Document of the International Development Association (acting as Administrator of the Interim Trust Fund) FOR OFFICIAL USE ONLY Report No. P-7097-UG REPORT AND RECOMMENDATION OF THE MANAGING DIRECTOR TO THE INTERNATIONAL DEVELOPMENT ASSOCIATION ON A PROPOSED INTERIM FUND CREDIT IN AN AMOUNT OF SDR 91.1 MILLION (US$ 125 MILLION EQUIVALENT) TO THE REPUBLIC OF UGANDA FOR A THIRD STRUCTURAL ADJUSTMENT CREDIT PROJECT May 1, 1997 Country Department for Uganda Macroeconomics II, AFTM2 Africa Region This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. GOVERNMENT FISCAL YEAR July 1 - June 30 (FY96 and 1995/96 = July 1, 1995 to June 30, 1996) CURRENCY EQUIVALENTS Currency Unit = Ugandan shilling (U Sh) Interbank Market mid-rate: US$1.00 = U Sh 1,025 (March 1997) ACRONYMS AND ABBREVIATIONS ASAC Agricultural Structural Adjustment Credit BFP Budget Framework Paper BOU Bank of Uganda COMESA Common Market for Eastern and Southern Africa CTL commercial transaction levy DIS Deposit Insurance Scheme ERC Economic Recovery Credit ESAF Enhanced Structural Adjustment Facility FSAC Financial Sector Adjustment Credit GDP gross domestic product GPT graduated personal tax KCC Kampala City Council LGTB Local Government Tender Board MOF Ministry of Finance MTEF medium-term economic framework NGO nongovernmental organization NTB non-tariff barrier OED Operations Evaluation Department PAC Public Accounts Committee PAYE pay as you earn PMU Parastatal Monitoring Unit PPA Priority Program Area PTA Preferential Trade Area SA Strategic Area SAC Structural Adjustment Credit SOE state-owned enterprise SSA Sub-Saharan Africa UCB Uganda Commercial Bank URA Uganda Revenue Authority Vice President Callisto E. Madavo Country Director James W. Adams Technical Manager : Roger Grawe Task Team Leader David E. Yuravlivker FOR OFFICIAL USE ONLY CONTENTS CREDIT SUMMARY ............i....................................... I. INTRODUCTION ..............1................................. I. RECENT POLITICAL AND ECONOMIC DEVELOPMENTS .............. 2 III. GOVERNMENT'S REFORM PROGRAM ............5............5 IV. MEDIUM TERM PROSPECTS AND FINANCING REQUIREMENTS .............. 6 V. BANK GROUP ASSISTANCE STRATEGY ............................ 7 VI. IMPACT AND LESSONS FROM PAST ADJUSTMENT LENDING ............... 9 VII. THE PROPOSED SAC III REFORM PROGRAM. ..................... 11 A. Objectives of the Program ..........................1 B . Toward Fiscal Sustainability................. ....... 12 C. Main Components of the Program ........................ 13 1. Revenue Mobilization and the Incentive Regime................ 13 2. Improving the Management and Efficiency of Public Expenditure .................................. 21 3. Reducing Quasi-Fiscal Deficits: Parastatals . ............ 27 4. The Fiscal Cost of Financial Sector Reform..... .......... 28 VIII. THE PROPOSED CREDIT ...................................... 31 1. CONDITIONALITIES: BORROWER-IDA CONTRACT ........... 31 2. SUPERVISION PLAN ..................................... 33 (a) Borrower's Leading Role........................... 33 (b) IDA Supervision ........................ 33 (c) Performance Indicators and Expected Results .......... 36 (d) Other Donors and Stakeholders Participation .......... 36 3. CREDIT ADMINISTRATION ................................ 36 IX. BENEFITS AND RISKS......................................... 37 X. RECOMMENDATION. ..................................... ..... 38 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  11 ANNEXES ANNEX 1: HIGHER IMPACT ADJUSTMENT LENDING DEFICITS.......................39 ANNEX 2: LIST OF TECHNICAL NOTES ......................................40 ANNEX 3: LETTER OF DEVELOPMENT POLICY .......................... .....41 ANNEX 4: POLICY MATRIX ...............................................59 ANNEX 5: STATISTICAL APPENDIX UGANDA AT A GLANCE .........................................63 TABLE 1: KEY ECONOMIC INDICATORS .......................65 TABLE 2: BALANCE OF PAYMENTS....................... ........66 TABLE 3: EXTERNAL FINANCING REQUIREMENTS ...................67 UGANDA RESOURCES AND EXPENDITURES PRIORITY POVERTY INDICATORS ...............................68 STATUS OF BANK GROUP OPERATIONS IN UGANDA . ..................70 STATEMENT OF IFC's COMMITTED AND DISBURSED PORTFOLIO..............71 ANNEX 6: TECHNICAL NOTES.............................................73 FIGURES FIGURE 1: FISCAL AND EXTERNAL DEFICITS ..................................12 FIGURE 2: TAX STRUCTURE 1985/86-95/96 .........................14 TABLES TABLE 1: REFORMS UNDERTAKEN UNDER PREVIOUS ADJUSTMENT OPERATIONS ......10 TABLE 2: EFFECTIVE PROTECTION IN SELECTED INDUSTRIES .................... 18 TABLE 3: ACTUAL AND PROJECTED SUBSIDIES TO SOES IN UGANDA, 1992-1998 .......27 TABLE 4: PERFORMANCE INDICATORS......................................35 This operation was prepared by a team led by Mr. David E. Yuravlivker (AFTM2). The team also included Ms. Ritva Reinikka (AFTM2, Trade Policy); Mr. John Short (Consultant, Trade Policy); Mr. Federico Changanaqui (AFTM2, Fiscal Sustainability); Mrs. Mimi Klutstein-Meyer (AFTM2, Statistics and Projections); Ms. Roboid Covington (AFTM2, Report Production); Mr. Roy Bahl (Consultant, Local Taxation); Mr. Iradj Alikhani (AFMUG, Public Expenditure); Mr. Sanjay Pradhan (WDR/PRDPE, Budgetary Reform); Mr. Andrew Bird (Consultant, Budgetary Reform); Ms. Cecilia Hermansson (Consultant, Budgetary Reform); Mr. Carlos Montes (Consultant, Budgetary Reform); Mr. Fitz Ford (TWURD, Decentralization); Mr. Stefano Migliorisi (AF2PE, Parastatals); Mr. Dan Mozes (FSD, Financial Sector). The Lead Advisor was Mr. Peter Miovic (EDINP). The Technical Manager is Mr. Roger Grawe (AFTM2), and the Country Director is Mr. James W. Adams (AFCO4).  THE REPUBLIC OF UGANDA THIRD STRUCTURAL ADJUSTMENT CREDIT Credit Summary Beneficiary: The Republic of Uganda Project Task ID: UG-PA-2987 Implementing Agency: Ministry of Finance IDA Amount: SDR 91.1million (US$125 million equivalent) Terms: Standard IDA terms: 40 years maturity with a 10-year grace period. Co-financing: Sweden (SEK 50 million) Disbursement: The proposed Credit will be disbursed through the Bank of Uganda. The initial tranche of US$45 million equivalent, including the allocation from the IDA reflows, will be available upon Credit effectiveness. Background: The Structural Adjustment Credit is based upon the Government's Policy Framework paper, distributed to the Board in November 1996, and related Letter of Development Policy which set out the Government's medium term strategy. This strategy has as its objectives the reduction in poverty through macroeconomic stabilization and the acceleration of economic growth. The principal objectives of the SAC are to foster longer-term fiscal sustainability, involving a more efficient tax system as well as better management of public expenditures. The program includes measures to reduce anti-export bias and it addresses the fiscal impact of 2 financial sector reforms within the overall macroeconomic framework. The medium term benefits of such measures would be higher exports and growth and improved delivery of public services, which would contribute to poverty reduction. This strategy is in line with IDA's Country Assistance Strategy to be submitted to the Board with this operation. Description: On the revenue side, the program includes measures to expand the tax base, improve tax administration and lower the level and dispersion of trade taxes. On the expenditure side, the focus is on improving the management of public expenditure consistent with the ongoing decentralization process. In addition, the program includes measures to reduce the fiscal burden of state-owned enterprises and to continue financial sector reforms within a sustainable macroeconomic framework. These actions, once they are fully implemented, would help maintain a growth rate of more than 7 percent per annum in the next three years and improve delivery of social services leading to a substantial decline in poverty. Rationale for IDA The SAC addresses policy areas that build on Involvement: operational and analytical work done in the context of several recent operations such as the Financial Sector Adjustment Credit and the Private Sector Competitiveness Project. The SA_ will capitalize upon and extend the agreements reached through the PFP/ESAF negotiations. The SAC emphasizes policies and processes. Capacity building and more detailed institutional change is being done as part of the ongoing Institutional Capacity Building Project and will be supported under a Technical Assistance Project currently under preparation. Government Commitment: Uganda has had a satisfactory track record of stabilization and adjustment over the last 10 years. The Government's resolve has been underscored by two previous SACs and by six 3 annual ESAF arrangements in 1989-96, with a further annual arrangement approved on November 18, 1996. Benefits: The proposed program will support the recent gains in achieving macroeconomic stability and will accelerate structural and institutional reforms in Uganda. The measures to improve tax collection and to reduce dependence of the public sector on external aid should result in a more sustainable fiscal position over the next few years. Efficiency gains are expected both in the private and public sectors, with the increased participation of the private sector in key areas of the economy and the expected improvement in the management and operation of the public sector. Risks: The main risks associated with this operation are related to the institutional capacity to carry out the necessary reforms. To minimize these risks, the Bank is currently implementing the Institutional Capacity Building Project and preparing a technical assistance credit to support economic and financial management.  REPORT AND RECOMMENDATION OF THE MANAGING DIRECTOR TO THE PRESIDENT OF THE INTERNATIONAL DEVELOPMENT ASSOCIATION (THE ASSOCIATION ACTING AS ADMINISTRATOR OF THE INTERIM TRUST FUND) TO THE REPUBLIC OF UGANDA FOR A THIRD STRUCTURAL ADJUSTMENT CREDIT I. INTRODUCTION 1. I submit for your approval the following report and recommendation on a proposed Interim Fund Credit to the Republic of Uganda for SDR 91.1 million, the equivalent of US$ 125 million, on standard IDA terms, with a maturity of 40 years to help finance a Third Structural Adjustment Credit (SAC). The proposed Interim Fund Credit of US$125 million would support Government efforts aimed at enhancing overall fiscal sustainability and the management of public expenditures, and encouraging supply response by reducing the anti-export bias in the economy. These elements support and are consistent with the Bank's assistance strategy and its main goal, poverty reduction, as spelled out in the Country Assistance Strategy (CAS) that is being presented to the Board jointly with this Credit, and with the Policy Framework Paper (PFP) distributed to the Board on November 8, 1996 (SecM96-1105). Many elements of the proposed program follow the recommendations of the most recent Country Economic Memorandum (report no. 14313-UG, June 1995), which also updates Uganda's poverty profile. SAC III will be complemented by a number of on-going investment projects. Proceeds from the proposed credit will help fill Uganda's projected external financing gap through 1997/98 and will provide budgetary support during a period of fiscal reforms supported by the program. 2. The timing and agenda of SAC III were fully determined by the Ugandan authorities. Moreover, the Public Enterprises component is based on a report of the Parastatal Monitoring Unit (PMU) of the Ministry of Finance, and the tax administration component was drafted by the Ministry of Finance. During preparation, the Ugandan Private Sector Foundation organized workshops to discuss the program, and SAC team members met with various private sector associations to continue these discussions. In addition, the Resident Mission organized meetings to brief donors as well as NGOs. The program was also periodically discussed with the IMF, which has had four successive annual arrangements and, since 1994, a three-year ESAF program with Uganda. 2 II. RECENT POLITICAL AND ECONOMIC DEVELOPMENTS 3. Political context. Uganda was politically stable and one of the most promising economies in Sub-Saharan Africa in the 1960s, but during the regimes of Amin and Obote in 1971-85 the civilian population suffered political turmoil, civil war and economic mismanagement had severe consequences. It is estimated that half a million people lost their lives during this period. Real GDP per capita declined by over 40 percent, and the formal economy became heavily regulated, while the share of the informal and non-monetary sectors increased substantially. 4. The government of the National Resistance Movement (NRM), which took power in January 1986, has succeeded in including a wide range of political views in the cabinet. The NRM has also established a system of elected Local (former Resistance) Councils to promote grass-root democracy. Following extensive debate by an elected Constituent Assembly (CA), a new Constitution was promulgated in October 1995. It maintains the ban on party political activity as the CA endorsed the 'movement political system' (individual merit as a basis of election) for five years, after which a referendum on the future system is expected to be held. The first direct presidential elections were held in May 1996, and President Museveni was elected with about 75 percent of the votes. Parliamentary elections were held in June 1996, and international observers certified that both elections were conducted both fairly and peacefully. Local elections are scheduled for second half of 1997. 5. Decentralization. As part of the long-term solution to political conflict and in support of efforts to increase democracy at the local level, the Government has embarked on a far-reaching decentralization program, under which much of the public service delivery is being devolved to local governments (districts and municipalities). Fiscal decentralization was initiated in a phased manner in the period 1993-96, while personnel decentralization was undertaken simultaneously for all districts. Following decentralization of expenditure assignment, an inter-governmental grant system, including block, conditional and equalization grants, is being designed (para. 31). In addition, there is a need to address the issue of tax assignment between the central and local governments. Successful implementation of the decentralization program will remain a major challenge because of scarce resources and limited manpower capacity particularly in the local governments. 6. Macroeconomic performance. As the interventionist stance initially adopted by the NRM proved unsustainable and inflation soared to triple digits, the course of economic policy was changed at first in 1987 but more fundamentally in early 1992. As a result, the Government has achieved a substantial transformation of economic policy by: (i) restoring fiscal discipline and monetary stability and improving public revenue mobilization; (ii) floating 3 the exchange rate and liberalizing the trade regime; and (iii) making progress in rehabilitation of the country's economic, social and institutional infrastructure. Notwithstanding rapid growth of close to 7 percent per annum, on average, since 1987, with a per capita income of US$240, Uganda still remains in the lower tier of low income developing countries. 7. Uganda has maintained macroeconomic stability since 1992, and its performance under the ongoing Fund's ESAF program is on track. In fiscal year 1995/96 real GDP growth (at factor cost) was 8.5 percent, and was broad based, mainly higher output of cash crops, industrial production, and construction. This improved growth performance was aided by favorable developments in world coffee markets, but reflected mainly the improved macroeconomic environment, the implementation of key structural reforms, political stability, and growing private sector confidence. Notwithstanding periodic fluctuations due to weather-related factors on food output, annual average inflation declined from an average of 35 percent in 1991/92-1992/93 to around 7 percent during the last three years. The real GDP and inflation targets for the current fiscal year 1996/97 are 7 and 5 percent respectively. 8. Fiscal and monetary policy. Fiscal discipline is the cornerstone of the stabilization program in Uganda. Given the limited number of instruments available to conduct an independent monetary policy, fiscal policy carries most of the burden of containing inflation. This reliance on fiscal measures is expected to decline as the development of financial markets increases the efficiency of monetary policy. The mechanism for fiscal control is monthly releases of funds to line ministries and local governments, depending on availability of tax revenue and aid inflows, taking into account prevailing pressures on the price level and other macroeconomic indicators at the time. As a result, government expenditure has fallen from 21.2 percent of GDP in 1991/92 to 16.3 percent of GDP in 1995/96, and Government has become a net creditor with respect to the domestic banking system. While government revenues-to-GDP ratio remains low, revenues rose by nearly one percentage point of GDP at market prices a year, from 6.8 percent of GDP in 1991/92 to 10.2 percent in 1995/96. The overall deficit on a commitment basis (excluding grants) fell sharply over this period from 14.4 percent of GDP at market prices in 1991/92 to 6.0 percent in 1995/96 and a further reduction to 5.2 percent of GDP is targeted in 1996/97. The improvement in the fiscal position helped reduce the growth of broad money from 54 percent in 1991/92 to 21 percent in 1995/96. Prudent macroeconomic policies also helped create confidence and increase monetization of the economy, which measured by the ratio of broad money (M2) to GDP has increased from 7.7 to 10 percent resulting from an increased demand for money. While financial sector reform is well underway, the financial sector remains inefficient, with a high percentage of nonperforming loans (leading to high lending rates) and a large number of problem banks. 4 9. On the external side, much progress was made in reducing external imbalances. This reflects increasing liberalization of the foreign exchange and trade systems as well as the rise in international coffee prices and increases in export volumes, a doubling of non-coffee exports, and a strong growth in private transfers, foreign investment and stronger aid inflows. As a consequence, the current account deficit (excluding official transfers) declined from 11.8 percent of GDP in 1991/92 to 8.6 percent in 1995/96. The overall balance-of-payments also improved sharply during this period, from a deficit of US$ 8 million to a surplus of about US$44 million, and foreign reserves rose from the equivalent of 1.5 months of imports of goods and nonfactor services to an estimated 3.6 months. In 1996/97, the current account deficit (excluding official transfers) is projected at about 8 percent of GDP, as the continued strength of non-coffee exports and private transfers and expected moderation of import growth offset the impact of lower coffee prices compared to the previous year. 10. Uganda's external debt has nearly tripled over the last decade to about US$3.4 billion by end-June 1996. The debt sustainability analysis (DSA) prepared in close collaboration with the GOU and IMF staff, shows that Uganda's debt indicators are expected to remain above sustainable levels in the medium term, even after taking into account the Paris Club Stock of debt operation on Naples terms and assuming comparable treatment by non-Paris Club bilateral and commercial creditors. In particular, during the next three years, Uganda's NPV debt to exports ratio are projected to be 254 percent in 1996/97, 247 percent in 1997/98, and 244 percent in 1998/99. These ratios are above or on the high side of the 200-250 percent range that is being defined to measure debt sustainability under the HIPC Initiative, which tries to resolve the problems of highly indebted poor countries (HIPCs). In addition, Uganda's debt service ratio is projected to remain above 20 percent during the next three years. Moreover, with coffee exports representing over two thirds of total exports and the scheduled external debt service representing over 29 percent of government recurrent revenues, Uganda's external situation continues to be vulnerable to external shocks. Hence its debt indicators could worsen if key assumptions of the baseline scenario do not hold, i.e. higher than projected imports, lower coffee prices, less concessional financing, and/or smaller private sector inflows. 11. Given its long track record of good performance and its vulnerability to external shocks, Uganda was the first country declared eligible under the HIPC initiative. On April 22, 1997, the Executive Board approved: (i) Uganda's eligibility under the Initiative; (ii) a decision point of April 1997; (iii) a completion point of April 1998, subject to meeting the recommended macroeconomic, structural and social performance criteria; and (iv) a target NPV debt-to-exports ratio of 202 percent, and a debt-service ratio of 20 percent, to be achieved by the completion point. This would translate into a total debt relief of US$ 338 million in net present value terms, of which US$160 million would be provided by IDA. 5 III. GOVERNMENT'S REFORM PROGRAM 12. Reducing poverty. Despite robust economic growth, rural poverty remains severe and wide-spread. The improvements in economic indicators have not yet spread across Uganda, nor have they been matched by equivalent improvement in social indicators. Poverty eradication is therefore the highest national priority. Given that 85 percent of Ugandans are self-employed, the Government continues to focus on improving incentives for farm and non-farm activities. The most important measures have been liberalization of domestic and export markets, and the exchange rate. In addition, Government has taken measures to improve the environment for private sector investment and is implementing reforms to improve the soundness of the financial system. Also, further trade liberalization will improve the terms of trade for agriculture and ensure that the assets of the poor are not heavily taxed through protection. 13. The Government is currently refocusing its policies and actions to ensure that the poor are able to participate in and benefit from economic growth more effectively. Its strategy is articulated in the Action Plan for Poverty Eradication, which is included in the Background to the Budget 1996-97. The Action Plan states that the best way to remove people from poverty is not to give them handouts, but to enable households to earn decent incomes. In order to raise incomes, people need the following: (i) secure access to productive assets, particularly land; (ii) macroeconomic stability, including fiscal stability, low inflation and predictable policy environment; (iii) infrastructure, particularly roads which give access to markets; (iv) information about how to use resources efficiently; and (v) adequate levels of health and education to enable the poor to work and use the information at their disposal effectively. 14. Promoting private investment. Since the beginning of the Economic Recovery Program in 1987, Government has been implementing far-reaching policy and institutional reforms designed to deregulate the economy, eliminate direct state involvement in all but selective public services, and improve efficiency in the public sector. The objective of these policies is to make the private sector the engine of growth and to reduce poverty by creating opportunities for the poor to participate in economic growth. Promotion of private investment is an important element of government policy. An Investment Code with a generous incentive package was promulgated in 1991, and it is being revised taking into account the improved business environment and the need to widen the tax base. The process of returning properties expropriated by the Amin regime to their former owners has been completed, thus reinforcing investors' confidence that property rights will be respected. Consequently, private investment increased from 7.8 percent of GDP (in constant prices) in 1987/88 to 13 percent of GDP in 1995/96. 6 15. Improving public sector management. Promotion of private investment has been accompanied by rehabilitation of infrastructure and by the rationalization of the public sector through army demobilization, civil service reform, and public enterprise (PE) reform. The first two phases of the planned army demobilization has been completed, employment in the civil service has been reduced by more than one-half, and substantial progress has been made in improving wages and incentives. Government is also implementing a privatization program with a target of privatizing 85 percent of all parastatals by the end of 1997. In the financial sector reform, however, progress has been slower than expected; Government is currently working towards privatizing the Uganda Commercial Bank; addressing other problem banks; recovering non- performing assets; and strengthening the Bank of Uganda. 16. Attaining the medium-term macroeconomic targets (para. 18) and ensuring a stronger supply response are essential elements of the Government's overall poverty reduction strategy (paras. 12-13 ). These will require addressing remaining institutional bottlenecks and improving resource allocation, including: (i) achieving a higher rate of domestic resource mobilization, both public and private; (ii) improving the management of public expenditures at both central and local levels; (iii) reducing anti-export bias in the economy; (iv) completing the privatization and PE reform program; and (v) implementing comprehensive financial sector reforms. The SAC program includes key policy measures in all these areas. IV. MEDIUM TERM PROSPECTS AND FINANCING REQUIREMENTS 17. The medium-term policy framework for 1996/97-1998/99 will seek to strengthen poverty reduction, particularly through continued strong economic growth, while maintaining macroeconomic stability in a liberalized environment. The Government's strategy to reduce poverty will seek to combine policies that encourage environmentally sustainable growth with public expenditures that will increasingly focus on the financing of programmes in priority areas such a rural roads and water supply, primary education, basic health services, and agricultural research and extension. This will require a policy framework that encourages export-oriented activity and improved agricultural productivity through the introduction of high value-added crops, and increased provision of financial and technical assistance to farmers. In agriculture, the greatest potential is in cash crops for export. Apart from coffee, cotton, and tea, non-traditional exports such as fish, horticultural products, and fruits and vegetables have a strong potential. There is also substantial scope for increasing yields and efficiency in food crop production, as well as for expanding agro-industry. 18. The objectives of the Government's medium-term policy framework are: (i) to sustain annual real GDP growth rates of 7 percent, which are regionally 7 balanced and broadly based, thereby allowing per capita GDP to rise by at least 4 percent a year; (ii) to contain the year-on-year inflation rate to 5 percent; and (iii) to diversify exports and reduce the dependence of the balance of payments in donor flows. 