Furthermore, article 23 of the law specifies the principles of IGFR; and article 24 specifies what forms of financial assistance can be provided from the central to local budgets. The key principle is that the central government seeks to guarantee a minimum level of budget security to local governments and to ensure a financing of minimal state social standards in all localities. According to the MOF and other interlocutors, in highly subsidized regions, subventions will only be sufficient to cover protected expenditure items – mainly salaries and utility payments – and will not be sufficient to cover items such as school textbooks or maintenance. Funding allocations (including the share of taxes to be retained in each oblast as own revenue) are decided annually during budget negotiations between the oblast chairmen (or rayon chairmen in the case of the RRS) and the MOF. Apart from the calculation of expenditure needs, there are no pre- determined rules governing these negotiations. SN governments (as well as other budget users – see also PI 20) can request additional funding from the MOF as well as from the presidential reserve fund during the year. The allocation of these additional funds appears to be largely on an ad hoc basis (see also PI 9). There are efforts to move to a per capita based system of allocations in the social sectors (health and education). This has been piloted in five rayons since 2005. 33 Furthermore, in the context of the MTEF strategy, efforts are underway to return to a greater level of centralization of responsibilities and management of funding (see also PI 12). Education has been selected as the pilot sector for these efforts (which are currently at very initial stages; see section 4). Dimension (i): C (This assessment is based on the judgment that while there are rules covering resource allocations (tax sharing, subventions and subsidies), the process of how these are annually defined is not sufficiently clear and transparent; and furthermore, the rules for determining expenditure needs are regarded as deficient to cope with current challenges. The assessment is not based on quantitative criteria regarding the share of transfers covered by certain rules). (ii) Timeliness of reliable information to SN governments on their allocations from central government for the coming year. During the budget preparation process, SN governments initially receive relevant information on their allocations via the budget circular which is issued in April/May (the circular contains indications what incremental increases can be requested for certain expenditure categories, but no overall ceilings). In August, the 17 main SN governments conduct their budget negotiations with the MOF, which results in indicative ceilings for their expenditures. Some changes may still occur 33 Per capita allocations change certain incentives for budget users, and they may lead to greater equalization of expenditures across localities, but specifying practical and fair formula also entails problems; as have been experienced in Tajikistan during the piloting phase. Page 35 35 d uring the government’s and parliament’s review of the budget. The main outcome of budget negotiations are top-down cuts to budget proposals that have been collected bottom-up. These top- down allocations are in turn cascaded down, as higher levels distribute the reduced totals among sub-ordinated localities and budget organizations. In recent years, budgets have been adopted well in time before the beginning of the new fiscal year. By the time that the allocations from central government are firmly known, the main local budget planning process has already taken place. Actual revenues that have substantially exceeded planned revenues in some of the past years, and additional allocations from the center to local levels during budget execution add some degree of unreliability; in particular with regard to the effective allocation and use of funds. Dimension (ii): C (iii) Extent to which consolidated fiscal data (at least on revenue and expenditure) is collected and reported for general government according to sectoral categories. Local treasuries collect and report consolidated information about sectoral spending. Quarterly and annual budget execution reports cover local level sector spending, consistent with central government fiscal reporting. However, local treasuries do not track or report on detailed information of funds received by individual budget organizations such as schools or health clinics. (see also PI-23). Dimension (iii): A Indicator Brief explanation Rating (M2) 8. Transparency of Inter-Governmental Fiscal Relations The current system of IGFR is not sufficiently transparent and rule-based. Sub- national governments receive largely reliable information on the allocations to be transferred before the start of the fiscal year, but not before the start of their budget planning process. Reporting of consolidated data on sectoral spending at local levels does not pose a problem. Scores: C, C, A B PI-9 Oversight of aggregate fiscal risk from other public sector entities Fiscal risks may emanate from autonomous government agencies (AGAs), sub-national governments (SNs) and Public Enterprises (PE). Central government should require and receive quarterly financial statements and audited year-end statements from AGAs and PEs, and monitor performance against financial targets. (i) Extent of central government monitoring of AGAs and PEs. The main area of fiscal risk is in respect of State Owned Enterprises (SOE). It has not been possible to obtain firm figures for the number of SOEs currently in operation. In 2005 it appears that the number exceeded 500. This lack of data is despite the fact that many Government Ministries/Committees interact with SOEs: a. Line Ministries should receive regular reports from the SOEs under their control; b. The MOF Department on State Debt and State Financial Assets receive quarterly and annual reports from each SOE; Page 36 36 c . The SFCC audits every two years the accounts of each SOE; d. The State Property Committ ee control the ongoing “privatization” process; e. The Tax Committee (formerly MSRD) maintains a tax registry that includes SOEs. However, no government organization undertake a consolidated exercise to establish the overall financial risk with which the government may be faced. Although data on poorly performing SOEs and ones that may face a risk of bankruptcy is available, no financial assessment is made of the risk to the Government. In accounting terms, SOEs are moving to commercial International Accounting Standards but some difficulties are apparently being experienced. There is no evidence that SOEs are obliged to use external audit services. A separate department of the SFCC undertake an audit (every two years) of each Joint Stock Company where the Government’s stake holding exceeds 50 per cent. Again it is a compliance type audit and no assessment of financial risk takes place. An active privatisation programme is in place. The Department of State Property claims that over 700 SOEs will have been switched to the private sector in the period 2003 to 2007. However, IMF figures suggest that 459 SOE’s are being privatised in the period 2003 to 2007. In 2007, the State Property Committee intends to prioritize the restructuring of 37 Natural Monopolies. The GOT hopes to attract foreign investment for the restructuring but intends to maintain substantial ownership (see also PI 17). 34 Furthermore, SOEs are among the enterprises with the largest tax arrears – by December 2006, Barki Tajik (energy company) had accumulated Somoni 63.6m in tax arrears, representing 28 per cent of the total stock of arrears. The overall tax collections for 2006 were Somoni 763.9m and total arrears at year end were 224m (see also PI-15). 35 As indicated above although several Government bodies have involvement with SOEs, there is no evidence of any co-ordination and consolidation of reports. A more focused approach with emphasis on monitoring any potential risk and identifying SOEs suitable for transfer to the private sector is required. Quasi Fiscal Activities (QFA). There is particular concern in respect of SOEs in the energy and transport sectors, where quasi fiscal deficits arise as a consequence of below-cost tariffs and poor collection rates. The quasi fiscal deficit in the energy sector was estimated in early 2005 by the IMF at about 10 per cent of GDP – largely because of under-pricing. Such issues are not covered in the Budget or in Budget Execution reports. It is noted that there was a 25 per cent increase in electricity tariffs enacted from February 1, and that gas prices are planned to be increased. However, consolidated reporting and monitoring of these issues remains insufficient. Agreement with National Bank. The GOT recently entered into a 20 year agreement with the National Bank to “service” defaulted loan credits of 153m Somoni issued by the NBT (not in accordance with the law) mainly to the agricultural sector in the late 1990s. This involves the quarterly payment of interest at the inflation rate plus 1 per cent. Dimension (i): C 34 It is understood that the State Property Committee is also seeking further financial assistance from the WB to examine the remaining SOEs with a view to collecting data on the capital assets being utilised and consideration of further privatisation. 35 These figures refer to revenue collectable by the Tax Committee only (i.e. does not cover tariffs and taxes collected by the Customs Committee or social protection collections). Page 37 37 (ii) Extent of central government monitoring of SN governments’ fiscal position. Transfers to sub national government are strictly controlled through the quarterly cash limit process and the budget smetas cannot be exceeded unless approval is given to an increase by the Government of the “subvention”, or by the Local Authority in respect of own revenues. The Local Authorities automatically receive a share of certain revenues, which are additional to the government subvention and, importantly, are empowered to spend any extra revenues that are collected above the budgeted figure. There is also a provision that allows local authorities to overspend by 3 per cent if sufficient financial resources are not available. 36 The extent of the additional revenues is of particular concern because, as can be seen from the table below, the budget plan significantly understates the level of expenditure that is eventually incurred. In 2005 and 2006 the position has been improved, but it remains a significant influence – an aspect further explored in PI 16. Table 16: Local Budget Revenues and Expenditures (mn Somoni) 2003 2004 2005 2006 plan actual % plan actual % plan actual % plan actual % total revenues 178.9 261.1 145.9 246.7 341.9 138.6 380.1 463.2 121.9 554.7 594.9 107.2 total expenditures 178.9 240.1 134.2 246.7 325.5 131.9 380.1 443.4 116.7 .. .. .. Source : MOF, Department of Local Budgets. The MOF is informed of the individual budgets approved and can monitor actual budget execution. There is a formal process which the Local Authorities must observe to obtain amendments to their Smeta. The local Treasuries will only allow local governments to spend in accordance with the approved Smeta. Information on local government’s budget execution position is prepared by the Treasury each quarter. This information includes details of total accounts payable not paid at the end of each quarter, provided by each local government. Monitoring of the budget execution process is strong. However, no consolidated report on overall fiscal risk from SN levels is being produced (as required for a rating of B or higher). Dimension (ii): C Indicator Brief explanation Rating (Method M1) PI-9. Oversight of aggregate fiscal risk from other public sector entities (i) Most major AGAs/PEs submit fiscal reports to central governments at least annually, but a consolidated overview is missing or significantly incomplete. Score C (ii) The net fiscal position is monitored at least annually for the most important level of SN government. Score C C PI-10 Public access to key fiscal information 36 According to the 2002 Law on State Finances, article 21: “The current expenditures of local budgets cannot exceed the cumulative incomes of given budgets. […] The size of the deficiency of local budgets cannot exceed three percents of the volume of their incomes without taking into account the financial assistance from the state budget of the Republic of Tajikistan.” Page 38 38 A ccountability to citizens has become an increasingly important issue in recent thinking about PFM reforms. Accessibility of information is a key precondition for accountability. The PEFA methodology defines six elements of documentation which should be publicly accessible. The 2002 Law on State Finances mentions “publicity (openness) of budgets” as one of the principles of TJ’s budget system (article 8). According to the constitution, article 25, government organs are under obligation to provide documents to individuals whose interests or rights are affected. 37 Tajikistan furthermore has a Law on Information (2002) and a Law on State Secrets (2003). Given capacity and resource constraints, little information on the government and its activities is currently available online. At the same time, demand for more information from civil society appears to be low. That said, the government is making some efforts to publish fiscal information in summary form mainly in special government press organs. The SFCC also mentioned that it holds quarterly press conferences at which it presents and discusses key audit findings. Press conferences on budget issues are said to also take place at local levels; and this practice has been endorsed by the president. The following map the information requirements according to the PEFA framework against current practices in Tajikistan. (i) Annual budget documentation: A complete set of documents (as according to PI-6) can be obtained by the public through appropriate means when it is submitted to the legislature. Æ In Tajikistan, basic budget indicators are published in newspapers after approval by parliament. Easier dissemination via an enhanced web-site is currently under preparation. It is currently difficult for the public to obtain a complete set of documents of annual budget documentation. Publishing budget documentation only after approval by parliaments means that public scrutiny of plans is not possible before decisions have been taken. (ii) In-year budget execution reports: The reports are routinely made available to the public through appropriate means within one month of their completion. Æ Quarterly budget execution reports are published in the government newspaper Giumhuriyat ; and in a magazine, Moliya ; however, the published reports made available to the study team only concern health and education, and not the budget in its entirety. (iii) Year-end financial statements: The statements are made available to the public through appropriate means within six months of completed audit Æ Currently, no audited year-end financial statements are prepared on an annual basis (see indicator PI-26); and therefore, no statements are published. (iv) External audit reports: All reports on central government consolidated operations are made available to the public through appropriate means within six months of completed audit. 37 Constitution, article 25: “Governmental organs, social associations, and officials are obligated to provide each person with the possibility of receiving and becoming acquainted with documents that affect her or his rights and interests, except in cases anticipated by law.” Page 39 39 Æ Only summary reports submitted to parliament by the SFCC; and only excerpts and summaries are available to the public. (v) Contract awards: Awards of all contracts with value above approximately USD 100,000 equivalent are published at least quarterly through appropriate means. Æ Contract awards are currently not being published. The Procurement Agency plans to produce publications in future. (vi) Resources made available to primary service units: Information is publicized through appropriate means at least annually, or available on request, for primary service units with national coverage in at least two sectors (such as elementary schools or primary health care clinics). Æ Such information is currently not published. Indicator Brief explanation Rating 10. Public access to key fiscal information Some information is being made public, but not to the extent or in ways compatible with the requirements of the PEFA framework. However, the government is taking some steps at informing citizens better. D C. BUDGET CYCLE C (i) Policy based budgeting PI-11 Orderliness and participation in the annual budget process There have been improvements over time. The 2002 Law on State Finance has contributed to clarifying roles and processes (although further legal development is necessary in future as budget processes continue to be modernized). Efforts to strengthen fiscal forecasting are underway and renewed efforts to develop a medium term framework are planned (see also PI 12). In preparing the 2007 budget, ministries were asked to submit sector strategies – however, not all have done so in practice. Recent budget circulars appear rather comprehensive, but their clarity and emphasis on more strategic rather than incremental issues is clearly an area for further improvement. The involvement of the cabinet in the budget preparation process is still weak; which implies that budget preparation is currently not rooted in a clear and structured discussion about policy priorities and trade-offs. A further key weakness is the fact that the budget circular does not contain explicit budget ceilings. Moreover, due to the fragmented system of government, political responsibility for managing funds and achieving results is likewise fragmented. 38 (i) Existence of and adherence to a fixed budget calendar Two key mileposts of the budget preparation calendar are permanently fixed in the 2002 Law on State Finance: submission to the government, Sept. 20; and submission to parliament: by Nov. 1. 39 The detailed calendar for the annual budget process is defined in a government decree that is being 38 See ProGramma, (Horizontal functional analysis of executive state organs of the Republic of Tajikistan) (December 2006) – Presentation of findings for the World Bank’s proposed Public Sector Reform project. 