Abstract
This report assesses the Observance of Standards and Codes on Fiscal Transparency for Bangladesh. Bangladesh has made significant progress in recent years toward addressing longstanding problems of fiscal reporting and meeting basic requirements of fiscal transparency. Measures are in hand to present audited accounts from 2002/3 onward within 12 months of year-end, and monthly reports are being produced about 5–8 weeks after month-end. Much remains to be done to improve data reliability, particularly with respect to reporting of foreign-financed projects, but processes are in place to establish sustainable improvements in accounts reconciliation.
I. Introduction1
1. This draft report provides an assessment of fiscal transparency practices in Bangladesh against the requirements of the IMF Code of Good Practices on Fiscal Transparency. The assessment has two parts. The first part is a description of practice, prepared by the IMF staff on the basis of discussions with the authorities and their responses to the fiscal transparency questionnaire, and drawing on other available information. The second part is an IMF staff commentary on fiscal transparency in Bangladesh.
II. Description of Practice
A. Clarity of Roles and Responsibilities
2. General government is defined broadly in accord with international practice, but the boundaries between general government, the rest of the public sector, and the rest of the economy are blurred in practice. General government consists of central government and local and municipal governments (see Box 1). However, some of the units of central government cover activities of a commercial nature, such as the national railway and the Bangladesh Telegraph and Telephone Board (BTTB), and some general government activity carried out by autonomous and semi-autonomous bodies (such as universities) are shown in the budget only to the extent of the budget subsidies provided. Annual Development Plan (ADP) spending includes public enterprise as well as general government investment spending in the annual budget. Moreover, many fiscal activities are carried out by other parts of the public sector, as discussed below.
3. Government continues to be heavily involved in all sectors of the economy and support of enterprise activity is complex and nontransparent. The government remains heavily involved in commercial and financial sectors of the economy.2 However, as indicated in Box 1, support to the enterprise sector is provided through a complex nexus of direct subsidies and quasi-fiscal activities of nonfinancial public enterprises (NFPEs) and the nationalized commercial banks (NCBs). Lack of clarity in reporting either direct or indirect support to enterprises activity makes it difficult to assess either industry policy or the sustainability of fiscal policy. In addition to lack of transparency in reporting, the lack of clarity of managerial roles between commercial activity and provision of subsidized services or goods to the public is a major factor contributing to widely acknowledged failures of management, corruption, and poor industrial relations in these sectors. However, as part of the World Bank’s Development Support Credit and subsequent to the mission, the authorities have agreed on periodic price adjustments for gas, power, water and petroleum based on economic pricing formulas, which should reduce the costs of quasi-fiscal operations in these sectors.
General Government in Bangladesh
The general government in Bangladesh consists of the central government and local and municipal governments. Central government is organized into divisions, self-contained administrative units with separate budget heads; divisions or groups of divisions are constituted as ministries (Rules of Business, 1996). As of the 2002/03 budget there were 53 ministry/divisional budget heads. Activities are carried out by departments and directorates—as of June 2002, there were [178] departments.
Administratively, Bangladesh is divided into six divisions plus four metropolitan (city) areas. Each division is subdivided into districts, which are in turn divided into thanas (for rural areas) and pourashavas (urban areas). Thanas are divided into unions, and each union into villages. As an extension of central administration, officials are appointed to District Councils (Zila Parishads), of which there are 64, and Thana(Upazilla) Parishads, of which there are 489. Although the constitution provides for representative local government bodies at every administrative level, local representative government at present consists of 4 city corporations (Dhaka, Chittagong, Khulna and Rajshahi), 198 municipalities, and 4,472 union councils covering some 85,500 villages. These bodies spend Taka 5-10 billion a year of public funds. In September 2002 the Cabinet subcommittee on local government made a recommendation to proceed with the establishment of elected bodies at the upazilla level.
Subsidies and Public Sector Quasi-fiscal Activities
The government pursues a range of social and economic objectives through various forms of direct and indirect subsidies. These interventions have complex interactions, and the policy objectives are generally not explicitly stated in the budget documents. The budget impact is both substantial and, to a large extent, unpredictable, and the extent to which (implicit) policy objectives have been achieved is not monitored or reviewed.
Direct budget subsidies are provided to farmers for (among other things) fertilizers, seeds and agricultural loans (amounting to Tk 1.0 billion or 0.04 percent of GDP, in FY 2000/01); open market sales of food (carried out to stabilize domestic prices) and food subsidies to police and defense personnel (Tk 3.1 billion, or 0.14 percent of GDP, in FY 2000/01); and subsidies to export industries, mostly jute products (Tk 0.9 billion, or 0.04 percent of GDP, in FY 2001/01). Other direct subsidies include subsidies to the Rural Electrification Board and for internal water transport. The budget documents show aggregate spending on subsidies and transfers, but no summary of subsidies by purpose nor any discussion of subsidy policy is provided.
Some specific purpose subsidies are provided to NFPEs, such as export subsidies to Bangladesh Jute Mills Corporation. However, the main (indirect) cost to the budget arises from government policies of subsidizing consumers through administered prices, particularly in the gas and energy sectors, which result in inter-enterprise arrears and inefficiencies that compound eventual losses from the public sector balance sheet. The World Bank (June 2002) has estimated that subsidies from the major power companies were around Tk 21 billion (0.8 percent of GDP) on a cash break-even basis (more on a long run marginal cost basis) in FY 2001/02. The World Bank also estimates that, in the petroleum sector, due to the absence of a mechanism that allows periodic adjustments to align domestic and import prices, Bangladesh Petroleum Company has incurred heavy losses (in FY 2001/02 around Tk 10 billion, or 0.4 percent of GDP).
Financial pressures in the NFPEs lead to inter-enterprise arrears in payments and tax arrears to the government. Periodically also, deficits of the enterprises have to be financed directly by lending or equity injection from the budget. The deficit of the ten main NFPEs amounted to Tk 71 billion (2.8 percent of GDP) in FY 2000/01; about half was covered by the budget, and the remainder by commercial borrowing and accumulation of arrears. The government also assists enterprises by guaranteeing commercial loans, adding to government contingent liabilities. It is important to note that these loans to NFPEs carry a high risk of default.
This situation is further compounded by the difficulties created by the quasi-fiscal operations of the NCBs. In particular, lending policies of these banks in support of loss-making NFPEs has increased the level of nonperforming loans and this in turn leads to eventual pressure for bank recapitalization. In FY 2001/02, Tk 1.25 billion was spent directly on bank recapitalization.
Policies of support through such off-budget mechanisms lack clarity, achievement of the intended policy objectives are not monitored or reviewed, and they give rise to substantial, but nontransparent, budget pressures and risks to the government’s intended fiscal policy position.
4. Privatization procedures are transparent, but coordination of overall policy needs further clarification. The Privatization Commission3 was set up under the Privatization Act 2000. It is a statutory body, which under Rules of Business is administratively part of the Prime Minister’s Office. It has overall responsibility, under the direction of the Cabinet Committee on Economic Affairs (chaired by the Minister of Finance), for privatizing those state-owned enterprises identified for privatization. The Commission has full autonomy in the divestment of listed enterprises with a sales price of under Tk 250 million; government final approval is required for all above that limit. The Commission lists a wide range of enterprises for privatization, many of which have been previously tendered.4 There may be a need to re-examine and clarify policies across the range of options for dealing with loss-making enterprises. Principles of transparency should apply also to closure and asset sale operations; disclosure of contingent liabilities that may arise are a particular concern in this regard. Current policy is addressing some of the major impediments to privatization, such as strengthening social safety net for laid-off workers and improving the regulatory environment (see paragraph 7 below).
5. Regulation of nonbank private sector is being simplified, but still appears subject to discretionary official ruling. Private sector activity has been regulated by a variety of outdated financial and commercial laws, with consequent opportunities for discretionary interpretation. The government has taken steps to simplify the regulatory framework, both with respect to unifying the framework in the energy and telecommunications sectors and establishing a “one-stop shop” for regulation of new investment through the Board of Investment (BOI). However, staff discussions indicate that BOI procedures remain complex, and continuing weaknesses in the judicial and law enforcement5 system and limited institutional capacity continue to provide official and private rent-seeking opportunities and impose significant transaction costs on entrepreneurs. This environment adds to the difficulties of establishing a strong basis for tax compliance. In this context it should be noted that, in April 2003, Parliament passed the Energy Regulatory Commission Act, creating an independent regulator for the sector, and the Independent Regulatory Commission for Telecommunications is now operational.