19. The maintenance of this growth path is feasible since: (i) essential infrastructure will be coming on stream as a result of past and current public investment programmes; (ii) financial sector reforms currently under way will increase the mobilization of financial savings; and (iii) the public sector reform and privatization programmes will improve the quality and utilization of public assets. Moreover, these favorable factors are taking place against a background of continued political stability. However, economic growth could be constrained by infrastructure and institutional bottlenecks, such as power shortages, the congestion of the legal system, and ambiguities over land ownership. Another factor is the security situation which has necessitated higher defense spending and has prevented certain regions from realizing their economic potential. 20. Following a marked rise in national savings in 1995/96, a further increase to nearly 14 percent of GDP is expected (in the aftermath of the coffee boom) over the next three years. This will be primarily due to increased public sector savings, which are projected to rise from 1 percent of GDP (at factor cost) in 1995/96 to 4 percent in 1998/99 reflecting the continuing efforts to mobilize domestic resources. The current weak state of the financial sector could constrain growth in private financial savings until the benefits of financial sector reforms are realized. On the other hand, the maintenance of macroeconomic stability, and a more conducive investment environment are expected to increase private investment from 12 percent to 16 percent of GDP (at factor cost). This will support higher growth rates as well as continued development of the private sector. 21. Consistent with these objectives and reflecting the budgetary requirements for quick-disbursing aid, the total external financing requirements (excluding the clearance of the total stock of arrears of US$ 246 million owed to non-OECD creditors) for 1996/97 are estimated at US$2,060 million. Exports, private transfers, project aid and direct foreign investment are expected to provide US$1,806 million. This leaves a financing gap of US$254 million. IDA and IMF adjustment operations would disburse about US$113 million, and donor cofinancing the remainder. (Table 3 in Annex 5 contains details of the external financing requirements for fiscal years 1996/97 and 1998/99). V. BANK GROUP ASSISTANCE STRATEGY 22. The primary objective of the Bank's assistance strategy for Uganda is to reduce poverty. The key elements of this strategy are: 8 * to achieve higher sustainable levels of economic growth, which is broad- based; * to strengthen physical and economic infrastructure to support growth; and * to strengthen social infrastructure and human resource development, both as a means of increasing income earning prospects for the poor and to improve service delivery to the population, especially to the poor. 23. A necessary condition for this strategy to work is continued maintenance of macroeconomic stability. Also, IDA's assistance strategy supports the Government's decentralization process and its efforts to strengthen local capacity, and includes direct actions to protect the environment. IDA lending to Uganda has been in the range of US$150million per annum, and on-going projects include most strategic sectors, such as agriculture, roads, water supply, power, education, health, prevention of AIDS, institutional capacity building, environmental protection and management of natural resources. In addition, IDA is financing a regionally targeted multi-sectoral project in the North. The new CAS proposes to increase lending levels to about US$250 million per year in FY98-2000. 24. IFC and MIGA. The objective of IFC's activities in Uganda is to support the private sector development through project financing, capital market development and advisory services. IFC has investments in agro-business (sugar and tea), manufacturing (glassworks) and in financial services and has been working with the Bank on the restructuring and privatization process of PEs. Through the Africa Enterprise Fund, IFC has supported a number of smaller projects (fishing, flower production, sack production, private schools, hotel apartments and office blocks). Advisory services to date include cotton ginneries, telecommunications and the stock market. IFC's portfolio in Uganda is US$2.6 million in equity, and US$25 million in loans; in addition, pending commitments for 1996/97 amount to US$3.6 in equity and US$18 million in loans. MIGA guarantees totaling over US$20 million in coverage have facilitated foreign private investment in Uganda of over US$50 million for cobalt extraction, fish processing, coffee processing and two telecommunications projects. In addition, MIGA has approximately 17 applications in Uganda for guarantees of prospective investments in manufacturing, mining, tourism, agribusiness, and infrastructure. Under its Investment Marketing Services, MIGA has worked closely with the Uganda Investment Authority and the Government to assist them in their investment promotion activities. This has been done through training sessions, application of communication systems such as IPA net and CD Rom technology as well as by MIGA sponsored sectoral conferences in which the Government of Uganda has been an active participant. 9 25. Relations with the IMF and other donors. The Bank and the IMF have collaborated closely over a number of years in the development and monitoring of the adjustment program in Uganda. More recently, discussions with Government have focused on the structure of taxation, tax administration, restructuring of the financial system, and debt sustainability, and the IMF and the Bank have worked closely with Government on the identification of macroeconomic and sectoral reforms as outlined in the 1996/97-1998/99 Policy Framework Paper. A number of other donors have collaborated with the Bank on the Public Expenditure Reviews, and several bilateral and multilateral donors have participated in various sector programs, e.g., road maintenance initiative, AIDS prevention and other health sector programs, water and sanitation. A number of bilateral donors have either cofinanced or parallel financed the SAC II. The Bank organizes Consultative Group meetings annually, the last of which was held in October 1996 in Paris, and chairs monthly donor meetings in Kampala. During the past year the Bank's collaboration with NGOs has increased considerably. VI. IMPACT AND LESSONS FROM PAST ADJUSTMENT LENDING 26. IDA has made two economic recovery credits (ERCs) and two structural adjustment credits to Uganda since 1987. In addition, there has been a sector adjustment credit for agriculture (ASAC) and for the financial sector (FSAC). Key lessons from the OED performance audit report for ERC II were: (i) the need to reduce the large fiscal deficit and continuing aid dependence, and (ii) that further measures are required to increase the responsiveness of the private sector to the reforms. The two subsequent structural adjustment credits began to address the two key problems identified by OED. Table 1 below summarizes the reforms implemented under the two SACs. 27. Response to these reforms has been mixed: annual GDP growth has averaged nearly 7 percent since 1986/87 and private investment has increased from 7.8 percent of GDP in constant prices 1987/88 to 13 percent in 1995/96, but the share of exports in GDP has not increased as rapidly. That may be partly because of anti-export bias in the economy, particularly the high and widely dispersed rates of effective protection (para. 16). Therefore, the proposed SAC III includes measures to reduce tax distortions and deepen trade liberalization, both in consultation with the private sector. The FSAC, addressing distortions in the financial sector, has been instrumental in the ongoing privatization process of UCB, recovery of its bad loans, dealing with other problem banks and improving the financial base of Bank of Uganda. SAC III takes these reforms a step further and incorporates their fiscal costs into the overall macroeconomic framework. 10 TABLE 1: REFORMS UNDERTAKEN UNDER PREVIOUS ADJUSTMENT OPERATIONS AREA OF SAC I SAC II REFORM Incentive and Import licensing replaced by Liberalization of coffee exports; Regulatory System; import certificates; New legislation on cotton and Business Climate Unification of the exchange restructuring of ginning industry; rate; Review of the Investment Code Elimination of Open and operations of the Uganda General License and Special Investment Authority; Import Program as means to Liberalization of petroleum prices. allocate foreign exchange; Repeal of Industrial Licensing Act; Establishment of Uganda Investment Authority. Custodian Board Overhaul of the Board to Completion of repossession of speed up the disposition of expropriated properties and sale of expropriated properties, the non-citizen properties with including their complete pending compensation claims. listing, gazetting and repossession Domestic revenue Improve tax administration Development of taxpayer by establishing the Uganda identification number system; Revenue Authority; Curtailing of tax exemptions. Reduction in a number of tax exemptions and income tax threshold Public expenditure Release of all allocations for Protect budgetary allocations for the high-priority recurrent high priority recurrent and and development development expenditures. expenditures. Civil service reform Design of plan for the civil Reduction in the number of civil service reform; servants (from 340,000 to about Initiate ministerial reviews 150,000); Updating of teachers' and design retrenchment payroll; Monetization of housing policy, and transport benefits and divestiture of these assets; Ministerial restructuring based on the ministerial reviews. 11 28. Assessing the impact of adjustment on poverty is more difficult.1 Available data do not yet show any substantial fall in the overall incidence of poverty in the period 1989-93, but hard-core poverty has been reduced. The impact of the reforms under the two previous SACs and the ASAC is most visible in the improvement of the domestic terms of trade for cash crops: real incomes in cash crops have grown at an annual rate of 24 percent in 1987-96. This allowed the poor to participate in the benefits of growth, since they grow proportionately about the same amount of cash crops as the non-poor. Also, emerging export demand for some food crops (for regional consumption and food aid) has led to a prompt supply response, similarly benefiting the poor as well as the non-poor. However, subsistence production has grown much slower than the overall economy. Given that poverty is predominantly rural, the proposed SAC III focuses on reducing the bias against rural areas inherent in the protection of domestic industries, promoting trade liberalization in agricultural products in order to gain better access to the regional market, and improving the effectiveness and technical efficiency of public spending in agriculture and social services. VII. THE PROPOSED SAC III REFORM PROGRAM A. Objectives of the Program 29. The Background to the Budget 1996/97 and National Development Strategy 1996/97-1998/99 (June 1996), spells out the link between management of the budget, macro stability, economic growth and poverty reduction. Within this context, the economic reform programme focuses on long-term fiscal sustainability, involving a more efficient tax system as well as better management of public expenditures. On the revenue side, the programme includes measures to expand the tax base, improve tax administration, and lower the level and dispersion of trade taxes in order to reduce anti-export bias throughout the economy. On the expenditure side, the focus is on improving the management of public expenditure consistent with the on-going decentralization process. Improving the budgetary process is a necessary condition for successful implementation of the Government's Action Plan for Poverty Eradication. In addition, the Program includes measures to: (i) reduce the fiscal burden of state-owned enterprises (SOEs); and (ii) address the fiscal impact of financial sector reforms within the overall macroeconomic framework. I The first nationally representative household survey, which was carried out in 1992/93, shows that 61 percent of Ugandans were below the poverty line, while 86 percent spent as much as 70 percent of their total expenditures on food. 12 B. Toward Fiscal Sustainability 30. Fiscal sustainability is defined as a level of fiscal deficits that can be financed from domestic or external sources and which are consistent with macroeconomic targets. Before 1986, the possibilities of Uganda financing the fiscal deficit by borrowing from domestic or external sources was very limited, so the budget deficits were low (Figure 1). Higher levels of foreign aid, which became available to the government after 1986, allowed the financing of expenditure levels that were well beyond its domestic revenues. To reduce these fiscal imbalances, the government has been making sustained efforts to both increase revenues and contain the growth of expenditures as a fraction of GDP. These policies have been particularly successful since 1991/92. Figure 1: Fiscal and External Deficits 18 15 .. - 12 6 - . 3 Co r - 00 0) 40 N Co) V ) C Co 00 Do Do CD0) 0) 0) a) 0) 0) LL L. U .. LL I L L L L L> L LL L Fiscal Year . .. Curr Acct Deficit Overall Fiscal Deficit ....... Total Aid 31. In 1995/96, about one-half of the central government's total expenditures was covered by foreign aid (grants and loans)2. Excessive reliance on external loans would aggravate Uganda's heavy debt burden (see para. 10). To improve its external and debt sustainability over the long term, the Government intends to facilitate the expansion of exports and to contract new borrowing only on highly concessional terms (Background to the Budget, para. 6.4.21). The Government's goal in the short term is to reduce Uganda's vulnerability to external shocks, particularly coffee price shocks. It is expected that, as the financial sector develops in the medium term, periodic suriges in coffee revenues would be absorbed smoothly by the private sector through normal portfolio adjustments. With respect to aid shocks, the Government's view is that it has already reduced its vulnerability since domestic revenues can cover the core recurrent expenditures. 32. While there has not been an explicit long term target for revenue increases or decision on the optimal level of public expenditure, the main thrust 2 Excluding IMF financing to BOU. 13 of fiscal adjustment has been to reach a situation where Uganda's own domestic revenues cover the normal functions of Government and its main expenditure priorities in order to reduce aid dependency (1995/96 Budget speech). In this regard, the fiscal program for the next three years will seek to reduce the budget deficit (excluding grants) from 5.8 percent of GDP at factor cost in 1996/97 to 3.6 percent by 1998/99 (PFP 1996/97-1998/99, para. 13). The program envisages a substantial buildup of net deposits in the banking system, which would contain inflationary pressures while enhancing private sector access to bank credit. To achieve the desired fiscal adjustment, the growth of government expenditures will be kept under control, while improving the allocation and efficiency of public resources. At the same time, revenues are programmed to increase at an average annual rate of one percentage point of GDP through improvements in tax administration and by expanding the tax base. C. Main Components of the Program 3 33. Following the Program's objectives (para. 29), its main components are: (1) revenue mobilization and improving the incentive regime; (2) improving the management and efficiency of public expenditure; (3) reducing para-fiscal deficits of parastatals; and (4) the fiscal costs of financial sector reform. 1. Revenue Mobilization and the Incentive Regime 34. This section discusses: (a) medium and longer term issues of the tax structure; (b) increasing tax revenues; (c) reform of trade taxes and reduction of the anti-export bias; and (d) local taxation. (a) Tax Structure - Medium and Longer Term Issues 35. Background. Tax revenues increased from 7.0 percent of GDP in 1989/90 to 9.5 percent in 1995/96, but they are still below the average in SSA (18-19 percent of GDP). The most substantial contribution to revenue growth has been the rapid expansion of the main tax bases. Increases in some tax rates, in particular excise and petroleum taxes, have also been important. 36. Throughout the reform program to date, revenue growth targets have outweighed considerations of optimal tax structure. The rush for growth in revenue, while justified in terms of macroeconomic adjustment, public expenditure needs and strong Government commitment to reduction in aid dependency, has also had costs in terms of distorting the allocation of resources. In particular, private sector surveys have signaled considerable concern on taxation, at least as much arising from administration as tax rates. A significant The Annex includes a list of available Technical Background Notes on components of the Program. 14 FIGURE 2: TAX STRUCTURE 1985/86-95/96 12 10 TOTAL RE%ENLIE 8 *Aher mwirect and non ii% C 6 sales laxand C T 2 in,xme lax 0 85/6 86/7 87/8 88/9 89/90 90/1 91/2 92/3 93/4 94/5 95/6 part of the problem for the private sector lies in uncertainty about future taxation, which has been a significant disincentive to investment. While the revenue effort is continuing and Uganda is developing a view on its long-term goals on revenue levels and the size of Government, the Ministry of Finance will ascertain what long term tax structure should be developed on the way. 37. Present Tax Structure. The broad pattern of changes in tax structure over the last ten years can be seen in Figure 2. There are several distinctive features. Firstly, the rapid displacement of export taxes can be seen during the period up to 1990/1, with the share of total revenue falling from 44 percent in 1985/6 to zero in 1991/2. Secondly, the contribution from petroleum taxes rose rapidly from 1988/9, rising to almost 30 percent of total revenue in 91-93 and then falling to around 23 percent in the last three years. Other movements are less dramatic. The overall percentage contribution from non-petroleum customs has risen over most of the period, settling at around 14 percent in the last four years. The contribution from sales tax and commercial transaction levy (CTL), grouped together as the taxes now replaced by VAT, has grown over the period as a whole, although the share rose sharply to 1988/9, fell rapidly over the next three years (partly arising from reclassification of sales and excise tax) and has since risen slowly from 26 percent in 1991/2 to 32 percent in 1995/6. Since the rate increases of 1992/3, excise tax has steadily contributed about 10 percent of revenue. The share of income tax has not risen much from the 10 percent share 15 achieved in 1988/9. However, Government has just approved a proposal for a new income tax code with a view of increasing the share of income tax in total revenues. 38. The Tax Structure in the Medium and Longer Term. During at least the latter few years of this period, the Government has frequently indicated a commitment to a strategy of reduction in trade taxes, steady growth in income tax and greater use of the sales tax/VAT. 39. Trade Taxes: (i) Customs. Dependence on trade taxes has increased significantly over the period as a whole and shows no sign of falling in recent years. A high level of dependence on trade taxes is typical of developing country tax structures, driven by the relative administrative ease of collecting such taxes. The main problem for export expansion is the non-petroleum customs. Reducing their share has to be a more explicit object of policy; (ii) Petroleum. The petroleum tax contributes to anti-export bias by raising exporter costs, both directly and indirectly, through the tax passed on in domestically produced inputs. Also, a tax on transport falls more heavily on regions distant from Kampala, including many of the poorest districts. In practical terms, it is difficult for Uganda to sustain petroleum taxes more than double the level in Kenya, for example. Thus, it seems both desirable and prudent to project a declining share of revenue from petroleum over the medium term, as other sources of revenue emerge. 40. As mentioned above, there will be an increasing role to income tax, VAT and excise taxes: (i) Income tax. Income tax is presently underachieving , both by international comparison and by comparison with the achievement on other taxes such as sales tax/CTL. Income tax at 1-1.5 percent GDP is low even for low income developing countries (for example, it's 3.5-4 percent in Tanzania). While leaving aside issues of composition between corporate tax and personal income tax at local and central level, it seems clear that in aggregate terms, income tax should be rising steadily over the medium term, arising from accelerated improvements in tax administration; (ii) VAT. There is no doubt that the effectiveness and acceptability of the tax is the most critical condition for desirable revision of the tax structure. The VAT should both contribute an increasing share of a rising level of aggregate revenue, and also play the role of primary instrument of short term macroeconomic adjustment on the revenue side. (iii) Excise. On the excise tax, it would be desirable to maintain its present contribution more or less unchanged. (b) Increasing Tax Revenues 41. Over the next three years, in order to reduce the fiscal deficit, the Government intends to increase further revenues at an average annual rate of one 16 percentage point of GDP. This target would be met by improving tax administration and by expanding the base rather than hiking rates. 42. Tax Administration. Lack of confidence in tax administration has featured prominently among negative perceptions of Uganda in surveys of the private sector, especially potential investors. In particular, lack of an independent tax appeals mechanism has undermined credibility of the system. However, the Ministry of Finance has made proposals for the required legislation. Recent studies have indicated that there is significant scope for increases in tax revenue by widening the tax net to reach the self-employed and other hitherto lightly taxed sectors, better assessment of tax payers as well as better valuation of imports and a reduction of tax evasion. The role of the Uganda Revenue Administration (URA) Board needs to be clarified so that management is protected from undue interference to perform its duties. Since further growth in revenue will be heavily dependent on administrative improvement, it is essential for the Ministry of Finance to set appropriate targets for URA performance and monitor closely its achievements. That would allow rapid identification of problems in these areas, while also providing a reliable basis for an efficient incentive system for the URA. 43. The Government is committed to continuing to increase revenue as a proportion of GDP. This will be achieved primarily through further improvements in tax administration involving: * better monitoring of URA performance enabling targets to be set for genuine tax administration improvements; * review of the composition and clarification of the role of the URA Board so that URA management has sufficient independence to deliver improved revenue performance with the Board's powers to intervene in the day-to-day management strictly curtailed, particularly with regard to appointments and case work; * enhancing public perceptions of the Board's integrity by reviewing and setting appropriate safeguards for Board members to declare their assets throughout their tenure and to declare any interest for example in tender discussions; * reviewing and strengthening existing procedures for investigating allegations of corruption; * submitting draft legislation to Parliament for the establishment of an independent mechanism for the resolution of disputes in tax assessments and pressing for this mechanism to be operational by FY97/98; and * strengthening the capacity of the Tax Policy Department in the Ministry of Finance (MOF) to: (i) develop medium term revenue 17 strategies that take into account wider economic considerations, and (ii) set appropriate targets for URA performance. 44. Expanding the tax base: elimination of tax exemptions. Importer and firm-specific tax exemptions have been pervasive in Uganda for a long time, distorting incentives and increasing the burden on tax administration. During the past year, Government has made remarkable progress in curtailing exemptions from import duties by codifying discretionary exemptions under the Investment Code into the Customs Code, and by cutting exemptions for a wide variety of institutions and organizations and replacing them by ministerial votes that are limited and explicit.4 The Minister of Finance now presents periodic reports to Cabinet on such expenditures. In addition, all Government, parastatal and donor construction projects have to be tendered on a tax inclusive basis. As a result of these measures, revenue forgone through exemptions has dropped from an estimated USh 45 billion in 1993/94 to USh 8-9 billion in 1995/96, and the MOF Vote to cover tax expenditures where there is no prior agreement should be no more than USh 1 billion this fiscal year. A recent review of direct taxation recommends that tax holidays for investment be removed and replaced by more favorable depreciation allowances, and continued loss carry over. This will streamline administration and have a positive revenue impact in the longer term. 45. Proposed Actions: (i) clarify the role of URA Board and set appropriate safeguards; (ii) submit draft legislation to Parliament for the establishment of an independent tax appeals system; and (iii) Government will not approve new discretionary tax or duty exemptions throughout the program. (c) Reform of Trade Taxes and Reduction in the Anti-Export Bias 46. Uganda's external trade taxes have been substantially reduced since the late-1980s both in the context of regional integration and unilaterally as part of the broader liberalization of the economy. An 80-percent regional preference is currently granted to imports from the COMESA countries, while the shift to a market-based allocation of foreign exchange has also been trade liberalizing. There are various constraints to export development, such as high transport costs and scarce management skills, but trade policy still has an unfinished agenda: although import duties and charges for most products are below 40 percent, the effective protection of industries producing for the domestic market is over 90 percent on average, and protection rates are widely dispersed between industries and firms (Table 2). 