39 The ex-post calendar specified in the Law on State Finance is as follows: BO’s have to submit budget execution reports by March 1; and MOF has to prepare an annual budget execution report by May 1 to submit to the Government; and by June 1 to parliament. Page 40 40 i ssued in February by the Government to the Ministry of Finance; and subsequently in the budget circular issued by the MOF (usually in April-May) to MDAs. The dates set out in the Law on State Finance have been met in recent years. However, schedules established by the February Government Decree and by the MoF Budget Circular have not all been observed, for example, during the 2007 budget preparation process (the Budget Circular was distributed to key budget organizations with a 2 week delay on May 30 rather than May 15; and several forward estimates – of revenues, of utility prices, and projections of socio-economic development were submitted after the established schedule of June 10. Dimension (i): B (ii) Clarity/comprehensiveness of and political involvement in the guidance on the preparation of budget submissions A Government Budget Commission has been established in 2004; 40 but its impact has remained limited. The main decisions during the budget preparation process are taken during the budget negotiations between the MOF budget commission and the MDAs and SN governments concerned; and through instructions issued by the president. There are no explicit budget ceilings at the outset of the budget preparation process (only some indication of expected overall resource increases is given). Consequently, many MDAs and SNs submit unrealistic budget requests. Dimension (ii): D (iii) Timely budget approval by the legislature or similarly mandated body In the last three years, the budget has been approved in time. Annual Budget Law dates are as follows: 2003 – December 2, 2002; 2004 – December 16, 2003; 2005 – December 17, 2004; 2006 – December 9, 2005. Dimension (iii): A Indicator Brief explanation Rating 11. Orderliness and participation in the annual budget process The budget process is broadly orderly in terms of timely adoption of the budget; however, there are still delays during the internal budget preparation process. No ceilings are issued to MDAs. B PI-12 Multi-year perspective in fiscal planning, expenditure policy and budgeting (i) Preparation of multi-year fiscal forecasts and functional allocations To date, the multi-year perspective in fiscal planning is weak, but evolving. An initial attempt to introduce an MTBF was made in the early 2000s. Currently, a renewed effort at developing an MTEF – which would entail a range of major changes to the budgeting process – is under discussion between donors and the government. 41 New regulations to develop the MTEF as an instrument to implement Tajikistan’s strategic plans – the NDS and the PRSP – have been adopted in September 2006. Macro-economic projections are developed by the MOET. The MOF prepares forecasts of revenues initially in March/April, which are subsequently updated several times. These forecasts include a three-year perspective, which is being produced on a rolling basis. The MOF also prepares expenditure estimates in total and by main functions. However, participants in the budget 40 Government Decree no. 294, June 30, 2004 “On the permanent budget commission”. 41 This effort comprises several changes; see section 4. Page 41 41 p reparation process agree that these forward estimates currently do not meaningfully guide the annual budget process.. Dimension (i): C (ii) Scope and frequency of debt sustainability analysis Debt emerged as a major policy challenge in Tajikistan in the course of the 1990s. Existing debt is mainly external. A thorough Debt Sustainability Analysis has been carried out in 2005, and an update is currently under way. As noted in section 2 above, Tajikistan has benefited from the cancellation of substantial shares of its external debt in recent years. The main responsibility for debt management rests with the Government Debt and Government Financial Assets Department in the Ministry of Finance (for further detail see PI 17). The Debt Sustainability Analyses which have been undertaken to date only cover external debt; therefore only a score of C can be allocated. A debt strategy was under development in 2005; but then progress stalled when new major loans were negotiated with China. Currently, a renewed effort at developing a strategy (covering both external and internal debt) is underway and is being supported by TA from the IMF. Dimension (ii): C (iii) Existence of sector strategies with multi-year costing of recurrent and investment expenditure The 2006 budget circular asked ministries to submit sector strategies; and in the 2007 budget circular, this is reinforced by demanding that budget organizations have to link their priorities to those set out in the National Development Strategy, and to provide costing estimates, including estimates for investment projects. However, the process of developing sector strategy is still in its infancy. Some ministries are moving ahead and are receiving external support. 42 The only sector for which a detailed and costed sector strategy exists at the moment is education. 43 Education accounts for about 20% of primary expenditure. For the PRSP2 strategies have been developed for the main other sectors (health, social protection, agriculture, public administration, energy, etc.), but these are not substantively and realistically costed yet. In general, the process of developing sector strategies will take time and will require substantial support for capacity development to sector ministries – especially as these have played limited roles in Tajikistan’s system of government to date. The process of designing meaningful sector strategies in principle will require a clearer structure of sectoral responsibilities. For example, in the education sector, many republican level educational institutions (PTUs – Professional Technical Schools) have been under the responsibility of the Ministry of Labour rather than of Education. In the re-organization of government that was undertaken in December 2006, some but not all of these issues of fragmented responsibilities have been addressed. Dimension (iii): C (iv) Linkages between investment budgets and forward expenditure estimates. Although initial steps are being undertaken to link investment and recurrent budgeting (annual updates of the PIP, budget circular demands), the two still remain rather separate. Moreover, estimates of future recurrent costs are not part of the process of planning and selecting investment projects at the moment. 44 As noted above, the domestically financed and the externally financed investment plans are not linked. There seems to be little awareness at the moment about the need for forward expenditure estimates. Dimension (iv): D 42 The EC has supported strategy development and budget submissions by the Ministry of Labour and Social Protection (MLSP); the Ministry of Education is working on developing its sector strategy and sector MTEF. 43 National Strategy for Educational Development of the RoT, Ministry of Education, August 2005; an English version is available at www.untj.org 44 See also IMF ROSC 2007, p.13. Page 42 42 Indicator Brief explanation Rating (M2) 12. Multi-year perspective in fiscal planning, expenditure policy and budgeting The fiscal framework, sector strategies, and linkages between investment budgets and forward expenditure estimates are still weak. DSA is good for external debt, but domestic debt is currently not covered. All these areas are currently subject to further reform. D+ C (ii) Predictability and Control in Budget Execution Revenues: policy, collection, and administration A new Tax Code and a new Customs Code came into effect January 1, 2005. 45 Implementing regulations for these codes are still under development. As discussed in indicators PI-13 to 15, basic tax administration capacity has been developed However, on the one hand, tax arrears are high; while on the other hand, tax policy and administration are not sufficiently geared towards creating a good investment climate; and frequent tax inspections have created unnecessary burdens for taxpayers. Also, significant parts of the system are currently rather non-transparent, including the presence of substantial tax exemptions and apparently considerable scope for negotiating tax payments between tax inspectors and taxpayers. 46 These problems contribute to low levels of revenue collection as a share of GDP. The tax administration has a total of 1,630 staff in 78 tax inspectorates. 47 A Large Taxpayer Unit was set up in 1999. Functional departments for registration and audit have been established. Reforms of tax policy and tax administration have been supported by donors (USAID, ADB, IMF). Until the restructuring of the government in December 2006, tax and customs administration were joined in the Ministry of State Revenue and Duties (MSRD); this ministry has now been dissolved and separate Tax and Customs Committees have been established “under the government”. This leaves accountability for weaknesses in revenue administration and for reform processes in this area rather unclear. Significant problems with tax arrears are concentrated around large taxpayers. Substantial tax arrears and their occasional clearance also affect the reliability of revenue forecasts (see PI 3). Furthermore, significant corruption problems are reported to continue to exist in the tax and customs administrations; a problem which Tajikistan shares with many countries in the region. 48 PI-13 Transparency of taxpayer obligations and liabilities (i) Clarity and comprehensiveness of tax liabilities The 2005 Tax and Customs Codes set out liabilities in general and these are generally clear. They include provisions on the rights and obligations of taxpayers. However, the fact that implementation regulation has not been completed means that there is procedural uncertainty for taxpayers 45 Tax Code – Law no. #61 and Customs Code – Law no. #62, December 3, 2004. 46 This is also supported by available survey evidence; see SECO/IFC, Business Environment in Tajikistan as Seen by Small and Medium Enterprises , Preliminary Survey Results, June 29, 2006. 47 As of April 2007; according to information by the Tax Committee. 48 See SECO/IFC, Business Environment in Tajikistan as Seen by Small and Medium Enterprises , Preliminary Survey Results, June 29, 2006. Page 43 43 ( preparation of further implementing regulations is currently under way). In particular, uncertainty is reported with regard to application of the Simplified Tax for SMEs. Furthermore, the high numbers of inspections and also the widespread imposition of additional payments for “landscape” improvements reported in surveys of SMEs indicate that de facto, there are still substantial discretionary powers (see also PI-14, iii). 49 Furthermore, according to the IMF “exemptions from direct and indirect taxes are extensive”; and existing exemptions are not reflected in budget documents. 50 No estimation of the cost of tax exemptions has been carried out thus far either by the GoT or the IMF. The Tax Committee has been asked by the MOF to provide such an estimate, but is still in the process of defining a suitable approach for this task. Dimension (i): C (ii) T axpayers’ access to information on tax liabilities and administrative procedures Tax service areas have been created in local tax administration offices. Currently, no information on tax obligations and procedures is accessible over the phone; however, in February 2007, the tax administration launched a website (www.andoz.tj). In the service areas, taxpayers can in principle receive all necessary information; however, there may still be problems with the actual training of tax administration staff, and the reliability of advice to taxpayers. In July 2006, amendments to the Tax Code were adopted (Law no. 193, July 28, 2006 “On the introduction of changes and amendments to the Tax Code of the Republic of Tajikistan”), which strengthened the rights of taxpayers to demand information from the tax authorities with regard to any questions related to taxation and other obligatory payments to the budget. Tax authorities have also conducted some information campaigns. Dimension (ii): C (iii) Existence and functioning of an appeals mechanism Key provisions for the appeals mechanisms available to taxpayers are set out in chapter 11 of the Tax Code. If taxpayers want to dispute a tax obligation, they can do so in written form to the tax inspectorate that specified the tax in the first instance; and the inspectorate is obliged to respond within a month. If the taxpayer is not satisfied with the response, he/she can turn to the next level (e.g. oblast level tax inspectorate); and from this level on to the Ministry of State Revenue and Duties (MSRD) where there is a special tax dispute forum staffed by specialists. Since the introduction of the 2005 Tax Code, taxpayers can also immediately take their appeal to a court (but the lack of courts that are well qualified to deal with such matters was pointed out as a problem both by representatives of the tax administration and of taxpayers). At the Ministry of Revenue (now: the Tax Committee), a tax dispute forum was established which is staffed by specialists for the various taxes. The forum can issue a recommendation on how to solve a tax dispute. Interviewed MSRD staff indicated that there were no more than 20-30 disputes countrywide per year of any form. 51 Information about tax disputes (and their resolution) is not published. Dimension (iii): C Indicator Brief explanation Rating 13. Transparency of taxpayer obligations and liabilities The Tax Code constitutes an important improvement; however, implementing regulations still need to be completed; and C 49 SECO/IFC (2006). 50 IMF ROSC (2007), p. 9. 51 Due to the ongoing restructuring process during the main mission, tax authorities did not provide any further detail on this issue. It is not clear whether elements of the dispute mechanism have been changed subsequent to the restructuring of the MSRD. Page 44 44 discretionary powers of revenue authorities still exist. The Tax Committee is undertaking laudable efforts to improve access to information, but this remains hampered by the incompleteness of implementing regulations. Basic access to a complaints mechanism exists; but further development is needed. PI-14 Effectiveness of measures for taxpayer registration and tax assessment (i) Controls in the taxpayer registration system Taxpayers are registered with their local tax office only, and no overall database for registering taxpayers is currently operational. USAID has supported the development of a central database with limited functionality, covering legal persons registered as taxpayers, but this remains far from completed, and is currently not electronically accessible either by central or by local tax organs, and is not regularly updated. Local tax offices issue taxpayer identification numbers (TINs) which are required for business licensing and for using financial services. 52 Since 2005, TINs are required to make tax payments (but it is not clear whether this is consistently enforced). Inter-departmental checks are limited, and no routine exchange of information between tax and customs authorities, nor between tax authorities and other institutions (business registry, land committee, etc.) is taking place. By the end of 2006, approximately 27,500 legal persons and 102,060 individual taxpayers were registered in Tajikistan, but of the latter category, only 23,103 were included in the pilot database. Tax inspectors conduct checks of compliance with tax registration. Dimension (i): D (ii) Effectiveness of penalties for non-compliance with registration and tax declaration Penalties for non-compliance were taken out of the new Tax and Customs Code and were inserted as amendments to an older Law on Administrative Offences . 53 Penalties are regarded as relatively low by MSRD staff. Fines for non-registration (or late registration) are defined in terms of multiples of the minimum wage (up to a maximum of 30 minimum wages for legal persons who fail to register for more than 90 days 54 ), while fines for declarations lowering the tax obligation and non- payment of taxes are in the range of 10 to 20 per cent of the amount that has not been paid. Given the level of interest rates charged by the banking sector to enterprises the MSRD considered these fines to be insufficient to deter taxpayers from avoiding taxes. The MSRD also raised the issue that it has insufficient qualified staff for taking taxpayers to court. The tax police becomes active only in cases of deliberate tax evasion (keeping of two sets of books, etc.); but not when taxpayers declare their obligations but fail to pay. Particularly in the case of large taxpayers, the MSRD’s options for ensuring the payment of overdue tax payment are limited. Submitting cases to courts is time-consuming; and there is a lack of specialist legal staff in most tax 52 The process is based on Instruction no. 141 (2000) on the registration of taxpayers which is currently being revised to bring it into conformity with the 2004 Tax Code. The system has been developed with support from USAID. 53 Article 165, point 1-13, as amended by law 74 of December 9, 2004. 54 The minimum wage in 2006 stood at Somoni 20, so the maximum fine was in the range of USD170. Page 45 45 i nspectorates. In the government restructuring in late 2006, the tax police was dissolved and its functions were distributed between the Tax Committee and the newly established State Committee on Financial Control and Fighting Corruption. Dimension (ii): C (iii) Planning and monitoring of tax audit programs A tax audit department has been created in the MSRD; and there are 45 dedicated tax auditors who have received some training. The main legal base for tax audits is chapter 4 of the 2004 Tax Code. Audit (or rather verification/inspection – Russian proverka ) plans are not based on risk assessments, but are based on schedules. The Tax Code envisage three types of verification: (i) cameral/documentary inspections; (ii) on-site audits; and (iii) time-study surveys (Tax Code, art. 48). A particular problem is that regular tax inspectors also carry out frequent on-site visits, according to surveys among SMEs. While a new Law on Inspections was recently adopted (Law no. 223, July 20, 2006) to reduce the burden on businesses, this law does not cover tax inspections and audit. The number of tax audits was 6975 in 2005 and 6186 in 2006 according to the Tax Committee. Tax inspectors as well as dedicated auditors can participate in audits, but must be registered in the audit plan. There is still a question as to why surveys of SMEs reflect a much higher number of inspections by tax officials than recorded by the tax committee. The audit department of the Tax Committee (formerly MSRD) is responsible for planning tax audits. Dimension (iii): C Indicator Brief explanation Rating 14. Effectiveness of measures for taxpayer registration and tax assessment Systems for registering taxpayers currently lack effective controls. The system of penalties for non-compliance with registration and tax declaration has weaknesses de jure and in actual practice. The audit role appears not yet fully established and distinguished from the tax assessment; and it is not based on a risk assessment but on schedules. D+ PI-15 Effectiveness in the collection of tax payments Tajikistan’s tax revenue ranks among the lowest in the CIS region. Its average ratio of tax revenue to GDP for the 1997-2002 period came to 13.2 percent of GDP. In recent years, this ratio has increased to around 16 per cent in 2004 and 2005. This ratio is low by regional standards, although still in line with other low income countries (e.g. in SSA, Latin America – many of which also suffer from tax collection problems). There is likely to be scope for increasing revenues without creating an unduly high tax burden by improving tax administration and reducing tax exemptions. 