6. The central bank is not yet fully independent, and it supports quasi-fiscal activities of PFIs. The Bank of Bangladesh (BB) is administratively under the Ministry of Finance and its nine member Board, including the Chairman, consists of government nominees, including three government officials. However, the government has given priority to improving the legal framework and upgrading prudential standards in the banking sector. Under the Bangladesh Bank Order of 1972, as amended, BB had limited autonomy with respect to the design and implementation of monetary and exchange rate policies. This Order was amended in March 2003 to provide greater operational autonomy to BB. The Banks (Nationalization) Act and the Banking Companies Act were amended at the same time to bring the Nationalized Commercial Banks (NCBs) and the Development Finance Institutions (DFIs) under closer BB supervision. Moreover, a new Money Loans Court law has been enacted that establishes specialist courts to deal with loan defaults and streamline the recovery process. About 100 loss-making branches of NCBs have been either closed or merged in the current financial year. However, the new banking law is still not clear on procedures for the winding-up of problem banks, and BB still requires the approval of the Finance Ministry to enforce corrective measures over the NCBs and DFIs.
7. Despite recent reforms, Public Financial Institutions (PFIs) continue to carry out a range of quasi-fiscal activities. While ownership of the banking system has become more diversified in the last decade or so, government involvement in the financial sector remains extensive.6 The four NCBs accounted for almost 60 percent of overall banking sector deposits at the end of 2001, and an even higher share of risk-weighted assets. The law (Articles 60–62) explicitly permits the BB to provide credit to specialized credit institutions (rural credit, export credit, and industrial credit), partly financed by direct appropriation from bank profits, although the BB has ordered banks to not exceed a maximum rate on credit for exporters. Interventions such as directed lending to NFPEs, as well as the channeling of subsidized credits thorough the NCBs and DFIs to favored sectors and enterprises such as agriculture and jute, have undermined their profitability and soundness and led to a significant accumulation of non-performing loans.7 The potential fiscal costs stemming from the granting of implicit government guarantees on loans of questionable quality are not discussed in the budget documents.8
8. Government holdings of equity and financial assets are very extensive and are partially reported. As well as its ownership of NFPEs, the government also retains holdings in joint-venture companies and residual holdings of privatized companies. However, subsequent to the mission’s visit, the government decided to privatize or close down, over the course of the next three fiscal years, all manufacturing units in the SOE sector except those in the sugar and fertilizer industries; this should substantially reduce the role of the state in the economy. All NFPEs are monitored by the MOF Monitoring Cell, but minority equity holdings are not systematically reported. A consolidated account of the NFPEs is prepared each year. It is approved by the MOF, but is not presented to parliament. NFPEs go through a somewhat similar budgeting, reporting, and audit cycle to government departments, though not subject to parliamentary appropriation control. All, including BTTB, the Post Office, and the Railways, use commercial accounting, although accounting practices as regards basic reconciliation and asset valuation are acknowledged as being extremely weak.9
9. Local government is subject to considerable central direction, administrative capacity is weak, and local government data are not systematically reported. Control over general government finances is complicated in part by the exclusion of local government data from the fiscal management system, and weak administrative capacity at the local levels. Moreover, the structure of intergovernmental fiscal relations in Bangladesh has served to discourage local governments from raising their own revenues. Data on local government budgets and accounts are provided by local authorities to the Ministry of Local Government, Rural Development and Cooperatives, but these are not systematically compiled or analyzed. The authorities advise that steps are being taken to establish a consolidated report from 2003/04.
10. Stable principles of decentralized management are not fully formulated. The 1983 Local Government Ordinance, and subsequent modification in 1990, reformed the government structure by devolving expenditure responsibilities in the areas of law and order, infrastructure development and maintenance, health and education to local governments. Local governments are, however, heavily dependent on revenue from the central government in the form of grants and transfers, which have not been based on a clearly defined set of principles. The negotiated and non-transparent basis for transfers has discouraged local governments from raising their own revenues (which consist of locally imposed fees and property taxes) and has created an over-dependence on the central government budget. Moreover, the mismatch between spending responsibilities and revenue sources, and lack of effective oversight by the center, have eroded incentives to maintain fiscal discipline at the local government level. Some steps have been taken, however, to develop more systematic principles of distributing grants. From 2002/03, grants to union parishads, for instance, are distributed on the basis of 15 percent allocation according to performance, the remainder according to area and population. For district grants, 25 percent of the allocation depends on performance.
11. The government is committed to further decentralization, but there is no clear consensus on the direction of reform and considerable capacity constraints at the local level. Over the past decade a number of attempts have been made to strengthen local government.10 While a continuing emphasis has been given to developing union and district level institutions, different administrations have given different emphases to the relative roles of village, upazilla/thana, and regional levels of government. Regulations defining responsibilities at different levels have been issued, but fiscal authority has not been developed to match these responsibilities. Moreover, the parallel operation of central administration and local parishads has meant that resources remain largely under central direction and control. It is also widely recognized that capacity at the local level is weak and corruption is prevalent. As noted earlier, the Cabinet Committee on Local Government has recommended that the government proceed to establish elected upazila parishads.
12. Fiscal roles of the executive, legislative and judicial branches are clearly defined in the constitution, but many aspects of the relationship need strengthening and are still evolving in practice. The 1972 constitution defines the roles of the executive, legislative, and judicial branches of government clearly, and Articles 81–92 of the constitution give clear basic principles for fiscal management (see following paragraph). Articles 127–132 established the role of the Comptroller and Auditor-General (C&AG) to carry out the auditing functions of government. The Additional Functions Act of 1974 (as amended in 1975 and 1983) had also allocated the accounting function to the C&AG. The accounting function, however, has now been transferred to the Finance Division of the MOF (hereafter, “Finance Division”) of the executive branch. However, in practice, the relationship between the different branches (particularly between the executive and legislature) is widely seen as confrontational11 and parliamentary consideration of economic reforms and development of effective watchdog institutions remains weak (see further discussion of audit arrangements in paragraphs 38–40 below).
13. The fiscal management legal framework is clear and based on sound basic principles, but has some key structural weaknesses. Article 84 of the constitution establishes the fundamental budget principle of universality, requiring that all receipts of government be paid into the consolidated fund or (for moneys of a trust or administrative nature) the public account of the republic; and Article 85 requires that all payments shall be by parliamentary act or presidential rule. Budget preparation is governed by detailed regulations issued under the authority of the President that define key responsibilities of various government entities, including the Finance Division; secretaries of ministries or divisions (who are designated as the principal accounting officers (PAOs) and held accountable for overall ministry or divisional management); the Controller General of Accounts (CGA); chief accounts officers; and other officials involved in handling public money. The system, while sound in many respects, does not provide sufficient assurance of regular monitoring and reporting, and, as discussed in the following paragraph, has contributed to some crucial failures of coordination. These aspects should however improve as a result of separation of audit and accounting functions and improved reconciliation practices.
14. The budget’s legal framework has impeded coordination both of development and recurrent spending and of budgeting and accounting. These key issues are now being addressed by the authorities. The separation of the consolidated fund transactions into recurrent and development spending has led, as in many other countries, to a failure to adequately match development project expenditures with adequate operating and maintenance provisions in the recurrent budget. The latter, moreover, is primarily administered by the Finance Division, while the Ministry of Planning is responsible for both planning and monitoring development spending. As discussed in paragraph 52, the government plans to address this issue by developing a medium-term expenditure framework (MTEF) which will incorporate a rolling three-year public investment program. Overall coordination has been further weakened by the lack of a comprehensive fiscal reporting system. Accounting is inadequate in terms of coverage, timeliness, and reliability, but has improved significantly in recent years (see paragraph 31). While problems remain, particularly with respect to foreign-financed projects (see following paragraph), accounting performance in the CGA controlled accounts has improved very substantially.