4 For NGOs, the Tax Exemptions Committee (TEC) at the Ministry of Finance has reviewed each agreement and placed them on tax inclusive basis with Government meeting the tax liability through Ministerial Votes. In addition, the Ministry of Finance's Vote includes a contingency allocation to cover the tax component of NGOs' and charities' imports where there is no agreement, but where the TEC has agreed that such support is legitimate 18 Table 2: EFFECTIVE PROTECTION IN SELECTED INDUSTRIES (percent) Industry Domestic Sales vis-h-vis Exports in World World Market Regional Market Market Soft Drinks 75 75 -19 Brewing 167 167 -24 Fish Processing 111 96 -9 Grain Milling 282 186 -26 Horticulture 24 24 -21 Sugar 34 11 -17 Clothing 220 99 -38 Leather 61 16 -2 Paints 181 130 -5 Plastic Goods 35 16 -2 Tobacco Products 266 26 -11 Misc. Manufacturing 53 296 -24 Unweighted average 93 64 -16 Nominal Av. Tariff 12 5 Average Collection' 8.9 5 ' July-December 1995 Source: World Bank (1996) 47. The 1996/97 Background to the Budget points out that" high taxation of imports is effectively a worsening of the terms of trade for exporters" (para. 6.2.22), and that "the immediate costs of protectionism are in a reduction in real incomes of the whole population" (para.6.2.21). Further, the Action Plan for Poverty Eradication states that "protectionist measures, which create a limited number of jobs in urban areas at the cost of hurting rural people (who are in general poorer than the beneficiaries of protectionism) must be phased out" (para 7.5.60). 48. Thus, Government will abolish all remaining non-tariff barriers (NTBs), and replace them with temporary import surcharges, which will be gradually phased out over a period of three years. In addition, Government intends to gradually consolidate import duties applied to extra regional trade into a low, single uniform rate by mid-1999. It is estimated that a uniform rate of about 12 percent, even with the current collection efficiency, would be revenue positive, resulting in an estimated increase of about 20 percent in tariff revenues. This revenue estimate could provide a "safety margin" during the trade reform period and while the new VAT is being put in place. Eventually, as the VAT takes hold, the rate of the uniform duty could be lowered according to the longer term tax strategy discussed above (paras. 38,39). Also, to increase transparency, the differential application of other taxes/rates to imports and domestic goods will be ended. To complement these reforms and to facilitate export growth, a 19 simple duty drawback5 would be implemented to reimburse exporters for import duties paid on their inputs, and export documentation will be streamlined, including, for example, elimination of the Export Certificate.6 49. In the case of PTA trade, which accounts for 30 percent of total imports, the Government consolidated the PTA/COMESA duty bands from 22 to 4 non- zero rates (2, 4, 5 and 6 percent) in the 1996/97 budget. However, as Uganda's regional tariffs are considerably lower than in the other COMESA countries, the Government introduced, as a countervailing measure, a temporary surtax of 10 percent on PTA imports competing with domestic production (about 50 items). The surtax is also applied to extra-regional imports. The Government has also adopted a more pro-active role in its regional trade policy to pursue a greater degree of regional trade liberalization of agricultural products and more efficient transit transport arrangements. 50. Proposed Actions. The trade liberalization program will be announced in its entirety in the FY97/98 budget. Specific measures to be proposed in the FY97/98 budget include: (i) reducing the maximum import duty to 20 percent (July 1997); (ii) abolishing all remaining non-tariff barriers (NTBs) (March 1998), with the exception of cigarettes; (iii) replacing these bans by temporary import surcharges (batteries 15%, soft drinks 15% and beer 30%), which will be gradually phased out over a period of three years; and (iv) reduce the maximum rate of excise taxes (other than on beer, soft drinks and cigarettes) to 10% (July 1997). In the FY98/99 budget, the Government will propose: (i) replacing the import ban on cigarettes by a temporary import surcharge which will be gradually phased out in a period of two years; (ii) reducing the maximum import duty to 15 percent (July 1998); (iii) establishing a minimum import duty in the range of 7-8 percent (July 1998); and (iv) abolish all excise taxes (other than on beer, soft drinks and cigarettes) that apply differentially to imports and domestic goods (July 1998). In FY99/2000, import duties applied to non-COMESA trade will be consolidated into a low, single uniform rate. Determination of this rate will take into account revenue considerations. 5 The current system has been revitalized, but it retains rigidities which previously rendered it inoperable. The MOF has recently established a working group to consider alternative schemes including a proposal based on input coefficients payable by tax credit, and a workable scheme, including its administration, is scheduled to be introduced in 1997. 6 Given the considerable improvements in the data collection of the Department of Customs with the installation of ASYCUDA, the Export Certificate has become redundant and should be discontinued. Further, given the changes in the Exchange Control Act and that the Customs Declaration Form contains all necessary information, the Bank of Uganda declaration form (known as CD3) should be abolished. 20 (d) Local Taxation 51. Decentralization is a central element in the 1995 Constitution. The long term success of decentralization will depend, to a large extent, on getting local governments to raise their own revenues, and shifting the accountability of locally elected officials from the central government to the local voters. This could be started by decentralizing revenue raising power to the larger urban governments and by raising the revenue potential of the Graduated Personal Tax (GPT) for all local governments.7 Although the current structure of the GPT makes it inelastic and regressive, its coverage is relatively broad, it raises a considerable amount of revenue and it is widely accepted by the population. Thus, it can meet the test of a good local tax. One direction for reform, consistent with good tax practice, is to convert the GPT to a flat rate tax, with a much higher ceiling than at present. 52. To enhance the revenue base of local governments further, the central government tax on rental income could be converted into an equally shared tax between the central and the local government. Collection would be done by URA but assessment would be carried out jointly by the URA and the local governments (who would draw on the tax roles from the property tax and the GPT). Revenues of both the central and the local governments would be enhanced. This would produce revenue primarily for Kampala. As local governments already raise significant revenues from various license fees, market dues and other charges (rural local governments about 15 percent of locally raised revenues, and the Kampala City Council -KCC- about one-third), these fees would be expanded to cover the cost of local services being provided to businesses. There is also need for greater information sharing between URA and local authorities; computerizing the two rolls; and moving towards assessment (and collection) of rental income tax by the local government. 53. The new grant system (para. 76) must be consistent with this revenue reform. And if district revenues are harmed by GPT reform, they could be compensated by a Central Government grant. To continue with the decentralization process, the government would have to implement a formula grant that would encourage fiscal self-sufficiency and move to introduce the equalization grant. 54. Proposed Actions. In consultation with local authorities, Government will prepare an action plan designed to strengthen intergovernmental finance, The GPT is the primary source of local government revenue in Uganda. In 1995/96, it accounted for about 40 percent of locally raised revenue in Kampala and between 60 and 80 percent in most districts. The GPT is a hybrid of a modern PAYE income tax, a presumptive income tax based on notional assessment, a wealth tax and a poll tax. It is levied without exemption or deduction. It has 36 bands resulting in rates between 5 and 10 percent up to an income of UShs 820,000. Above that income, the tax is fixed at UShs 80,000. 21 including (i) converting the GPT to a flat rate tax, which will simplify its administration and collection; (ii) raising the GPT ceiling up to the floor of the income tax, which will reduce its regressiveness; and (iii) evaluating options for sharing revenues of selected taxes with local governments and for greater information sharing between URA and local authorities to improve collection efficiency. 2. Improving the Management and Efficiency of Public Expenditure 55. In order to achieve Government's poverty reduction objectives with its scarce resources, the SAC III program includes far-reaching reforms to the management of public expenditures. It reorients the budgetary process to focus on outcomes and outputs to be achieved rather than on conventional inputs, and to make it consistent with the decentralization process. The proposed reforms stress hard budget constraints coupled with increased flexibility and accountability at the ministry and district levels. While this is a long term endeavor, the program includes actions in three pilot sectors to start implementing these reforms. 56. The budgetary system will be geared towards three interrelated objectives: (i) aggregate fiscal discipline; (ii) strategic prioritization of spending among competing sectors and programs; and (iii) technical efficiency in the use of budgeted resources. Measures introduced since 1992 - particularly, expenditure control through monthly cash management - have resulted in considerable progress on the first dimension. In addition, the government has taken steps to improve expenditure prioritization by: (i) instituting a medium- term Budget Framework Paper (BFP), which includes a medium-term expenditure framework (MTEF); and (ii) ensuring increased funding to Priority Program Areas (PPAs) in key development sectors. It has also sought to improve technical efficiency by reducing the size of an overstaffed civil service and raising civil service wages. 57. The measures to be taken under the SAC III program include: (a) improving fiscal discipline by limiting supplementary budgetary provisions; (b) introducing an outcome-orientation to the budgetary process; (c) strengthening the poverty focus of public spending; (d) improving expenditure planning and management under decentralisation; and (e) enhancing accountability and transparency. (a) Limiting Supplementary Budget Provisions 58. While Government has been very effective in staying within its aggregate fiscal limits, substantial supplementaries have been incurred during budget execution. This creates two fundamental problems: (i) the budgeted priorities and the budgetary process in general are undermined; and (ii) agencies face tremendous uncertainty about the flow of resources, which undermines 22 technical efficiency. In order to address these problems, the Government will limit the total amount of supplementaries, unless there are national emergencies, as described below. Having achieved this, and when appropriate control mechanisms are in place, it should be possible to move gradually away from monthly cash budgeting, in order to increase efficiency in the management of expenditures by line Ministries and Districts. 59. Proposed Actions: the Government will limit the total amount of supplementaries (approved and/or actually spent) to 5 percent of Government funded expenditures in FY96/97 and 3 percent thereafter, except for national emergencies (throughout the program). (b) Introducing an Outcome-Oriented Budgetary Process 60. The government budget is divided between: (i) the recurrent budget of central government ministries and institutions; (ii) the budgets of local governments; and (iii) the development budget. This three-way split makes it difficult to determine levels of expenditure on key public services, particularly as a substantial share of recurrent financing is provided through the development budget and leads to a lack of transparency and accountability that is accentuated by budget practices which concentrate on the detailed budgeting of inputs rather than on levels of service delivery. The resulting lack of a significant "outcome orientation" to the budgeting process has contributed to inefficiencies in the public expenditure program. 61. Under the SAC III program, a more strategic focus will be brought to the expenditure planning and budgeting cycle. The main emphasis will be on enhancing the role of MTEF exercise in determining financing priorities within each sector, setting resource ceilings for each program, and identifying the split in resource allocations between central and local government budgets. This will allow preparation of the annual budget to become primarily concerned with the "technical efficiency" of public expenditures by focusing on ensuring an appropriate input mix within programs rather than on deriving overall program resource allocations. The specific issues to be addressed include: 62. Introducing an outcome orientation to the public expenditure program. This will involve: (i) strengthening sectoral policy and strategy development so that resource allocations are better linked to government policy objectives; (ii) improving program definition by requiring for each sub-sectoral programme a clear statement of objectives, activities to be funded from the budget, and achievement indicators; and (iii) direct linkage of budgetary allocations to service delivery outputs where this is feasible. 63. Maintaining a strong sectoral perspective in the planning of public expenditures. This will emphasize: (i) the role of line ministries in coordinating sectoral expenditure programs and in determining service standards and funding 23 norms; and (ii) an integrated sector programme approach in the MTEF that analyses together expenditure allocations from the recurrent and development budgets, and from central and local government budgets. 64. Ensuring a comprehensive district focus to the budget. This will require: (i) disaggregation of the development budget between central and district allocations so that total resource allocations to each district can be determined; and (ii) better consultation between line ministries and districts in reviewing sub-sectoral programs and determining funding requirements for the annual MTEF exercise. 65. Rationalizing the functional classification in the budget so that total resource allocations on each broad function within a sector can be determined. This will involve: (i) redefinition of "programs" in the recurrent budget in order to reflect the main areas of service delivery and ensure compatibility with the program classification used in district budgets; and (ii) introduction of the same program classification into the development budget and the disaggregation of project funding allocations by sub-sector program. 66. Rationalizing the item classification of expenditures by economic category so that expenditure allocations by broad category within programs can be determined. The most immediate requirement is to make explicit allocations in the development budget for capital expenditures, wage and salary costs, non- wage recurrent and technical assistance, so that total recurrent spending on each sub-sectoral programme can be determined. In the longer-term, a common itemization classification should be introduced. 67. Improving budgetary management and monitoring. This will involve measures to: (i) increase the "integrity" of the budget by restricting the diversion of funds between subsector programs; (ii) ensuring that donor financed expenditures are accounted for against the itemization in the development budget; and (iii) putting greater emphasis on the review of program performance in the annual MTEF exercise. 68. Implementation of these measures will be initiated through the preparation of "integrated sector programmes" for three pilot sectors - agriculture, health and education. This will involve: (i) sector strategy development in order to provide a strong policy basis for the restructuring of sector expenditure programs; (ii) institutional and program restructuring to reflect the new sector strategies and changed institutional roles post- decentralisation; (iii) the development of integrated sector budgets that incorporate funding available from the recurrent, development and district budgets; and (iv) the preparation, using these integrated programmes, of sectoral proposals for the MTEF that incorporate a review of program performance and reflect a realistic assessment of available budgetary resources. Technical 24 assistance to support the integrated sector programming exercise is available under on-going sectoral support projects and also from bilateral sources. Additional assistance will be available from the Bank's Second Economic and Financial Management Project, if required. 69. Concurrently with the integrated sectoral programming exercise, the Government will initiate a program for more comprehensive reform of budgetary systems and procedures. This will cover: (i) measures aimed at closer integration of recurrent and development budgets with the eventual objective of introducing a unified budget; (ii) associated improvements in the budget program and item classifications; (iii) greater computerization in budgeting processes; (iv) streamlining of bu iget release and implementation procedures; and (v) improved monitoring of budget implementation. The development of this programme is expected to be supported by bilateral technical assistance. 70. Proposed Actions: (i) develop a program for comprehensive reform of budgetary systems and procedures by September 1997; (ii) complete the integrated sector programming exercise in the three pilot sectors by February 1998; and (iii) based on the results of this exercise, rationalize the budgets for the three sectors in the 1998/99 annual budget. (c) Strengthening the Poverty Focus of Public Spending 71. The Government is currently preparing an Action Plan for Poverty Eradication. As the first step, the Background to the Budget 1996-97 includes a chapter describing the Government's Strategy for Poverty Eradication. The Action Plan will prioritize and refocus Government's policies and actions to ensure that the poor are able to participate in and benefit from economic growth. In order to realize this objective in practice, it is essential that the Plan be reflected in budgetary priorities and that public funds are spent efficiently in pursuit of these priorities. 72. Thus, improving the management and efficiency of public expenditures is essential for implementing the Government's anti-poverty strategy. It will support Government's effort to improve service delivery to the poor which - as shown clearly by the recent pilot survey of nine districts - suffers from poor quality and lack of transparency. Priority areas for service delivery are primary health care and education, agricultural extension, road maintenance and demand- driven water development. These Strategic Areas (SAs) are defined in the 1996/97 budget as a focal point for prioritization of public expenditures. 73. To strengthen the anti-poverty focus of the budget, Government will increase recurrent budget allocations to agriculture research and extension, to primary health and to primary education at least as much as the rate of growth of nominal GDP, and it will implement the outcome oriented budgetary process described above to improve the efficiency of these expenditures. The 25 Government will also incorporate the main recommendations of the anti-poverty Action Plan into the BFP exercise for the FY97/98 budget, and it will ensure adequate budgeting for PIP projects and that all allocated funds are fully released. 74. Proposed Actions: (i) increase budgetary allocations for agriculture research and extension, primary health and primary education at least at the rate of increase of nominal GDP, and (ii) ensure adequate budgeting for PIP projects and ensure that funds are fully released. (d) Expenditure Management Under Decentralization 75. There are several aspects in which the present budgetary process is inconsistent with the decentralization programme that the Government has initiated. In particular, the division of responsibilities in the new Constitution specifies that the central government will be responsible primarily for policy formulation and monitoring, maintaining implementation responsibilities only for national public goods, such as defense and trunk roads. However, substantial budgetary resources are still being allocated to central government ministries for functions that have been decentralized. Within the context of budgetary reform, local government budgets will also have to become outcome-oriented. The budget cycles of the two levels of government will be synchronized beginning in FY97/98, in order to facilitate the budgetary preparation process and be able to consolidate government accounts. 76. Given the extent of decentralization, substantial resources will have to be transferred from the central to district governments through conditional and equalization grants. Pilot ministries will work closely with the Districts in the preparation of sectoral budgets for programmes of national priorities. These budgets will be based on district-level plans for achieving agreed upon outputs and outcomes and on costing of programmes. Jointly with the Districts, the pilot ministries will design the rules and mechanisms for the operation of conditional grants, including appropriate monitoring requirements. These resource shifts and procedures will be implemented in FY98/99 budget. 77. Finally, maintaining fiscal discipline is important at the district level as well. In particular, the Government will regulate borrowing by districts according to the following criteria: (i) borrowing will be linked to own- generated revenues such that repayments do not exceed a pre-determined fraction of these revenues; (ii) borrowing will be published; (iii) districts will be required to report loans taken to the MOF; (iv) violation of any of these rules will be punishable under the Local Government Act; and (v) unless explicitly signed by the MOF, all district loans are non-recourse to the central government. 26 78. Proposed Actions: In accordance with the decentralization program, Government will: (i) ensure that there are no constraints in the use of unconditional block grants and delink their allocation from District personnel establishment positions as of FY97/98; (ii) modify the rules for conditional grants and develop the rules for equalization grants, including monitoring programs; (iii) adopt these rules in the 1998/99 budget; (iv) include these rules and regulations in the annual Appropriations Bill; and (v) regulate borrowing by Districts reflecting prudent guidelines. (e) Accountability and Transparency 79. Preliminary results from a recent study on tracking of public expenditures indicate that public spending is not reaching the primary education and health facilities as intended. The following bottlenecks were identified: (i) substantial delays in submitting and auditing district final accounts and monthly expenditure reports prevent any corrective measures when resources are diverted or misused; (ii) many Local Government Tender Boards (LGTB) lacked integrity, resulting in substantial overpricing of contracts and supplies; and (iii) budgetary resources are not always distributed to the departments of district administrations according to their approved work plans with adverse consequences for service delivery. 80. By and large, necessary accountability mechanisms such as the Local Government (Financial) Regulations (1995) are decreed. The issue is implementation and increased transparency. To address these problems, the Government will: (i) introduce a uniform system of budgeting, monitoring and accounting for all local governments by FY97/98; (ii) publish monthly cash releases of block grants and conditional grants to all districts in the main local newspapers beginning immediately; (iii) ensure publication of a monthly summary of all tenders awarded by the LGTB, with copies sent to the Inspector- General of the Government as proposed under the new Local Government Act; (iv) deconcentration of the activities of the Auditor-General's office to the district level by FY97/98 to enable timely and credible auditing of district accounts; (v) amend during FY97/98 the Local Government Financial Regulations (1995) to ensure that advances can be made only for developmental purposes; and (vi) ensure that appropriate disclosure, including district budgets, program outcomes and tender awards, is required of districts, and that commensurate sanctions be applied for non-compliance. 81. The Government will ensure: (i) compliance of the provisions of the Local Government Act in three randomly selected districts among the 13 districts included in the first phase of fiscal decentralization by FY98/99; and (ii) design and implementation of an Action Plan for improving financial accounting and auditing systems for central and local governments by FY97/98. 27 82. Proposed Actions: (i) public dissemination of intergovernmental transfers at the district level; (ii) satisfactory preparation of the Local Government (Decentralization) Act, and compliance with the provisions of the Act in three randomly selected districts among the 13 districts decentralized in the first phase of fiscal decentralization; and (iii) implementation of improved system of financial accounting and auditing systems for central and local governments. 3. Reducing Quasi-Fiscal Deficits: Parastatals 83. Most Ugandan state-owned enterprises (SOEs) are inefficient and represent a serious constraint to the development of the private sector. Inefficiencies in the SOE sector also have a significant fiscal impact: indirect and direct subsidies amounted to USh 208 billion in 1994 (nearly 5 percent of GDP), which was equal to five and half times government recurrent expenditure on health (USh 37.5 billion), more than twice the expenditure on education (USh 88.4 billion), and over 50 percent of the entire recurrent budget (USh 375 billion). Table 3 presents a summary of subsidies to SOEs since 1992 and projections through 1998, assuming no corrective measures are undertaken. As Table 3 clearly shows, the fiscal implications of this passive scenario are huge. Table 3: ACTUAL AND PROJECTED SUBSIDIES TO SOES IN UGANDA, 1992-1998 Type of Subsidy <-- Actual in USh billions-> <----- Projections in USh Billions ----> (1993 prices) 1992 1993 1994 1995 1996 1997 1998 Direct Subsidies 9.4 15.2 19.4 20.1 22.0 18.0 20.0 Equity Support 34.2 55.1 78.8 135.0 210.0 285.0 230.5 Financing Terms 16.9 27.3 57.0 62.7 69.0 75.9 83.5 Fiscal Terms 28.4 46.8 18.0 15.1 18.0 11.5 11.0 Monopoly situations 8.8 20.2 35.3 37.4 15.0 8.0 4.0 Total 97.7 164.6 208.5 270.3 334.0 398.4 349.0 84. Government's Strategy. Government is strongly committed to privatization and SOE reform and has made significant progress since early- 1995. Government is now using three instruments - sector reform, privatization and financial controls - to address the two sets of problems. First, provision of key services will be improved through sector reform and 8 In 1993, the public enterprise sector comprised over 130 firms, including trade and commerce, agro-production and processing, manufacturing, finance, insurance, and other services. These enterprises had a debt stock of U Sh 968 billion (out of a national debt stock of U Sh 3,600 billion) and an estimated employment of 78,000 (formal employment being 275,000). The sector was estimated to be operating at 25-30 percent capacity utilization and contributing to about 5 percent of GDP. The privatization program started in 1992/93 but at a very slow pace. A review of the program in 1995 resulted in targeting 85 percent of all SOEs to be privatized by the end of 1997. At the end of 1995, 48 SOEs had been divested (about one-half of the total). 