55 As a consequence of low levels of revenue collection, the government’s resources to provide public services have been very limited over a number of years – especially in contrast to the pre-1991 period when the public sector enjoyed transfers from the Soviet Union’s budget. 55 Tajikistan currently has rates on main taxes which are similar to those of other CIS countries; but alongside Georgia and Armenia has the lowest actual collection rates in the region. Page 46 46 (i) Collection ratio for gross tax arrears, being the percentage of tax arrears at the beginning of a fiscal year, which was collected during that fiscal year The ratio of tax arrears (stock) to tax revenues is significant. When calculated for taxes under the responsibility of the tax department of the MSRD the ratio exceeds 20 per cent; when calculated for revenue raised by the tax and the customs departments it exceeds 10 per cent (for 2004 and 2005) (see table 17). 56 In early 2007, tax arrears amounted to 30 per cent of all taxes collected by the tax department in 2006 (736.9m Somoni); and arrears for social contributions stood at around 90m Somoni, compared to total SPF collections for 2006 of 189.4m Somoni. 57 Records on the non-payment of taxes are kept by tax type and by tax payer, but not by year when the tax debt was incurred (i.e. there is no age profile). Therefore, the annual revenue debt collection ratio cannot be determined. However, year-to-year changes in the stock of tax arrears suggest that the collection ratio is below 60 per cent. According to the 2005 Tax Code, tax payments are to be written off after 6 years (art. 86); however, since records do not include an age profile, currently this provision cannot be put into practice. Tax amnesties have repeatedly been used in a number of CIS countries in recent years. The last major tax amnesty in Tajikistan occurred in 2003; and there may be expectations of further amnesties. Following recommendations by the IMF, a list of the largest tax debtors is published quarterly. According to this list, somewhat less than half of outstanding tax debts is owed by large taxpayers. Table 17: Tax debt as a share of total tax revenue (in %) year to Jan 04 year to Jan 05 year to Jan 06 Tax arrears stock/ annual tax revenue 37.8 23.4 27.2 Large taxpayer arrears stock/total annual tax revenue 18.7 10.9 10.7 Arrears stock for tax and customs collections/total revenue collected by tax and customs 25.3 14.7 16.7 Source : MSRD Dimension (i): not assessable, due to lack of age-profile of tax arrears (ii) Effectiveness of transfer of tax collections to the Treasury by the revenue administration. Taxpayers pay taxes via commercial banks; which then transfer tax payments to the treasury account of the local government and to transit accounts of the central Treasury respectively. Funds in transit accounts are forwarded to the central treasury account daily. There have been delays in the forwarding of tax payments from commercial banks to the accounts operated by the treasury, but these are now reported to be short (no longer than 2 days). 58 Dimension (ii): B (iii) Frequency of complete accounts reconciliation between tax assessments, collections, arrears records and receipts by the Treasury 56 The Custom’s department collects not only customs dues, but also sales taxes on imports, and this constitutes a substantial share of combined revenues (>40 per cent). 57 Again, these are accrued arrears which have accumulated over several years, but an age profile is not available. 58 The Law on Administrative Offences, article 165.6 envisages a fine in the order of 10 per cent of the sum not transferred in time. Page 47 47 A t the central level, a weekly reconciliation process between Treasury and the MSRD (now: the Tax Committee) has been established, covering the level of actual collections received. At local levels, the reconciliation process is due to take place on a monthly basis (by the 3 rd of the following month) and records of the reconciliation are submitted to the central Treasury. This is reconfirmed by a monthly reconciliation process between the central Treasury and the Tax Committee that is undertaken by the 10 th of the following month. However, there is no effective reconciliation between tax assessment and actual collections; only a monthly update of tax arrears. Tax records are maintained manually in most places which limits opportunities for centralized control and monitoring. Dimension (iii): D Indicator Brief explanation Rating (M1) 15. Effectiveness in the collection of tax payments High tax arrears constitute a key area of concern. The effectiveness of transfers of tax collections to the Treasury appears to be in a relatively good state. The overall reconciliation between tax assessments, collections, arrears records and receipts by the Treasury has important gaps and does not meet the PEFA criteria; reflecting important weaknesses in the systems of the tax administration. No overall rating is possible D (i): not assessable D (ii): B D (iii): D PI-16 Predictability in the availability of funds for commitment of expenditures Effective execution of the budget, in accordance with the work plans, requires that the spending ministries, departments and agencies (MDAs) receive reliable information on availability of funds within which they can commit expenditure. This indicator assesses the extent to which the MOF provides reliable information on the availability of funds to MDAs. (i) Extent to which cash flows are forecast and monitored The overall budget strategy is based upon the level of estimated revenues determining the affordable expenditure levels. The MOF assesses the quarterly revenue flow for the year ahead and determines quarterly cash limits for KBOs and subventions to local governments. Higher level BOs will notify those BOs subordinated to them of the annual budget approved and the quarterly cash limits. BOs prepare their smetas (quarterly expenditure estimates) for submission to the MOF (Budget Department and Treasury), either directly or through the respective higher-level institutions. 59 The MOF will verify that the overall Rospis (total quarterly breakdown of expenditures) agrees with the functional breakdown of the budget and quarterly cash limits. The control figures based on the approved smetas are then entered into the Treasury computer system for each institution, by detailed economic classification. They are rigidly applied and cannot be exceeded. If BOs require any changes there are established procedures (described in dimension (iii) below) to amend the set limits. 59 Many BOs de facto have had a choice of agreeing the smeta directly with the MOF or with their line ministry first. Page 48 48 T he Treasury monitors and updates cash flow in respect of the republican budget on a weekly basis. However, because a conservative approach has been taken to forecasting revenues and in recent years actual revenues have exceeded the planned levels, the strength of cash flow monitoring in the face of strain has not been tested. Cash flow forecasting and monitoring will become more important once budgeted expenditures are again close to or by a degree exceed revenues Dimension (i): B (ii) Reliability and horizon of periodic in-year information to MDAs on ceilings for expenditure commitment The process described above gives heads of budget organizations an expenditure plan at the beginning of the fiscal year – a process in which the head will have had some involvement, but probably little influence (with variation depending on the level of the BO). It is, nonetheless a reliable and understood process, which appears to give predictability for budget users at least for a quarter in advance. 60 As will be illustrated below (see also PI-3) there have been no in-year reductions of budget allocations for some years – only in-year increases. There may at times be issues with the predictability of the schedule for releasing capital funds. Dimension (ii): B (iii) Frequency and transparency of adjustments to budget allocations, which are decided above the level of management of MDAs The present arrangements allow the MOF to impose reductions in quarterly cash limits, in the event of a significant revenue shortfall. Priority for payment is established centrally – with staff payments, communal services and debt repayment taking precedence according to existing rules. There are clear guidelines for the priority areas, and these appear to be adhered to. Any reductions to quarterly cash limits imposed by the MOF would be extremely disruptive to BOs – but very few reductions if any have been imposed in the last 4 years. Instead, we have a situation where upward adjustments have been regularly made; and the ways in which such adjustments have been made have been too numerous and overall not sufficiently transparent: a. Central and SN governments have allocated additional funds in-year. BOs, at both republican and local levels can “bid” for such additional funds. In the current interpretation of the Law on Public Finance, approval by the national parliament is regarded as only required if the Republican share of the budget is exceeded by more than 10 per cent (see also indicator PI-1). b. According to the Law on State Finance, article 42, 2 per cent of the state budget is allocated to the Presidential Fund and can be used for additional allocations. c. There is the ability for individual BOs to spend the EBR they collect. As can be seen from the table in PI 7, in 2005 ‘excess’ EBR amounted to 8 per cent of overall revenues. d. Finally there are the virement procedures for amending Budget/Cash figures at Republic and Local levels. These are commendable “virement” procedures that allow budget 60 This refers primarily to higher level budget organizations. Heads of budget organizations at the lowest levels, e.g. headmasters of schools may not be fully informed about the budget allocation they are due to receive for the year. This issue could not explored in-depth as part of the PEFA assessment. Page 49 49 o rganizations to move monies from one budget head to another(s). 61 There are restrictions imposed on protected items, salaries, utility payments, etc. Apart from these protected items, virement limits are wide (up to 20 per cent of a BOs budget allocation; and this limit can be exceeded if so agreed between the MOF and the BO). The overall consequence is that the budget is not treated as a binding financial plan for the year. The credibility of the Budget is undermined and it cannot possibly provide a firm base for developing a MTEF. A culture now exists whereby Line Ministries and Budget Organizations do not regard the budget as a limit – they expect to be able to bid for increases and receive approval for some of those bids. The Government’s freedom to allocate excess revenue up to a limit of 10 per cent without parliamentary approval is of particularly concern. The process is not transparent and there is no way of determining whether the extra budgets are allocated on a rational and fair basis. The substantial allocations from the Presidential Fund (amounting to 2 per cent of total expenditures) are also outside normal parliamentary controls and transparency is an issue. Dimension (iii): D An IMF report was issued in October 2005 and contained a number of significant recommendations. The Treasury has made progress towards implementing some of these recommendations. A Cash Management Unit was restructured in early 2006, but it is not yet producing viable cash flow forecasts/monitoring. A further re-structuring is envisaged. [Sentence deleted] The Treasury is currently negotiating with the WB on the design process for a new treasury system (see section 4 below). Indicator Brief Explanation Rating (Method M1) 16. Predictability in the availability of funds for commitment of expenditures (i) A cash flow forecast is prepared for the fiscal year, and updated at least quarterly, on the basis of actual cash inflows and outflows. Score B (ii) MDAs are provided reliable information on commitment ceilings at least quarterly in advance. Score B (iii) Significant in-year budget adjustments are frequent and not done in a transparent manner. Score D D+ PI-17 Recording and Management of Cash Balances, Debt and Guarantees Debt management, in terms of contracting, servicing and repayment, and the provision of government guarantees are often major elements of overall fiscal management. Poor management of debt and guarantees can create unnecessarily high debt service costs and create significant fiscal risk. (i) Quality of debt data recording and reporting 61 When MDAs want to change their allocations (from the smeta originally approved by the MOF) they have to submit a written request to the MOF. A kurator from the treasury department checks the current payment situation of the BO concerned and informs the budget department (ensuring all protected items have been appropriated). Only the budget department can change entries into the automated treasury system at the level of the Central Treasury. The Budget department then issues a prikaz (order), which is signed by the first deputy minister of finance. The procedure takes about 2-5 working days. No statistic on the frequency of such requests is kept by the MOF; but an estimate of several dozen requests per month was given. Page 50 50 G OT is using DTS (Debt Tracking Systems) software to monitor its external debt. There have been considerable improvements to the external debt position since 2003, as indicated by the Table below (see also table 4 in section 2). Table 18: External Debt of Tajikistan 2000 2001 2002 2003 2004 2005 Debt – US$ m 1027.1 1018.5 984.5 1030.7 871.4 894.9 Debt/GDP % 108.2 97.0 82.5 66.3 42.2 38.9 The MOF has strengthened its Department and received technical assistance from the IMF. Data quality appears overall fair; there are concerns that debt and deficit data are currently not yet reconciled (IMF Fiscal ROSC 2007). Improvements in transparency include the submission by government of quarterly reports to Parliament on its external debt position. These reports include details of debt service obligations and accumulated arrears. The data is also incorporated in to the Annual Budget Execution Report. A Law on State Borrowing is annually submitted to Parliament, together with a program on foreign borrowing. In August 2006 the MOF produced a comprehensive “Status report on External Debt”. It states that: “to consolidate the gains, a long term debt management strategy that would quantity sustainability benchmarks, is in an advanced stage of preparation”. However, due to the decision to proceed with the loans from China, the preparation of this strategy was disrupted and no strategy has been adopted. The preparation of a long term strategy should be re-considered by the GOT. There is at present minimal domestic debt, being restricted mainly to Lottery Bonds with the Amonat Bank. The only other domestic debt identified is the 153m Somoni, already referred to in PI 9, where the government agreed to take over responsibility from the National Bank. No comprehensive figures on the extent of domestic debt were available. In 2007, however, some domestic borrowing is envisaged and the MOF are working with the IMF to establish formal “Securities Market” arrangements before embarking on any domestic fund raising. Dimension (i): C [Note: in a strict interpretation, the rating would be ‘D’ – as complete, updated and reconciled data on domestic debt is not produced even on an annual basis; however, in light of the fact that there have been hardly any ongoing domestic debt operations we see a ‘C’ rating as appropriate.] (ii) Extent of consolidation of the government’s cash balances The concept of a Single Treasury Account has been in place for some time. This involves the daily/weekly consolidation of cash balances included in the Single Treasury account. However, the consolidation process only applies to the Single Treasury Account maintained at the National Bank and covers only the Republican Budget and EBR in respect of Republican level BOs. Transit accounts are still maintained with the Amonat Bank for each of the Local Treasury offices, in respect of Republic budget revenues, which are cleared daily. Separate accounts are maintained for sub-national level revenues and expenditures. Until late 2006, a separate bank account with NBT was used for funds managed by the SPF. In late 2006, the SPF was replaced by a Social Protection Agency which now has a separate bank account with the Amonat Bank (rather than with the NBT). The Treasury has thus far not moved towards incorporating all local treasury accounts into the Single Account as recommended by the IMF. Such a move can only be achieved if two vital Page 51 51 c onditions can be met. First is the practical need for communications networking and the introduction of computerisation in all local treasuries. Secondly will be political support from the Oblasts and Rayons who may view this move as a considerable centralisation of control, and, therefore, a possible threat to their budget autonomy. At present, therefore, the majority of the Sub National cash balances are excluded from the single Treasury Account. Dimension (ii): C (iii) Systems for contracting loans and issuance of guarantees. The External Debt Status Report issued in August 2006 states that the 4 th principle of present public debt management policy is: “Guarantees are centralized in the Ministry of Finance and are provided only in exceptional circumstances to public sector borrowings. This restrictive policy is due to past experience of indiscriminate issue of guarantees, which have created a backlog of disputed cases and an effort to minimize contingent risk to the budget”. The existing legal framework was established by the Law on State and State Guarantees, Lending and Debt in December 1999 with the MOF being the sole authority for the issue of loans and guarantees. There are some ongoing Government Guarantees, and according to the statistics produced as at 31 December 2005 the position is as follows. Table 19: Debt of Public Enterprises No Debt of Public Enterprises (SOE’s) $m 1 With Government Guarantees 33.5 2 Without Government Guarantees 14.4 Total 47.9 Finally it is necessary to draw attention to two