15. Coordination of externally financed projects with overall fiscal management needs strengthening. The External Relations Division (ERD) of the Ministry of Finance is responsible for monitoring resource flows for externally financed development projects and for external debt management. The Implementation, Monitoring, and Evaluation Division (IMED) of the Ministry of Planning monitors project implementation on the basis of financial and physical data from project directors. The CGA gets financial data on project expenditures through CAOs for local funded components, but relies on reports from project directors for the externally financed component. All external financing flows through various forms of special accounts (as prescribed by different donors) and these accounts are managed by project directors rather than the CAOs. At present, reconciled statements of expenditure and receipts cannot be produced because of incomplete reporting. One fundamental problem is that the CGA records only expenditures12 based on project directors’ reports. Foreign aid disbursements are entered in the CGA accounts simply as an offsetting counterpart item to the recorded expenditure. Actual receipts data cannot be reliably obtained from the ERD because it comes from donor reports, and these do not provide appropriate breakdowns. A second major problem is that bank statements are not provided by project directors to the CGA. Moreover, only some receipts flow through the consolidated fund accounts.
16. Tax laws and regulations are published, but there continues to be scope for discretion on the part of tax officials. The need to reduce scope for official discretion was strongly recognized in the FY 2002/03 budget speech. The National Board of Revenue (NBR) has taken steps toward strengthening tax and customs administration, including inter alia the establishment of a Large Taxpayers Unit, a mandatory pre-shipment inspection system, a unique taxpayer identification number, and in June 2002 the establishment of a Revenue Reform Commission. However, the legal basis for taxes continues to be undermined by the practice of the executive branch of granting exemptions to protect or to foster specific commercial activities and, more generally, to influence resource allocation. In particular, non-transparent tax incentives and open-ended subsidies are granted to certain industries in the private sector, particularly in the export sector. These extensive specific and ad hoc exemptions, as well as open-ended tax holidays, have reduced tax collection, impaired the equity and transparency of the tax system, provided discretionary power to tax officials, and encouraged rent seeking. As noted further below, there is no systematic reporting to the public on the extent and cost of such concessions.
17. Tax regulations and procedures are clearly documented, but are ineffectively applied because of inadequate monitoring of their implementation. Compliance with tax law is difficult to assess, but available evidence on tax arrears and overall tax performance indicates that current legislation on penalties for late filing and payment are not regularly enforced and that tax evasion is common. Moreover, the lack of effective supervision of tax officials allows for some taxpayers’ rights to be infringed; although taxpayers have the right to go to tax appellate tribunals to get redress for their grievances, the process is regarded as highly politicized and ineffective.
18. A public sector ethics code and disclosure of assets rules exist, but are not adequately enforced. Bangladesh has long-established institutions and laws for curbing corruption. The principal laws governing anti-corruption initiatives are: (1) Prevention of Corruption Act, 1947; (2) Anti-Corruption Act, 1957; (3) Criminal Law Amendments Act, 1958; and (4) Anti-Corruption Tribunal Ordinance, 1960. While the laws and the institutions seem to be in place, their effectiveness in curbing corruption is limited in practice, in part because of the widespread prevalence of non-transparent practices in the public sector. Allegations against politicians are seldom independently investigated. Some steps have been introduced this year against lax treatment of officials with regard to taxation by enforcing the requirement for submission of a TIN as part of salary payment procedures. An Anti-Corruption Bureau has been established in the Prime Minister’s Department. The need to make such an institution clearly independent of the executive is widely supported.13
B. Public Availability of Information
19. The budget provides a comprehensive coverage of central government fiscal activity. Line ministries/divisions prepare detailed budgets covering activities within their responsibilities, and these budgets are further broken down by individual agencies within each ministry.14 Data for local government budgetary operations are not easily available and are not presented with the budget. Detailed data on the estimated expenditures of the Ministry of Defense are provided in the budget documents. Ministry of Defense expenditures are treated in the same way as other ministries in the budget. Accounts for defense are departmentalized and, while consolidated by the CGA, transactions are not directly brought to account through the CGA. The government accounting and audit rules and regulations apply to defense transactions.
20. The budget document discloses the main fiscal aggregates for the budget year and the preceding year’s revised budget estimates. Data on central government operations are compiled according to constitutional principles and government regulations. Comprehensive data on actual budget outturns for the central government have not, however, been produced and disseminated on a timely basis, owing to extensive delays in identifying government transactions and problems with reconciliation of fiscal data with information from bank accounts. Revised budget estimates are only disseminated on an annual basis at the time of presentation of the next year’s budget. Revised budget numbers for the preceding fiscal year rather than final year-end data on government revenues and expenditures are published in the summary budget document, but, in the Demands for Grants and Appropriations of 2002/03, preliminary accounts data for two years preceding the budget (2000/01) were presented. No projections for future years are given. As discussed further in paragraph 52, the authorities propose to introduce a medium-term expenditure framework (MTEF), in which fiscal projections for two years beyond the budget year would be included. As discussed further in paragraph 52, however, it is critical that implementation of such proposals adequately emphasize accountability of forward estimate projections.
21. No estimates of tax expenditures, contingent liabilities or the extent of quasi-fiscal activity are included in the budget documents or annual accounts. Changes to revenue measures are listed in full in the annual budget speech, but no estimate of the cost of nonstandard concessions is given. The NBR authorities indicated that data are available to compile estimates of revenue foregone through concessions. No estimates are made of contingent liabilities, such as government guarantees on debt contracted by state-owned enterprises. In addition, the impact of quasi-fiscal operations of the government (as described above) is not systematically reviewed in the context of the budget, nor is any official estimate of potential costs provided.
22. Information on gross external public debt is published, but domestic debt is not reported to parliament and financial assets are not reported. The ERD compiles and disseminates comprehensive annual data on external government debt, but data for domestic debt are produced by the MOF for internal government use only. Data on government domestic borrowing and debt are published in the BB’s Annual Report and made available on its website (http://www.bangladesh-bank.org/) about nine months after the end of the fiscal year. Public debt information is not included in the audited final accounts of the government. As noted above, data is compiled for NFPEs but these data are not used for reporting on government financial assets.
23. Formal commitments for more regular publication of fiscal data have not been made and advance release date calendars are not announced. Reporting of fiscal data has suffered from significant weaknesses and currently falls short of GDDS recommendations with regards to timeliness, comprehensiveness and reliability. However, as a result of a long process of reforms through the Reforms in Budget and Expenditure Control (RIBEC) family of projects,15 the Finance Division has begun producing quarterly reports on the central government fiscal position drawing from the Controller General of Accounts (CGA) monthly reports and ERD data on external financing. Subsequent to the mission’s visit, the practice of preparing monthly (and quarterly) reports on the government’s fiscal position has been established; the reports are published on the MOF website (http://www.gobfinance.org/).
C. Open Budget Preparation, Execution, and Reporting
24. The annual budget process is clearly defined, but the process is geared to incremental additions to existing commitments and lacks an open and clearly disciplined priority setting mechanism, particularly for development projects. The annual budget call circular provides clear directions to ministries for preparing budget estimates. There is not a good linkage between recurrent and capital budget preparation, nor of their implementation. The Finance Division is responsible for overall budget preparation and detailed administration of the revenue budget; the Ministry of Planning is responsible for planning and implementation of the development budget. There is no adequate mechanism for overall review of spending and revenue priorities. The Annual Development Plan (ADP) budget process is not clearly disciplined and has been undermined by the incorporation of a large number of projects mid-year, often with questionable priority and rationale for public sector investment, resulting in scarce budgetary resources being spread over a large number of projects. Most such projects are not subject to the evaluation and scrutiny that apply under normal budgetary mechanisms, making the approval process more vulnerable to political influence. Some improvements have been made in procedures for release of funds during the year for development projects. These changes are aimed at more orderly spending and avoidance of year-end rush of spending. The lack of coordination between development and revenue budgets has been noted above (paragraph 14). Provision for operations and maintenance spending in the revenue budget following project completion is made by block allocation in the year of completion, and this expenditure is subsequently incorporated in the regular item-wise provisions.16 The authorities indicated that maintenance allocations through this mechanism are generally inadequate, mainly because of concerns about wasteful use of maintenance provisions. The extensive use of block allocations reduces transparency of the budget process and provides considerable discretionary power for reallocation during the year.