28 privatization in power, telecom, water, and transport. This will involve: (i) breaking up the UPTC monopoly through the splitting of Posts and Telecommunications, the liberalization of non-core telecommunication services, and the introduction of a duopoly in core services. The Government intends to undertake similar reforms in the other sectors, starting with the power sector. Sector reform plans will be prepared for power, water and transportation in the by December 1997; and (ii) increasing competition by reducing barriers to entry and by allowing private participation in management and/or ownership of SOEs. Second, the drain on the budget will be addressed by: (i) imposing strict financial controls on all SOEs to eliminate all direct and indirect subsidies to commercial SOEs and to phase out subsidies to non-commercial SOEs over a four-year period (except for subsidies relating to social missions); and (ii) privatizing 85 percent of all SOEs by the end of 1997. 85. SAC III will focus on one of the three instruments used by Government - financial controls. Sector reform is being supported through investment operations (e.g., Power III, EDP), while Technical Assistance for privatization is being provided through the EDP project. The recently established Parastatal Monitoring Unit (PMU) within the Ministry of Finance, which records financial flows to and from SOEs, will have the necessary data to monitor implementation of the SAC program. 86. Proposed Action: A 20 percent reduction in overall subsidies to the SOE sector in 1996/97 compared with 1994/95 (excluding subsidies to and from the Uganda Commercial Bank dealt with under a separate component of this project). The reduction will be measured using each year's report on Financial Flows Between Government and Public Enterprises and Amongst Public Enterprises. 4. The Fiscal Cost of Financial Sector Reform 87. Strategy for the Financial Sector. Uganda's growth targets require a sound and more efficient financial system that facilitate fast and secure payments, provide saving facilities for urban and rural population and loans to credible businesses. In the last two years, the government and the Bank of Uganda took steps to rectify weaknesses in the financial sector. Actions taken included: (i) placing UCB non-performing portfolio with a specially created Non Performing Assets Recovery Trust (NPART); revamping UCB's management and providing foreign technical assistance to improve management; preparing UCB for privatization, with the assistance of an international investment bank; and evaluating the consequences of UCB privatization on providing banking services in rural areas; (ii) intervening 2 private banks, placing BOU nominated care-taking management in them, and providing BOU funds to cover negative capital positions; both banks were sold recently to new investors; (iii) monitoring support provided by donors or owners in three other problem banks; (iv) announcing a minimum capital requirement for banks to be 29 reached by December 31, 1996; (v) declaring a moratorium on the licensing of new banks, except for banks with special benefits to the Uganda financial sector; (vi) strengthening the capacity and skills of the BOU supervision department; and (vii) creating a deposit insurance scheme, financed by the commercial banks and the Government, to guarantee deposits up to USh 3 million per depositor. 88. UCB and NPART. Under the policy agreed with IDA under the FSAC and based on the work performed by an international investment banker, the Memorandum of Offering for the sale of UCB was issued in December 1996. The investment banker is currently evaluating notices of interest received from potential buyers. Actual materialization of the sale of UCB to private investors will depend on certain conditions: (i) the buyer should be a reputable investor with recognized capacity in managing financial institutions and should provide both the capital required for UCB to meet BOU capital requirements and the management for UCB in a marner acceptable to BOU; (ii) the strategic buyer, by himself or together with other private sector investors as a group, should buy 100 percent of the bank shares. Investors proposing to buy a majority of more than 51 percent in the bank capital will also be considered if no 100 percent sale is possible. (iii) UCB sale is being facilitated by the issue of government bonds covering the negative capital of the bank; the government will further support the sale with guaranties against contingent liabilities, if so demanded, as long as they will be reasonable, in type and amount. 89. If successful sale is achieved, the new owners will decide which of the existing branches they wish to operate and which they would rather close. However, 18 rural branches of UCB were identified as essential for the provision of banking services in remote districts where UCB is the only operating bank. If UCB new owners decide to close any of these identified branches, their continued operations will be secured by auctioning their management for 2 years among financial institutions licensed in Uganda, with a budgetary subsidy by the government. At the end of the 2 years contract the branches will be offered for sale. If no buyer is found, the need to keep each branch open will be reviewed and if essential, a new 2 years contract will be auctioned for the remaining unsold and essential branches. 90. In case no private investor acceptable to the government and BOU is found, or if the demands of potential buyers for government compensation and guaranties are not acceptable, by June 30 1997, an alternative plan for the restructuring of UCB will be pursued. Under such plan: (i) the management of UCB will be contracted to a reputable international banking group for 3 years; (ii) upon handing over the management, the government will provide UCB with enough government securities to cover the bank minimum capital requirements; (iii) the management contract will aim at reducing, over the management contract period, UCB staff and branch network (with the exception of 18 rural branches identified by the government as essential for the provision of banking 30 services in remote districts) to a level that will secure efficient and profitable operations. To maximize the range of options available to Government, the management contract will provide a fee structure based on achieving both profitability and down-sizing. At the end of the management contract period, the bank will be offered again for sale to private investors. 91. The success of NPART is essential as a relief to the budgetary cost of recapitalizing UCB and as a provider of correct signals to all parties about the need to repay loans. To achieve these goals, the government is committed to support NPART activities in collecting UCB non-performing loan portfolio, with the objective of collecting at least USh 25 billion. The government will submit to Parliament an amendment to the NPART Act, to cover (i) extending NPART mandate for at least an additional year to compensate for the delay in starting its operations; and (ii) limiting grounds for appealing the special tribunal decisions to points of law only. 92. Proposed Actions: (i) complete sale of UCB, acceptable to IDA, or implement back-up plan agreed with IDA; (ii) implement policy with regard to UCB's 18 essential rural branches, agreed with IDA; and (iii) provide all necessary support to NPART to collect at least USh 25 billion in non- performing loans. 93. Treatment of Problem Banks. A clear policy statement on future treatment of problem banks, when such problems are identified, was prepared by the government and BOU, in consultations with IDA, and will be published in May 1997. The policy is intended to secure the health of the financial system while limiting to the largest degree possible the cost of such actions to the government of Uganda budget or to BOU. 94. Proposed Action: Finalize and publish a policy paper on future bank intervention agreed with BOU, in consultation with IDA. 95. Uganda Development Bank (UDB). UDB, a fully owned parastatal, was established to provide long term financing. Over the years, the Bank has continuously accumulated losses, thus showing a substantial negative net worth. Currently, UDB's portfolio contains almost in its entirety non-performing loans. Therefore, UDB is not a viable intermediary and cannot be considered for privatization. Nor will UDB be considered as an eligible intermediary by most international lenders, as it does not meet minimum requirements for solvency and capital ratios. As UDB sources of funds are foreign lenders, and all its loans are guaranteed by GOU, the government is paying UDB's debt, which is estimated to amount to USh 60 billion within the next 6 years. 96. While UDB is placing a heavy burden on the GOU budget, it provides no benefits to the economy: in 1994 and 1995, it had very little activity, as total new disbursements amounted to less than its own operational costs. Moreover, 31 two development banks - the East African Development Bank (EADB) and the Uganda Development Finance Corporation (UDFC), as well as the commercial banks utilizing lines of credit provided through the Development Finance Department (DFD) at the BOU, are providing long and medium term loans for viable investments 97. Proposed Action: The government will review the recommendations of a restructuring task force, presented in April 1997, and decide on UDB future in consultation with IDA before the end of FY96/97. The collection of UDB portfolio could be added to the tasks of NPART, through an amendment to the NPART Act to be submitted to Parliament. Government intends to implement policy regarding UDB to stop losses and minimize budgetary costs, in agreement with IDA, early in FY97/98. 98. BOU Recapitalization. Capitalization of BOU to a level acceptable for successful operations of BOU, both in Uganda and vis-A-vis international and foreign financial institutions, is essential for BOU functioning. The Government took the initial action of placing USh 60 billions of government bonds against BOU shares. A study by a reputable international auditing firm was finalized recently and established the BOU capital discrepancy. The study results were accepted by both BOU and the government. Both parties are now discussing ways to bring BOU to the desired level of capitalization (by issuing additional government bonds or by changing the terms and conditions of the government debt to BOU) and assessing what further restructuring is required at BOU, with the view of accelerating the recapitalization process by end-June 1997. VIII. THE PROPOSED CREDIT 1. CONDITIONALITIES: BORROWER-IDA CONTRACT 99. Agreements and Understandings to be Reached and Monitorable Actions. The Government's reform program will be set out in the Letter of Development Policy and summarized in the Policy Matrix attached to it (Annexes 3 and 4). The Government would be expected to honor its commitments under previous Bank Adjustment operations. It will also continue to maintain a sound macroeconomic framework as set out in the Policy Framework Paper. 100. The following specific actions are expected to be included in the Government reform program and implemented before the proposed Credit is presented for Board approval: * Submit draft legislation to Parliament for the establishment of an independent tax appeals system. 32 * Adopt rules for conditional grants following constitutional criteria; establish monitoring programs; and delink block grants from District personnel establishment positions. * Regulate borrowing by Districts. * Prepare timetable for auditing of local governments. * Submission of the draft final report on Financial Flows Between Government and Public Enterprises and Amongst Public Enterprises for 1994/95. * Finalize and publish policy paper on future bank intervention agreed with BoU and acceptable to IDA. 101. The following condition would need to be fulfilled prior to the Credit becoming effective: * Announce tariff reform plan and timetable. 102. The following conditions would need to be fulfilled prior to second tranche release: * Submit draft legislation to clarify the role of URA Board and set appropriate safeguards. * Reduce maximum import duty to 20 percent; end differential use of other taxes/rates for imports and domestic goods; and abolish remaining NTBs (except tobacco). * Implement integrated sector programming and budgeting exercise for three pilot sectors (Education, Health and Agriculture) in conjunction with Districts. * Increase budgetary allocations for agriculture research and extension, primary health and primary education at least at the rate of increase of nominal GDP. Ensure adequate budgeting for PIP projects and ensure that funds are fully released. * Commence implementation of action plan for improving financial accounting and auditing for central and local government. * Timely publication and dissemination of: (1) District budgets and Local Government Public Accounts Committee Reports; and (2) audits of district accounts on completion. * Achieve 20 percent reduction in overall subsidies to the SOE sector in 1996/97 compared with 1994/95 (excluding UCB and UDB, dealt with separately). * Complete sale of UCB, acceptable to IDA, or implement back-up plan agreed with IDA. * Implement policy regarding UDB to stop losses and minimize budgetary costs in agreement with IDA. 103. The following conditions would have to be fulfilled prior to the release of each of the floating tranches: 33 * Implement a low uniform duty rate on all non-regional imports. * NPART to collect at least Ush 25 billion in non-performing assets. 104. In addition, all the following conditions would have to be fulfilled throughout the program: * Stabilization program on track. * Limit supplementaries to no more than 5 percent of Government funded expenditures in FY96/97 and no more than 3 percent thereafter, except for national emergencies. 105. Government Commitment and Participatory Approach. As mentioned above (para. 2) the preparation of the proposed operation has already benefited from extensive Government participation, including determination of its main components. To broaden ownership during the forthcoming months, the Government intends to hold sensitization seminars for government officials, parliamentarians, NGOs, the private sector and donors to explain the adjustment program and build broad based support for it. This initiative would benefit from Bank assistance, and is likely to include support from EDI. Dissemination of the program has already been started on a limited basis in March and July 1996, when the proposed public expenditure reform component was discussed with a wide cross section of public officials. During appraisal (October 1996) the main aspects of the proposed program were discussed with key stakeholders, including private sector representatives (the Private Sector Foundation, PSF). 2. SUPERVISION PLAN (a) Borrower's Leading Role 106. The overall coordination and implementation of SAC III will be the responsibility of MOF. This ministries and others involved in implementation - including the Ministries of Planning and Economic Development, Public Service, Local Government, Education, Health and Agriculture - will work jointly with IDA in-country supervision missions. In addition MOF would be responsible for preparing the groundwork for supervision missions as well as organizing quarterly project reviews with various stakeholders and submitting regular progress report to IDA. (b) IDA Supervision 107. Given the complex nature of the proposed operation, adequate supervision resources would need to be provided by IDA. Requirements are estimated at about 57 SWs (including consultants) over three fiscal years. A more detailed break-down can be found in the table below. These supervision 34 SUPERVISION PLAN (in staff weeks) Fiscal Year FY97 FY98 FY99 HQ Supervision 2 6 3 RM Supervision 2 6 3 Missions 5 20 10 TOTAL 9 32 16 Note: See Table on In-Country Supervision. resources will be financed out of IDA's administrative budget, not from the proceeds of the IDA Credit. 108. As indicated in the above table, routine supervision and monitoring will be shared between Headquarters and Resident Mission staff in equal shares. Significant in-country supervision resources would also be needed during FY98, between effectiveness and second Tranche release. Details of composition of the in-country supervision team are in the attached table. TIMETABLE FOR IN-COUNTRY SUPERVISION Approximate Date Activity Expected Skills (Staff Weeks) May /June 1997 Supervision Mission to assess Task Manager (2) compliance with conditions for Board Country Lawyer (1) presentation and agree with Public Expenditure Economist (1) Government on next steps. Financial Sector Expert (1) September 1997 Supervision mission to assess progress Task Manager (2) in implementation of the program, Financial Sector Expert (1) particularly the financial sector. Trade Expert (2) January 1998 Supervision Mission to assess Task Manager (2) progress towards meeting second Public Expenditure Economist (1) tranche conditions. Decentralization Specialist (1) Financial Sector Expert (1) Tax Expert (1) Public Enterprise Reform Expert (1) June 1998 Supervision Mission to evaluate Task Manager (2) whether second tranche release Country Lawyer (1) conditions have been met, and assess Public Expenditure Economist (1) progress towards meeting floating Decentralization Specialist (1) tranche conditions. Financial Sector Expert (1) Tax Expert (1) Public Enterprise Reform Expert (1) December 1998 Supervision Mission to evaluate Task Manager (2) whether floating tranche release Country Lawyer (1) conditions have been met. Trade Economist (2) June 1999 Final supervision mission and Task Manager (2) Implementation Completion Review Fiscal Economist (1) Preparation Mission. Public Enterprise Reform Expert (1) Financial Sector Expert (1) 35 Table 4: PERFORMANCE INDICATORS Policy Indicators Intermediate Indicators Outcome Indicators Macroeconomic Policy Fiscal deficit/GDP declines from Inflation target 5% Annual GDP growth 7% in 1996- 7% in FY96 to 4.7% in FY98 2000 Revenue/GDP Central government savings from 11% in FY96 to 13% in from 0.6% of GDP to 2% FY98 (factor cost) Structural Reforms Average and spread of tariff rates. Average level and spread of the Effective Share of exports in GDP Rate of Protection; decline in the level at from 11.8% in FY96 to 13% in least by 1/3 in FY98 compared to FY95 FY98 Revamping the import duty Share of non-coffee exports drawback scheme for exporters from 36% in FY96 to 45% in FY98 (assuming a normal coffee year) Public Expenditure Supplementaries/total budget Deviation between budget and releases at Adherence to public expenditure Max 5% in FY97; max 3% the vote level decline from 27% in FY95 priorities set in the budget. thereafter. to 15% or less. Expenditure on Strategic Areas During FY96-98, textbooks supplied to Literacy levels, primary enrollment (excl. defense)9 schools increased by at least 20%; and drop-outs; immunization coverage up by at least 5%; infant mortality rates; productivity extension visits increase by at least 20%; in agriculture; access to markets. Launching of outcome-oriented Sector programming exercise implemented Improved public service delivery; budgetary process in three pilot progress to be measured by service ministries delivery surveys. Decentralization Allowing increased taxing power Local tax collection increases Improved service delivery to local to local authorities by 20% from FY96 to FY98 communities to be measured by service delivery surveys. Finalizing design and Share of public spending (non-statutory, Higher share of expenditure implementing of transfer system to non-debt related) at the district level reaching beneficiaries to be districts from 25% in FY96 to at least 50% in measured by expenditure tracking in FY98 selected areas. Putting in place systems to Timely dissemination of information Improved accountability and improve information and within government and to the civil society transparency in the public sector dissemination of central and local government accounts State-owned enterprises Monitor of financial flows to! from 20% reduction in subsidies by end June Reduction in fiscal deficit as above. Government and SOIs 1997 (from June 1995) Progress in privatization Financial Sector Reform Complete privatization of UCB Spread between lending and savings rates Private investment increases from in banking system declines at least by 1/3 13% in FY96 to 15% of GDP in FY98. Produce Policy Paper Increased monetization of economy; Larger private savings in the M2/GDP from 10% in FY96 to 11% in banking system. FY98. Complete recapitalization of BOU Central Bank finances 100% of its operating costs. mThese indicators are also influenced by development expenditures. 36 (c) Performance Indicators and Expected Results 109. In accordance with the program objectives described in para. 8, the following results are expected from the SAC III policy and institutional reforms: (The corresponding performance indicators are presented in Table 4.) * maintenance of fiscal and macroeconomic stability, including reduction of the fiscal burden of state-owned enterprises and the financial sector reform; * reduction in the anti-export bias and higher export growth; * improved budgetary process and more efficient use of public resources; * progress in decentralization and improved service delivery; and * a shift in the public expenditure pattern to match the decentralized expenditure assignment, including a larger share for spending at the district level and on basic services within sectors, consistent with the Government's poverty reduction strategy. (d) Other Donors and Stakeholders Participation 110. Building on the consultative approach pursued during project preparation, key stakeholders will be either asked to participate in supervision mission or briefed by such missions. These groups include the donor community, selected Districts, NGOs and private sector representatives. 111. Expected Complementary Donor Financing. As in the case of previous adjustment operations, the program is being prepared in consultation with other donors. Additional financing for the program, some of which might be through contributions to the Ugandan multilateral debt fund, is expected from a number of donors, including the EU, ODA (UK), the Netherlands, Japan and USAID, as well as possibly Norway and Switzerland. In this way, SAC III can act a catalyst for attracting additional grants, with IDA being the lender of last resort. 3. CREDIT ADMINISTRATION 112. Credit Amount, Use of Funds, and Tranching. The proposed Credit would be SDR91.1 (US$125 million equivalent). The borrower would be the Republic of Uganda. The counterpart funds generated by the SAC will be used to finance part of the Government's budget. The Credit would be disbursed in two regular tranches of US$45 million and US$40 million, and two floating tranches of US$20 million each. The fixed tranches would help Uganda fill its budgetary financing gap, essential to maintain a sound financial and macroeconomic program. The floating tranches would be disbursed when key specific measures are implemented. This design would provide the flexibility needed to implement fairly complex reforms, while protecting the program from 37 implementation delays. The first regular tranche would be disbursed at effectiveness, and the second regular tranche would be disbursed about a year later, in mid - 1998. The floating tranches are expected to be disbursed by end 1998 (see attached Policy Matrix). 113. Disbursement. Disbursement arrangements will follow the simplified procedures approved by the Board on February 1, 1996. The Borrower will open an account in the Bank of Uganda. Upon IDA notification of tranche release, proceeds of the credit will be deposited by IDA in this account at the request of the Borrower. If after deposit in this account, the proceeds of the credit are used for ineligible purposes (i.e., to finance items imported from non- member countries, or goods or services in the standard negative list), IDA will require the Borrower to either (a) return that amount to the account for use for eligible purposes; or (b) refund the amount directly to IDA, in which case IDA will cancel an equivalent undisbursed amount of the loan. Although a routine audit of the account will not be required, IDA reserves the right to require it. 114. Management, Monitoring and Accounts. The Ministry of Finance will have overall responsibility for the management of the program. The Government would establish a Coordinating Group made up of key staff from the ministries and agencies which will actually carry out the various proposed actions to be supported by the credit. This group would work closely with IDA staff and will provide the focal point for the implementation of the program. The Resident Mission will play an important role in program supervision. The Government's capacity to maintain adequate records and accounts in respect of quick-disbursing credits has shown considerable improvement. It is therefore in a better position to comply with the standard covenants relating to accounts and audit reports. IX. BENEFITS AND RISKS 115. Benefits. The proposed program will support the recent gains in achieving macroeconomic stability and will accelerate the structural adjustment process in Uganda. The measures to improve tax collection and to reduce dependence on budget support should result in achieving a sustainable fiscal position with a lower share of aid over the next few years. Efficiency gains are expected both in the private and public sectors, with the increased participation of the private sector in the key areas of the economy and the expected improvement in the management and operation of the public sector. This would help sustain high levels of economic growth, while encouraging greater efficiency in public service delivery and improved equity. 116. Risks. The main risks associated with this operation are political as well as related to the institutional capacity to carry out the necessary reforms. A related risk is continued insecurity in the North and resulting high level of non- 38 developmental expenditures. These could undermine some of the proposed budget reform measures. To alleviate such problems, there would be enough budgetary flexibility to ensure that national emergencies can be addressed. Another issue is that Uganda is highly dependent of foreign aid for its public investment and certain recurrent programs. There may be a risk of underfinancing which might slow down reform implementation. This risk is mitigated by the country's good track record and expected reductions in aid dependency. Slow implementation of certain key reforms, particularly those affecting the financial and public enterprise sectors, might also undermine the program. However, given that the pace of these reforms has picked up recently, and a participatory approach is being pursued to ensure broad based ownership, this risk should be manageable. Finally, to minimize the risks related to institutional capacity, complementing support being provided by other donors, the Bank is currently financing the Institutional Capacity Building Project and preparing, in parallel with the SAC, a technical assistance project -EFMP II- to support improved economic and financial management. X. RECOMMENDATION 117. I am satisfied that the proposed Interim Fund Credit would comply with Resolution No. 184, adopted by the Board of Governors of the Association on June 26, 1996, establishing the Interim Trust Fund and I recommend that the President approve it. Gautam S. Kaji Managing Director Attachments Washington, D.C. May 1, 1997 39 Annex 1: Higher Impact Adjustment Lending 1. What is the demonstrated commitment of the Government to the program? The Government has a good track record of implementing reforms since 1987, in particular, since 1992. The proposed components of the program were suggested by the Government, who also participated in putting together this report. Moreover, the Government initiated extensive consultation with the private sector to take into account its concerns and to widen the support base of the program. 