future transactions which are likely to lead to the GOT re-commencing the issue of SOE Guarantees. Firstly there is the proposed re-structuring of 37 Natural Monopolies in 2007. The government’s intention is to attract investment, but retain substantial ownership. Brief discussions on the outline proposition for Tajik Telecom indicate that loan approval from the European Bank of Reconstruction will be sought in the sum of 21m. In these circumstances any lender will wish to seek guarantees, from the owner (GOT) on the repayment of such sums. If the other re-structuring proposals also involve significant borrowing by the State Enterprises, then the level of guarantees given could rise substantially. Secondly, the recent commitment to Loans from China in respect of the Barki Tajik energy SOE (in itself direct support for an SOE) also envisages the GOT “guaranteeing” the servicing of the element of debt which is the direct responsibility of the SOE. Score of B has been allocated on the basis of the current position. The area appears to be very well managed, and is currently under control. However, clarification on the major proposals for 2007 would perhaps eliminate possible future concerns. Indicator Brief Explanation Rating (Method M2) PI-17 Recording and management of cash balances, debt and guarantees. (i) Foreign debt records are complete, updated and reconciled quarterly. Data quality is considered fair. Comprehensive management and statistical reports (cover debt service, stock and operations) are produced at least annually. Domestic debt is reported as minimal, but no consolidated figures are available. Score C (ii) Calculation and consolidation of most government cash balances take place at least monthly, but the system does not allow consolidation of cash balances. Score C (iii) Central government’s contracting of loans and issuance of guarantees are made within limits for total C+ Page 52 52 debt and total guarantees and always approved by a single responsible government entity. Score B PI-18 Effectiveness of Payroll Controls The wage bill is usually one of the largest items of government expenditure and can be susceptible to weak control and corruption. The GoT has made significant improvements in the payroll arena recently. Pay for the core civil service has been simplified by removing all but two pay supplements. A similar approach has been negotiated in the Education sector, again reducing pay supplements to just two as from April 2007. A personnel register is being established by the Civil Service Department under the President, based on data provided by MOF. Currently this is updated annually on the basis of information supplied by Line Ministries. The system has been computerized since 2005, and an upgrading of the software is planned. The register referred to above only covers Civil Service staff and no other employees of the GOT. The Civil Service Reform program includes plans for a networked personnel database covering all Civil Service employees to replace the manual register (this would cover 17,300 civil servants provided for in the 2007 Budget). There are no plans at this point in time to extend the database to cover other major public sector employee groups such as teachers and health workers. In 2005 total public sector employment stood at 355,021; the education sector accounted for close to 200,000 public employees, and the health sector for around 100,000. (i) Degree of integration and reconciliation between personnel and payroll data. All personnel and payroll records are manually maintained and all staff are paid in cash. There are no computerized payroll software facilities operative within the GOT and no firm plans to establish such facilities. This limits the scope for effective internal control mechanisms, and also the scope for the PEFA team to validate the observance of GOT personnel procedures. Basic personnel files have to be maintained by each Budget Organization. Files should contain all relevant data, including the initial appointment form of the employee. No computerized personnel databases are maintained in budget organizations. Importantly, there is an annual exercise which establishes the maximum number of staff that can be employed by the BO in the year. This Staff Limit and detailed staff data is attached to the smeta and forms the basis for payroll validation. The monthly payroll procedures that apply for civil servants are set out in brief below (the arrangements in lower level BOs may vary): a. The accounting office, not the Human Resource (HR) section, will prepare the payroll documentation for submission to the Treasury. b. The office will use the smeta data to ensure that approved staff levels have not been exceeded, and should receive updated information from the HR section of any changes made. c. A detailed Treasury form, which provides for all staff to be listed, must be completed. d. The payroll documentation is certified by an authorized officer. Page 53 53 e . Upon receipt in the Treasury, the appropriate Treasury kurator will check the submission against the smeta (which is also available in the Treasury) – and raise queries if necessary f. The documentation is again certified by an authorized Treasury official and paid. There are, therefore, some basic checks in place to ensure that staff limits are not exceeded, and that individuals receive their correct pay. Furthermore, the “internal control” procedures appear to be soundly based in the one area examined (MoF). However, there is little evidence of specific reconciliation between the personnel records and the payroll records taking place. Moreover, validation is not possible. Within the MOF there are four Departments placed with responsibility for controlling, monitoring and preparing staff budgets. The areas covered are Civil Servants; Education; Health and SOEs. The Civil Servants area is the longest established, and the department prepares staff budgets for the Annual Budget exercise, and staff number/lists limits for each budget organisation which are attached to the smeta . These procedures are less developed in the areas of Education and Health, but the same principles apply. There is no evidence of any ‘ghost workers’ being uncovered in the area of (core) civil servants. At least on recent case of organized ‘ghosts’ on the pension payroll has been acknowledged, however. There are no computerized payroll software facilities operative within the GOT and no firm plans to establish such facilities. The Treasury intends to raise this issue in the review of Treasury systems which is being sponsored by the WB. Good budgetary controls do exist for civil servants and are being developed for other public employees. However, given that payroll and personnel records are not reconciled at all on a regular basis, it is only possible to give a score of D. Dimension (i): D (ii) Timeliness of changes to personnel records and the payroll. There are clearly laid down procedures relating to all staff changes, which involve the HR section being responsible for forwarding a copy to their appropriate accounting office. The procedures are clearly set out and interviews reveal a confidence that records are updated promptly in the areas examined. However, it was clearly not possible to validate compliance with the procedures throughout the whole government sector. The audit evidence available is limited. The focus of audit is on compliance and on verifying the correctness of payroll calculations. At local level “ghost workers” 62 have been identified, albeit in small numbers and in a few instances only, and so have instances of payments being drawn for vacant posts and then being utilised to pay existing staff and used for other purposes. There is little doubt that some of the issues raised by audit will involve “errors” arising because of a failure to update payroll records on time. However, no specific analysis of this type of error is maintained. Neither is there any specific data on the need for retrospective adjustments to pay. The procedures in place are strong and suggest that compliance will also be strong, but the hard evidence is limited. In the circumstances a score of C is considered appropriate. Dimension (ii): C (iii) Internal controls of changes to personnel records and the payroll. 62 Ghost workers in the sense of fictitious names being included in the payroll. Page 54 54 T he procedures set out for the maintenance of personnel files and the communication of changes to payroll are clear and, as indicated above, appear to be followed effectively. It must be acknowledged that the scope for adequate internal controls in small institutions is limited. At local level, for instance, payroll requests are initiated by small institutions, but the personnel records are maintained in rayon-level departments (e.g. the local education departments). At the republican level, the controls are likely to work more effectively as collusion between staff would be required if fraud is to be perpetrated. It is also likely that the scope for fraud is significantly limited both by the budgetary constraints and the fixing of staff budgets at the beginning of each year. However, the 2007 IMF fiscal ROSC report finds that “the control mechanisms to ensure the correctness of payroll records is weak”. The Audit of payroll also identify numerous errors, although these may largely arise as a result of “poor calculations” rather than a failure of personnel changes to be reflected in the payroll. In the time allocated for the PEFA review it has not been possible to verify the effectiveness of internal controls where manual systems are operative. Dimension (iii): C (iv) Existence of payroll audits to identify control weaknesses and/or ghost workers It is the intention of the MOF ICAD and the SFCC to visit most major BOs once every two years. In the compliance checks that are undertaken payroll features prominently. Discussions with the external audit indicate that payroll features prominently in the audit work undertaken. Emphasis is given to verifying the calculations of pay to individuals. It was indicated that a sum of Somoni 1.8m was recovered in 2006 in respect of over-payments made to staff (this figure included SOE audits). The external auditors also undertake sampling checks to establish that staff are physically present at work, and to establish that Personnel files are available for all staff on the payroll. No “ghosts” have been identified by external audits in recent years. The ICAD has produced no detailed analysis on payroll audit. Interviews indicated that “ghost” staff on the payroll have been identified, but again no specific details have been supplied Given the manual nature of the procedures there are system weaknesses to be exploited and it would be of considerable concern if no errors were being discovered. The compliance nature of the audits undertaken fail to address systems weaknesses and do not give sufficient emphasis to the “ghost worker” issue. Nonetheless, the audits undertaken do comply with the criteria for a score B. Dimension (iv): B The GOT are to be commended for the recent initiatives undertaken in the payroll area. However those improvements do not impact directly upon the specific criteria set out in this Performance Indicator (which focus on reporting and monitoring processes). Only manual procedures are operative and all staff are paid by cash. Apart from the Civil Service staff, staffing functions are de- centralised, although central budgetary control is being exercised by the MOF. The personnel/payroll procedures, therefore, rely heavily upon the integrity of the head of the institution and the individuals responsible for human resources and payroll. The larger the organization, the greater is the scope for internal control checks. Audit visits by the MOF ICAD and the external audit office provide a measure of assurance. Nonetheless, it is extremely difficult to give the degree of assurances implied in the PEFA indicator. The arrangements that exist appear to be sound and effective, and are well implemented in the limited areas examined. Indicator Brief Explanation Rating (method M1) PI-18. Effectiveness of payroll controls (i) Integrity of the payroll is significantly undermined by lack of complete personnel records and personnel database , or by lacking reconciliation between the three D+ Page 55 55 lists. Score D (ii) There is an annual complete updating process. The MOF also receives updates on changes to personnel records continuously. However, there appear to be still practices of pay being drawn for vacant positions. The indicator could not yet be fully assessed based on available evidence. (iii) Controls exist, but are not adequate to ensure full integrity of data. Score C (iv) Payroll audits have been undertaken by the MOF ICAD and by the SFCC. These have covered all or nearly all central government entities. Score B PI-19 Competition, value of money and controls in procurement Significant public spending takes place through the public procurement system. A well-functioning procurement system ensures that money is used effectively and efficiently. Open competition in the award of contracts has been shown to provide the best basis for achieving efficiency in acquiring inputs for value for money in delivery of programs and services by the government. (i) Evidence on the use of open competition for award of contracts that exceed the nationally established monetary threshold for small purchases (percentage of the number of contract awards that are above the threshold) The position in Tajikista n is complicated by the fact that a new Law “On Public Procurement of Goods, Works and Services” was introduced on 3 rd March 2006. The law closely follows models provided by the WB, the ADB and is similar to the model now operational in the Kyrgyz Republic. It is considered UNCITRAL compliant. However, the practical implementation of this law is expected to be delayed by more than a year from its date of adoption which creates some difficulties in that the new law revoked the previous legalisation “On Public Procurement of goods, production, works and services” dated 12 December 1997. 63 Under the new Law the Procurement Agency (PA) becomes an oversight, advisory, training and regulatory organization. The new Law is aimed at the decentralisation of procurement arrangements, which is obviously an exercise that needs to be carefully planned, and will involve considerable guidance and training from the PA, who are hopeful of donor aid support to cover such costs. Until these new, decentralized arrangements become operational, the PA remains the only institution through which procurement is undertaken above the specified minimum threshold (3,000 somoni) within Tajikistan. 64 Presently there are a number of concerns; contracts are always awarded on the basis of lowest price; there is inadequate competition because thresholds for bidding are too low and this results in too few competitors; there is no appropriate mechanism for advertising publicly procurement opportunities, which again limits competition. The PA has introduced some basic computerized “tracking” procedures from the beginning of 2006. The table below is based upon this data and shows tenders above the 3,000 somoni minimum threshold, and separately tenders above the 50,000 somoni maximum threshold (for tenders above 50,000 somoni, stricter rules apply than for those between 3,000S and 50,000S.) It should be noted that under Article 15 of the new Law the minimum and maximum thresholds are set annually by the Government of Tajikistan. The current thresholds were set by a Government Decree dated 1 August 2006. 63 It is expected to be implemented from mid-2007; but there is uncertainty whether this will be achieved. 64 The New Law has stipulates a Minimum Threshold, whereby all contracts estimated to exceed this threshold must be subject to procurement procedures. It also specifies a Maximum Threshold above which it imposes more stringent rules. Page 56 56 Table 20: Procurement Procedure – 12 months to 31 December 2006 Procurement procedures Number of contracts in % TJ Somoni in % Republican budget 623 24.3 71723108 59 Local budget 1944 75.7 49597913 41 Tenders from 3000 to 49999 Open competitive tender 213 10 5972769 16 Request for proposals 1774 80 29053248 78 No-competitive tender 225 10 2267963 6 Total 2212 37293980 Tenders for more than 50000 Open competitive tender 277 78 56853065 68 Request for proposals 68 19 23341256 28 No-competitive tender 10 3 3832720 4 Total 355 84027041 All tenders for 2006 Open competitive tenders 490 19 63038688 52 Requests for Proposals 1842 72 52181650 43 No-competitive tenders 235 9 6100683 5 Total 2567 100 121321021 100 Source : Public Procurement Agency The Approved Budget for 2006 indicates that a sum in the region of Somoni 730m was included for the purchase of “Goods and Services” and “Domestically Financed Capital Expenditures”. Compared to the figure of Somoni 121.3m shown in Table 20 above as being “subject to tendering arrangements” this indicates that about 17% is subject to procurement arrangements. Dimension (i): C (ii) Extent of justification for use of less competitive procurement methods In Articles 24 to 30 of the existing procurement law clear guidelines for the use of less competitive methods are set out. Evidence of compliance with these rules is not readily available; but it appears that currently, justification of less competitive methods is not regularly recorded. The external audit department has indicated that they audit the “Goods and Supply” expenditure areas and specifically validate that contract prices have been charged where appropriate. They also make comparisons with “market prices” to identify over-charging. Importantly, they check to ensure that the BO has complied with Procurement Arrangements. The Procurement Agency was subject to four audits in 2006: by the Prosecutor’s office, by the World Bank, by the Executive Office of the President, and by the external audit office. The external audit office mentioned that it had little evidence to suggest that BOs were not observing procurement rules; but was unable to provide any specific data on the number of “non- compliances” incidences, nor any other data in relation to contracts. Neither was any specific data on procurement forthcoming from the ICAD (internal control and audit). Page 57 57 T able 20 above demonstrates a significant gap between the areas where contract arrangements should apply and the level of procurement undertaken. 