25. Economic assumptions underlying the budget estimates are not shown systematically or comprehensively and there is no statement of fiscal risks. The annual Budget Speech provides a broad economic background and indicates some of the key assumptions underlying the estimates. These are not, however, presented in a tabular or other systematic form and do not facilitate an analysis of budget realism. No sensitivity analysis of budget assumptions is undertaken and there is no statement of overall macroeconomic risk. Revenue projections are based predominantly on past trends and examination of specific factors that influence various taxes.17
26. The overall deficit of central government is the main focus of fiscal policy. Fiscal data are classified in a way that is compatible with the 1986 GFS conventions on analytical reporting. Although past years’ data have not been of sufficient quality to report in the GFS Year book, improvements underway in accounting and reporting are designed to permit tracking of the overall deficit of the central government; further improvement in above-the-line data is needed to make this fully operational. The prevalence of quasi-fiscal activities described above, however, suggests that it would be appropriate to develop some form of augmented general government balance once data sources are significantly improved.
27. New policy proposals are estimated separately from existing policy expenditures during budget preparation. Ministries and divisions submit their revenue budget estimates of requirements over and above existing policy (“permanent”) expenditures (new item statements) by end-November of the pre-budget year. New projects are distinguished from ongoing projects in the ADP. Budgeting and accounting are input-based, with no systematic definition of objectives or outputs for performance tracking.
28. Internal control and internal audit procedures are weak. Currently an internal audit arrangement is absent in almost all ministries, divisions and departments except the Ministry of Works, Ministry of Land, Ministry of Roads and Highways, and Department of Education. Some CAO offices have pre-audit arrangements carried out by the staff of the C&AG. Most NFPEs, however, have internal audit departments. No systematic attention has been given to assessing effectiveness of the existing internal audit functions, and few of the staff of the internal audit departments have adequate professional training.18 At the aggregate control level steps are being taken to improve coordination of monitoring of the development and revenue budgets by regular, high-level meetings of the Finance Division, Planning Ministry, and ERD, and through the newly established Public Expenditure Review Committee. The authorities plan to integrate the internal audit function into all activities of the Finance Division and to set up a pilot internal audit unit in the MOF.
29. Procurement lacks a comprehensive framework and systematic procedures. Although tender committees are established for major procurement, oversight of these procedures is inadequate.19 In recent years ADP projects in transport, shipping and telecommunications have been increasingly financed through suppliers’ credit obtained on non-transparent terms that circumvent competitive bidding processes.20 There is no legal framework for procurement and the government has not established standard bidding documents, though a set of guidelines has been issued by ERD for externally financed procurement. Procurement processing is inordinately prolonged, often due to interferences by interest groups. Lack of knowledge and skills at the working level is another contributing factor. These matters are to be addressed through a World Bank-supported project to prepare a procurement law, standard bidding documents, and training of government officials.
30. Civil service recruitment and pay procedures remain rigidly structured and are not open. The civil service is overmanned21 and is not open to competitive employment practices.22 The basis of promotion remains largely based on district quotas and seniority, and this has reduced the overall quality of the civil service. The current government has, however, introduced a greater degree of merit-based promotion, but it is limited to a particular base/year of recruitment of civil servants. Pay structures are rigid, and have little differentiation between grades. According to the PARC report salaries are far below the private sector at higher grades and overgenerous at the lower grades.
31. The accounting system has improved significantly, but does not yet produce complete and accurate in-year reports on central government budget outturn. Reliability and timeliness of accounting reports have improved dramatically (see Box 3) in recent years as a result of the RIBEC project, although significant problems remain.
RIBEC and Reforms in Public Sector Financial Management
Following the 1989 recommendations of the Committee on Reforms in Budgeting and Expenditure Control (CORBEC), the UK’s DFID has supported public sector financial management reform in Bangladesh through the RIBEC family of projects, which started in 1993 and ended in September 2001. So far approximately £16 million has been spent on this family of projects.
Good progress has been made in the field of public accounts, which have been reclassified according to international standards and successfully computerized with RIBEC support. RIBEC has helped the CGA to capture summarized transactions data at the ministry and regional level, and to improve the timeliness and quality of aggregate monthly accounts. Prior to RIBEC, final accounts were not presented until 2–3 years after year-end, within-year accounts were not prepared for 10–14 weeks after month end, and reconciliation processes were highly inadequate for both final and within-year reports. The main achievements under RIBEC have been to:
establish a working reconciliation process for CGA accounts;
attain more timely within-year reporting of fiscal data to 5–8 weeks after month-end, and to improve the timeliness of submission of audited accounts to parliament;
separate the accounting and audit function in July 2002;
improve the quality and timeliness of audit reporting; and
establish FMUs in ministries with heavy ADP spending.
No information is available through CGA on outcomes for most non-ADP capital spending and the entire food account23 during the budget implementation period; such data only becomes available after the end of the fiscal year with a lag of a few months.24 Accounts are maintained on a cash basis and the system does not provided data on expenditure arrears—although these are not generally perceived to be a major issue at present. As noted in paragraph 15, foreign-financed project spending is also a source of problems for timely and reliable fiscal reporting. The RIBEC project has helped address this issue to some extent through the FMUs that have been introduced in seven line ministries (all of which have heavy ADP expenditure). Overall, however, the RIBEC project appears to have attained a critical momentum in establishing an effective reporting system: as well as achieving respectable levels of improvement in timeliness and data quality, the authorities are committed to addressing the remaining problems in a systematic and disciplined way.
32. The government is introducing a mid-year fiscal review process. As standard practice the MOF conducts an internal mid-year review of the budget (scheduled for completion by end-February), and the government prepares revised budget estimates25 toward the end of the fiscal year. In 2002/03, the government plans to present the legislature with a mid-year fiscal review to assess economic and fiscal trends and indicate any revisions to its fiscal policies.
33. The national tax administration is not given separate legal authority. The NBR is not given independent status within the administrative structure and its authority derives from tax legislation and Rules of Business. It does not directly provide any report to the legislature other than the overall departmental reports compiled by the Cabinet Secretariat.
34. The audited final accounts are not yet available within twelve months of the end of the fiscal year. Up to the present, the central government accounts26 have been prepared and audited by the C&GA. The latest such documents presented to parliament is for FY 1999/2000. However, FY 2000/01 accounts are nearly finalized, and it is anticipated that the improvements in the accounting system will permit submission of the FY 2001/02 accounts to the C&AG by end-January 2003—and, thereafter, regular submission of annual accounts to parliament within 12 months of year-end, as currently required by executive order.
D. Assurances of Integrity
35. Indicators of the reliability of budget data are not made publicly available, and the reliability of budget data needs to be improved. Until recently, it has not been possible to prepare an up-to-date comparison of actual outturns with original budgets because of the nonavailability of final accounts data. Data for FY 2000/01, now available, indicate some weakness in the reliability of the original budget estimates—principally in estimation of nontax revenue for that year.27 Publication of assumptions underlying revenue and expenditure estimates would also help strengthen budget data reliability.
36. Statements of accounting policy are not included in the budget and final accounts documents. Government accounts are prepared on a cash basis. No statement of the basis of accounts, or of key elements of accounting policy with regard to recognition and disclosure of data, are provided in the budget or accounts documents.
37. The processes of accounts reconciliation and fiscal reporting are not yet fully effective. Technical improvements achieved through the RIBEC project have improved reconciliation in all areas directly under the control of the CGA. These improvements are being institutionalized both through regular follow-up procedures by the CGA28 and, on a broader level, by an interdepartmental Task Force on Monitoring Macroeconomic Data, which is looking at improving data reconciliation from foreign-financed projects as well as CGA data (see paragraph 15).