2. What is the country specific evidential basis for the proposed structural reforms, including their impact on the poor? The Bank's Poverty Assessment (1993) and the Country Economic Memorandum (1995), which also updates the Poverty Assessment, highlight the importance of acceleration of growth and export development as a necessary condition for reduction of poverty. Low level of revenue mobilization, anti-export bias and remaining inefficiencies in the public sector were found to be the main constraints to faster growth. The above reports used data from household surveys and a domestic firm survey as the basis of the analyses. Specific actions draw also on the annual Public Expenditure Reviews. Also, the Background to the Budget 1996/97 and National Development Strategy 1996/97-1998/99, which includes the Government's Action Plan for Poverty Eradication stresses inter-alia the need to: (i)improve the management of public expenditure in order to improve delivery of social services; and (ii) reduce anti-export bias to stimulate exports and GDP growth. 3. Are too many institutional development aspects of public, privatization and financial system reform being loaded on to the adjustment operation? No. Necessary institutional development support is already provided for in a number of on-going IDA financed projects, such as Enterprise Development Project (privatization and public enterprise reform), Institutional Capacity Building Project and Economic and Financial Management Project; SAC III will focus on implementation of policy-level commitments. 4. Are there explicit measures on the level and composition of the fiscal balance? The proposed Credit has revenue mobilization and improved management of public expenditures at center stage. It is set in a medium term fiscal framework with explicit fiscal targets to be monitored and monitored by the Bank. The IMF's ESAF arrangement (in place until September 1997) could help in the monitoring. The Credit has explicit social expenditure conditionality as well as specific targets for curtailing the fiscal burden of SOEs and financial sector reforms. 5. What would happen to the macro resource gaps if one or more of the tranches were held up? The total size of the credit is less than 10 percent of the projected annual medium and longer term capital inflows to Uganda in 1997 and 1998. Thus, delay in tranche release would adversely affect but not necessarily derail the Program. If a delay triggers a similar reaction from all other donors, then there could be a danger of derailment.  40 Annex 2 List of Technical Notes (See Annex 6) Technical Note 1: Tax Revenue Issues: A Longer Term Perspective Technical Note 2: Tax Exemptions: Past and Present Technical Note 3: Adjusting Trade Policy for Export Growth Technical Note 4: Estimates of Effective Protection and Manufacturing Efficiency Technical Note 5: A Proposal for an Effective Duty Drawback System Technical Note 6: Issues in Local Taxation Technical Note 7: Outcome-Oriented Budgetary Process Technical Note 8: Improving the Management and Efficiency of Public Expenditures Consistent with the Decentralization Program Technical Note 9: Reducing Quasi-Fiscal Deficits: Parastatals  ANNEX 3 Page 1 of 17 Telephones: Minister; 2 4 3iniStry of Finance, Kampala 243054 & 232370 Mnsr fFnne Office; P.O. Box 8147, Kampala 23470019 (10 lines) Telex: 61170 Telegrams: "FINSEC" Uganda. In any correspondence on THE REPUBLIC OF UGANDA this subject please quote No. EU/A/44 Letter of Development Policy Mr. Gautam Kaji Managing Director International Development Association 1818 H Street, NW Washington, DC 20433 Dear Mr. Kaji: 1. I am writing to request, on behalf of the Government of Uganda, a Credit of US$ 125 million equivalent from the International Development Association (IDA) in support of our program of Structural Adjustment. The proposed Credit will support Government efforts to enhance overall fiscal sustainability and the management of public expenditure, as well as to encourage supply response by reducing the anti-export bias in the economy. It will support the achievement of the Government's main goal of poverty reduction, and is consistent with the Policy Framework Paper dated November 8, 1996, agreed between the Government, the Bank and the IMF. Proceeds from the proposed credit will help fill Uganda's projected external financing gap through 1998/99 and will provide budgetary support during a period of on-going fiscal reform. BACKGROUND AND RATIONALE 2. Since 1987, following 20 years of political turmoil and civil war, Uganda has been implementing an economic recovery program supported by the World Bank, the IMF and other multilateral and bilateral donors. The main focus of the program has been: (i) restoration and maintenance of macroeconomic stability; (ii) liberalization of the exchange, trade, price and marketing systems; (iii) improvement of the incentive structure and business climate to promote savings and investment for productive activities by the private sector; and (iv) improvement of the country's economic, social, and institutional infrastructure. 3. The results of the program have been impressive, particularly since 1992. Real GDP has increased at an average annual rate of about 7 percent, while the annual inflation rate has declined from 35 percent in 1991 to around 7 percent in each of the last ANNEX 3 Page 2 of 17 2 three years. Government revenues have risen from 7 percent of GDP in 1991/92 to over 11 percent in 1995/96, resulting in a drop in the overall fiscal deficit (on a commitment basis and excluding grants) from 15 percent to 7 percent of GDP over the same period. The improvement in the fiscal position has helped cut the level of external current account deficit by over half to an estimated 7 percent of GDP in 1995/96. 4. Notwithstanding these achievements, further sustained effort is necessary to strengthen the economy in general, and the fiscal accounts in particular. While the Government has been very successful in controlling macroeconomic aggregates, longer term fiscal sustainability will depend on measures to increase revenues and improve the management of public expenditures. The Government has taken a major step by introducing a Value Added Tax (VAT) at a rate. of 17 percent. To further improve the fiscal situation there is need to reduce the burden of state owned enterprise on the budget, and to minimize the fiscal cost of on-going financial sector reforms. 5. Although private investment has increased from 7.8 percent of GDP in 1987/88 to 13 percent in 1995/96, the share of exports in GDP has not increased nearly as rapidly. This is partly due to the anti-export bias in the economy, particularly the high and widely dispersed rates of effective protection resulting from the current system of trade taxes. Further tax reforms aimed at greater trade liberalization will therefore be needed. Such measures will not only encourage exports but, by improving the terms of trade for agriculture, will facilitate the Government's poverty eradication strategy. 6. As part of a long term government policy to enhance political and social harmony, and in support of efforts to increase local democracy, the Government has embarked on a far-reaching decentralization program, under which much of the responsibility for public service delivery has been devolved to local governments (districts and municipalities). Following decentralization of recurrent expenditure, an inter-governmental grant system including block, conditional and equalization grants is being introduced. Fiscal control and the management of public expenditures have become much more challenging in this newly decentralized political and economic environment. Although there is considerable evidence of improved service delivery, recent studies have shown that resources do not always reach the intended recipients. Further reforms will therefore be needed to budgetary and expenditure management systems in order to reflect better the decentralization of service responsibilities, ensure continued sound fiscal management, and address policy objectives under the Poverty Eradication Action Plan. MEDIUM TERM POLICY FRAMEWORK 7. The medium-term policy framework for 1996/97-1998/99 will seek to strengthen poverty reduction, particularly through continued strong economic growth, while maintaining macroeconomic stability in a liberalized environment. The Government's strategy to reduce poverty will seek to combine policies that encourage environmentally sustainable growth with public expenditures that will increasingly focus on the financing of programs in priority areas such a rural roads and water supply, primary education, ANNEX 3 Page 3 of 17 3 basic health services, and agricultural research and extension. This will require a policy framework that encourages export-oriented activity and improved agricultural productivity through the introduction of high value-added crops, and increased provision of financial and technical assistance to farmers. In agriculture, the greatest potential is in cash crops for export. Apart from coffee, cotton, and tea, non-traditional exports such as fish, horticultural products, and fruits and vegetables have a strong potential. There is also substantial scope for increasing yields and efficiency in food crop production, as well as for expanding agro-industry and other manufacturing production and exports. 8. The objectives of the Government's medium-term policy framework are: (i) to sustain annual real GDP growth rates of 7 percent, which encompass all regions and are broadly based, thereby allowing per capita GDP to rise by at least 4 percent a year; (ii) to contain the year-on-year inflation rate to 5 percent; and (iii) to diversify both agricultural and manufacturing exports and reduce dependence on donor flows. 9. The maintenance of this growth path is feasible since: (i) essential infrastructure will be coming on stream as a result of past and current public investment programs; (ii) financial sector reforms currently under way will increase the mobilization of financial savings; and (iii) the public sector reform and privatization programs will improve the quality and utilization of public assets. Moreover, these favorable factors are taking place against a background of continued political stability. However, economic growth could be constrained by infrastructure and institutional bottlenecks, such as power shortages, the congestion of the legal system, ambiguity over land ownership. Another factor is the security situation which has necessitated higher defense spending and has prevented certain regions from realizing their economic potential. 10. Following a marked rise in national savings in 1995/96, a further increase to nearly 17 percent of GDP is expected (in the aftermath of the coffee boom) over the next three years. This will be primarily due to increased public sector savings, which are projected to rise from 1 percent of GDP in 1995/96 to 4 percent in 1998/99 reflecting the continuing efforts to mobilize domestic resources. The current weak state of the financial sector could constrain growth in private financial savings until the benefits of financial sector reforms are realized. On the other hand, the maintenance of macroeconomic stability, and a more conducive investment environment are expected to increase private investment from 13 percent to 16 percent of GDP. This will support higher growth rates as well as continued development of the private sector. THE ECONOMIC REFORM PROGRAM Objectives of the Programme 11. The Background to the Budget 1996/97 and National Development Strategy 1996/97-1998/99 (June 1996), spells out the link between management of the budget, macro stability, economic growth and poverty reduction. Within this context, the economic reform program focuses on long-term fiscal sustainability, involving a more ANNEX 3 Page 4 of 17 4 efficient tax system as well as better management of public expenditures. On the revenue side, the program includes measures to expand the tax base, improve tax administration, and lower the level and dispersion of trade taxes in order to reduce anti-export bias throughout the economy. On the expenditure side, the focus is on improving the management of public expenditure consistent with the on-going decentralization process. Improving the budgetary process is a necessary condition for successful implementation of the Government's Action Plan for Poverty Eradication. In addition, the Program includes measures to: (i) reduce the fiscal burden of state-owned enterprises (SOEs); and (ii) address the fiscal impact of financial sector reforms within the overall macroeconomic framework. Towards Fiscal Sustainability 12. As mentioned above, the Government has been making sustained progress in both increasing revenues and containing the growth expenditure. Nevertheless, in 1995/96, around half of the total expenditure of central government was still covered by foreign aid (grants and loans). In order to reduce aid dependency in the medium term, the Government intends to: (i) continue increasing the share of domestic public revenue in financing the budget in a way which is efficient, transparent and conducive to economic growth; (ii) keep the growth of the overall level of government expenditures in check, so that tax revenues increase more rapidly than government expenditures until the fiscal deficit reaches sustainable levels; and (iii) improve the allocation and the efficiency of utilization of public sector resources. Revenue Mobilization and the Incentive Regime 13. In spite of rapid revenue growth during the last ten years, aggregate revenue as a proportion of GDP is still low by comparison with other African countries. The most substantial contribution to revenue growth has been the rapid expansion of the main tax bases. Increases in some tax rates, in particular excise and petroleum taxes, have also been important. Tax Administration 14. The Government is committed to continuing to increase revenue as a proportion of GDP. This will be achieved primarily through further improvements in tax administration involving: * better monitoring of URA performance enabling targets to be set for genuine tax administration improvements; * clarification of the role of the URA Board so that URA line management has sufficient independence to deliver improved revenue performance with the Board's powers to intervene in the day-to-day management strictly curtailed, particularly with regard to appointments and case work; ANNEX 3 Page 5 of 17 5 * enhancing public perceptions of the Board's integrity by reviewing and setting appropriate safeguards for Board members to declare their assets throughout their tenure and to declare any interest for example in tender discussions; * reviewing and strengthening existing procedures for investigating allegations of corruption; * submitting draft legislation to Parliament for the establishment of an independent mechanism for the resolution of disputes in tax assessments and pressing for this mechanism to be operational by FY97/98; and * strengthening the capacity of the Tax Policy Department in the Ministry of Finance (MOF) to: (i) develop medium term revenue strategies that take into account wider economic considerations, and (ii) set appropriate targets for URA performance. Elimination of Tax Exemptions 15. During the past year, Government has made considerable progress in curtailing exemptions from import duties by codifying discretionary exemptions under the Investment Code into the Customs Code, and by cutting exemptions for a wide variety of institutions and organizations and replacing them by ministerial votes that are limited and explicit. Periodic reports on such expenditure are now presented to Cabinet by the Minister of Finance. In addition, all Government, parastatal and donor construction projects have now to be tendered on a tax inclusive basis. As a result of these measures, revenue forgone through exemptions has dropped from an estimated USh 45 billion in 1993/94 to around USh 9 billion in 1995/96, and the MOF Vote to cover tax expenditures where there is no prior agreement should be no more than USh I billion in FY96/97. The Government will not approve new discretionary tax or duty exemptions throughout the Program. Local Taxation 16. Decentralization is a central element in the 1995 Constitution. The long term success of decentralization will depend, to a large extent, on getting local governments to raise their own revenues, and increasing the accountability of local governments to their electorates. In conformity with the Constitution and in consultation with local authorities, Government intends to further increase the revenue raising power of local governments by raising the revenue potential of the Graduated Personal Tax (GPT). This could be done by: (i) converting the GPT to a flat rate tax, which will simplify its administration and collection; and (ii) raising the GPT ceiling up to the floor of the income tax, which will reduce its regressiveness. In addition, collection of GPT by withholding could be merged with the central government income tax collection apparatus and the revenues distributed to the appropriate local government. This will require greater information sharing between URA and local authorities and the computerization the two rolls. In addition, license fees, market dues and other charges will be expanded to cover the cost of local services being provided to businesses. ANNEX 3 Page 6 of 17 6 17. The new grant system will be consistent with this revenue reform. As the higher GPT ceiling increases revenue for KCC and other urban councils, councils could become less dependent on the grant system. To further the decentralization process, the Government is providing block grants that encourage fiscal self-sufficiency and will introduce equalization grants by the start of FY98/99. Reform of Trade Taxes and Reduction in the Anti-Export Bias 18. While Uganda's external trade taxes have been substantially reduced since the late-1980s, further work needs to be done on trade liberalization. In particular, the average effective protection of industries producing for the domestic market is over 90 percent with actual protection rates varying widely between industries and firms. As the 1996/97 Background to the Budget points out, Government recognizes that high taxation of imports is effectively a worsening of the terms of trade for exporters and that the immediate costs of protectionism are in a reduction in real incomes of the whole population. Further, our Action Plan for Poverty Eradication states that protectionist measures, which create a limited number of jobs in urban areas at the cost of hurting rural people (who are in general poorer than the beneficiaries of protectionism) must be phased out. Thus, Government intends to gradually reduce the spread and level of import taxes and to end the differential application of other taxes/rates to imports and domestic goods. 19. Measures to be proposed in the FY97/98 budget include: (i) reducing the maximum import duty to 20 percent (July 1997); (ii) abolishing all remaining non-tariff barriers (NTBs) (March 1998), with the exception of cigarettes; (iii) replacing these bans by temporary import surcharges (batteries 15%, soft drinks 15% and beer 30%), which will be gradually phased out over a period of three years; and (iv) reduce the maximum rate of excise taxes (other than on beer, soft drinks and cigarettes) to 10% (July 1997). In the FY98/99 budget, the Government will propose: (i) replacing the import ban on cigarettes by a temporary import surcharge which will be gradually phased out in a period of two years; (ii) reducing the maximum import duty to 15 percent (July 1998); (iii) establishing a minimum import duty in the range of 7-8 percent (July 1998); and (iv) abolish all excise taxes (other than on beer, soft drinks and cigarettes) that apply differentially to imports and domestic goods (July 1998). In FY99/2000, import duties applied to non-COMESA trade will be consolidated into a low, single uniform rate. Determination of this rate will take into account revenue considerations. The above trade liberalization program will be announced in its entirety in the FY97/98 budget. To complement these reforms and to facilitate export growth, a simple duty drawback would be implemented to reimburse exporters for import duties paid on their inputs, and export documentation will be streamlined. The Government will pursue a more pro-active role in its regional trade policy in order to allow a greater degree of liberalization in the trading of agricultural products and more efficient transit transport arrangements ANNEX 3 Page 7 of 17 7 Improving the Management and Efficiency of Public Expenditure 20. Since 1992, substantial progress has been made in the management of public expenditures. The monthly cash release system, for example, was instrumental in attaining aggregate fiscal discipline. In addition, the Government has taken steps to improve expenditure prioritization by instituting a medium-term Budget Framework Paper (BFP), with increased funding to Priority Program Areas (PPAs) in key sectors. It has also sought to improve efficiency by reducing the size of an overstaffed civil service and raising civil service wages. 21. In order to implement its poverty reduction strategy and improve efficiency in the utilization of public resources, the Government will undertake strategic reforms to the management of public expenditures. These reforms will aim at: (i) improving fiscal discipline and protecting budget allocations by limiting supplementary budgetary provisions; (ii) introducing a stronger outcome-orientation to the budget process; (iii) increasing the poverty focus of public spending; (iv) strengthening public expenditure management under decentralization; and (v) improving accountability and transparency. Appropriate incentives in the civil service will be critical to the effective implementation of these measures. Limiting Supplementaries 22. While Government has been very effective in staying within its aggregate fiscal limits, substantial supplementaries have been incurred during budget execution. This creates two fundamental problems: (i) the budgeted priorities and the budgetary process in general are undermined; and (ii) agencies face tremendous uncertainty about the flow of resources, which undermines the technical efficiency of their programs. In order to address these problems, the Government will limit the total amount of supplementaries (approved and/or actually spent) to 5 percent of Government funded expenditures in FY96/97 and 3 percent thereafter, except for national emergencies. Outcome-Oriented Budgetary Process 23. Currently, Ministries budget in an incremental way, and there is a lack of consistency between the allocation of resources and government priorities. Recent inter- governmental reforms have reduced sectoral perspectives in the budgeting process. There is also a lack of budget transparency since the development budget does not clearly identify the substantial amounts of donor funding being provided to meet recurrent expenditures and to support district-based programs. In the last two years the Government has sought to address these issues through the introduction of a sectoral expenditure review covering both central and local government budget as part of the BFP exercise. This initiative needs to be developed further and embedded into the budget cycle. 24. In order to address these issues and introduce a stronger outcome orientation to the budget process, the Government will design and implement an Action Plan to carry ANNEX 3 Page 8 of 17 8 out an integrated sectoral expenditure programming exercise in the health, education and agriculture. This exercise, which will be led by the line ministries, will include the following elements: * Introducing a strong outcome focus to the budget by emphasising the objectives, activities and intended outcomes of sectoral expenditure programmes. This will be backed up by setting and monitoring of service delivery targets for key areas of public expenditure. * Restructuring sector expenditure programmes consistent with a strategic sectoral approach and with the decentralisation process. * Preparing, under a hard budget constraint, an "integrated sector expenditure programme" for each pilot sector which combines recurrent and development budget at programme level and distinguishes between central government and district-level expenditures. * Strengthening the BFP as an expenditure planning tool by setting resource ceilings at the programme level to provide a strategic framework for the subsequent annual budget. * Achieving a more efficient input mix by allowing greater flexibility in the budgeting of line items within budget programmes. In particular, increase the incentives for efficiency and the reduction of overheads by allowing line Ministries to adjust their staff complement within their wage ceilings at their discretion, and to reduce staff levels and costs below that ceiling without losing the resulting savings. 25. Implementation of these measures will be coordinated by the MOF and will be linked to the 1998/99 BFP and the preparation of the FY98/99 Budget. The exercise to develop integrated budgets for the pilot sectors will be completed by February 1998 and will provide the basis for the preparation of the line ministry submission for the sector expenditure component of the BFP exercise. The FY98/99 Budget for the three pilot sectors will reflect the conclusions of the BFP sector expenditure review. 26. The Government also intends to introduce more comprehensive reforms to its budgetary systems and procedures including: (i) integration of the recurrent and development budgets; (ii) budget classification and itemization; (iii) computerization; (iv) streamlining of budget release and implementation procedures; and (v) improved implementation monitoring. An Action Plan for the reform of budgetary systems and procedures will have been prepared by September 1997. Anti-Poverty Focus 27. Despite robust economic growth over the last decade, rural poverty remains severe and wide-spread. The improvements in economic indicators have not yet spread across Uganda, nor have they been matched by equivalent improvement in social indicators. Poverty eradication is therefore the highest national priority. Given that 85 ANNEX 3 Page 9 of 17 9 percent of Ugandans are self-employed, the Government continues to focus on improving incentives for farm and non-farm activities. The most important measures have been liberalization of domestic and export markets, and the exchange rate. In addition, Government has taken measures to improve the environment for private sector investment and is implementing reforms to improve the soundness of the financial system. Also, further trade liberalization will improve the terms of trade for agriculture and ensure that the assets of the poor are not unduly taxed through protection. 