65 Moreover, there is not yet a well established practice of providing justifications for the use of less competitive procurement methods. Given these circumstances a score of C is allocated. Dimension (ii): C (iii) Existence and operation of procurement complaints mechanism Chapter 7 of the 1997 Procurement Law provided comprehensive review procedures for participants of tendering proceedings. Under this legalization there have been no complaints received according to the PA. The new legalization includes clear procedures for complaint both to the contracting entity and to the “Authorized Body” (the Procurement Agency) as well as provisions for Judicial Review. However, until the new legalization is operational in a practical sense, the Procurement Agency is the only organization empowered to implement procurement procedures and contracts. Currently, the PA arranges procurement contracts for all entities that have not acquired “Qualified Procuring Entity Status”. The complaints legalization does not appear to clearly cover the situation where the PA is both “procuring entity” and “authorized body”. This is a somewhat unsatisfactory position. Given the concerns about the PA acting in two capacities score of C has been allocated. This rating will improve once the new legislation is made operational. The new Procurement Law will only be capable of proper review in about 18 months time. It is, however, clear that there is a major task ahead of the GOT if they are to ensure that the legalisation is to effectively bring the benefits sought. Developing procurement entities in all localities, for instance, will involve extensive training, and require careful monitoring to ensure compliance. There must also be some concern about retaining the minimum threshold at 3,000 somoni, when this figure does not produce adequate competition at present. Finally there is the question of how long it will take for all entities to acquire “Qualified Procuring Entity Status”, and how long, therefore the PA will continue to directly procure contracts. Indicator Brief explanation Rating (Method M2) PI-19 Competition, value of money and controls in procurement. (i) Available data shows that less than 50% of contracts above the threshold are awarded on an open competition basis, but the data may not be accurate. Score C (ii) Justification for use of less competitive methods is weak or missing. Score C (iii) A principally well-designed process for submitting and addressing procurement process complaints is defined by legislation, but is not yet operative. Score C C 65 A more detailed comparison would be necessary to establish the real size of the ‘gap’ between expenditures that should be covered by procurement procedures and are not; as not all items included in the classification ‘Goods and Services’ and ‘Domestic Capital Expenditures’ would formally be required to be subject to procurement. However, the size of the gap (>80%) between these expenditures and the procurement figures provided by the PA suggests that there are purchases being made outside of existing procurement rules. Page 58 58 PI-20 Effectiveness of internal controls for non-salary expenditure An effective internal control system is one that (a) is based on an assessment of risks and the controls required to manage the risks, (b) incorporates a comprehensive and cost effective set of controls (which address compliance and rules in procurement and other expenditure processes, prevention and detection of mistakes and fraud, safeguard of information and assets, and quality and timeliness of accounting and reporting, (c) is widely understood and complied with, and (d) is circumvented only for genuine emergency reasons. (i) Effectiveness of expenditure commitment controls GOT operates a tight control system in respect of expenditures met from the approved budget, based upon a Smeta for each BO. However, as detailed in PI 16 there are a number ways in which the annual budget and/or the quarterly cash limit can be amended. In the last four years, therefore, significant in-year adjustments to the approved budget were made. In the majority of instances, these adjustments have been increases. Because the systems are manual, the administrative procedures are a little cumbersome, but they work effectively and each BO’s Smeta is updated to reflect the new budget limits/quarterly cash limits. Current payment procedures involve the BO submitting a payment request voucher to the treasury, which is checked both in terms of its compliance with the authorized procedures, and to ensure the availability of funds on the BO’s account. Protected items such as salaries and communal services are given payment priority. If the payment request is in respect of a procurement contract, already recorded in the Treasury systems (see below) the commitments will automatically be cancelled when the payment is made. The Central Treasury now use an “Electronic Payment System” to instruct the National Bank to make the payment to the creditors’ bank accounts. Specific Commitment Data. The current Treasury systems have the capacity to hold contract details, and all BOs are required to supply all contract information, for contracts above the minimum threshold of 3,000 somoni, to either the Central or Local Treasury. All data collected is entered into the Central Treasury computerized database (for local levels, the entry is done via the central Treasury). However, this commitment data is only shown in the particular quarter when it is expected to be spent. In financial management terms its usefulness is, therefore, limited. Nonetheless, the Treasury is able to produce monthly statements comparing approved budget with the total of the executed budget and the outstanding contractual commitments. The Treasury will not allow the Budget/Cash Limits to be exceeded. Moreover, as indicated in PI 4 the Treasury does receive data on the level of outstanding payments at the end of each quarter, so it is a position to monitor outstanding expenditure commitments. In practice, the main focus of the Treasury is on keeping payments within approved limits, rather than closely supervising outstanding expenditure commitments. The budgetary control system is tightly operated and there is no possibility of a budget being exceeded. There is some “commitment” data in respect of procurement contracts above the minimum threshold, but is it not comprehensive. In view of the absence of all commitments being covered, a score of B has been allocated. (ii) Comprehensiveness, relevance and understanding of other internal control rules/ procedures The functions of the Treasury are governed by the Treasury Law (enforced by Presidential Decree no. 19, 12 May 2001) and by the Instruction on Treasury Accounting and Reporting on the State Budget Execution of the Republic of Tajikistan (MOF Decree no. 110, 20 October 2005). Page 59 59 T he existing internal controls/rules that are in place are capable of improvement, but they are basically sound and in the absence of any computerized systems should be considered as adequate. There is tight control of the quarterly cash limits, but equally some flexibility in the virement rules in operation which is positive. In terms of the role of the Treasury detailed checking may be somewhat excessive, and there is little emphasis on the whether the expenditure incurred has successfully delivered the “budget goals”. However, given the manual nature of procedures and the current budget policies the situation is acceptable. In terms of an understanding of the internal rules, the limited data received from the SFCC and ICAD does not suggest that there any difficulties in respect of staff understanding the procedures. A 2006 EU report on”Developing Public Internal Financial Control” (with modifications accepted as a reform strategy by the GoT) indicates that there is a need to update and further develop the legal basis of the Treasury functions (see section 4 of this report regarding envisaged reforms) which should lead to (possibly substantial) changes in internal control rules. Dimension (ii): B (iii) Degree of compliance with rules for processing and recording transactions The SFCC (respectively its successor office) and ICAD both aim to visit major BOs and local authorities once every two years. Visits are focused in ensuring compliance with the Law, and with internal controls/rules set out in legislation. Where the audit staff identifies serious malpractice, they will inform the appropriate authorities. The procedures for both bodies are similar: an act of audit will be “signed” and audit findings are discussed with the audited entity resulting in an agreement on measures to be taken. Serious violations arising are reported to the Prosecutor’s office. The only evidence available on compliance is that supplied by the two audit bodies. This indicates that compliance does not appear to be a significant issue. There are legal powers that enable the MOF to impose penalties for “violations” by BOs. These are aimed at the recovery of any assessed loss arising from neglect, fraud or misuse. In the time allocated to the PEFA exercise there was no opportunity to verify the extent of compliance in the field. The study team has relied upon feedback from the two audit bodies and unfortunately the volume of evidence shared with the team is limited. The control procedures are sound and effective for manual processes However, given the inability to verify the extent of compliance and the uncertainty on the effectiveness of audit arrangements, with little emphasis on ‘systems’ and ‘internal control’ verification, a score C has been allocated. The 2006 EU report on “Developing Public Internal Financial Control” within the public sector of Tajikistan has been agreed by the GOT and tender specifications for a project in support of the reform are being prepared. This report outlines a long-term strategy for developing Financial Management IT systems, ensuring that financial accountabilities are properly allocated, that the budget/budget execution focus moves towards the delivery of objectives/goals and suggesting the introduction of a modern auditing approach. Consequently, current “internal control” procedures would be substantially reformed in the context of such changes. Indicator Brief explanation Rating (Method M1) PI-20. Effectiveness of internal controls for non - salary (i) Expenditure commitment controls are in place and effectively limit commitments to actual cash availability and approved budget allocations for most types of expenditure with minor areas of exception. Score B C+ Page 60 60 expenditure (ii) Other internal control rules and procedures consist of a basic set of rules for processing and recording transactions, which are understood by those directly involved I their application. Some rules and procedures may be excessive, while controls may be deficient in areas of minor importance. Score B (iii) Rules are complied with in a significant majority of transactions, but use of simplified/emergency procedures in unjustified situations is an important concern. Score C PI- 21 – Effectiveness of Internal Audit Regular and adequate feedback to management is required on the performance of the internal control systems, through an internal audit function. Such an audit function should meet international standards and should focus on systemic issues in relation to: reliability and integrity of financial and operational information; effectiveness and efficiency of operations; safeguarding of assets and compliance with laws, regulations, and contracts. (i) Coverage and quality of the internal audit function The Internal Control and Audit Department (ICAD), of the Ministry of Finance, and its branches in the local Treasury offices directly perform ex-post reviews of the financial operations of budget entities. There are 58 Internal Audit staff in total, 7 based at MOF and 51 at oblast/rayon level. There was a concern that the Internal Control and Audit Department (ICAD) is not independent of the MOF’s line management. From April 2006, the central ICAD has been moved under the direct responsibility of the first deputy Ministers of Finance (previously it was subordinated to the Treasury; see the MOF organizational chart in appendix 3). The focus of audit is almost exclusively “legal compliance”. The Head of the ICAD summarized the main features of the audit process as: i. A formalized audit plan is developed before the financial year ii. It is not possible to cover all BOs given the limited audit resources available (e.g. 7 central ICAD staff have to cover approximately 800 BOs at the Republican level) iii. Focus is on those BO’s with an annual budget allocation of more than 100,000 somoni iv. The aim is to visit these larger BOs once every two to three years v. Audit staff in Oblasts/Rayons provide reports to the central ICAD office in the MOF. A report is produced annually for the Deputy Minister of Finance on the activities of the ICAD. The ICAD has provided some data on activities in 2005, when 589 audits were undertaken and 2006, when 742 audits were undertaken. This data focuses on the sums involved in violations identified by the audit staff. To support the assessment of PEFA indicators, specific information on the types of violation identified would be required, particularly in respect of payroll audit (PI 18), procurement procedures (PI 19) and verification of internal controls (PI 20). The absence of such specific evidence also has negative implications for the assessment of the internal audit function as such. The EU report on “Developing Public Internal Financial Control” (2006) analyses the current legal position and finds it to inhibit effective audit. The report indicates that the by-laws and guidelines related to the State Finance Law do not provide a sufficient basis for the establishment of a sound Public Internal Financial Control system due to the inconsistency and lack of structure in the Page 61 61 a ssignment of duties, responsibilities and accountability for implementing governmental programs between key organizations. The coverage and quality of Internal Audit is clearly not adequate. There are many difficulties: perhaps most importantly the institutional issues identified by the EU where they feel it inhibits the ability of GOT to develop a modern systems audit approach. The ICAD appears to fulfill its duties in a responsible manner, and no doubt assists in the observance and compliance of rules by staff in the major BOs. However, it is faced with inadequate resources and staff who are not specifically qualified as auditors that seriously limit its scope and coverage. In the circumstances a D score has been allocated. Dimension (i): D (ii) Frequency and distribution of reports As indicated above the ICAD is required to visit all major BOs once every two years. The audit activities cover the whole period since the previous visit. At the end of a visit the ICAD issues an audit act report which outlines shortcomings and violations discovered. This report is submitted to the BO, the appropriate higher authority (e.g. a LM) and to the Deputy Minister of Finance. The project team was assured that the BO has the right to challenge the audit report, but this has never happened in fact. Serious violations will be referred to the Prosecutor’s Office. The Budget Department of the MOF (under Article 50 of the Law on State Finance) can impose reductions in approved budget allocations equivalent to the cost of any misappropriated/misused amount. The SFCC (respectively its successor) receives a copy of the ICAD quarterly report submitted to the Deputy Minister of Finance, which covers the activities of the ICAD for the quarter. However, there is no evidence of active co-operation between the two audit organizations, although the SFCC indicated that (prior to the government restructuring) there had been discussions to co-ordinate annual audit plans in 2007. Dimension (ii): C (iii) Extent of management response to internal audit findings The focus of the activity of ICAD is on compliance with existing legislation and in many cases requires resolution by regulatory or judicial bodies. Action on the findings of ICAD appears to be prompt and complete. Statistical data is produced on the work of the ICAD. ICAD is confident that its work has contributed to improving the management of public finances. However, as indicated in Dimension (ii) above, there appears to be no feedback from BOs, so there is uncertainty about the response by the management of BOs. Given this uncertainty a score C is allocated. Dimension (iii): C Major reforms of the internal audit system are in the planning stage, based on the agreement reached on the reform strategy. A decentralization process that will involve the establishment of internal audit units in pilot ministries by the end of 2007 is one of the key proposals. Furthermore, a new law on internal audit is expected to be developed. These proposals are to be welcomed. They should bring significant improvements to the Internal Audit approach in Tajikistan, which is currently seriously under-developed. Indicator Brief Explanation Rating (Method M1) PI-21. Effectiveness of Internal Audit (i) There is little or no internal audit focussed on systems monitoring. Score D . (ii) Reports are issued regularly for most government entities, but may not be submitted to the ministry of finance and the SAI. Score C D+ Page 62 62 (iii) A fair degree of action taken by many managers on major issues but often with delay . Provisional Score C C (iii) Accounting, Recording, and Reporting PI-22. Timeliness and regularity of accounts reconciliation Reliable reporting of financial information requires constant checking and verification of the recording practices of accountants – this is an important part of internal control and a foundation for good quality information for management and for external reports. (i) Regularity of bank reconciliations Article 9 of the Treasury Law and its corresponding instructions stipulate the arrangements for bank reconciliation. The National Bank of Tajikistan services the single treasury account; while branches of the state- owned Amonat Bank service the local treasury accounts and the account for the Social Protection Agency. The banks send daily bank statements covering all daily revenues received and expenditures outgoing to the Central Treasury and local treasury departments. The National Bank of Tajikistan and the Amonat Bank branches also send monthly statement on treasury accounts revenue and expenditures with the breakdown by days to the Central Treasury and local treasury departments on 3rd day of every month. At local treasury level they manage the transit accounts (for Republican budget revenues) which should have a zero balance at the end of each day, and the Local Government accounts (for local revenues and expenditures). However, as discussed under PI-15, there appear to be cases where delays occur in transferring tax levies by commercial banks to the treasury accounts (this factor is considered under PI-15, and is not further considered here). The Central Treasury reconciles its expenditures and revenues records with the bank statements on a daily basis. According to the Central Treasury there are virtually no problems arising in this area. In respect of revenues there is a separate weekly exercise to ensure that collections are reconciled with the detailed