38. External audit is functionally independent, but administratively part of the executive. The Auditor General’s position is established under the constitution, which specifies that he shall not be subject to direction or control of any other person or authority in the exercise of his functions (Article 128 (4)). Under the Rules of Business, however, the C&AG is a division of the MOF, and is treated like any other government division in terms of budget and establishment review. Constitutionally, the Auditor-General is appointed by the President, but Rules of Business requires that the Finance Division present the case for appointment, resignation, removal, conditions of service, making provision for additional functions (under Article 28(3) of the constitution), and the form and manner of keeping public accounts. In practice, the Finance Division does not influence the operational work of the C&AG. The perception of close administrative supervision by the principal agent of the executive government subject to C&AG scrutiny, however, appears more limiting of independence of the external audit function than the general international practice.29
39. Strengthening of audit capacity is required. The C&AG has limited capacity to produce high quality audit reports over the full range of government entities to be covered. Of 3500 established posts, only 2500 are filled, and a high proportion are relatively low level staff. The proportion of cadre officers is low, at only around 2 percent (in the 9 Audit Directorates, there are only 54 cadre officers). Audit teams are headed by an officer but most of the other team members are junior level staff and lack expertise in modern system-based audit. Manuals are available and codes have been updated, but capacity to implement them remains weak and there is a need for more training and development in their practical use. Other weaknesses are the lack of systematic working files and absence of systematic peer review processes. Auditing practice gives too little emphasis to materiality of findings and provides instead detailed reports on financial and regulatory compliance without assessing systemically important deficiencies. The FMRP will help to train audit staff, improve the quality and focus of audit reports, and improve capacity, with the overall aim of establishing the C&AG as an independent, effective service.
40. Neither the legislature nor the executive adequately follow up external audit reports. In Bangladesh, the cycle of accountability remains incomplete in most cases, as the Public Accounts Committees (PACs) of parliament take an excessively long time to consider audit recommendations. The unfocused detail of audit reports also contributes. Partly as a consequence of such delays, the executive arm of the government is very slow to take actions on the recommendations of the C&AG. No PAC has yet been formed under the present government, apparently due to non-participation of the opposition. The last PAC appointed had started reviewing the audit reports of 1993/94 and also some special audit reports prepared by the C&AG on different agencies and ministries. No permanent secretariat or mechanism has been established to review the implementation of the recommendations of the PAC.
41. External scrutiny of macroeconomic assumptions is not undertaken. As discussed above, macroeconomic and technical assumptions underlying the budget are not presented in a sufficiently comprehensive and systematic way to allow detailed external scrutiny.
42. Fiscal statistics are compiled on solely technical criteria, and institutional performance is being strengthened. The Bangladesh Bureau of Statistics is the national statistical agency, but its role in the compilation and dissemination of fiscal data is limited. As indicated earlier, fiscal reports are compiled by the Finance Division drawing from CGA and ERD, with increasing emphasis on reconciliation with departments and revenue agencies (particularly NBR and the National Directorate of Savings) and BB.
III. IMF Staff Commentary
43. Bangladesh has made good progress in basic areas of fiscal transparency. In recent years, the standards of accounting and fiscal reporting are improving toward meeting a number of the basic requirements of the fiscal transparency code. A number of significant gaps (particularly in coverage) remain, but the processes of reconciliation and reporting by the CGA have been transformed from a highly inadequate and uncoordinated mechanism to a broadly integrated system capable of producing reports that will meet international standards of timeliness, reliability, comprehensiveness, and analytical capacity. This achievement, in turn, should provide a basis for radically improving transparency and decision-making across a wide range of fiscal management functions. It is critical that this momentum be sustained.30
44. Highest immediate priority should be given to consolidating and fully institutionalizing the improvements that have been made. As described in the previous section, the reporting system has improved considerably compared to past years, but adequate coverage of the system and complete and timely reconciliation are yet to be achieved. The problems are well recognized by the authorities and action is being taken to resolve them, but the standards and procedures for their continued application must be firmly established to ensure sustainability of the reforms. At the same time, there are a number of relatively easy gains in transparency that can be made simply by publishing and actively disseminating information that is already in the system. Greater openness in this regard will further help to strengthen fiscal decision-making and allow more participation and broader commitment to government policies. Specific suggestions toward these ends are made in paragraphs 47–51 below.
45. Fundamental improvements in transparency of budget management and of the broader public sector are central to financial management reform. The Bangladesh government is actively pursuing a wide range of public sector reforms with the support of the donor community.31 Successful implementation of this and other related reforms will improve fiscal transparency. But the authorities’ commitment to long-term improvement of fiscal transparency will also improve the likelihood of successful implementation of these reforms and the realization of real improvements in fiscal transparency and fiscal management. This approach is particularly important in addressing the nexus of issues associated with reform of the state-owned enterprise sector, as well as in the difficult task of introducing an MTEF. Some specific suggestions in both of these areas are given in paragraphs 52 and 53 below.
46. Strengthening of watchdog institutions and a sustained effort to improve governance and fight corruption are critical to the long-term improvement of fiscal transparency and fiscal management. Improvement in fiscal transparency will, in itself, contribute to reducing corruption. However, it is but one element in a much broader set of reforms, and the contribution of improved fiscal transparency may be negated if efforts to improve government regulation and reduce rent-seeking activity are not also very actively pursued. A common theme in this area is the need to establish watchdog institutions that are credibly seen as independent of the executive government. In the staff’s view, action should be taken to strengthen the role of the C&AG in this regard. Similar principles can be applied more broadly to ensure mutually reinforcing reforms aimed against the debilitating effects of corruption. Some further suggestions in these areas are made in paragraphs 54–58 below.
Consolidating recent improvements in fiscal transparency
47. Measures to establish a fully reconciled, timely, and comprehensive fiscal reporting system must be sustained and clear assurances of data quality put in place. The government should implement immediate measures to embed improved practices—achieved through RIBEC and the Task Force for Monitoring Macroeconomic Data—in its routine operations in a way that can be sustained and open to critical review. Suggested measures are as follows:
Establish the Task Force as a high-level Economic Monitoring Committee with the regular functions of reviewing the monthly report on the government fiscal position in the context of other macroeconomic data sets, and, until clear standards of reliability have been established, of reviewing data quality and action taken to reconcile CGA and other source data.
Specifically, the Economic Monitoring Committee should oversee the incorporation of clear assurances of data quality in the monthly, quarterly, and annual fiscal and financial reports of the government. Such assurances should include:
A statement of reconciliation between CGA accounts and BB/commercial bank balances—including, as soon as possible, reconciliation of ADP foreign financing flows (see paragraph 48);
A statement of reconciliation between CGA and departmental reports of transactions during the period;
A statement of uncleared suspense and remittance account balances and, for year-end accounts, an assurance that all balances have been cleared;
An assurance for year-end accounts by the CGA that no transfers of authority from the consolidated fund to the public account have occurred;
For monthly reports and all fiscal reports in GFS format, a brief reconciliation showing the relationship between the GFS deficit and budget or CGA accounting statement aggregates; and
A statement of reconciliation at the macroeconomic level between the fiscal, monetary, and external data sets.
Establish clear standards for timeliness of monthly, quarterly, and annual fiscal/financial reports—GDDS should be treated as a minimum in this regard—and publish monthly reports on the MOF website as soon as is practical.
Both external and domestic debt data should be included in the monthly and quarterly reports by the Finance Division; and annual financial statements of the government should include debt data as a memorandum record. The CGA should provide a statement of assurance that debt records are complete and consistent with transactions records.
The C&AG, in certifying the annual accounts of government, should comment specifically on government’s compliance with the standards of data quality and timeliness. It should also comment on adequacy and consistency of the debt records maintained by the CGA.32
48. Special efforts are required to improve coordination, monitoring, and reporting of foreign-financed projects. As outlined in paragraph 15, some changes in current reporting practices are required to provide a satisfactory assurance of reliable data on ADP spending and receipts. The following measures are suggested:
Project directors should be required to provide prompt monthly reports to the CAO, covering receipts and bank statements for project accounts as well as spending data;
A small External Financing Unit should be set up within CGA to monitor reporting and investigate major discrepancies between ERD, CGA, and IMED data on ADP spending, and amend CGA accounts accordingly; and
All receipts of foreign aid or loans should be recorded in a memorandum account by BB before being transferred to project accounts, and balances of the memorandum account provided to CGA, ERD, and IMED.