28. The Government is currently refocusing its policies and actions to ensure that the poor are able to participate in and benefit from economic growth more effectively. Its strategy is articulated in the Action Plan for Poverty Eradication, which is outlined in the Background to the Budget 1996-97. The Action. Plan states that the best way to remove people from poverty is not to give them handouts, but to enable households to earn decent incomes. In order to raise incomes, people need the following: (i) secure access to productive assets, particularly land; (ii) macroeconomic stability, including fiscal stability, low inflation and predictable policy environment; (iii) infrastructure, particularly roads which give access to markets; (iv) information about how to use resources efficiently; and (v) adequate levels of health and education to enable the poor to work and use the information at their disposal effectively. 29. Improving the management and efficiency of public expenditures is essential for implementing this strategy. It will support Government's effort to improve service delivery to the poor which -- as shown clearly by the recent pilot survey of nine districts - - suffers from poor quality and lack of transparency. Priority areas for service delivery are primary health care and education, agricultural extension, road maintenance and demand-driven water development. These Strategic Areas (SAs) are defined in the 1996/97 budget as a focal point for prioritization of public expenditures. To strengthen the anti-poverty focus of the budget, Government will increase recurrent budget allocations to agriculture research and extension, to primary health and to primary education at least as much as the rate of growth of nominal GDP, and it will implement the outcome oriented budgetary process described above to improve the efficiency of these expenditures. The Government will also incorporate the main recommendations of the anti-poverty Action Plan into the BFP exercise for the FY97/98 budget, and it will ensure adequate budgeting for PIP projects and that all allocated funds are fully released. Expenditure Management Under Decentralization 30. There are several aspects in which the present budgetary process is inconsistent with the decentralization program that the Government has initiated. In particular, the division of responsibilities in the new Constitution specifies that the central government will be responsible primarily for policy formulation and monitoring, maintaining implementation responsibilities only for national public goods, such as defense and trunk roads. However, substantial budgetary resources are still being allocated to central government ministries for functions that have been decentralized. Within the context of budgetary reform, local government budgets will also have to become outcome-oriented. ANNEX 3 Page 10 of 17 10 The budget processes of the two levels of government will be synchronized beginning in FY97/98, in order to facilitate the budgetary preparation process and be able to consolidate government accounts. 31. Given the extent of decentralization, substantial resources will have to be transferred from the central to district governments through conditional and equalization grants. Pilot ministries will work closely with the Districts in the preparation of sectoral budgets for programs of national priorities. These budgets will be based on district-level plans for achieving agreed upon outputs and outcomes and on costing of programs. Jointly with the Districts, the pilot ministries will design the rules and mechanisms for the operation of conditional grants, including appropriate monitoring requirements. These resource shifts and procedures will be implemented in FY98/ 99 budget. 32. In accordance with the decentralization program, Government will: (i) ensure that there are no constraints in the use of unconditional block grants and de-link their allocation from district personnel establishment positions as of FY97/98; (ii) modify the rules for conditional grants, develop the rules for equalization grants, and adopt these rules in the FY98/99 budget; and (iii) include these rules and regulations for districts under the Public Finance Act. Accountability and Transparency 33. Preliminary results from a recent study on tracking of public expenditures indicate that public spending is not reaching the primary education and health facilities as intended. The following bottlenecks were identified: (i) substantial delays in submitting and auditing district final accounts and monthly expenditure reports prevent any corrective measures when resources are diverted or misused; (ii) many Local Government Tender Boards (LGTB) lacked integrity, resulting in substantial overpricing of contracts and supplies; and (iii) budgetary resources are not always distributed to departments according to the approved work plans with adverse consequences for service delivery. 34. By and large, necessary accountability mechanisms such as the Local Government (Financial) Regulations (1995) are in place. The issue is implementation and increased transparency. To address these problems, the Government will: (i) introduce a uniform system of budgeting, monitoring and accounting for all local governments by FY97/98; (ii) publish monthly cash releases of block grants and conditional grants to all districts in the main local newspapers beginning immediately; (iii) ensure publication of a monthly summary of all tenders awarded by the LGTB, with copies sent to the Inspector-General of the Government as proposed under the new Local Government Act; (iv) deconcentration of the activities of the Auditor-General's office to the district level by FY97/98 to enable timely and credible auditing of district accounts; (v) amend during FY97/98 the Local Government Financial Regulations (1995) to ensure that advances can be made only for developmental purposes; and (vi) ensure that appropriate disclosure, including district budgets, program outcomes and tender awards, is required of districts, and that commensurate sanctions be applied for non-compliance. ANNEX 3 Page 11 of 17 11 35. The Government will ensure: (i) compliance of the provisions of the Local Government Act in three randomly selected districts among the 13 districts included in the first phase of fiscal decentralization by FY98/99; and (ii) design and implement an Action Plan for improving financial accounting and auditing systems for central and local governments by FY97/98. Reducing Subsidies to State Owned Enterprises 36. Most Ugandan state-owned enterprises (SOEs) are inefficient and represent a serious constraint to the development of the private sector. Inefficiencies in the SOE sector also have a significant fiscal impact. For example, indirect and direct subsidies amounted to USh 208 billion in 1994 (nearly 5 percent of GDP), which was more than half of the entire recurrent budget. The drain on the budget will be addressed by: (i) imposing strict financial controls on all SOEs, eliminating all direct and indirect subsidies to commercial SOEs and phasing out subsidies to non-commercial SOEs over a four-year period (except for subsidies relating to social missions); and (ii) privatizing 85 percent of all SOEs by the end of 1997. At the same time sector reform and privatization will be pursued in the power, telecommunications, water and transport sectors with the objective of improving the delivery of key services to the economy. 37. The Government will implement the Action Plan prepared by the Parastatal Monitoring Unit (PMU) to achieve a 20 percent reduction in overall subsidies to the SOE sector in 1996/97 compared with 1994/95 (excluding subsidies to and from the Uganda Commercial Bank and the Uganda Development Bank which will be dealt with under a separate component of the Program) and to continue subsidy reduction in the following years. The reduction will be monitored through the annual report on "Financial Flows Between Government and Public Enterprises and Amongst Public Enterprises". The Fiscal Cost of Financial Sector Reform Uganda Commercial Bank (UCB) 38. Under the policy agreed with IDA under the Financial Sector Adjustment Credit, the Memorandum of Offering to sell UCB was issued in December 1996. The Government intends to sell UCB to private investors subject to the following conditions: * The strategic buyer is a reputable investor with recognised capacity in managing financial institutions and will provide the capital required for UCB to meet Bank of Uganda (BOU) capital requirements, and undertake the management of UCB in a manner acceptable to BOU. * The strategic buyer, by himself or the strategic buyer together with other private sector investors as a group, will buy 100 percent of the bank shares. If no outright sale is possible, the Government will also consider sale to investors proposing to purchase more than 51 percent in the bank's capital. ANNEX 3 Page 12 of 17 12 * The Government supported the UCB sale by covering the negative capital of the bank as of September 1996 with government securities. It will also provide guaranties against contingent liabilities, if so demanded, as long as they are reasonable in type and amount. Such support will be undertaken in consultation with IDA. 39. In case no private investor acceptable to the Government and BOU is found, or if the demands of potential buyers for government compensation and guaranties are not acceptable, the Government will, by June 1997, initiate an alternative plan for the restructuring of UCB. This alternative plan will have the following elements: * The management of UCB will be contracted to a reputable international banking group. The management colntract will be for three years and will provide the contracted manager with full independence in the conduct of business. * Upon handing over management, the Government will provide to UCB sufficient government securities to cover the bank's minimum capital requirements and including any redundancy payments, not provided for, but required to be paid under Uganda's employment legislation. * The management contract will aim at reducing UCB staff numbers and its branch network (exept for a list of 18 rural branches identified by the Government as essential for the provision of banking services in remote districts) to a level that will secure efficient and profitable operations. In order to maximize the range of options available to Government, the management contract will provide a fee structure based on achieving both profitability and downsizing within the three year duration of the contract. * It is the Government's intention to review its policy regarding UCB at the end of the management contract period. At that time a new effort to sell the restructured bank will be pursued. However, if the efforts to sell the bank at that time are unsuccessful, the Government, in consultation with IDA, will decide on future policy. Essential Financial Services in Rural Areas 40. If successful sale is achieved, the Government will allow the new owners to decide which of the existing branches they wish to operate and which they would rather close. However, a list of 18 rural branches of UCB was identified by a recent study as essential for the provision of banking services in remote districts where UCB is the only operating bank branch. If the new owners of UCB decide to close any of these identified branches, it is the Government's intention to secure their continued operations with a budgetary subsidy. This will be done through auctioning the management for each branch separately among financial institutions licensed in Uganda. The Government will prepare a list of essential services to be provided by these branches and will sign a two year management contract with the bidder demanding the lowest subsidy to operate each branch subject to it being judged by BOU as capable of managing the branch. If there are ANNEX 3 Page 13 of 17 13 no bidders for a specific branch, BOU will approach commercial banks to find a bank to undertake the management of the branch. At the end of the management contract, the branches will be offered for sale, with first refusal being given to the managing bank. If no buyer is found, the need to keep each branch open will be reviewed and if considered essential, new two year contracts will be auctioned for the remaining unsold and essential branches. Non-Performing Assets Recovery Trust (NPART) 41. The Government places great value in the success of NPART, in mitigating the budgetary cost of recapitalizing UCB and in providing a clear signal concerning the need to repay loans and meet financial and other obligations. The Government will therefore support NPART activities in collecting UCB's non-performing loan portfolio. It is the intention to achieve collections by NPART of at least USh 25 billion. 42. Government will make its best efforts to ensure that NPART collects outstanding debts. In particular, it will submit to Parliament an amendment to the NPART Act, to cover the following issues: (i) extending the mandate of NPART for at least an additional year to compensate for the delay in starting its operations; and (ii) limiting grounds for appeal to points of law only. Treatment of Problem Banks 43. It is the Government's intention to prepare, in close cooperation with BOU, a clear policy statement on the future treatment of problem banks that addresses the need to secure the health of the financial system while limiting to the extent possible the cost of such actions to the Government budget or to BOU. A draft policy memorandum is attached to this letter. It is the Government's intention to finalize the policy statement in collaboration with BOU and in consultation with IDA and to publish the policy statement by May 1997. Uganda Development Bank (UDB) 44. The Government recognizes the heavy budgetary burden placed on it by the continued losses of UDB, while the continuation of its activities provide no benefits to the economy. It is the Government's intention to adopt and implement a clear policy with regard to UDB before the end of FY96/97, in consultation with IDA. This policy will aim at minimizing the Government budgetary costs resulting from UDB conduct of business, as presented in the recent report prepared by DFC (which was commissioned by PMU). The Government will review the recommendations of the restructuring task force and decide on UDB's future in consultation with IDA. The collection of UDB portfolio could be added to the tasks of NPART, through an amendment to the NPART Act to be submitted to Parliament. ANNEX 3 Page 14 of 17 14 Recapitalization of Bank of Uganda (BOU) 45. It is the Government's intention to capitalize BOU to a level acceptable for successful operations of BOU, both in Uganda and vis-h-vis international and foreign financial institutions. The Government took the initial action of placing with BOU a total of USh 60 billion of government bonds against BOU shares. A study by a reputable international auditing firm was finalized recently and established the BOU capital discrepancy. The study results have been endorsed by both the Government and the BOU. Agreement will be sought with BOU on the ways to bring it to the desired level of capitalization (by issuing additional government bonds and/or by changing the terms and conditions of the Government debt to BOU). The Government and the Bank of Uganda will review the auditors' report and agree on what further restructuring is required at the BOU so as to accelerate the recapitalization process by end-June 1997. CONCLUSION 46. Over the past 10 years, the Government of Uganda has implemented a program of far-reaching macroeconomic and structural reforms with the aim of attaining sustainable economic growth and poverty reduction. We are strongly committed to continuing and accelerating this process under the Structural Adjustment Program. It is our intention to expedite the program and take whatever complementary measures necessary to ensure its success. The Government views IDA support as an important input to this effort. Yours sinc ely, J.S. Mayapja-Nkangi Minister of Finance ANNEX 3 Page 15 of 17 15 APPENDIX INTERVENTION POLICY BY GOU AND BOU In recent years the financial system in Uganda suffered severe problems from the poor performance of several major banks. The actions taken included: (i) Placing UCB non-performing portfolio with a specially created NPART; providing UCB with foreign technical assistance to improve management; preparing UCB for privatization, with the assistance of an international investment bank; re-capitalizing UCB with GOU bonds; and, evaluating the consequences of UCB privatization on providing banking services in remote districts. (ii) Intervening in 2 private banks, Sembule and Nile; and facilitating their sale to new strategic investors with BOU loans that covered negative capital positions. (iii) Monitoring support provided by donors or owners in three other problem banks (USAID supporting COOP Bank, European Catholic organizations supporting Centenary Bank, and the Libyan government committing to replenish Tropical Africa Bank capital). (iv) Enforcing a new minimum capital requirement for banks which was reached by December 31, 1996. (v) Declaring a moratorium on the licensing of new banks, except for banks with special benefits to the Uganda financial sector. (vi) Strengthening the capacity and skills of the BOU supervision department. (vii) creating a deposit insurance scheme to protect depositors The deposit insurance fund collects an annual premium from all banks and received an initial grant from GOU. The GOU and BOU are determined that future problems in commercial banks will be treated in a way that will: (i) provide for a healthy financial sector; (ii) take action if and when necessary immediately when a problem is identified; and (iii) minimize the budgetary costs of future such intervention. ANNEX 3 Page 16 of 17 16 To reach these goals the following policies will be adopted and implemented: * The moratorium on the licensing of new banks (except when special benefits are expected, as in the case of the entry of a major international banking group) will be initially extended for 2 additional years, at the end of which the Government and BOU will review the need for a further extension. Such extensions will prevent the proliferation of small, undercapitalized banks, and ease the supervision and monitoring of commercial banks. * It is the Government policy to require banks to have sufficient capital in order to strengthen the financial system. The Government will submit to Parliament an amendment to the Financial Institutions Act to authorize the Governor of BOU, in consultation with the Minister of Finance, to update the minimum capital requirements. It is the Government's intention to reach, at a minimum, a doubling of the minimum capital requirement by end- 1999. * BOU will continue to strengthen its supervision and monitoring capacity, and to implement all internationally accepted standards for banks' financial reporting and financial position. Special effort will be undertaken to direct the banks' external auditors to follow and to certify the audited banks treatment of their portfolio classification, the adequacy of their provisioning for bad-debt, and the suspension of all interest in arrears. * A public declaration will be made stating that in the future, banks that do not meet the BOU standards will be intervened and, where a bank will be liquidated, the GOU and BOU commitment to depositors will be limited to the U Sh 3 million per depositor covered under deposit insurance scheme. * A contingent intervention plan will be prepared and followed when needed. This plan relates to intervention due to deterioration of a bank's financial position. Other types of intervention may be temporary, with the bank remaining open or closed at the discretion of BOU, and with the bank returned to the management of its owners once the purpose of the intervention is achieved Such contingent plan will include the following elements: (i) Intervention will take place if a bank fails to meet BOU requirements and to agree with BOU on corrective measures to be taken within an agreed time- table, not to exceed 6 months. Failure to meet interim targets during such period will justify immediate intervention. (ii) Once intervention takes place, the intervened bank will be closed and remain closed while an auditing firm nominated by BOU reviews the bank financial position and identifies its assets and its liabilities. (iii) The list of deposits eligible for protection under the deposit insurance scheme, excluding depositors that have debts to the bank, shareholders (other than holders of a small number of shares with no influence on the ANNEX 3 Page 17 of 17 17 bank conduct of business), the bank directors and officials, will be established and steps taken to repay depositors their protected deposits within 90 days from intervention. (iv) The collection of all debt and sale of assets will be undertaken by a liquidator that will then repay residual amounts of deposits and other liabilities (according to the priorities decreed by Uganda laws) of the intervened bank in instalments when the revenue from the sale of assets and collections of loans will provide enough funds, after all funds provided from the deposits insurance scheme are repaid. (v) If the liquidator will find that the policy of selling the bank as a going concern to new owners is feasible arid will result in better financial outcome, the bank will be sold and opened for business under the management of the new owners, once the sale and the new owners is approved by BOU. (vi) Prompt legal action will be taken against the intervened bank major shareholders, directors, officials and auditors for any fraudulent actions for which they are responsible.  Uganda - Third Structural Adjustment Credit (SAC III) - Policy Matrix OBJECTIVE ACTIONS ALREADY TAKEN RECOMMENDED MEASURES TIMETABLE MACROECONOMIC FRAMEWORK 1. Maintain macroeconomic Government has increased current revenues from Stabilization program on track. Throughout the program stability. 7.6% of GDP in 1992/93 to 11.4% in 1995/96, and has reduced the fiscal deficit (exc. grants) from 11.8% to 6.4% GDP in the same period. The inflation rate has been in single digits in the last 4 years. REVENUE MOBILIZATION AND EXPORT INCENTIVE REGIME 1. Expand the tax base and Government has made substantial progress in Submit draft legislation to Parliament for the Board Presentation improve tax administration. curtailing discretionary exceptions under the establishment of an independent tax appeals investment code and Statutory Instruments. system. Submit draft legislation to clarify the role of Second Tranche URA Board and set appropriate safeguards. 2. Reduce Anti-Export Bias by Trade liberalization to date has reduced the Announce tariff reform plan and timetable. Effectiveness harmonizing and reducing impor maximum tariff to 30 percent. A 5-percent duty tariffs. rate was introduced for previously exempted raw Reduce maximum import duty to 20%; end Second Tranche materials in July 1995. PTA trade has a 80 differential use of other taxes/rates for imports percent preference over inter-regional trade since and domestic goods; and abolish remaining July 1996. NTBs (except tobacco). Implement a low uniform duty rate on all non- Floating Tranche COMESA imports. PUBLIC EXPENDITURE MANAGEMENT 1. Strengthening Fiscal In 1992 Government established a committee Limit supplementaries to no more than 5 percent Throughout the Program I Management. responsible for ensuring that monthly of Government funded expenditures in FY96/97 0 expenditure releases are consistent with cash and no more than 3 percent thereafter, except for flow and overall macroeconomic indicators. national emergencies. Uganda - Third Structural Adjustment Credit (SAC III) - Policy Matrix OBJECTIVE ACTIONS ALREADY TAKEN RECOMMENDED MEASURES TIMETABLE 2. Adopting an outcome- Introduced 3-year Budget Framework Paper. Adopt rules for conditional grants following Board Presentation oriented budgetary process Government has started design of Results constitutional criteria; establish monitoring consistent with the Oriented Management (ROM). programs; and delink block grants from District decentralization program. Study underway to track public expenditure in personnel establishment positions. health and education. First phase of fiscal decentralization initiated in 1993/94; all districts Regulate borrowing by Districts. Board Presentation are in the unconditional (former block) grant system by 1996/97. Implement integrated sector programming and Second Tranche budgeting exercise for three pilot sectors (Education, Health and Agriculture) in conjunction with Districts. 3. Enhancing the anti-poverty Since FY93, GOU has been protecting from cuts Increase budgetary allocations for agriculture Second Tranche focus of the budget. non-wage recurrent expenditure in social sectors research and extension, primary health and and other priority program areas. primary education at least at the rate of increase Government is implementing long-term program of nominal GDP. Ensure adequate budgeting for road maintenance. for PIP projects and ensure that funds are fully released. 4. Improving governance and Financial accounting and auditing systems and Prepare timetable for auditing of local Board Presentation accountability. regulations for central and local government are governments. mostly in place but their implementation is lagging. Commence implementation of action plan to Second Tranche improve financial accounting and auditing for central and local government. Government has started publishing monthly Timely publication and dissemination of: (i). Second Tranche cash releases to districts in the main newspapers. district budgets and Local Government Public Accounts Committee Reports; and (ii) audits of district accounts on completion. Uganda - Third Structural Adjustment Credit (SAC III) - Policy Matrix OBJECTIVE ACTIONS ALREADY TAKEN RECOMMENDED MEASURES TIMETABLE PARASTATALS 1. Limit para-fiscal Establishment of a Parastatal Monitoring Unit Submission of PMU report on financial flows Board Presentation expenditure. (PMU) within the Ministry of Finance and to/from SOEs for 94/95. Economic Planning, which completed database on financial flows to/from SOEs for FY92-95. Achieve 20% reduction in overall subsidies to Second Tranche Privatized 41 SOEs as of June 1996. the SOE sector in 1996/97 compared with 1994/95 (excluding subsidies to and from UCB and UDB). FINANCIAL SECTOR 1. To enhance the soundness of Merchant bank has completed due diligence and Complete sale of UCB, or implement back-up Second Tranche the banking system within a prepared a divestiture action plan for sale of plan agreed with IDA. sustainable macroeconomic UCB. Study of implications of UCB sale on framework. rural finance has been completed. 2. To enhance financial As of September 30, 1996, NPART has NPART to collect at least USh 25 billion in non- Floating Tranche discipline and to create an recovered approximately USh 6 billion in non- performing loans. appropriate policy and performing loans. regulatory framework. Implement policy regarding UDB to stop losses Second Tranche and minimize budgetary costs in agreement with IDA. Finalize and publish a policy paper on future Board Presentation bank intervention agreed with BOU and acceptable to IDA.  