Debtor records held by the Tax and Customs offices. Reconciliation problems do arise, but these are usually resolved. Local Treasuries are expected to reconcile records with their Amonat bank branch on a daily basis, for both expenditures and revenues. They are also required to reconcile each month the revenue figures with the local tax office. There is a formal process whereby bank statements have to be stamped and signatures of the staff undertaking the reconciliation have to be added. Within the scope of the PEFA exercise it was not possible to verify the extent of compliance with the procedures. However, procedures do look sound, and do facilitate the production of the Statement on the Movement of Reserves and Balances, so a Score B has been allocated. Dimension (i): B (ii) Regularity of reconciliation and clearance of suspense accounts and advances. The MOF informed us that there are no suspense accounts in operation, and that advance accounts are controlled manually by BOs. There are well documented procedures for advance payments and there is no evidence to suggest problems in this area. There is also a requirement that all balance Page 63 63 s heet accounts are reconciled annually at 31 December. Given that there are no suspense accounts to be reconciled. the same score of B given for Dimension (i) has been allocated. Dimension (ii): B Indicator Brief explanation Rating (Method M2) 22. Timeliness and regularity of accounts reconciliation (i) Bank reconciliation for all Treasury managed bank accounts take place at least monthly, usually within 4 weeks from end of the month. Score B (ii) Reconcialation and clearance of suspense accounts and advances takes place annually within 2 months of the end of period. Some accounts have uncleared balances brought forward. Score B B PI 23 Availability of information on resources received by service delivery units Problems frequently arise in front-line service delivery units providing services at the community level in obtaining resources that were intended for their use. These front-line service delivery units, being the furthest in the resource allocation chain, may be the ones to suffer most when overall resources fall short of budget estimates. Tracking of resource allocation is vital to ensure that PFM systems effectively support front-line service delivery. (i) Collection and processing of information to demonstrate the resources that were actually received (in cash and kind) by the most common front-line service delivery units (focus on primary schools and primary health clinics) in relation to the overall resources made available to the sector(s), irrespective of which level of government is responsible for the operation and funding of those units Primary and secondary schools are financed out of local budgets (see 2004 Organic Law of the Republic of Tajikistan on Local State Power, art. 32). Schools’ financial management is the responsibility of rayon education departments who, for instance, maintain the personnel records of all school staff. There is some general guidance of budget allocation on a “norms” basis (which takes account of assessed teaching hours and class sizes, etc.). However, there does not appear to be any regulation issued to Rayons on how funds should be allocated between the various schools in a locality. There is scope for “ad hoc” allocations. It is also apparent that there are substantial regional differences in funding levels for schools. Currently there is an experiment underway which is considering the adoption of a “pupil based formula” for the allocation of resources. The intention is to extend the pilot to a total of 12 Rayons in 2007 based on a per capita formula. A similar approach is under consideration for the funding of primary health facilities. PET assessments for the health and education sectors have been initiated, but have not yet been finalized. A fiduciary assessment of the education sector has been undertaken; but the assessment primarily focuses on the need for capacity building to improve significantly both the quality and understanding of “funding” at local school level. The initial findings of the PETS assessment indicate a consistent use by Rayons of the “norms” to calculate the budget allocations for individual schools. However, individual schools are generally not “informed” of their budget allocations. Neither is there any real evidence that relevant budget execution data is being distributed either by the Education Office of the Rayon or the Local Treasury. Both schools and primary health facilities are reliant on the rayon for accounting services. Page 64 64 For the Education Pilot Rayons some accountant posts are being allocated to schools level, and the post occupants will maintain manual records on Budget and Budget Executions for each pilot school. The PET Survey of the health sector that is in process informed the PEFA study team that (a) most service delivery units do not know how much budget they get allocated; (b) there is no information available at the front line service delivery, especially on goods and services due to in-kind payment for goods and services and a lack of accounting reports. Local Treasuries issue daily statements (covering both expenditures and revenues) to BOs – to enable them to update their accounting records. The Local Treasuries have indicated that they do not issue the quarterly budget execution statements to schools unless requested to do so. However, there is little evidence to suggest that the Treasuries hold individual School/Health unit budgets within their systems, nor that they allocate budget execution detail to individual unit level. The processes in Local Treasuries ensure that both the local and central accounting data are reconciled. Some Local Treasuries may have the IT capacity to produce statements for individual institutions, but do not appear to do so because both budgets and budget execution details are not maintained at unit level within the Treasuries. It appears that front-line service delivery units have little information about their budget allocations and capacity to compare these to actual receipts (which reduces the scope for them to demand the full disbursement of allocated expenditures). Also, to date, there has been no special survey yet completed and there is no routine monitoring of actual receipts of front-line service delivery units. However, in light of the fact that PET surveys are near completion in two sectors, this indicator is rated as C. Indicator Brief explanation Rating (Method M1) PI-23. Availability of information on resources received by service delivery units No comprehensive data collection on resources to service delivery units in any major sector has been collected and processed within the last 3 years. However, PETS for two major sectors – health and education – are close to completion. C PI 24. Quality and timeliness of in-year budget reports The ability to successfully achieve budget execution depends upon timely and regular information on actual budget performance being available, throughout the year. This information is required at all management levels, from Government down to managers of BOs, to ensure that effective monitoring is undertaken throughout the year. (i) Scope of reports in terms of coverage and compatibility with budget estimates The Treasury’s “Automated Computer Management System” produces regular monthly reports comparing actual expenditures/revenues with approved budgets. Given the procedures operative within the Central Treasury it is reasonable to assume that these monthly reports contain accurate information. However, it must be borne in mind that the Treasury Systems are single entry software, and do not give the same assurance as double entry software. The BO budget smeta is the key control document and budget execution follows the same format so that there is compatibility. Page 65 65 S tatements are not automatically distributed to all republican level BOs by the Treasury – they are only distributed to those BOs who request statements. The statements can be “tailored” to accommodate the requests of the individual BO. For instance information on contractual commitments still outstanding can be included. Additionally, the Central Treasury supplies the BOs that it deals with directly with daily statements of transactions processed (expenditures and revenues). 66 At both republican and local budget levels the comparison is always with the latest updated budget approval. There is no detailed information showing the movement from the original budget to the latest position and this is a clear weakness, as is the position whereby only partial commitments data is collected. In terms of the more formal process of reporting to parliament and to local executive bodies, more detailed information is set out in (ii) below. The treasury systems can provide reliable information comparing the “adjusted” budget with the actual payments made to date, so a Score C has been allocated. Dimension (i): C (ii) Timeliness of the issue of reports Line Ministries, agencies and BOs are required to submit quarterly reports on revenues and expenditures execution to the MOF. At local level there is a similar requirement to submit quarterly reports on local budgets execution to the executive body of local authorities. The reports must include information on any outstanding debts due for payment at the end of the quarter. The reports for the MOF should be produced by the 15 th of the following month, and this timetable appears to be met, although clearly it was not possible to verify for all areas. The MOF drafts quarterly reports on a commitment basis according to the results of state budget and Republican budget execution. Information is verified according to Treasury figures. The reports should contain information on initial plan, revised budget and actual budget execution. The MOF will submit quarterly reports prepared on a commitment basis (including invoices outstanding but not yet paid) to the GOT. At local levels the reports are submitted to the executive body of local authorities. GOT approves the quarterly report and further submits it to the Parliament. The Treasury monthly and quarterly reports, where requested by the BO, appear to be produced within 4 weeks of the end of the financial period. Dimension (ii): A (iii) Quality of information The procedures in place provide confidence that the financial data produced is accurate. There are processes in place to ensure reconciliation between the BO manual records and the Treasury records. However, there has been little scope to examine the data entry procedures that are in operation – to ascertain if there are weaknesses that would impact upon the quality of the allocation of budget classification data entered. Furthermore, as indicated above the Treasury Automated 66 At the local level the arrangement differ, since the local treasury has no networked access to the Central Treasury computer systems. It is understood that again regular monthly reports are provided when requested. Usually, these will utilise simple software facilities such as Excel. At the local level it is not possible to include the contractual commitments information. Page 66 66 C omputer Management system is a “Single Entry” process and cannot provide the extra security offered by a “Double Entry” system. Nonetheless, the reconciliation procedures with the manual “double entry” records maintained by the BO’s give a good degree of assurance on the quality and accuracy of the records. There is a good degree of assurance on the quality and accuracy and a Score B has been allocated. Dimension (iii): B The proposed new Treasury Systems should offer significant advantages in relation to this PI. The Treasury indicates a long term exercise is planned supported by the World Bank. It is understood that a FMIS approach is envisaged which will address the needs of the Treasury and the future management information requirements of spending sectors, LMs and BOs. Indicator Brief explanation Rating (Method M1) PI-24. Quality and Timeliness of in-year budget execution reports (i) Comparison to budget is possible only for main administrative headings. Expenditure is captured either at commitment or at payment stage (not both). Score C. (ii) Reports are prepared quarterly or more frequently, and issued within 4 weeks of end of period. Score A (iii) There are some concerns about accuracy. Nonetheless, overall the reconciliation procedures give a good degree of assurance on the quality and accuracy of the records. Score B C+ PI 25. Quality and timeliness of annual financial statements Consolidated year-end financial statements are critical for transparency in the PFM system. To be complete they must be based on details for all ministries, independent departments and deconcentrated units. Additionally the ability to prepare year-end financial statements in a timely fashion is a key indicator of how well the accounting system is operating. (i) Completeness of the Financial Statements The MOF is required under Article 57 of the Law on State Finances to draft an annual Report on the GOT State Budget Execution and submit it to the Government by 1 May. The MOF also has to prepare a Public Sector Annual Financial Report (which forms part of the annual report of Budget Execution). This covers information on revenues and expenditures, treasury account balances, state loans and debts, budget deficit and the financing of the budget. There exists a strong legal foundation for preparing annual budget execution reports and financial statements (the annual budget execution reports are part of the annual financial statements). Article 57 of the Law on State Finance also stipulates the contents of the annual state budget execution report to be prepared at the end of each fiscal year. The annual budget execution reports are consolidated on a cash basis in accordance with the budget classification. Ministries, agencies and other budget organizations submit their annual financial statements, which are reconciled with the Treasury data for preparation of the consolidated annual financial statements. The annual financial statements prepared by the Government include information on: revenues and grants; expenditures according to budget classifications with comparisons between budget and outturns; Treasury bank account balances; report on state borrowing transactions and stock of debt and size of budget deficit and financing. The Law then requires the Government to submit these Annual Accounts to the Parliament and the SFCC by 1 June. As set out above the Annual Accounts provide good coverage of the Budget Page 67 67 E xecution, but it must be recognized that the accuracy of the accounts cannot be guaranteed given the “single entry” system of accounting currently operative within the Treasury. Furthermore, the Accounts do not comply fully with IPSAS cash accounting standards or any other recognized international accounting standards. Information on certain financial assets/liabilities such as outstanding arrears, other payables, receivables, commitments and contingent liabilities are not reported in a comprehensive manner. The IMF ROSC report also points out that information on the stock of domestic debt, loans, advances and ownership are not included in the annual statements. The consolidated financial statements do not yet meet the stipulated requirements of IPSAS in terms of coverage, method and quality. 67 There is no consolidated Balance Sheet for the GOT included in the Accounts, although there is a Statement on the Movement of Balances and Reserves which is commendable given the single entry nature of the accounting records. Despite the technical deficiencies the Treasury appears to be well organised and effective in performing its role. Score of C is considered appropriate. Dimension (i): C (ii) Timeliness of submission of the Financial Statements The Law requires the Government to submit these Annual Accounts to the Parliament and the SFCC by 1 June. The deadlines for submission to Parliament have been met for the last 3 financial years. However, the Annual Accounts do not appear to be submitted formally to the SFCC. Despite the legal requirement that the SFCC should submit its audit of the annual execution report to the President, the Government and Parliament by 1 October, this does not represent current practice. The SFCC has carried out a full audit of Budget Execution once every two years, when it undertakes an audit of the entire two year period. Moreover, in terms of compliance with this indicator, the SFCC (respectively its successor organization) currently does not meet the characteristics of an independent external audit body as defined by international standards. Dimension (ii): D (iii) Accounting Standards Used As indicated above the Annual Accounts are not fully compliant with recognized International Accounting Standards. However, they do have the benefit of a strong legal base and do, therefore, have the merit of consistency. Unfortunately the accounting standards used are not disclosed in the published documentation. Again, therefore, it is only possible to score D Dimension (iii): D Indicator Brief Explanation Rating (Method M1) PI-25. Quality and timeliness of annual financial statements. (i) A consolidated government statement is prepared annually. Information on revenue, expenditure and bank account balances may not always be complete, but the omissions are not significant. Score C. (ii) If annual statements are prepared, they are generally not submitted for external audit within 15 months of the end of the fiscal year. Score D (iii) Statements are not presented in a consistent format over time or accounting standards are not disclosed. Score D D+ 67 From the point of view of the quality of financial reporting, the four principal qualitative characteristics according to the International Public Sector Accounting Standard (IPSAS) are understandability, relevance, reliability and comparability. See IFAC Public Sector Committee, Cash Basis IPSAS. Financial Reporting under the Cash Basis of Accounting , pp.87-89 Page 68 68 C (iv) External Scrutiny and Audit PI-26: Scope, nature and follow-up of external audit A high quality external audit is an essential requirement for creating transparency in the use of public funds. Key elements of external audit include: the scope/coverage of the audit; the independence of the external audit; focus on significant and systemic PFM issues; reliability of the financial statements. In December 2006 the SFCC was dissolved by Government Decree. Its successor is the “Agency of State Financial Control and Combating Corruption”. The changes made reduced rather than increased compliance with international standards on the independence of the audit institution; and the combination of the external audit function with the function for corruption control does not conform with international practices. There is no indication that the changes