Over the longer run, the staff support the suggestion made in the Bank CFAA that special accounts for foreign aid and loans be brought within the government accounting system. To achieve this objective, however, the CGA accounts must be seen as reliable and expenditure control as efficient and effective. Negotiations with donors toward this end should be initiated at some point, after improvements in the government’s reform programs have been more substantially realized.
49. Coordination between ministries and CAOs needs to be strengthened. The establishment of FMUs in 7 ministries was intended to create a strong support for financial management in the ministries and to allow CAOs to provide more effective support to the ministry PAOs. In the staff’s view, although this mechanism has improved reporting on foreign-financed projects to some extent, the FMUs do not appear to have achieved their primary goal of providing overall financial management analysis and advice to the PAOs. Several factors appear to have contributed to this failure: (i) the continuation of existing planning, budgeting, and control mechanisms, often dominated by line divisions—and the nonfunctioning of budget committees within ministries; (ii) the location of CAOs outside the ministries; and (iii) the relatively junior status (deputy secretary or senior assistant secretary) of CAOs. Establishing strong and cohesive financial management within the ministries will play a critical role in the next phase of reform through the FMRP, both in terms of fiscal reporting and the implementation of an MTEF (see paragraph 52 below). Additional measures need to be initiated as soon as possible to clarify the role of the FMUs and to seek structural change within ministries to facilitate their operation. Actions could include the following:
Prepare a discussion paper to clarify the broad role envisaged for FMUs and outline the steps that will be required within ministries to achieve these objectives;
A series of seminar/workshops with the PAOs and senior officials of ministries that are setting up FMUs to discuss the planned development and to examine issues across ministries; and
Upgrade selected CAO positions and promote appropriately qualified candidates to these positions.
Such measures will serve to initiate debate and raise awareness among PAOs. However, it will take a considerable and sustained effort to establish operational systems and a consciousness of the importance of financial management at ministry level. Ministries with established FMUs could be an appropriate focus for initial efforts.
50. Internal audit mechanisms should be strengthened. As noted in paragraph 28, improvements have been made to aggregate monitoring and control processes; there is also, however, a need to upgrade the quality of management within individual ministries. Building the internal audit function to meet accepted international standards is generally recognized as an important element of reform. The last PAC report recommended that an Internal Audit Cell should be created within each Ministry. However, creating effective units throughout government will take time and training. Strengthening the internal audit function would take considerable pressure from external audit, allowing the C&AG to focus on its primary function. In addition, internal audit would provide assurance of the strength of systems control, and should result in an improvement in the quality of financial information. Recognizing the long-term nature of effort required, the staff suggest that this work could be initiated by creating an Internal Audit Cell in the Ministry of Finance on a pilot basis. Such a unit would require a relatively small number of highly qualified professional staff with a strong background in internal audit, and would work as a ‘center of excellence’. It would develop internal audit standards, guidelines, manuals and training materials to develop the quality of the internal audit functions in the existing government ministries and departments. Based on experience in the MOF, similar units would be rolled out in other government ministries and divisions.
51. More analytical information on the budget and accounts should be disseminated to the public in an accessible form. The current budget presentation is insufficiently analytical and is not therefore sufficiently open to external scrutiny nor to policy accountability. However, much of the information needed is already available in the system and presentation could be improved (both internally and for external publication) over time at relatively little cost. The following measures should be considered:
Basic macroeconomic, technical, and policy assumptions underlying the budget estimates should be presented systematically in the budget documents, possibly initially as an annex to the Budget in Brief or the Budget Speech. The assumptions should be presented as far as possible in tabular form showing clearly the linkage between assumptions and their revenue or expenditure impact;
Contingent liabilities arising from government guarantees, or other contractual arrangements that could give rise to future liabilities of the government, should be disclosed in an annex to the budget documents (including contingent liabilities that may arise as a result of enterprise closure—see paragraph 4);
A statement of tax expenditures (revenues foregone as a result of concessions to specific activities) should be prepared as an annex to the budget documents; and
A statement of fiscal risks should be compiled to indicate the possible impact of changes in key assumptions or contingent event risks on the overall fiscal position.
Not all of these can be put in place in full in the short run. However, a number of elements could be introduced for the 2003/04 budget and progressively improved and integrated into the budget decision-masking processes in subsequent years. In the longer run, these reporting requirements and more general issues of government reporting and clarity of fiscal policies could be embodied in a revised legal framework for managing the public finances. The current legal framework is in need of streamlining and updating. A number of countries, including most recently India and Pakistan, have introduced - or are introducing - fiscal responsibility laws to codify such requirements in law.
Fiscal transparency and the establishment of an effective MTEF
52. Assurances of fiscal discipline and policy accountability should be given priority in implementing an MTEF. The authorities are committed to the introduction of outputoriented budgeting in a medium-term time horizon based on a sound macroeconomic framework. It is recognized, however, that establishing such a system takes a long time even in advanced economies. Experience in these economies, moreover, suggests very strongly that an effective MTEF depends crucially on an open and disciplined budget planning and implementation process. An MTEF involves much more than a set of rolling macroeconomic projections over a three-year period; it involves significant changes to annual budget procedures, and, in particular, an integrated approach to development and recurrent spending. The staff suggest, therefore, that work on the establishment of an MTEF proceed in a carefully phased manner and that key components of fiscal policy discipline and accountability be established as prerequisite for full implementation of the MTEF across all government ministries.33 The following measures should be introduced:
The macroeconomic framework being developed for the poverty reduction strategy paper (PRSP) be developed as the overall resource envelope for the MTEF;
Existing policy expenditures should be carefully costed and clearly separated from new policies in budget preparation and forecasting future commitments (forward estimates);
New policy submissions should encompass both proposals for the rolling public investment program and additions to the revenue budget above existing policy commitments, and these proposals should be reviewed in an integrated process;
A high-level committee should be established to coordinate annual review of all new policies (encompassing both ADP and additional revenue budget proposals)—and strict procedures must be established to control admission of new policies in the budget and to contain total spending within the overall resource envelope;
Ministries should be made responsible for their detailed rolling forward estimates of expenditures and receipts as a basis for annual budget submissions and negotiations—and FMUs or an appropriately modified unit within the ministry should be primarily responsible for maintaining these estimates; and
Once the preceding conditions have been satisfied, forward estimates of expenditures and receipts under existing policy (both revenue budget and ADP) should be published in the annual budget document to reveal the continuing costs of government policy and provide a basis for policy accountability.
Fiscal transparency and reform of the state-owned enterprises
53. Systematic measures to strengthen transparency and reforms of the state-owned enterprises sector are urgently required. As indicated in Box 1, rundown of state resources through the nexus of direct subsidies and quasi-fiscal activities of NFPEs and PFIs is the single greatest source of long-term loss and fiscal risk. Data collection in this area, however, is inadequate and is not focused on clearly representing government policies and their costs. It will clearly take time to introduce all the reforms needed in these sectors, but measures should be initiated now to establish a strong unit within the MOF that is capable of systematic analysis of this aspect of the overall government balance sheet. Measures that could be taken include the following:
Strengthen the role of the MOF monitoring cell by adding skilled staff to enable it to produce within-year reports on state-owned enterprise activity and more effective monitoring of income and losses of major enterprises in this sector;
A longer-term objective of such a unit should be to develop a stronger analytical focus in order to quantify the costs of NFPE noncommercial activity along the lines developed by the World Bank and highlighted in this report. Building skills to achieve this objective should be explicitly part of the unit’s terms of reference at the outset, however;
As reporting on NFPEs improves, a regular report on the consolidated public sector balance should be produced—along the lines suggested by the World Bank. In the staff’s view, such a balance should be used to inform fiscal policy management, but could not be treated as a key policy target variable without a much stronger database and detailed analysis of quasi-fiscal activity of the NFPE and PFI sectors;
Develop an annual analytical report on subsidies and quasi-fiscal activities as an annex to the budget documents. As with other suggested annexes (paragraph 51), such an analysis could be initiated in a preliminary form quite quickly. Over time, however, it is vital that its analytical quality be strengthened and that it be considered actively in budget decision-making; and
In light of further advice from the FSAP mission, action needs to be taken to strengthen analysis of the costs of quasi-fiscal activities by PFIs and to strengthen the banking supervision role of the BB.