ANNEX 5 Page 1 of 9 Uganda at a glance 4/24/97 Sub- POVERTY and SOCIAL Saharan Low- - - Uganda Africa income Development diamond* Population mid-1995 (millions) 19.2 583 3,180 GNP per capita 1995 (US$) 240 490 430 Life expectancy GNP 1995 (billions US$) 4.7 288 1.3B2 Average annual growth, 1990-96 Population (%) 3.2 2.6 1 7 GNP Gross Labor force (%) 2.7 2.6 1.7 per - pnmary Most recent estimate (latest year available since 1989) capita enrollment Poverty: headcount index (% of population) Urban population (% of total population) 13 31 29 Life expectancy at birth (years) 42 52 63 Infant mortality (per 1,000 live births) 98 92 69 Accesa to safe water Child malnutrition (%A of children under 5) 23 34 41 Access to safe water (% of population) 42 47 53 JIlliteracy (% of population age 15+) 38 43 34 Uganda Gross primary enrollment (% of school-age population) 91 72 105 Male 99 78 112 -Low-income group Female 83 65 98 KEY ECONOMIC RATIOS and LONG-TERM TRENDS 1975 1985 1995 1996 Economic ratios* GDP (billions US$) . 3.5 57 6.0 Gross domestic investment/GDP 7.6 8.7 16.4 16.7 Openness of economy Exports of goods and services/GDP 8.2 13.7 12.0 12.1 Gross domestic savings/GDP 5.5 7.5 7.4 6.3 Gross national savings/GDP 5.9 7.5 12.2 12.3 Current account balance/GDP .. -2.5 -7.9 -8.6 Interest payments/GDP .. 0.5 0.6 0.7 Savings - - Investment Total debt/GDP (DRS Data) .. 35.0 63.0 57.4 Total debt servicelexports 6.6 38.1 25.2 22.4 Present value of debt/GDP .. .. 30.0 28.8 Present value of debtlexports (current yr) . 254.7 238.0 Indebtedness 1975-85 1986-98 1995 1996 1997-05 (average annual growth) Uganda GDP .. 6.7 11.5 9.8 5.7 GNP per capita .. 3.4 8.6 7.2 Low-income group Exports of goods and services .. 8.2 27.4 30.S 6.6 STRUCTURE of the ECONOMY 1975 1985 1996 1996 (% of GDP at Factor Cost) Growth rates of output and Investment (%) Agriculture 722 527 497 45.6 40 Industry 8.2 9.9 14.0 16.0 30 Manufacturing 6.4 5.8 6.5 7.2 20 Services 196 37.4 36.2 38.4 in 0 Private consumption .. 78.0 84.3 84.3 -10 91 92 93 94 95 9 General government consumption .. 14.5 9.6 9.8 GDI GDP Imports of goods and services 10.3 15.0 21.0 22.4 1976-86 1986-96 1995 1996 (average annual growth) Growth rates of exports and imports (%) Agriculture .. 4.2 5.9 4.2 s0 Industry . 105 20.3 197 Manufacturing .. 11 169 18.1 25 Services . 76 13.1 94 Private consumption 6.1 13.4 77 o --- General government consumption .. 5.6 9.2 86 92 4 5 5e Gross domestic investment .. 7.9 39.7 14.7 -25 Imports of goods and services .. 5.4 43.6 4.9 -Epots Imports Gross national product .. 68 11.9 10.3 Note: 1996 data are preliminary estimates. The diamonds show four key indicators in the country (in bold) compared with its income-group average. If data are missing, the diamond will be incomplete. ANNEX 5 Page 2 of 9 Uganda PRICES and GOVERNMENT FINANCE 1975 1985 1995 1996 Domestic prices inflation (%) (% change) Consumer prices, Kampala CPI .. 102.5 6.5 8.6 Implicit GDP deflator .. 1203 8.3 4.9 25 Goverment finance (% of GDP) o - Current revenue - 9.1 9.7 10.2 91 92 93 94 95 96 Current budget balance .. 03 0.3 0.4 - GDP def -0-- CPI Overall surplus/deficit . -4.3 -7.6 -6.0 TRADE 1974i 1985 1995 1996 (milhons, US$) Export and import levels (mill. US$) Total exports (fob) . 383 595 590 1,so0 Coffee 353 457 404 Cotton .. 3 13 Manufactures I.DO0 Total imports (cif) 404 1,086 1,218 Food . Fuel and energy 76 64 90 5 Capital goods .. i Export price index (1987=100) 103 95 66 90 91 92 93 94 95 96 Import price index (1987=100) 0 136 140 E :l Imports Terms of trade (1987=100) 129 70 47 BALANCE of PAYMENTS 1975 1985 1995 1996 - (milions USS) Current account balance to GDP ratio (%) Exports of goods and services 248 408 667 726 - Imports of goods and services 313 484 1,383 1,601 92 93 94 9 Resource balance -65 -76 -716 -875 Net factor income -4 -53 -58 -46 Net current transfers 13 40 330 407 Current account balance, L before official capital transfers -56 -89 -444 -514 Financing items (net) 56 128 613 606 Changes in net reserves 0 -39 -169 -92 -5 Memo: Reserves including gold (mill. USS) 31 85 388 480 Conversion rate (locaIUVS$) .. 5.1 932.5 1,012.0 EXTERNAL DEBT and RESOURCE FLOWS From DRS Data 1975 1985 1995 1996 (millions US$) Composition of total debt, 1996 (mill. US$) Total debt outstanding and disbursed 209 1,232 3,564 3,445 IBRD 4 37 0 0 F IDA 40 285 1,792 1,952 16 43 E Total debt service 17 155 137 167 620 IBRD 1 3 12 0 IDA 0 4 21 25 Composition of net resource flows 447 B Official grants 30 47 399 D T1aO2 Official creditors 11 140 171 201 Private creditors -3 6 -10 -7 Foreign direct investment 2 -4 5 33,7 Portfolio equity 0 0 0 World Bank program Commitments 0 45 94 0 A - IORD E - Bilateral Disbursements 3 92 160 171 -IDA D5-2Otermultlateral F Pdvate Principal repayments 1 2 20 11 C -IMF G - Short-term NTtflows 2 90 140 160 620 Interest payments 1 4 13 14 Net transfers 2 86 126 145 International Economics Department 4/24/97 ANNEX 5 Page 3 of 9 Table 1: Key Economic Indicators - HIPCs/PFP Case 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 Growth Rates: GDP at Factor Cost 5.2 3.1 8.4 5.3 10.6 8.5 7.0 7.0 7.0 Gross Domestic Income 4.7 2.6 7.9 8.3 17.2 7.8 4.3 7.4 7.2 GDY per Capita 1.7 -0.3 4.0 4.9 13.7 4.8 1.4 4.5 4.3 DOD, incl IMF and Arrears: DOD, US$ million 2,592 2,648 2,637 2,992 3,388 3,448 3,432 3,651 3,884 DOD/Exports G&S, percent 1283.0 1333.1 1090.1 869.9 496.3 461.5 441.5 438.8 425.3 DOD/GDP at MP, percent 78.0 92.7 81.9 75.2 59.9 57.4 52.6 49.4 46.4 Debt Service Obligations: Debt Service, US$ million 191 249 175 179 172 167 181 185 192 Debt Service/Exports G&S, percent 94.4 125.3 72.5 52.1 25.2 22.4 23.2 22.3 21.0 Debt Service/GDP at MP 5.7 8.7 5.4 4.5 3.0 2.8 2.8 2.5 2.3 Ratios to GDP at MP, Current Prices: Gross Domestic Investment 15.2 15.9 15.2 14.7 16.4 16.7 18.5 19.2 19.8 Private Consumption 92.0 90.6 89.1 85.4 84.3 84.8 82.2 81.1 80.3 Gross Domestic Savings 0.7 0.4 1.1 4.1 7.4 6.3 4.2 5.3 6.3 Gross National Savings 1.4 2.1 3.6 10.4 12.2 12.3 10.2 11.1 12.3 Government Revenue 1/ 7.5 6.8 7.3 8.3 10.0 10.3 11.0 11.9 12.6 Government Expenditure 14.7 21.2 18.6 18.7 17.6 16.3 16.2 16.1 15.8 Overall Deficit, excl Grants -7.2 -14.4 -11.3 -10.4 -7.6 -6.0 -5.2 -4.2 -3.2 BOP Statistics: Exports Merchandise, growth rate -15.7 22.2 -1.4 24.2 30.5 30.5 21.3 4.0 4.3 Imports Merchandise, growth rate -10.2 -19.9 15.4 33.0 42.7 4.9 2.5 7.3 6.7 Exports GNFS/GDP at MP 6.0 6.8 7.4 8.4 11.8 12.1 11.4 10.9 10.3 Imports GNFS/GDP at MP 20.2 20.4 21.7 22.4 24.5 26.7 25.7 24.8 23.8 Current Account excl Grants/GDP at MP 2/ -13.5 -11.8 -11.8 -7.8 -7.9 -8.6 -8.3 -8.0 -7.4 Current Account incl Grants/GDP at MP 2/ -5.6 -4.6 -3.5 -1.5 -2.5 -4.0 -3.6 -3.8 -3.5 Current Account excl Grants, US$ million 2/ -449 -338 -381 -309 -444 -514 -540 -591 -622 Current Account incl Grants, US$ million 2/ -187 -132 -112 -59 -142 -237 -232 -280 -295 Miscellaneous: Annual Average Inflation (Composite), percent 32.9 42.2 30.0 6.5 6.1 7.4 6.3 5.0 5.0 Terms of Trade Index, 1991 = 100 108.1 91.6 88.3 106.9 199.6 135.5 108.7 109.1 109.1 Gross Reserves, months of imports GNFS 0.9 1.5 1.9 2.9 3.4 3.6 4.4 4.9 5.5 Source: Government of Uganda, IMF and staff estimates. Notes: 1/ Includes windfall from Coffee Stabilization Tax in 1994/95 and 1995/96. 2/ Beginning in 1995/96, Private Transfers have been adjusted to estimate Direct Foreign Investment separately. ANNEX 5 Page 4 of 9 Table 2: Balance of Payments (In Millions of US Dollars) 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 Exports (g+nfs) 199 195 238 333 667 726 742 804 866 Merchandise (fob) 175 172 157 254 595 590 588 638 685 o/w Coffee 127 117 99 172 457 404 332 361 382 Non-Factor Services 23 23 81 79 72 135 154 166 182 Imports (g+nfs) 671 582 699 893 1,383 1,601 1,675 1,829 1,994 Merchandise (fob) 545 451 531 718 1,085 1,218 1,276 1,393 1,518 Non-Factor Services 125 131 168 176 298 383 400 437 476 Resource Balance -472 -387 -461 -560 -716 -875 -933 -1,026 -1,128 Net Factor Income -58 -87 -49 -61 -58 -46 -31 -39 -19 o/w Net Interest -58 -87 -49 -43 -34 -33 -15 -22 -3 Current Private Transfers 81 136 129 312 330 407 424 473 525 C/A Balance (excl Grants) -449 -338 -381 -309 -444 -514 -540 -591 -622 C/A Balance (incl Grants) -187 -132 -112 -59 -142 -237 -232 -280 -295 Official Transfers 262 206 269 250 303 277 308 312 326 o/w lmport Support 87 75 111 75 96 86 108 96 96 o/w Project Aid 175 131 158 175 207 190 200 216 230 Net M< Loans 122 38 138 179 282 210 220 252 260 Disbursements 214 163 242 294 377 282 293 324 341 Project Loans 115 94 158 175 253 233 244 264 281 Import Support Loans 99 69 84 119 124 50 49 60 60 Repayments 92 125 104 115 96 72 73 72 81 Foreign Investment/Kenya Comp. 1 2 4 5 2 113 160 193 230 Short-Term, net -37 -31 -22 -55 -7 -69 -24 0 0 Errors and omissions 0 1 -16 36 -18 27 0 0 0 Overall Balance -101 -121 -8 106 117 44 124 165 195 Financing. 101 121 8 -106 -117 -44 -124 -165 -195 Monetary Authorities 37 -2 -29 -90 -146 -73 -124 -196 -229 Gross Reserve Changes -15 -24 -38 -107 -169 -92 -134 -133 -167 IMF, net 52 22 10 18 23 19 10 -63 -62 Other,net 0 0 0 0 0 0 0 0 0 Short Term/Commercial -2 -4 -3 0 -1 0 0 0 0 Change in External Arrears 65 98 -330 -59 14 11 -246 0 0 Exceptional Financing 1 29 370 43 15 18 30 31 34 Residual Finance Gap * 0 0 0 0 0 0 216 0 0 Source: BOU, IMF and staff estimates. Note: Beginning in 1995/96, Private Transfers have been adjusted in order to estimate Foreign Direct Investment separately. These figures could change based on new findings. * Includes the clearance of the total stock of arrears of US$246 million. ANNEX 5 Page 5 of 9 Table 3: External Financing Requirements (In Millions of US Dollars) 1990/91 1991/92 1992/93 1993/94 1994/95 1995/96 1996/97 1997/98 1998/99 Requirements 848 787 1,264 1,312 1,742 1,931 2,060 2,164 2,370 Imports GNFS 671 582 699 893 1,383 1,601 1,675 1,829 1,994 Scheduled Debt Service 191 249 175 179 172 167 181 185 192 Interest 61 91 53 54 49 54 50 50 49 Amortization 92 125 104 115 96 72 73 72 81 IMF Repurchases 37 33 18 10 28 41 57 63 62 ReservesBuild-up 15 24 38 107 169 92 134 133 167 Short-Term, net 37 31 22 55 7 69 24 0 0 Factor Service Payments (excl Interest) 0 0 0 18 24 13 16 16 17 Settlement of Arrears* -65 -98 330 59 -14 -11 30 0 0 Available Financing 282 334 356 697 996 1,294 1,362 1,498 1,668 Exports GNFS 199 195 238 333 667 726 742 804 866 CoffeeExports 127 117 99 172 457 404 332 361 382 Noncoffee Merchandise Exports 49 55 58 82 139 186 256 277 303 Non-Factor Services 23 23 81 79 72 135 154 166 182 Interest Received 3 4 4 11 16 21 35 28 47 Private Transfers 81 136 129 312 330 407 424 473 525 Foreign Direct Investment 1 2 4 5 2 113 160 193 230 Other, net -2 -2 -19 36 -18 27 0 0 0 Foreign Financing 566 453 909 615 746 636 698 666 702 ImportSupport 185 144 195 195 220 136 157 156 156 Project Aid 291 225 315 350 460 423 444 479 512 [MF Purchases 89 55 28 28 50 60 68 0 0 Debt Relief** 1 29 370 43 15 18 30 31 34 Residual Gap* 0 0 0 0 0 0 0 0 0 Source: BOU, IMF and staff estimates. Note: Beginning in 1995/96, Private Transfers have been adjusted in order to estimate Foreign Direct Investment separately. These figures could change based on new findings. * Excludes the clearance of the total stock of arrears of US$246 million. The residual gap differs from that in the Balance of Payments (Table 2) by US$216 million, or the remaining stock of arrears after a settlement of US$30 million. ** Includes assumed debt relief from non-Paris Club creditors on Paris Club terms. ANNEX 5 Page 6 of 9 Uganda Most Same regionncome group Nea Latest single year recent SAO. higher Unit of estimate Sahara Low- income Indicator measure 1970-75 1980-85 1989-94 Africa income group Resources and Expenditures HUMAN RESOURCES Population (mre=1994) thousands 11,228 14.134 18.592 571,902 3.182,221 1.096,881 Age dependency ratio ratio 1.00 1.02 1.05 0.94 0.66 0.63 Urban % of pop. 8.3 9.9 12.2 30.6 28.3 55.9 Population growth rate annual % 2.6 2.0 3.1 2.8 1.7 1.3 Urban 3.3 4.2 5.4 4.9 3.2 2.7 Labor force thousands 5,942 7,844 10,190 254,250 1,590.533 488.647 Agriculture % of labor force 89 86 85 65 67 36 Industry 4 4 5 9 14 26 Female 48 48 48 41 39 40 Labor participation rates Total % of pop. 53 56 55 44 50 45 Female 25 27 26 37 41 36 NATURAL RESOURCES Area thou. sq. km 235.88 235.88 235.88 24,273.83 40,391.42 40,594.43 Density pop. per sq. km 47.60 59.92 76.42 22.90 77.44 26.66 Agricultural land % of land area 36.10 42.07 42.93 50.61 52.42 41.05 Change in agricultural land annual % 2.02 1.20 0.00 0.01 0.16 -1.38 Agricultural land under irrigation % 0.06 0.11 0.11 0.86 17.84 11.40 Forests and woodland thou. sq. km .. 69.91 63.46 5.323.14 7,632.00 5,969.25 Deforestation (net) % change. 1980-90 .. .. 0.96 INCOME Household income Share of top 20% of households % of income 47 .. 48 Share of bottom 40% of households 17 .. 17 Share of bottom 20% of households 6 .. 7 EXPENDITURE Food % ofGDP .. .. 46.8 Staples .. .. 25.4 Meat. fish, milk; cheese, eggs 1.. .. 1.0 Cereal imports thou. metric tonnes 6 16 76 14.051 36,922 68.936 Food aid in cereals 0 31 59 5,079 8,516 5,771 Food production per capita 1987 =100 161 103 104 102 115 102 Fertilizer consumption kg/ha 0.2 0.0 0.3 5.3 58.5 46.3 Share of agriculture in GDP % of GDP 66.6 48.4 45.7 19.5 27.6 14.0 Housing % of GDP .. .. 12.0 Average household size persons per household 4.8 .. 4.8 Urban .. .. 4.1 Fixed investment: housing % of GDP .. .. 2.6 Fuel and power % of GDP .. .. 1.7 Energy consumption per capita kg of oil equiv. 43 25 23 251 373 1,602 Households with electricity Urban % of households ., .. 34.9 Rural .. .. 1.8 Transport and communication % of GDP .. .. 4.3 Fixed investment: transport equipment .. .. 2.6 Total road length thou. km 26 28 29 INVESTMENT IN HUMAN CAPITAL Health Population per physician persons 9.302 22,679 22,987 .. .. 3,064 Population per nurse ., 2.049 ... Population per hospital bed 607 661 1,091 1,316 1,034 592 Oral rehydyration therapy (under-5) % of cases .. .. 45 37 38 Education Gross enrollment ratios Secondary % of school age pop. 4 10 13 24 48 63 Female 2 3 10 23 42 62 Pupil-teacher ratio: primary pupils per teacher 34 36 .. 40 39 Pupil-teacher ratio: secondary " 23 23 .. .. 20 Pupils reaching grade 4 % of cohort .. .. .. Repeater rate: primary % of total enroll 10 10 .. Illiteracy % of pop. (age 15+) .. 57 38 53 35 Female % of fem. (age 15+) .. 71 50 54 46 Newspaper circulation per thou. pop. 7 2 5 12 .. 236 World Bank International Economics Department, April 1996 ANNEX 5 Page 7 of 9 Uganda Most Same re' n Come y roup Next Latest single year recent Sub- higher Unit of estimate Saharan Low. income Indicator measure 1970-75 1980-85 1989-94 Africa income grap Priority Poverty Indicators POVERTY Upper poverty line local curr. .. 6,000 Headcount index % of pop. .. .. 55 Lower poverty line local cur. .. .. 3,000 Headcount index % of pop. .. 19 GNP per capita USS 160 200 500 390 1,670 SHORT TERM INCOME INDICATORS Unskilled urban wages local curr. .. .. Unskilled rural wages Rural terms of trade Consumer price index 1987=100 13 1.446 Lower income Food I .. .. .. Urban Rural .. 15 .. SOCIAL INDICATORS Public expenditure on basic social services % of GDP .. .. 5.5 Gross enrollment ratios Primary % school age pop. 44 70 91 71 105 104 Male 53 56 99 77 112 105 Female 35 43 83 64 98 101 Mortality Infant mortality per thou. live births 104 116 122 92 58 36 Under 5 mortality .. .. 206 161 101 47 Immunization Measles % age group .. 17.0 74.0 51.4 86.2 77.4 DFT .. 14.0 77.0 53.5 89.1 82.0 Child malnutrition (under-5) " .. .. 23.3 .. 38.2 Life expectancy Total years 51 48 42 52 63 67 Female advantage 1.6 1.6 1.2 3.5 2.4 6.4 Total fertility rate births per woman 7.1 7.3 7.1 5.9 3.3 2.7 Maternal mortality rate per 100,000 live births .. 300 550 Supplementary Poverty Indicators Expenditures on social security % of total gov't exp. .. 1.8 .. Soctal security coverage % econ. active pop. .. .. .. Access to safe water total % of pop. 22.0 15.0 .. Urban 88.0 45.0 .. Rural " 17.0 12.0 .. Access to health care .. 42.0 .. Population growth rate GNP per capita growth rate Development diamondb 6 (average annual, percent) (average annual, percent) 10+ Life expectancy 445 20I GNP Gross per primary I capita enrollment -2 -10 1970-75 1980-85 1989-94 1970-75 1980-85 1989-94 Access to safe water 0 Uganda - Uganda - Low-income - Low-income a. See the technical notes, p.387. b. The development diamond, based on four key indicators. shows the average level of development in the country compared with its income group. See the introduction. MOP Schedule D Generated: 04/23/97 Status of Bank Group Operations in Uganda IBRD Loans and IDA Credits in the Operations Portfolio Difference Original Amount in US$ Millions Between actual Loan or Fiscal and expected Project ID Credit Year Borrower Purpose IBRD IDA Cancellations Undisbursed Disbursements a/ No. Number of Closed Loans/credits: 58 UG-PE-2969 C21240 1990 GOVERNMENT WATER SUPPLY II 0.00 60.00 0.00 18.38 13.98 UG-PE-2929 C22680 1991 GOVERNMENT POWER III 0.00 125.00 0.00 71.75 74.52 UG-PE-2933 C22060 1991 GOVERNMENT URBAN I 0.00 28.70 0.00 6.50 4.93 UG-PE-2930 C21760 1991 GOVT. LIVESTOCK 0.00 21.00 0.00 9.60 7.74 UG-PE-2981 C23620 1992 GOVERNMENT NORTHERN RECONSTRUCT 0.00 71.20 0.00 19.67 -2.01 UG-PE-2968 C23150 1992 GOVERNMENT ENTERPRISE DEVELOPME 0.00 65.60 23.75 22.42 28.58 UG-PE-2962 C24960 1993 GOVERNMENT FINAN SECT ADJUST CR 0.00 100.00 0.00 47.07 43.91 UG-PE-2953 C24930 1993 GOVERNMENT PRIMARY EDUC. & TEAC 0.00 52.60 0.00 31.79 10.09 UG-PE-2938 C24460 1993 GOVERNMENT AGRIC. RES & TRG. 0.00 25.04 0.00 10.03 -3.44 UG-PE-2991 C24240 1993 MINIS. AGRIC ANINMAL AGRIC. EXTENSION PRO 0.00 15.79 0.00 4.17 2.21 UG-PE-2975 C24180 1993 GOVERNMENT ECON & FINANCIAL MAN 0.00 29.00 0.00 4.43 -9.74 UG-PE-2977 C26090 1994 GOU COTTON SECTOR DEVELO 0.00 14.00 0.00 9.63 4.01 UG-PE-2963 C26030 1994 GOVERNMENT SEXUAL.TRANS.IN 0.00 50.00 0.00 42.64 10.56 UG-PE-2923 C25870 1994 GOVT. TRANSP. REHAB. 0.00 75.00 0.00 64.95 33.16 UG-PE-2957 C25830 1994 GOVERNMENT SMALL TOWNS WATER 0.00 42.30 0.00 32.79 9.89 UG-PE-2976 C27360 1995 GOVERNMENT INST. CAPACITY BLDG 0.00 36.40 0.00 29.55 4.00 UG-PE-2971 C26790 1995 GOVERNMENT DISTRICT HEALTH 0.00 45.00 0.00 39.34 2.75 UG-PE-37582 C28450 1996 GOV. OF UGANDA AG SEC MGT PROJECT 0.00 17.90 0.00 16.74 1.79 UG-PE-35634 C27980 1996 GOU PRIV. SECTOR COMPETI 0.00 12.30 0.00 10.20 .54 UG-PE-2978 C27770 1996 GOVERNMENT ENVIRONMENT MANAGEMT 0.00 11.80 0.00 8.87 2.64 UG-PE-46836 C29090 1997 GOVT. OF UGANDA LAKE VICTORIA ENV. 0.00 12.10 0.00 11.62 0.00 Total 0.00 910.73 23.75 512.14 Active Loans Closed Loans Totl Total Disbursed (IBRD and IDA): 385.61 1,466.39 1,852.00 of which has been repaid: 0.00 102.56 102.56 Total now held by IBRD and IDA: 886.98 1,292.30 2,179.28 Amount sold : 0.00 25.82 25.82 Of which repaid . 0.00 25.82 25.82 Total Undisbursed : 512.14 1.72 513.86 a. Intended disbursements to date minus actual disbursements to date as projected at appraisal. b. Rating of 1-4: see OD 13.05. Annex D2. Preparation of Implementation Summary (Form 590) . Following the FY94 Annual Review of Portfolio performance (ARPP) , a letter based system will be used (MS = highly Satisfactory, S = satisfactory, U = unsatisfactory, HU = highly unsatisfactory) : see proposed Improvements in Project and Portfolio Performance Rating Methodology (SecM94-901), August 23, 1994. c. Following the FY94 ARPP, "Implementation Progress" will be reported here. Generated by the Operations Information System (OIS) Page 1 D 2 00~ ANNEX 5 Page 9 of 9 Uganda STATEMENTOF IFC's Committed and Disbursed Portfolio As of 31-Mar-97 In Millions US Dollars Committed Disbursed ------------ IFC----------- IFC---------- FY Company Loan Equity Quasi Partic Loan Equity Quasi Partic Approval 1983 Uganda Sugar 8.22 0.00 0.00 0.00 8.22 0.00 0.00 0.00 1984 DFCU 0.00 .38 0.00 0.00 0.00 .38 0.00 0.00 1992 AEF Clovergem .73 0.00 0.00 0.00 .73 0.00 0.00 0.00 1992 DFCU 0.00 .60 0.00 0.00 0.00 .60 0.00 0.00 1993 AEF Nile Roses .19 0.00 0.00 0.00 .19 0.00 0.00 0.00 1993 AEF Rwenzori .75 .19 0.00 0.00 .75 .19 0.00 0.00 1993 Jubilee 0.00 .10 0.00 0.00 0.00 .10 0.00 0.00 1994 AEF Polypack .80 0.00 0.00 0.00 .80 0.00 0.00 0.00 1994 AEF Skyblue .51 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1994 Celtel 3.43 .64 .80 0.00 3.43 .64 .80 0.00 1995 AEF Rainbow .79 0.00 0.00 0.00 .79 0.00 0.00 0.00 1995 Uganda Leasing 0.00 .33 0.00 0.00 0.00 .33 0.00 0.00 1996 AEF Agro Mgmt .60 .40 0.00 0.00 .38 .40 0.00 0.00 1996 AEF Govinda Kiri .23 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1996 EAGW 6.50 0.00 0.00 0.00 0.00 0.00 0.00 0.00 1996 Uganda Leasing 2.00 0.00 0.00 0.00 .30 0.00 0.00 0.00 Pending Commitments 1997 * AEF CONRAD PLAZA 1.50 0.00 0.00 0.00 1997 * AEF SAMBIYA .37 0.00 0.00 0.00 1996 * KASESE COBALT 16.00 3.58 0.00 5.00 Generated by the Operations Information System (OIS) Page 1  ANNEX 6 Technical Notes Technical Note 1: Tax Revenue Issues: A Longer Term Perspective .................1 Technical Note 2: Tax Exemptions: Past and Present ...........7........7 Technical Note 3: Adjusting Trade Policy for Export Growth...................13 Technical Note 4: Estimates of Effective Protection and Manufacturing Efficiency......23 Technical Note 5: A Proposal for an Effective Duty Drawback System...... .....32 Technical Note 6: Issues in Local Taxation ........................ ......41 Technical Note 7: Outcome-Oriented Budgetary Process .................51 Technical Note 8: Improving the Management and Efficiency of Public Expenditures Consistent with the Decentralization Program.........74 Technical Note 9: Reducing Quasi-Fiscal Deficits: Parastatals ........... .....83  � � GA � О � � � � у � и .� и r.,► � v � � С� � � � � �, а � � � � �� . � � � о � � е� � .� � � � а� Е�  TAX REVENUE ISSUES: A LONGER TERM PERSPECTIVE' Revenue Planning 1. In spite of rapid growth in revenue during the last ten years, aggregate revenue as a proportion of GDP is still low by comparison with other African economies and remains one of the most important issues for future economic policy. The growth that has been achieved so far has not been the outcome of detailed medium term revenue plans. Beyond a broad strategy of displacing the earlier dependence on export taxes and the gradual lowering of average tariffs, both PTA and external, there are not many medium term themes in tax policy changes so far. While there have been contributions form improvements in tax administration , and increases in some key tax rates (excise taxes, petroleum), the most substantial contribution to the growth in revenue has been rapid growth in the main tax bases, generally significantly more rapid than total GDP. 2. Looking forward, there are a number of reasons why greater investment in planning is going to be needed on tax policy in future. Firstly, throughout the reform programme to date, aggregate revenue has been so far below desirable levels that revenue growth has outweighed considerations of optimal tax structure. While revenue remains low, Uganda now needs to have a clearer objective of long term goals on revenue and the size of Government. While far below the revenue levels in many African economies, the level of revenue in some Asian economies is not so far off. If the rate of growth of revenue over the past five years is maintained , Uganda's revenue/GDP ratio would exceed Indonesia in six years, Thailand in seven and Korea in eight. With these landmarks not so far over the horizon, it would be wise to pause and take bearings on where to stop, and what long term tax structure should be developed on the way. 3. Secondly, the rush for growth in revenue, while justified in terms of macroeconomic adjustment, public expenditure needs and strong Government commitment to reduction in aid dependency, has also had costs which now need to be more carefully considered. Private sector surveys have signaled considerable concern on taxation, at least as much arising from administration as tax rates. A significant part of the problem for the private sector lies in uncertainty about future taxation. Attempting to achieve a fixed annual target for revenue growth predominantly from tax rate changes rather than gradual improvement in administration has led the Government into frequent ad hoc tax adjustments, often with very little information on likely impact. Private sector resistance to tax measures has often gained strength from these weaknesses, while the uncertainty generated has been a significant disincentive to investment. While Government faces very obvious and well publicised future expenditure commitments (roads, defence, and now education, being most prominent) and yet is seen to have no plan for generating the necessary revenue growth, it is clear that investors must view the future of taxation with some disquiet. This is likely to abate only if Government produces a medium term revenue plan which is available to the public and developed on the basis of extensive consultation with the private sector. This note was prepared by Mr. Allister Moon (The World Bank). 2 Present Tax Structure : A Review 4. In looking at the scope for achieving the long term tax structure as a planned outcome, it is useful to consider what developments in tax structure have occurred in the recent past. 5. The broad pattern of changes in tax structure over the last ten years can be seen in the attached graph. There are several distinctive features. Firstly, the rapid displacement of export taxes can be seen during the period up to 1990/1, with the share of total revenue falling from 44% in 1985/6 to zero in 1991/2. Secondly, the contribution from petroleum taxes rose rapidly from 1988/9, rising to almost 30% of total revenue in 91-93 and then falling to around 23% in the last three years. Other movements are less dramatic. The overall percentage contribution from non-petroleum customs has risen over most of the period, settling at around 14% in the last four years. The contribution from sales tax and CTL , grouped together as the taxes now replaced by VAT, has grown over the period as a whole, although the share rose sharply to 1988/9, fell rapidly over the next three years (partly arising from reclassification of sales and excise tax) and has since risen slowly from 26% in 1991/2 to 32% in 1995/6. Since the rate increases of 1992/3, excise tax has steadily contributed about 10% of revenue. The share of income tax has not risen much from the 10% share achieved in 1988/9. Other indirect and non tax revenue has been contributing a steady at around 6 or 7%, (excluding a blip in 1986/7 which appears to arise from data problems in non-tax revenue). Tax Structure 1985/6-95/6 12 ----------- ---------- ---------....