made have addressed the SFCC’s deficiencies outlined below. Tajikistan’s development partners have repeatedly emphasized their concern about the absence of an independent external audit institution and did so again after the restructuring of the government in late 2006. In early 2007, the World Bank and the GOT reached an agreement that a Chamber of Accounts reporting to parliament will be developed. The current time horizon for this is March 2008. (i) Scope/nature of audit performed (incl. adherence to auditing standards) Until December 2006 the SFCC had its mandate from the Law on State Financial Control. It was responsible for governmental control over budgetary compliance, the use of government property, and compliance with laws and acts regulating government property, economic, financial and fiscal relations. The SFCC’s leadership was appointed by the President. The SFCC has three local branches with chairmen appointed by the head of the SFCC and approved by the President. The Committee had significant powers to access information about budgetary activities at both the state and local levels. It also reviewed the activities of the Social Protection Fund and other (public and private) entities that use public funds. The SFCC primarily reported to the President, but the reports were also shared with the Council of Ministers and the Parliament (with the latter in summary form only). Whilst the SFCC had administrative autonomy, it did not operate as an independent external audit body. In common with several other CIS countries, it acted as an internal control body reporting primarily to the executive, and concentrating on compliance testing. There is little evidence of auditing standards that focus on significant or systemic issues. The SFCC did not use INTOSAI standards for auditing and many of the staff were not professionally qualified. The SFCC aimed to audit the activities of every BO once every two years, on the principle that the accounts for a two year period will be examined at the one audit. This appears to cover all areas of budget execution including the use of EBRs. The SFCC also reviewed the financial transactions of government agencies. However, it did not provide any certification of final accounts, neither for individual agencies nor for the government as a whole. There was very little evidence of co-operation with the MOF’s ICAD. Mention was made that in 2007 they would be coordinating their annual Audit Plans, and the SFCC indicated that they did use MOF audit staff on SFCC audits. Page 69 69 S ome improvements arose from the recent work undertaken with the National Audit Office of the UK which has involved visits of SFCC staff to the UK and trainings carried out in Tajikistan, as well as the preparation of an Institutional Development Plan. In 2005 the SFCC undertook 1254 audits and 189 cases were referred to Law Enforcement Agencies. In 2006 1522 audits were undertaken, but no figure on cases referred is available. Unfortunately, these figures include audits of SOEs, so they do relate solely to Republic and Local Government bodies. The information provided has similar limitations as that referred to in PI-21 in respect of the MOF ICAD – it is not specific in terms of the types of audit work undertaken and the types of errors/violations uncovered. Available evidence suggests that the SFCC has covered at least 50% of total expenditures through its audits per annum. However, the team did not receive sufficient evidence that audits highlight significant issues. Therefore, a score of D has been allocated in respect of the audits performed by the SFCC. Dimension (i): D (ii) Timeliness of submission of audit reports to legislature The SFCC submitted quarterly reports on all its regular audit activities to the President and in summary form to parliament. In terms of the audit of the annual budget execution report, this exercise was undertaken every two years, and the audit covered two years of annual accounts. At the completion of the audit a detailed report was prepared and submitted to the President. A report was also submitted to the Parliament, but it is understood that this was in a summary format, less detailed than the Presidential report and was only sent for information purposes (see PI-28). It is only possible to allocate a score of D. Dimension (ii): D (iii) Evidence of follow up on audit recommendations The SFCC appears to have followed consistent practices in its approach to each audit. It has “signed” an act of audit and then discussed its findings in a meeting of all interested parties, with a view to reach agreement on measures to be taken. Any serious violations arising would be reported to the Prosecutors office. Where violations were identified the SFCC was empowered under Article 17 of the Law on State Financial Control to issue instructions so any damage and misuse was rectified. The SFCC indicated that it would initiate a follow up of significant issues within 3 to 4 months, and automatically follow up all issues at the time of the next audit (2 years time). Issues referred to the Prosecutor’s office would be specifically followed up to ensure that action is taken. Within the scope of the PEFA exercise it was not possible to verify the extent of compliance with these procedures. The economic and budget committee of Parliament discusses the annual summary of audit reports, but does not appear to “follow up” on any audit findings. Equally there is little evidence of “follow up” action being taken by the MOF and individual Budget Organizations. While the process described includes a formal response by the audited BO, there is overall little evidence of effective follow-up. Dimension (iii): C Indicator Brief Explanation Rating (Method M1) Page 70 70 PI-26. Scope, nature and follow-up of external audit. (i). Audits cover central government entities representing less than 50% of total expenditures or audits have higher coverage but do not highlight the significant issues. Score D (ii). The parliament does not receive full audit reports. Score D (iii) While the process described includes a formal response by the audited BO, there is overall little evidence of effective follow-up. Score C D+ PI-27. Legislative scrutiny of the annual budget law The Parliamentary Budget Committee consists of 7 MPs (1 chairperson plus 6 members) out of a total parliament of 63 MPs (and 9 committees in total). It is supported by 2 assistants and 2 senior inspectors. There are no sub-committees of the Budget Committee. Tajikistan’s system of government is strongly presidential; and consequently the role of parliament is limited. Tajikistan’s parliament – the Majlisi Oli – consists of two chambers, the Majlisi Milli and the Majlisi Namoyondagon . The Majlisi Milli (upper house) is based on indirect elections by regional councils (3/4) and presidential appointments (1/4). The Majlisi Namoyondagon (lower house) is directly elected by popular vote. Only the lower house is directly involved in the scrutiny of the budget. In the current lower house, 49 out of a total of 63 deputies are members of the pro-presidential party, the People’s Democratic Party of Tajikistan. According to interviews with the head of the committee, budget documentation received from the Ministry of Finance has improved in recent years; and the committee can pose questions and receive more detailed information on demand. Discussions on the draft budgets mainly focus around additional funds; which are requested by regionally elected MPs. There appears to be no in- depth review of budget execution. (i) Scope of the legislature’s scrutiny Parliament reviews the proposed budget once it has been finalized and adopted by the government; the deadline for forwarding the proposed budget to parliament is November 1 (see PI-11). As discussed under PI-6, the government does not submit analytical material on fiscal policies to parliament for discussion (in the sense of explaining budget implications of new policy initiatives and estimating the budgetary impact of all major revenue policy changes and/or changes to expenditure programs). As discussed under PI-5, currently, no administrative classification of the budget is used during the budget preparation process and in budget execution reports. Hence, parliament’s review is limited to a review of the economic and the functional classification. According to its head, the budget committee can request further detail from the government; but the fact that no presentation of an administrative breakdown is provided significantly limits the ability of the parliament to scrutinize the budget in detail. Dimension (i): C (ii) Extent to which the legislature’s procedures are well-established and respected A specialized budget committee exists, even if it is relatively small, and it has some support staff, which is positive. The parliament’s reglament (internal procedures) contains rules for the review of the budget (chapter 15). The draft budget is reviewed by all committees and commissions of the Majlisi Namoyandagon . Page 71 71 The reglament was adopted in 1997, and is not well known among current parliamentarians (elected in 2005) and thus are not always respected. Support st aff, and staff in the parliament’s secretariat change together with MPs and hence do not provide a base for institutional memory and continuity. Dimension (ii): C (iii) Adequacy of time for the legislature to provide a response to budget proposals both the detailed estimates and, where applicable, for proposals on macro-fiscal aggregates earlier in the budget preparation cycle (time allowed in practice for all stages combined). According to the 2002 Law on State Finances, the parliament has two months to review budget proposals (from November 1 to December 31). Parliamentary review has taken less time than legally possibly in recent years (e.g. in 2005, the annual budget law was adopted in early December; and in 2006, by Nov. 22). Dimension (iii): A (iv) Rules for in-year amendments to the budget without ex-ante approval by the legislature According to the 2002 Law on State Finances (article 46): “In case of decrease or increase of budget revenues, that can lead to the change of financing in comparison with approved budget for more or less then 10% of yearly prescription, the Government of the Republic of Tajikistan in this case submits to the Majlisi namoyandagon Majlisi Oli of the Republic of Tajikistan the draft Law of the Republic of Tajikistan about introduction of amendments and additions to the Law of the Republic of Tajikistan on State budget of the Republic of Tajikistan for the next fiscal year together with necessary documents.” No supplementary budgets were adopted by the national parliament in recent years; they were issued by government decree ( postanovlenie ). Supplementary budgets have been adopted by legislative councils at local levels where revenues exceeded planned levels by more than 10 per cent. 68 However, given the overall size of the excess of actual to planned revenue that in recent years, a review also by the national parliament would have been called for (particularly in 2003 and 2004); as this situation has potentially important implications for the overall budget and for spending in sectors that are of national importance. Dimension (iv): C Indicator Brief explanation Rating (M1) 27. Legislative scrutiny of the annual budget law Parliament has sufficient time to review the budget. However, there are substantial concerns in particular about the scope of parliamentary scrutiny in the absence of an administrative classification and about the involvement of parliament in in-year amendments. C PI-28 Legislative scrutiny of external audit reports Dimension (i): Timeliness of examination of audit reports by the legislature (for reports received within the last three years) The legislature does not receive full audit reports, it receives only summary versions. The SFCC must send its report on the performance of the State budget to the President, Government and the MN before October 1. 69 Currently, the Law does not require Parliament to approve the SFCC’s 68 We could not confirm whether this rule has been followed consistently in all local entities. 69 According to article 60, Law of State Finances (2002). Page 72 72 a nnual report on budget execution, nor does it empower Parliament or its committees to summon the Chairman or SFCC staff to meetings to discuss the audits or budget execution reports (see CFAA 2004, 23). Dimension (i): D Dimensions (ii) and (iii): extent of hearings on key findings undertaken by the legislature; issuance of recommended actions by the legislature and implementation by the executive Given that the legislature is not among the main recipients of external audit reports, it currently does not play a role in reviewing and acting on these reports. According to information received from the parliamentary budget committee, the summary versions of the audit reports which are received are discussed informally within the committee, but parliament has no formal role in reacting to external audit reports. Tajikistan is receiving some technical assistance in the field of external audit (see PI 26), and as part of this assistance, the budget committee has been exposed to discussions about reforming the legal base for external audit and increasing parliament’s role. However, this process is still at a very early stage. Dimensions (ii) and (iii): D, D Indicator Brief explanation Rating (M1) 28. Legislative scrutiny of external audit reports There is currently no formal legislative scrutiny of external audit reports. D Page 73 73 D. Donor Practices Harmonization and alignment are still limited in Tajikistan; and the use of new aid modalities, such as sector and direct budget support is very limited thus far. Although Tajikistan is a signatory of the Paris Declaration, no official baseline survey was undertaken in 2006. 70 However, donors have agreed to undertake a survey among themselves. This exercise is led by SIDA; by November 2006 9 donors had submitted questionnaires, including some, but not all of Tajikistan’s largest donors. A problem presented for the assessment of these three indicators (and in particular with respect to indicator D 2) is that different sources of data show considerable variation with regard to aid disbursements. The three potential sources of data are (i) the OECD’s Creditor Reporting System, (ii) the Donor Financial Tracking system maintained by the UN in Tajikistan, and (iii) the GoT Aid Coordination Unit. Table 21: Total disbursed ODA, by source of data, in mn USD current 2003 2004 2005 2006 OECD CRS 83.6 94.6 149.2 .. UN/Principals Tajikistan 255.3 304.0 333.2 179.6* ACU 182.3 192.3 179.6 .. Notes: *the database shows an additional USD199.4 million as outstanding obligations (signed contracts and agreements). ACU data for 2006 was not yet available by May 2007. Table 22: Disbursed ODA 2003-2005; largest donors (in USD mn) according to UN Financial Tracking 2003 2004 2005 GDP in USD 1,555 2,076 2,298 Total ODA disbursed 255 304 333 ODA in % of GDP 16.2 14.7 14.5 USA 108.4 50.5 91.5 ADB 24.6 22.8 29.4 EC 13.7 38.5 45.4 WB 13.4 55.0 33.1 IMF 11.9 14.6 28.8 IsDB 10.0 22.9 21.6 Russia .. 12.5 12.9 Source : Principals Co-ordination in Tajikistan The World Bank prepared an Aid Effectiveness Review of Tajikistan in 2006. 71 An Aid Coordination Unit was established in 2002 in the presidential administration. With capacity building support from the ADB, the ACU developed a database on donor support covering aid from all sources and of all types. However, the PITAPU in the MOET rather than the ACU has been the main gatekeeper regarding the inclusion of investment projects in the PIP. The ACU had a permanent staff of 12, and an additional staff of 12 people funded on a project basis by UNDP by fall 2006. In late 2006, the ACU was dissolved and integrated as a department into the State Committee on Investment and Management of State Property (SCIMSP). The previous staff was not transferred to the newly created department, however; and in the new department, the staff dedicated to aid coordination issues is smaller. The drafting of new rules for aid management is 70 On the Paris Declaration baseline survey, see http://www.oecd.org/document/52/0,2340,en_2649_15577209_36162932_1_1_1_1,00.html. 71 Available at http://web.worldbank.org/WBSITE/EXTERNAL/PROJECTS/STRATEGIES/CDF/0,,contentMDK:20919987~menuPK :2540090~pagePK:139301~piPK:139306~theSitePK:140576,00.html. Page 74 74 p lanned. The intention is to improve the database, as well as reporting procedures for aid, especially with regard to reporting on planned projects. However, as of May 2007, it was yet unclear if and when a new system will materialize. D-1 Predictability of Direct Budget Support To date, Tajikistan has received DBS from two donors only, the World Bank and the EU. In 2006, the World Bank started providing budget support through a Development Policy Grant (USD 10.3). The European Commission has been proving budget support within the framework of its Food Security Program. Between 2001 and 2005, the EU has also provided balance of payments support under its Macro-Financial Assistance (MFA). Table 23: Direct Budget Support, planned and actual 2004-2006 2004 2005 2006 Planned Actual Planned Actual Planned Actual EU FSP (‘000 ) 8,000 7,000 7,000 6,887.5 9,000 5,025 WB DPG (‘000 $) -- -- -- -- 10,300 10,300 Source: EU and WB (i) Annual deviation of actual budget support from the forecast provided by donor agencies at least six weeks prior to the government submitting its budget proposals to the legislature The EU has specified that it makes an estimate of the funding available for the following year in time for the budget preparation process. However, the conditions for the funding have been negotiated during the ongoing financial year in 2006 (from February to June 2006) and the release of funds occurred only at the very end of the fiscal year. Budget support by the WB and by the EU FSP has been allocated on an annual basis. In 2005, the WB provided an indication of the DBS to be received by August. The receipt of DBS from the EU is based on a process that does not include a fixed time-table for the transfer of funds. For 2007 and 2008, the EU is providing a bi-annual commitment to Tajikistan for the first time (of 17m for two years). In 2006, DBS to be received from the WB was received in the full amount. DBS to be received by the EU was received by 71 per cent. For part of the funds to be received from the EU (EUR1.25m) the government asked for an extension into early 2007 to meet the conditionality. In total, in 2006, the DBS actually received to DBS that had been forecast was 85 per cent. Dimension (i): C (ii) In-year timeliness of donor disbursements (compliance with quarterly estimates) No quarterly disbursement estimates were agreed with donors providing DBS; total funds from a donor have been released at one point in time. DBS to be received from the WB was delayed by one quarter. DBS to be received from the EU was received at year end. Combining the negotiation of conditionalities, and the review of conditionalities all into the same FY when DBS is to be disbursed necessarily makes the disbursement schedule unpredictable. Dimension (ii): D Budget support is a very valuable form of support in the context of aiming to achieve poverty reduction as it allows the financing of government priorities. As Tajikistan’s PFM system improves, more donors should be encouraged to provide DBS. A DBS is provided, donors and the government need to pay greater attention to the issue of predictability. Page 75 75 Indicator Brief explanation Rating (M1) D1. Predictability of Direct Budget Support DBS is currently being provided by few donors; and the way in which it is provided is not yet sufficiently predictable. Especially in-year predictability has been low. D+ D-2 Financial information provided by donors for budgeting and reporting on project and program aid (i) Completeness and timeliness of budget estimates by donors for project support Budget estimates of donor assistance only cover assistance included in the Public Investment Program (PIP) and direct budget support. Until 2006, only PIP domestic counterpart funding has been included in the budget. In the 2007 budget, external PIP funding is included in the budget breakdown for the first time. Project support outside the PIP is not included in the budget. The Aid Coordination Unit in the presidential administration (restructured in December 2006) tracks most project aid; but this is separate from the budget process, and tracking is largely done ex- post. An annual overview of aid provided by donors is also available at a UN sponsored website (see also indicator 7, table 14). 