Strengthening watchdog institutions and governance
54. Finally and, in many ways, most fundamentally, strong, independent watchdog institutions need to be fostered in Bangladesh. The reforms outlined above will not be sustainable, and the integrity of data is unlikely to be maintained, unless the overall public sector management environment becomes more open to scrutiny by organizations that are credibly independent of the executive branch of government.
55. The independence and capacity of the C&AG should be strengthened. The staff strongly support the proposals in the FMRP to strengthen capacity and quality of external audit by C&AG. The staff also consider it important that steps be taken to give the C&AG clearer administrative independence. In association with implementation of this component of the FMRP, the following suggestions should be considered:
The C&AG could be made an officer of the parliament; while still subject to review by the Finance Division (in the same way as the Parliament Secretariat), his budget and manpower could then be sanctioned by the Public Accounts Committee on behalf of the parliament;
To ensure independence of the C&AG on staff matters, as well as strengthening management by the CGA, the envisaged separation of the accounts and audit cadre should be implemented as soon as possible;
The C&AG accounts should also be subject to audit, possibly by an independent private sector auditor. In addition, a peer review process could be established; in line with International Organization of Supreme Audit Institutions (INTOSAI) practices, this could involve another Auditor General’s Office from a developed country in periodic review of audit practices; and
An Audit Act incorporating these and other elements should be drafted and enacted as soon as possible to establish a better enabling environment for independent and high quality audit.
56. Oversight of the public finances by parliament needs to be strengthened and focused. It is vital that steps be taken as soon as possible to reinstate the PAC mechanism as soon as possible. The FMRP is aimed at producing a series of outputs to improve C&AG reports and executive branch follow up. These measures should also be supported by strengthening of the capacity of the PAC to examine audit findings and to review follow up. The following measures could be considered:
The PAC to be provided with a permanent secretariat to follow-up on all audit recommendations;
Experts on financial management to be included in the public accounts committee;
PAC hearings to be made open to the public;
Audit Committees to be established in all government ministries including members from the civil society; and
The Standing Committees on Ministries to be given additional responsibility to review the status of the settlement of the audit recommendations based on the initial report by the C&AG Office.
57. Long-run improvements in transparency require a substantial, well-directed, and nonpartisan effort to fight corruption. Progress in improving fiscal transparency—and in many other areas—is impeded by the widely recognized prevalence of corruption in both public and private sectors. Failure to tackle this problem could jeopardize the gains that are being made in public sector management; conversely, joint efforts to tackle corruption and promote fiscal transparency and accountability should be strongly synergystic. It is beyond the scope of a fiscal ROSC to attempt any detailed recommendations in this area. However, a key recommendation that has been made in other reports34, and that enjoys widespread support, is that an Anti-Corruption Commission, including highly qualified members drawn from civil society, be set up directly under parliament or an ombudsman office. The staff support the implementation of such a proposal as soon as possible.
58. The role of local government should be clarified and its mechanisms of accountability strengthened. Decentralization and strengthening of local government are necessary elements of a strategy to provide citizens with greater participation in public decision-making, to encourage more effective delivery of poverty-reducing programs, and to deepen the structure of accountability. Government policy in this area is in the process of further clarification in terms of the precise relationship among the tiers of local government and the center, and mechanisms for ensuring transparency and accountability at the local levels need to be established. These tasks will take some time to address fully, but will need to be guided by firm political direction as early as possible. The consensus-building approach advocated by the UNDP35 could play an important role; however, establishing some form of local government commission with a clear mandate to strengthen local government institutions, and especially their financial accountability framework, should be considered as soon as possible. As a initial step, the following actions should be considered:
Financial management and internal audit of LGIs should be strengthened;
Extensive training should be provided to finance and audit staff of LGIs for capacity development;
The accounts of all city corporations and municipalities should be prepared on accrual basis and audited by private independent auditors; and
City and municipality annual reports and audited financial statements should be made available to the media and the public.
Discussions on fiscal transparency were held in Dhaka during August 11-21, 2002. The staff team comprised Mr. Allan (head) and Ms. Bhattacharya from the Fiscal Affairs Department, together with Mr. Hoque from the World Bank. Mr. Verhoeven, the IMF Resident Representative, participated in a number of meetings and discussions and provided substantive inputs and guidance in preparation of the report; Mr. Pritchett, a member of the IMF Statistics Department multitopic mission, also provided useful comment and inputs. The team met with officials from Bangladesh Bank; Ministry of Finance, including the Finance Division and the External Relations Division; Ministry of Health & Family; National Board of Revenue; Office of the Comptroller & Auditor General; Controller General of Accounts; the Implementation, Monitoring and Evaluation Division of the Ministry of Planning; Ministry of Food; Ministry of Establishment; Ministry of Local Government, Rural Development and Cooperatives; the Privatization Commission; the Metropolitan Chamber of Commerce and Industry; and Proshika (an NGO). The mission prepared an action plan to implement many of its initial recommendations and several were put in place after the mission’s visit. These and other changes subsequent to the mission’s visit were discussed with Mr. Verhoeven and are noted at appropriate points in the text.
There are around 200 NFPEs under the purview of 41 sector corporations, as well as the 3 quasi-commercial government departments included in the budget. Their operations cover all economic sectors and are dominant in key sectors such as railways, electricity, gas, and water. Nonetheless, the overall contribution of NFPEs to the economy is estimated at less than 3 percent of GDP for FY 2001/02. The largest number of public enterprises are in the manufacturing sector (constituting over 20 percent of manufacturing output), and include industries such as textiles, steel, engineering, food, chemicals, and jute.
The constitution (1971) gives emphasis to state ownership of the means of production as well as permitting cooperative and private ownership (Article 13)—all financial institutions and industries were nationalized following independence in 1971, though around 500 were sold or returned to the former owners during 1976-92. In the 1990s, action was taken to revitalize liberalization of the economy. A new industrial policy was introduced in 1991, with the objectives of setting up an improved regulatory framework for private investors, protecting consumers, and to establishing a commercial basis for state-owned enterprise operations. A Privatization Board was created in 1993 and was replaced by the Privatization Commission under the Privatization Act, 2000. Current policy, as required under the Act, is gazetted, and also published on the Commission’s website (http://www.bangladeshonline.com/pb/). The sectors to be privatized and areas where private investment is encouraged are also published on the website. The Commission is headed by a Chairman with State Minister rank, appointed by the government, and has two fulltime members also appointed by the government, six members of parliament, and one representative from any professional organization. Secretaries of the Ministries of Industries, Commerce, Finance, Textile, Jute, the Chairman of the Securities and Exchange Commission, and the President of the Federation of the Bangladesh Chamber of Commerce and Industry are ex officio members.
From 1972–1985 around 570 enterprises were privatized, initially smaller enterprises, but including some larger jute and textile mills in the latter part of the period. After these initial efforts, few privatizations occurred until the Privatization Board was established. Data supplied by the Commission indicates that the pace of privatization remained moderate over the period from 1993 to August 2002: 41 entities (including 3 banks) were partially or wholly divested, and total receipts over the period amounted to only Tk 4.5 billion (0.2 percent of average GDP). A new list of 53 enterprises to be privatized (on top of the 49 selected for privatization by the previous government), covering all economic sectors (previously confined only to manufacturing), has been prepared by the Commission for the present government.
Law enforcement institutions were rated by Transparency International in its Corruption Database 2001 (Executive Summary, July 9, 2002) as having the highest incidence of corruption in Bangladesh. See also Public Administration Reform Commission (PARC) Report, Chapter 6, June 2000.
Commercial banks in Bangladesh comprise four nationalized commercial banks (NCBs), 30 private domestic banks (PDBs), and 12 foreign banks. They are regulated by the Banking Companies Act 1991 and circulars issued by BB from time to time.
See IMF Selected Issues, 2002 (SM/02/114) for a discussion of systemic weaknesses in the banking sector in Bangladesh.
Detailed discussion of banking practices is beyond the scope of the fiscal transparency ROSC, but the fiscal management impact of these practices is very significant over the long-term (see Box 1). A number of aspects are discussed further in Bangladesh: Financial Accountability for Good Governance, a Country Financial Accountability Assessment(CFAA), The World Bank, Washington DC, 2002 (hereafter, “Bank CFAA”), chapter 8. It is noted that politicians and government officials regularly interfere in the operations of the NCBs and DFIs, while the management of the banks is not generally held accountable for performance. Insider lending is legally sanctioned in Bangladesh, and the soundness of most banks has been severely undermined by insider lending, fraud and negligence. Limits are set to such lending, however, and these are under review. The Banking Companies Act 1991 allows banks to make secured loans to their directors and their affiliates, and insider lending has become a common practice in most banks. Some efforts have been made by BB to reduce the scope for such activity, to enhance asset quality, to improve loan classification, and to introduce better monitoring and enforcement of the net open foreign positions. These matters should be pursued further in light of the recommendations of the Financial Sector Assessment Program mission (October 2002). In this context it is encouraging that a Money Laundering Prevention Act has been passed to stem illegal financial transactions.
The Bank CFAA describes NFPE reporting, budgeting and audit in more detail. This report provides considerable information on budgeting, accounting, and auditing structures and practices in Bangladesh.
Government of the People’s Republic of Bangladesh, Summary of the Local Government Commission Report (1997). The UNDP has also strongly recommended strengthening local government (UNDP, Role of UNDP in Promoting Local Governance and Decentralization in Bangladesh, July 12, 2002).
Taming Leviathan: Reforming Governance in Bangladesh, Chapter 2 (Washington DC: The World Bank, March 2002). Reforms to strengthen the role of parliament and the supreme court are described in chapter 3.
In principle, these records should also include payments made by the donor agencies directly to contractors which do not pass through the project accounts.
Subsequent to the mission’s visit, the authorities advise that a Cabinet decision has been taken to establish an Independent Commission for Anti-Corruption.
The main budget documents are: the Annual Budget: Budget Speech; the Budget in Brief, which provides a summary overview; Consolidated Fund Receipts, which provides estimates of revenue and capital receipts (including foreign grants and loans); Demands for Grants and Appropriations (Non-Development), which provides details by vote for each grant as well as all expenditures charged on the consolidated fund; the Development Budget, which covers appropriations for the Annual Development Plan, prepared by the Planning Commission; and the Annual Financial Statement, which provides an overview of government transactions on the consolidated fund and public account. An Economic Survey, which gives background information, is also distributed just prior to the Budget (available in Bengali, but not in English).
Consisting of RIBEC 2000 (aimed at developing the core budget and accounting systems); FIMA, (including establishment of the Financial Management Academy); and RIGA (Reforms In Government Accounting)—all supported by the U.K. Department for International Development (DFID). See paragraph 31 below.
Block accounts are general provisions in the budget that are subsequently allocated to specific line items. They cover such things as contingencies for emergency-related spending such as flood relief; amounts for election costs and for military spending; and the estimated costs of maintaining and operating completed ADP projects until these have been transferred from the Development Budget to the Revenue Budget.
Considerable detail on budget tax measures was provided in the FY 2002-03, but their impact on the revenue estimates was not spelt out in detail.
The last PAC report recommended the creation of Internal Audit Cell within each Ministry to reduce the burden of the external auditor to verify each and every documents and strengthen internal control arrangement. The C&AG has also stated that he requires external audit to spend more of its time auditing the financial statements of Ministries. The introduction of internal audit would take a considerable pressure from external audit, allowing it to focus on this primary objective. In addition, internal audit would provide assurance on the strength of systems control, resulting in an improvement in the quality of financial information. Such services, however, need to be seen as professionally skilled and capable of providing quality advice on financial control and effectiveness to the PAOs of the ministries and divisions.
These issues are discussed at more length in the World Bank CFAA (Chapter 3) and the Country Procurement Assessment Report (World Bank, 1999).
The absence of a centralized mechanism for authorizing and monitoring projects financed through foreign suppliers’ credit also poses significant macroeconomic risks that are not properly accounted for in the budgetary process.
The PARC report (chapter 4) says that a comprehensive and reliable database on the total number of staff in public organizations is not available. The Ministry of Establishment indicates that currently there are a total of 936,000 sanctioned posts (90,000 class I officers, 48,000 class II officers, 598,000 class III and about 200,000 class IV staff); in addition around 300, 000 staff are employed for fixed periods on development projects. PARC indicates that in 1996, the Ministry of Establishment estimates were 730,117 total sanctioned posts.
See the PARC report (chapter 3), which describes the closed entry system, whereby the key decision-making class I officers are recruited directly at entry level through open competitive examination into 29 cadres and other services. There are limited opportunities for promotion to class I positions from lower ranks.
The food account is an extrabudgetary account which records all transactions of the Ministry of Food in receiving (in cash or in kind) and distributing food aid. The food account is published with the budget and the food balance is incorporated in the budget accounts. The food account balance (the difference between the value of food distribution and gross expenditures incurred in obtaining, storing and distributing food) can be defined as government subsidies on sales of food plus expenditures on change in the stock of food.
In-kind budget transactions (for example, the Food For Work program) that go through the food account and ADP and non-ADP capital spending present major difficulties for effective tracking.
Revised estimates incorporate supplementary budget estimates, reviewed by the MOF around February/March, and generally approved by parliament by mid-May. It also includes reappropriations from block accounts for approved uses or emergency-related spending and those carried out by ministries within the original appropriation limits. Use of reappropriated block funds is shown against the relevant individual budget heads, but no overall summary is given.
Consisting of an overall financial statement, covering the consolidated fund and the public account (the Finance Accounts), and a detailed statement of actual expenditure versus authorized expenditure (the Appropriation Account).
Unaudited data for actual budgetary outturns are available for 2000/01. The data show a shortfall in total revenue of Tk 13 billion, or 5.4 percent (Tk 228 billion in actual collections, compared with Tk 242 billion in the original budget), entirely due to shortfalls in nontax revenues. Total budgetary expenditures were also lower by Tk 13 billion (Tk 371 billion estimated outturn versus Tk 384 in the original budget) or 3.5 percent, with higher current expenditures (+ Tk 7 billion) offset by lower spending on the ADP (- Tk 10 billion) and on non-ADP capital expenditure and net lending (- Tk 8 billion). With regard to current expenditures the data show higher spending on interest payments (+ Tk 4 billion) and on subsidies and transfers (+ Tk 10 billion) and lower spending on block allocations (- Tk 8 billion) relative to the original budget.
For instance, significant differences arise in the data on revenue receipts prepared by NBR (from Commissioners reports), CGA (from CAO accounting inputs), and receipts recorded by the BB/Sonali Bank. Discrepancies arise because of timing differences and by different coverage (for instance autonomous bodies not covered by the CGA). As a standard practice, the data from each source is compared, an explanation of differences is given, and remedial action is undertaken as appropriate to minimize avoidable differences.
INTOSAI, Independence of SAIs Project Final Task Force Report (2001) shows that, in over 60 percent of 113 countries surveyed, the head of the external audit office was primarily directly accountable to the legislature.
As indicated in footnote 1, all recommendations in this section of the report were substantially agreed with the authorities and an action plan developed during the mission’s visit. Subsequent to the mission, many of the more immediate recommendations have been implemented or are in progress.
The difficulties and the long-term nature of these reforms are well recognized, and it is encouraging to note that, particularly in the case of the recently initiated Financial Management Reform Project (FMRP), the donors and the authorities recognize the long-haul nature of the commitment.
Such an emphasis would be consistent with the recommended emphasis of C&AG audits on questions of materiality rather than simply compliance (see paragraph 39).
These matters could also be referred for consideration to the recently formed Public Expenditure Review Commission (terms of reference issued on June 13, 2002), as well as to the Revenue Reform Commission (terms of reference issued on July 2, 2002).
See World Bank CFAA (Chapter 3), Taming Leviathan, Chapter 6; PARC Report, Chapter 6.
UNDP op cit, July, 2002