72 PIP funding accounted for less than 30 per cent of total assistance received by Tajikistan in recent years (less than 50 per cent of total assistance according to ACU figures on total ODA). Current reporting procedures do not encourage donors to provide information on project support ex- ante at stages consistent with the government’s budget calendar and with a breakdown consistent with the government’s budget classification. As set out under PI-7, for investment funding, the government circulates a form among donors which these submit once they have approved a project. TA, such as provided by many bilateral donors, is reported twice annually to the ACU, but not to the MOF. Direct budget support is negotiated with and reported to the MOF. Dimension (i): D (ii) Frequency and coverage of reporting by donors on actual donor flows for project support Disbursements under externally funded and country-executed PIP projects (loans and grants) are tracked by the ACU and MoF on a monthly, quarterly, and annual basis. To date, the reports have not followed Tajikistan’s budget classification; and the PIP has been only partially integrated with the budget. Thus, the coverage of reporting that are included in the budget (i.e. those that are part of the PIP) is 100% (or close to it); but the reporting has not followed the budget classification until 2006. Furthermore, the ACU has maintained a substantial database that appears to have data on all or most ongoing projects by donor, including also major NGOs. The information is summarized and published in an annual report. 73 Dimension (ii): C 72 http://www.untj.org/principals/tracking/summary.php?year=2005. 73 See ACU, Foreign Aid Report 2004, Dushanbe 2005. Page 76 76 Indicator Brief explanation Rating (M1) D2. Financial information provided by donors for budgeting and reporting on project and program aid Basic reporting structures for aid have been established. However, these have not been closely linked to budget planning and budget reporting. Improvements are expected in 2007. D+ D-3. Proportion of aid that is managed by use of national procedures Only aid disbursed in the form of DBS currently fulfils the criterion of using national procedures (in full). 74 DBS accounts for less than 50 per cent of aid funds to central government. Confidence of donors in national procedures is still limited. Indicator Brief explanation Rating (M1) D 3 Proportion of aid that is managed by use of national procedures Use of national procedures by donors is currently limited to aid received as DBS. D 4. Government Reform Process 4.1 Summary of recent and on-going reforms This section aims to summarize ongoing reform efforts; pulling together the detailed information provided under different indicators in section 3. Four main reform processes are currently underway in the area of Public Financial Management in Tajikistan: 1) a reform aimed at introducing an MTEF, which is linked to wider efforts to restructure the government as well as inter-governmental fiscal relations, 2) a process of strengthening the internal control and audit system, 3) efforts to develop and automatize the Treasury system, and 4) efforts to establish (financial) management information systems for sectors (especially in health and education). Other recent or ongoing reform efforts related to PFM include capacity development at the SFCC, and the development of parliamentary scrutiny and oversight. Furthermore, PFM reforms are set in a wider context of a Public Administration Reform as outlined in the PAR Strategy which the GoT adopted in 2006. A modern public procurement system is being initiated but its effectiveness is hampered by weak capacity and governance. In March 2006, a new public procurement law based on international standards came into force (see indicator PI-19). To implement the new law, the Procurement Agency is currently implementing an action plan that is expected to be carried out with the financial assistance of an IDF grant. The action plan includes preparation of implementation regulations, essential standard bidding documents, dissemination of the PPL and the regulations, and initial training of public officials at all levels of government. The draft standard bidding documents have already been prepared and are expected to be made available to procuring entities by March 2007. It is expected that the implementation of the action plan will result in improved awareness of the new 74 See also: World Bank, Aid Effectiveness Review: Tajikistan , November 2006; and SIDA, Results of the Baseline Survey on Aid Effectiveness (presentation), 2006. Page 77 77 P PL among public officials at all levels of government; in development of a regulatory framework, and in strengthening the capacity of procuring entities to enable them to implement the law in an effective manner. The Government is making efforts to repeal the law on tender which exists in parallel with the new procurement law and may cause confusion among procuring entities. The ADB is currently implementing a comprehensive Customs Modernization Project with a loan of $10.7 million. The ADB had also planned on a comprehensive tax modernization program; but has decided not to proceed with it. Given the weaknesses in the tax administration diagnosed by PI- 13 to PI-15 it seems important for the GoT and for donors to develop a reform plan for this important area of Tajikistan’s fiscal system. The WB has indicated that it may seek to facilitate a dialogue between the government and interested donors (DfID, EU, Japan, ADB) on these issues. These current reform efforts build on a previous round of reforms, which inter alia aimed at introducing an MTBF, and at supporting the legal reforms in the area of PFM as set out in chapter 2 above. This previous round was initiated in a post-conflict situation in the late 1990s. It has contributed to building Tajikistan’s PFM system as it stands today, and to imbue it with some important elements of capacity as reflected in this assessment. At the same time, not all efforts have been successful; and some reforms were overly ambitious relative to the situation. 75 Moving forward, the introduction of an MTEF and the establishment of financial management information systems are the most ambitious of the reform efforts currently under discussion. A key issue of concern is that in the existing system of governance in Tajikistan, sector ministries have a relatively limited role in shaping and managing sector policies. Institutional structures at the republican level are fragmented. Furthermore, as a result of fiscal decentralization, central sector ministries’ control over local level institutions (e.g. primary education and health care) is limited. 76 The government and those donors likely to support the development of an MTEF (World Bank, DFID) are currently discussing the components as well as the sequencing of such a reform. The strengthening of the internal control and audit system is mainly supported by the EU. A reform strategy has been agreed with the MOF and is currently being reviewed by the government. The strategy contains proposals to introduce several legal changes, as well as changes to the MOF’s institutional structure and modus operandi . Proposed new legislation includes the following: a Public Sector Financial Management and Control Act, a Public Sector Internal Audit Law, and amendments to the Accounting Act, the State Financial Control Law, and the Law on State Finance . A key goal of the PIFC strategy is to strengthen internal control and audit functions, to develop such functions in main MDAs (while currently most MDAs are covered by the Internal Control and Audit Department of the MOF), and to allocate greater responsibility for internal control and audit to heads of MDAs. The proposed time horizon for the strategy is 2007 to 2010 (including the organizational and legal changes envisaged). 4.2 Institutional factors supporting reform planning and implementation 75 Some donors have analyzed the reform efforts and achievements from their respective perspectives. For example, difficulties were encountered in achieving conditionalities set by the EU for the provision of Direct Budget Support under the Food Security Programme. The most comprehensive analysis of past PFM reform efforts is available from the World Bank, see: World Bank, Implementation Completion Report of IBTA2 , April 2006. 76 Decentralization is a difficult process in any developing country; and some devolution of (fiscal) control can be desirable. Central level ministries still have some degree of control over standards via education departments at local levels. Page 78 78 G overnment leadership and ownership . The 2006 Public Administration Reform Strategy evinces a substantial general commitment of the GOT to undertaking further public sector reforms; and among these reforms of PFM. However, previous reform efforts indicate that there is a risk that initial reform efforts are not followed through in full (see WB April 2006). At this point in time, the various strands of PFM reform summarized in section 4.1 are under discussion or under preparation. It is expected that a PFM reform strategy will be formulated in 2007 (with support from the World Bank). Coordination across government . Several government departments are implicated in planned medium-term PFM reform efforts: the MOF, line ministries, the MOET, as well as potentially the SFCC, the tax and customs authorities, and others. There is an ongoing effort at public administration reform which seeks to consolidate the structure of government (see Pro-gramma 12006); and in turn has implications for coordination across government. Specific coordination with respect to PFM reforms will need to be established in the context of drawing up a PFM reform strategy. Sustainability of the reform process. At this stage, the sustainability of an overall PFM reform process cannot be assessed reliably. To date, discussions on individual reform components are more advanced than a potential overall strategy; and no coordination structures are yet in place. Page 79 79 Annex 1 List of documents used Bearing Point. Information about the execution of the project “Reform of Economic Policy of Central Asian Republics (Tajikistan) since January 2003 , no date. EU/Government of Tajikistan. Strategy for Developing Public Financial Control in the Republic of Tajikistan , July 2006. Government of Tajikistan. Public Administration Reform Strategy , 2005. Government of Tajikistan. Self-assessment on the effectiveness of public expenditure management , 2006. Ilolov, Mamdasho, and Khudoiyev Mirodasen, “Local Government in Tajikistan,” in: Muteanu, Igor and Victor Popa (eds.): Local Governments in Eastern Europe, in the Caucasus and Central Asia , Budapest: Open Society Institute, 2001. International Monetary Fund. Tajikistan: Public Financial Management Performance and Treasury Reforms . October 2005. International Monetary Fund. Tajikistan: A Program for Tax Administration Reform , November 2005. International Monetary Fund. Republic of Tajikistan: Sixth Review Under the Poverty Reduction and Growth Facility . Country Report no. 06/62, February 2006. International Monetary Fund. Republic of Tajikistan: Report on The Observance of Standards and Codes (ROSC), Fiscal Transparency Module , January 2007. International Monetary Fund. Republic of Tajikistan: 2006 Article IV Consultation—Staff Report. Country Report no. 07/144, April 2007. Neel, Gerard. Technical Report (EU: TAJIKISTAN: Streamlining Policy Planning Processes: PRSP and PIP), March 2006. Pro-gramma. Horizontal Functional Analysis of the State Executive Bodies of the Republic of Tajikistan , Preliminary Results (presentation) 2006. World Bank. Country Financial Accountability Assessment , 2004 World Bank. Country Procurement Assessment Report , December 2003 World Bank. Public Expenditure and Institutional Review , December 2005 World Bank. Implementation Completion Report: IBTA2 , April 2006. World Bank. Aid Effectiveness Review: Tajikistan , November 2006. Page 80 80 Annex 2 Tables for Indicator PI-2 Data for year = 2003 functional head budget actual difference absolute percent 1. Public administration 83 132 49 49 59.0% 2. Defence 56 47 -9 9 16.1% 3. Law enforcement 43 46 3 3 7.0% 4. Education 101 112 11 11 10.9% 5. Health 36 43 7 7 19.4% 6. Social protection 101 109 8 8 7.9% 7. Communal services (incl. utilities) 26 44 18 18 69.2% 8. Culture and sport 18 23 5 5 27.8% 9. Heating and energy complex 6 9 3 3 50.0% 10. Agriculture 20 23 3 3 15.0% 11. Industry 4 8 4 4 100.0% 12. Transport and communications 41 57 16 16 39.0% 13. Other economic activities and services 3.4 4 0.3 0.3 8.8% 14. Other services 33 12 -21 21 63.6% total expenditure 571.4 668.7 97.3 97.3 17.0% composition variance 571.4 668.7 157.3 27.5% Data for year = 2004 functional head budget actual difference absolute percent 1. Public administration 119 160 41 41 34.5% 2. Defence 75 79 4 4 5.3% 3. Law enforcement 59 71 12 12 20.3% 4. Education 154 164 10 10 6.5% 5. Health 59 58 -1 1 1.7% 6. Social protection 154 158 4 4 2.6% 7. Communal services (incl. utilities) 37 74 37 37 100.0% 8. Culture and sport 27 33 6 6 22.2% 9. Heating and energy complex 13 21 8 8 61.5% 10. Agriculture 32 34 2 2 6.3% 11. Industry 7 26 19 19 271.4% 12. Transport and communications 62 84 22 22 35.5% 13. Other economic activities and services 5 4 -1 1 20.0% 14. Other services 56 18 -38 38 67.9% total expenditure deviation 859 984 125 125 14.6% composition variance 859 984 205 23.9% Page 81 81 Data for year = 2005 functional head budget actual difference absolute percent 1. Public administration 160 204 44 44 27.5% 2. Defence 102 96 -6 6 5.9% 3. Law enforcement 92 96 4 4 4.3% 4. Education 245 253 8 8 3.3% 5. Health 89 82 -7 7 7.9% 6. Social protection 228 232 4 4 1.8% 7. Communal services (incl. utilities) 56 80 24 24 42.9% 8. Culture and sport 38 42 4 4 10.5% 9. Heating and energy complex 15 18 3 3 20.0% 10. Agriculture 41 38 -3 3 7.3% 11. Industry 18 23 5 5 27.8% 12. Transport and communications 76 79 3 3 3.9% 13. Other economic activities and services 5 9 4 4 80.0% 14. Other services 69 24 -45 45 65.2% total expenditure deviation 1234 1276 42 42 3.4% composition variance 1234 1276 164 13.3% Data for year = 2006 functional head budget actual difference absolute percent 1. Public administration 185 224 39 39 21.1% 2. Defence 97 104 7 7 7.2% 3. Law enforcement 141 118 -23 23 16.3% 4. Education 336 317 -19 19 5.7% 5. Health 106 105 -1 1 0.9% 6. Social protection 263 293 30 30 11.4% 7. Communal services (incl. utilities) 67 103 35.8 35.8 53.3% 8. Culture and sport 59 62 2.7 2.7 4.6% 9. Heating and energy complex 26 38 12.3 12.3 47.9% 10. Agriculture 50 44 -5.6 5.6 11.3% 11. Industry 12 47 35.5 35.5 308.7% 12. Transport and communications 74 87 12.7 12.7 17.1% 13. Other economic activities and services 5 4 -1 1 20.0% 14. Other services 81 72 -8.6 8.6 10.7% total expenditure deviation 1,501 1,618 116.8 116.8 7.8% composition variance 1501 1618 233.2 15.5% Page 82 8 2 A n n e x 3 : O r g a n i z a t i o n a l C h a r t , M i n i s t r y o f F i n a n c e M i n i s t r y o f F i n a n c e M i n i s t e r S e c r e t a r i a t 1 s t d e p u t y m i n i s t e r 2 n d d e p u t y m i n i s t e r 3 r d d e p u t y m i n i s t e r 4 t h d e p u t y m i n i s t e r C e n t r a l T r e a s u r y D e p a r t m e n t I n t e r n a l c o n t r o l a n d a u d i t o f f i c e B u d g e t D e p a r t m e n t D e p a r t m e n t f o r s t a t e d e b t a n d f i n a n c i a l a s s e t s B u d g e t o f f i c e f o r d e f e n c e a n d l a w e n f o r c e m e n t D i v i s i o n f o r b u d g e t r e l a t i o n s w i t h C I S c o u n t r i e s F i n a n c i a l o f f i c e f o r G B A O O f f i c e f o r o v e r a l l p l a n n i n g o f t h e s t a t e a n d r e p u b l i c a n b u d g e t O f f i c e f o r t h e s o c i a l s p h e r e O f f i c e f o r i n v e s t m e n t p l a n n i n g F i n a n c i a l o f f i c e f o r K h a t l o n o b l a s t B u d g e t o f f i c e f o r g o v e r n m e n t a u t h o r i t i e s a n d p u b l i c a d m i n i s t r a t i o n O f f i c e f o r a c c o u n t i n g a n d a u d i t i n g m e t h o d o l o g y C o m p u t e r a n d i n f o r m a t i o n d i v i s i o n O f f i c e f o r b u d g e t p o l i c y i n e c o n o m i c s e c t o r s D e p a r t m e n t f o r t a x p o l i c y a n d s t a t e d u t i e s F i n a n c i a l o f f i c e f o r R a y o n s o f R e p u b l i c a n S u b o r d i n a t i o n F i n a n c i a l o f f i c e f o r t h e c i t y o f D u s h a n b e O f f i c e f o r p a y m e n t e x e c u t i o n a n d m e t h o d o l o g y D i v i s i o n f o r i n v e s t m e n t f i n a n c i n g D i v i s i o n f o r f i n a n c i n g t h e s o c i a l s p h e r e D i v i s i o n f o r f i n a n c i n g t h e e c o n o m i c s p h e r e a n d p u b l i c a d m i n i s t r a t i o n D i v i s i o n f o r t h e e x e c u t i o n a n d o v e r s i g h t o f l o c a l b u d g e t s D i v i s i o n f o r t a x r e f o r m D i v i s i o n f o r m o n i t o r i n g t a x a n d n o n - t a x r e v e n u e s D i v i s i o n f o r s t a t e r e v e n u e f o r e c a s t i n g Page 83 83