----------- ----- 10 TOT.4L R EVENUE 8 other indirect and non tax 1-excise 4 -other cuitoms exports I0I 2 petroleum income tax 0 85/6 86/7 87/8 88/9 89/90 90/1 91/2 92/3 93/4 94/5 95/6 3 Customs 6. During at least the latter few years of this period, the Government has frequently indicated a commitment to a strategy of reduction in trade taxes, steady growth in income tax and greater use of the sales tax/VAT. The existing structure falls well short of this goal. 7. Firstly, dependence on trade taxes has increased significantly over the period as a whole and shows no sign of falling in recent years. This is not altogether surprising in the context of rapid growth in aggregate revenue. A high level of dependence on trade taxes is typical of developing country tax structures, driven by the relative administrative ease of collecting such taxes. Achieving rapid growth in aggregate revenue almost inevitably entails initially more rapid growth in trade taxes than other components of revenue. How much of a problem this presents for export policy depends on the nature of the trade taxes involved. Sales tax on imports is not a problem, assuming that the tax applies equally to domestic goods. The petroleum tax can also be considered neutral in this respect, although there are other problems with the high dependence on petroleum, discussed below. The main problem category is the non-petroleum customs. While not growing as rapidly as petroleum taxes, the share of non petroleum customs has nevertheless more than doubled over the period and has remained steady at about 14% of total revenue in recent years. Reducing this share has to be a more explicit object of policy. However, it will not be easy to replace 4% of current revenue, about 1.4% GDP. Sales tax/CTL amounted to 3.3% GDP in 1995/6. Replacing non petroleum customs would require a 42% increase in VAT as a proportion of GDP above the 1995/6, more than a 50% increase in real terms., roughly double the achievement of the past four years. Petroleum 8. The level of petroleum taxes raises different issues. It is doubtful how long Uganda can sustain petroleum taxes in excess of double the level in Kenya. Apart from the existing problems of smuggling, it is probable that this tax will receive increasing attention in discussions with Kenya and Tanzania on tax convergence. In the absence of alternative broad based taxes, the petroleum tax was an attractive instrument in the campaign for rapid revenue growth in recent years, both because of a relatively diffuse incidence and relatively low impact on activity, demonstrated in an estimated elasticity below 0.4. However, as Paul Collier has recently noted, a tax on transport falls more heavily on regions distant from Kampala, including many of the poorest districts. 9. The petroleum tax contributes to anti-export bias by raising exporter costs, both directly and indirectly, through the tax passed on in domestically produced inputs. In principle, the contribution of the petroleum tax to anti-export bias could be eliminated, if a means could be devised of capturing petroleum costs in the duty drawback scheme for exporters. This faces two difficulties. Administering the drawback on direct purchases by exporters would be impossible to administer without significant leakage to non-export uses. Secondly, a significant proportion of exporters fuel costs are likely to be incurred indirectly. The latter could be captured if the drawback system is administered on the basis of standard export production coefficients rather than reimbursement on purchased inputs, as John Short has recommended. At present this looks unlikely to work, facing difficulties both from the adverse revenue implications and possible contravention of international trade agreements. 10. On balance, it seems both desirable and prudent to project a declining share of revenue from petroleum over the medium term. Given that petroleum imports are likely to continue to 4 grow faster than GDP, a gradual decline in rate could be accommodated without an equivalent decline in tax/GDP ratio. Income Tax 11. Income tax is presently underachieving , both by international comparison and by comparison with the achievement on other taxes such as sales tax/CTL. Income tax at 1-1.5% GDP is low even for low income developing countries (cf.3.5-4% in Tanzania). The future of income tax in Uganda is clouded by two factors. Firstly, it is uncertain how far income tax or related tax instruments can be developed at the district level, either supplementing or partly replacing central collections. Secondly, there are diverse views on what concessions could sensibly be made on corporate income tax in the process of revamping the investment incentive regime. While noting these uncertainties and leaving aside issues of composition between corporate tax and personal income tax at local and central level, it seems clear that in aggregate terms, income tax should be rising steadily over the medium term, arising from accelerated improvements in tax administration. VAT 12. While any predictions on the future of VAT must appear somewhat foolhardy in the light of recent events, there is no doubt that the effectiveness and acceptability of the tax is the most critical condition for desirable revision of the tax structure. The hope must be that VAT will both contribute an increasing share of a rising level of aggregate revenue, and also play the role of primary instrument of short term macroeconomic adjustment on the revenue side. The outcome of the recent resistance to the tax may keep the rate steady for now, but will not encourage policymakers to risk any increases in the near term and may also discourage use of the tax in short term adjustment. Excise 13. On the excise tax, it would be desirable to maintain its present contribution more or less unchanged. There are several reasons why this might be difficult. Taxes on the main excise products are fairly high by regional standards, giving rise to some smuggling difficulties. The present excise rates were established after a period of intensive lobbying and considerable volatility in the rates in the early 90s. Policymakers tend to be reluctant to provoke a recurrence by increases in any of the excise rates, and rates since this settlement have only been adjusted downwards. The excise revenue can probably be maintained as a proportion of GDP - most of the tax bases are growing more rapidly than GDP - but is unlikely to hold its share in total revenue assuming the aggregate revenue/GDP ratio continues to rise. Aggregate Revenue 14. So far this review of the present tax structure has looked at possible and desirable changes in structure, without a view of the path of aggregate revenue. It seems evident that a trade off must exist between the rate of growth of aggregate revenue and the likelihood of achieving desirable changes in tax structure. At the limit, if aggregate growth dominates all other considerations, there is no chosen tax structure, only the maximisation of each available tax handle. Well short of this limit, it is clear that objectives such as the reduction of the dependence 5 on petroleum and other trade taxes cannot be achieved if too great an emphasis is placed on revenue growth. 15. Prima facie, it appears probable that aiming for revenue growth at the level attained in the past four years - more or less 1% GDP per year - is not going to leave room for much reform in structure. The past four years revenue growth have benefited from some significant tax rate increases (excise, petroleum) and probably one off step improvements in administration for some tax bases. The longer run growth (1986/7-1995/6) is more like 0.6% GDP per year, also including a number of step improvements. Even assuming unchanged weight given to aggregate revenue growth, it seems unlikely that the recent growth rate would be equaled in future. The implication is that a serious commitment to significant improvement in tax structure is likely to be incompatible with a planned growth rate in excess of 0.5% . This does not mean that growth outcome may not exceed this rate of growth, assuming that this comes from administrative improvements, the revenue outcome of which is always difficult to predict. Tax Administration 16. The reform of tax administration in Uganda is widely regarded as highly successful. The major measure of reform was the transfer of responsibilities for tax collection from the ministry of Finance to an independent Revenue Authority, combined with a screening of staff transferred to the new Authority, a significant increase in pay levels beyond civil service norms, and substantial donor financed technical support and provision of capital equipment. 17. The success of the reform has normally been measured in terms of the growth in aggregate revenue as a proportion of GDP since establishment of the URA. This was virtually the only aggregate performance measure assessed in the 1995 ODA review of the Authority, which has been the most substantial review of URA to date. 18. In retrospect, it can be seen that the steps taken in establishing the URA could usefully have been supplemented with other measures, mostly bearing on the environment within which the new Authority was supposed to operate. 19. Firstly, the Authority has operated without any clear and independent mechanism for resolution of disputes in tax assessments. Appointment of a Commissioner for tax appeals has been promised on several occasions. There has also been discussion of establishing a Tax Court. In the absence of implementation of these measures, tax payers have recourse only to procedures internal to the URA. 20. The issue of equity and probity in tax collection must be taken seriously, not only because of wider concerns on transparency in Government, but also because surveys show that the private sector takes the issue seriously. Tax administration problems, as distinct from problems of tax level, remain high on the list of private sector concerns with the investment environment in Uganda. 21. Lack of an effective appeals procedure tends to deflect private sector criticism of URA performance into more general lobbying of Government on taxation matters, often at high level and often mixed with a wide array of other concerns, including tax levels. At Efficient resolution of specific disputes would remove genuine tax administration grievances from the arena of tax lobbying. 6 22. Secondly, the Authority has lacked a satisfactory framework for assessment of its performance. As noted above, attention has mainly been focused on aggregate revenue collection as a proportion of GDP. While performance on individual taxes has been monitored by the Tax Policy Department, this has not contributed to an overall assessment of the contribution of improved tax administration to revenue growth, after allowance for growth arising from policy changes or growth in the underlying tax bases. 23. Broad growth in the revenue/GDP ratio is a misleading indicator of administrative improvements. When revenue growth is related to monetary GDP , which has been growing more rapidly than total GDP, or to the main sectors liable to taxation, such as manufacturing, which have been growing faster than monetary GDP, it is not obvious that the observed growth in revenue is indicative of significant gains in tax administration. For some taxes, revenues have simply matched growth in the tax base, as appears to be broadly the case for petroleum imports. In other cases, such as customs revenue from non petroleum imports, the growth in the tax base appears to be higher than revenue growth, implying some deterioration in tax administration, although this needs to be assessed in the light of policy interventions including the impact of exemptions. 24. Such an assessment needs to made routinely for each of the main taxes as a means of monitoring URA performance and possibly providing a basis for incentive payments. Annual targets for performance in each of the main tax categories should be formulated not only in terms of aggregate revenue growth but as a proportion of tax base growth, thereby setting targets for genuine tax administration improvements. While this framewc k would not address issues of collection efficiency - which probably need to be handled through the URA's business plan targets, monitored by the Board - it would represent a significant improvement in the existing incentive structure. 25. Thirdly, the Authority has operated with a persistent lack of clarity regarding the allocation of responsibilities for tax policy. As noted in the ODA review, URA have often expressed an understanding that tax policy falls within their responsibilities. There is some ambiguity in the terms of reference for the Authority, which partly accounts for this view, although the Ministry of Finance has on several occasions clarified the respective responsibilities, confining the URA role to tax administration. 26. However, this ambiguity stems less from the letter of URA's terms of reference than from the history of the Authority's establishment. Very little attention was given to tax policy functions at the time of separating the revenue functions from the Ministry of Finance. URA staff were drawn from the Ministry's revenue staff selectively, with the staffing of the remaining tax policy section (TIPD, later Tax Policy Department) determined as a residual. The role of advice on tax policy had in practice already been taken over by the Economic Analysis Unit within the Ministry of Finance. TIPD/TPD received none of the technical and material support provided by donors to the URA and has only begun to receive some attention within the last two years. 27. In retrospect, the neglect of tax policy capacity can be seen as a strategic error in the reform program for taxation, closely related to other problems in Uganda's tax strategy. The lack of capacity in the Ministry of Finance to drive taxation from a strategic perspective has contributed to the relatively ad hoc nature of tax policy in recent years. In lacking a strong tax policy capability, the Ministry also lacked a base for establishing effective scrutiny of URA performance and identifying supply side problems in tax policy and its implementation. Support to tax policy functions should be included in EFMP2. Technical Note 2: Tax Exemptions: Past and Present  TAX EXEMPTIONS: PAST AND PRESENT' Introduction 1. Uganda's tax system has been complicated by the widespread use of tax exemptions as a policy tool that confront the principles of equity and administrative efficiency in a tax system. As part of the work carried out by the Ministry of Finance discussion, significant and far reaching changes have been introduced as part of the overall reform process. This note examines * First, the position relating to exemptions prior to this work and a critique of them. * Second the reforms that have been enacted and the consequent (current) position. Tax Exemptions: Historical Perspective 2. There are three generic sources of tax exemptions outside of the normal tax allowances (such as the first 840,000 Shillings of income) embodied in the tax code. These are: * Conditional exemptions from import duties and sales tax under the second schedule of The Finance Statute; * Discretionary exemptions granted by Statutory Instrument; and * Exemption from corporation tax and other taxes under the provisions of the 1991 Investment Code. 3. These exemptions apply to the individual tax paying unit irrespective of on what the exemption is being granted. For example, the import may carry a tariff, but because of the exempted status of the importer, the tax is waived. 4. This waiver status is different from a fourth source of exemption prevalent in Uganda that is the pervasive use of zero-rating of many classes of imports. While zero-rating of imports is not really an exemption, it is clear that the majority of zero-rated items in Uganda can be classified as an exemption given the ad hoc granting of zero-ratings particularly with respect to import duties. 5. The Government is well aware of the issue of exemptions particularly from the revenue foregone perspective. 2 The 1993 and 1994 Budget speeches had sections that deal specifically with the topic and significant reforms have been introduced on paper to eliminate the abuse that This note was prepared by Mr. John Short, R.E.P.I.M., Northumberland, UK. 2 Estimates by the IMF suggest that that the total revenue foregone from exemptions is at least 10% of Government tax revenue, but could well be higher.  9 11. The system of Statutory Exemptions is administratively cumbersome as there is an apparent lack of criteria as to why an exemption should be granted. This, by default, makes almost each exemption discretionary when, instead, transparency and automaticity are desirable. Such procedures, in turn, tie up senior officials (including the Minister of Finance) in deciding the fate of individual applications for exemptions. 12. There would appear to be two generic classes of SI exemptions. One group covers the imports of NGOs and others. The apparent rationale for these exemptions is a humanitarian one. The other group can be classed as economic: a company requests an exemption from payment of duties or a temporary exemption from sales tax on its products due to trading conditions. 13. For those SI categories of a NGO humanitarian nature, Government may wish to provide the institution the funds that would have been waived through the exemption as a Government contribution to the project. In this way, the duty would be paid, but the budget would be increased. This would force both the Government and the NGO to address the nature of the expenditure and its justification. While the net effect on Government's finances would be zero, this system would discourage the culture of requesting and granting exemptions that is so prevalent. Government had implemented such a system for Government and donor funded projects in the 1994 Budget by terminating tax-exclusive tendering on the grounds of easier administration and elimination of abuse. However, it is not clear how effectively this has been implemented, if at all. 14. However, for this alternative system or the exemption system to work effectively, a clearly delineated list of accredited NGOs and other agencies has to be objectively determined. However, this will still be open to abuse. Clear and transparent guidelines for accreditation of such organisations and of the exemption status of the types of transactions are required. Once these have been determined, careful policing of the rules is required. 15. The policy towards the second category of SI exemptions - those of an economic nature - is much more clear-cut. The possibility of applying and receiving this type of exemption should be stopped forthwith. These exemptions are inequitable as they only apply to the individual company and introduce distortions to the competitive framework in which these companies operate. The Minister of Finance should announce that all sales tax, import duties and excise tax will apply equally to all companies and that any provision for altering these on a case-by-case basis shall be discontinued. Applications for exemptions shall no longer be entertained as the provision for such exemptions shall no longer be possible in law (once the law is amended, if necessary). 16. The third and fourth categories of exemption are Investment Code allowances and zero- rated tariffs and these are considered in reverse order. Zero Rated Tariffs 17. In the 1993 Budget, the Government imposed a 10% Customs Duty on all raw material imports ending much of the zero-rating in the Customs Code. This essentially created a two tier tariff for the same good. If the product was to be used as an input to the manufacturing process then its tariff became 10% even if it carries a higher tariff in the tariff book. The same good which could be used as an input or were not imported by a producer or used as a final goods did not carry the same tariff rate. This was part of a general reform of the duty system that was intended to provide relatively uniform and moderate protection across sectors. The top rate was 10 reduced to 30% with two non-zero rates of 20% and 10% with some zero rates for social reasons. The reasoning behind the reforms was based on producing a sound economic incentive system to reduce anti-export bias, but also to eliminate the import dependence of producers. The elimination of the zero-rating of raw materials was intended to encourage producers to use a domestic substitute for an import. It also encourages producers and investors to develop a domestic capacity to supply intermediate goods and iaw materials. 18. However, in 1994, after strong lobbying from the manufacturing sector, the Government effectively restored the zero-rating of those raw materials and intermediate goods that cannot be produced in Uganda by remitting the duty to 0% while it remained in the Customs Book at 10%. The list of these goods is extensive: 3 complete chapters, 41 sections at the 4 digit level and 41 sections at the 6 digit level and I at the 8 digit level of the HS Code. 19. There are four significant implications of this reversal of policy. The first is that the tariff system no longer gives any signal to the potential producer of raw materials and intermediate goods, however few they may be. 4 The second is continuing the anomaly that tariffs in some instances have different rates depending on use. The third is that Government are denied the windfall revenues (which may be small) from tariffs that it could use to eliminate the reasons that the manufacturers argue that they need protection. 20. The fourth implication is that the intended moderate and uniform protection is increased. Past studies on protection in Uganda 5 show that all import substituting sectors in Uganda are well protected with an average rate around 60%. The top tariff rates have been reduced from 50% and 40% to 30% since these studies were produced. The cascading nature of the tariffs where inputs carry a 10% or 20% tariff has meant that goods produced for the domestic market are still positively and well protected. A more recent study of effective protection carried out as part of the exercise to determine a Common External Tariff for the Common Market for Eastern and Southern Africa (COMESA) indicates the level of protection that is enjoyed by Uganda producers. 21. However, with the reversal of the 1993 Budget provisions eliminating most of the zero rates and these are on inputs means that the level of effective protection will have been raised by a significant amount. This, in turn, increases the anti-export bias as producing for the domestic market becomes more attractive relative to exporting. 22. The picture is complicated by the provisions of the PTA tariff agreement. An import from a PTA country that meets the PTA rules of origin 6 commands a tariff rate that is a fraction of the statutory rate. There is a possibility that Uganda's manufacturers import inputs from non- 4 Shortly after the list of zero-rated products was published, there was a report of complaints by potential investors in some of the products with respect to the removal of the limited protection. 5 Effective Assistance to Industry in Support of Structural Adjustment, REPIM June 1991; Study of the Effectiveness of Policies, Facilities and Incentives for Investment Promotion, Maxwell Stamp, June 1993; Effective Protection Study carried out by Louis Berger for EPADU, 1993 The Establishment of a Common External Tariff in the Common Market for Eastern and Southern Africa, Imani Development International and REPIM. Nov. 1995 6 Abuse by falsifying certificates of origin may be one of the reasons for the recent significant upsurge in imports from PTA countries. 11 PTA countries, while competing imports come from PTA countries such as Kenya. This will reduce the levels of protection quite significantly and in extreme cases could generate negative protection given the mix of inputs and the extent of the PTA import competition. 23. The evidence from the tariff studies that specifically examined the PTA phenomenon suggests that effective protection from PTA imports is high (though less than from non-PTA countries). Since 1991, the top rates have been reduced and the percentage discount on tariff rates on PTA imports has been increased. These changes will have reduced the level of protection from PTA imports. Likely levels of protection from PTA and non-PTA competition using the COMESA results with the remission of duties on imports are shown in the annexe. Protection from PTA imports is still high and protection from non-PTA imports is very high. 24. A case can be made for re-establishing the tariffs on inputs. Initially, this could be at 5% with the status quo of 10% the following year. The danger that Government faces in restoring the 1993 position is that its credibility in setting a stable tax regime will be eroded. However, where a mistake has been made it may be better to rectify the mistake rather than continue with it. The presentation of the reversal of policy and any accompanying measures that may assist in eliminating the reason for the policy will be critical in obtaining consensus. Exemptions under the 1991 Investment Code 25. The final set of exemptions to be considered relates to those given under the 1991 Investment Code for domestic investors committing $50,000 and foreign investors with a $100,000 investment. These cover: * Exemptions from import duties and sales tax on plant and machinery. Initially, import duties and sales tax on construction equipment were also exempted, but following abuses, this exemption was ended in the 1993 Budget. However, in practice, this exemption has continued. * Exemptions from taxes on profits and dividends (including withholding taxes) on approved investment projects for periods of 3 to 5 years depending on the size of the investment plus 1 additional year for 24 listed priority sectors. Actions Taken to Reform the System Government has made significant and substantial progress in curtailing exemptions during the course of 1995 and in the 1996/97 Budget. Perhaps the big difference in the 1996/97 budget compared to the previous budgets that attempted to deal with the issue is that the 1996/97 budget made provision for the removal of exemption to be made effective rather that just stating that they were to be removed. Removal of exemptions can have an expenditure consequence as Government and Government projects supported by donors and NGOs have to pay taxes that have to be budgeted for. In the past, these were in part accounted for by Treasury Credit Notes (TCNs) which were an accounting mechanism and were partially applied. There was no issue of hard budget constraint. As illustrated below, these were now budgeted for in cash terms. 12 The changes that have taken place are: * First, discretionary exemptions from import duties under the Investment Code have been replaced by either a zero or a low positive rate in the Customs Code. Transparency has thus increased considerably. * Second, exemptions from import duties for a wide variety of institutions and organizations have been eliminated. For NGOs, the MF's Tax Exemptions Committee (TEC) has reviewed each agreement and placed them on tax inclusive basis with Government meeting the tax liability through Ministerial Votes. All Government, parastatal and donor construction projects have to be tendered on a tax inclusive basis. In addition, the Ministry of Finance's Vote includes a contingency allocation to cover the tax component of NGOs' and charities' imports where there is no agreement, but where the TEC has agreed that such support is legitimate. The use of TCNs was discontinued. To further increase transparency and accountability, the Minister of Finance has started to present reports to Cabinet on such expenditures. * The exemptions contained in schedule 2 (as described above) have been removed and placed in the Customs tariff. * Imported inputs are no longer duty free but are subject to a 5% tariff. 26. Some of these actions have increased Government expenditures to compensate for the payment of taxes. A further USh 52 billion in expenditure has been budgeted in 1996/97 for Government tax payments (if included in the tax to GDP ratio this sum would increase it to 13.6%). However, net revenue is expected to increase by USh 4 bl. Revenue foregone through exemptions has dropped from an estimated USh 45 bl in 1993/94 to USh 24 bI in 1994/95 and to between USh 8-9 bi in 1995/96. For 1996/97, the Government has agreed with the IMF that the MF Vote to cover tax expenditures where there is no prior agreement should be no more than USh 1 bI (against a target of USh 12 bi for exemptions in the previous year). 27. These achievements now mean that virtually all exemptions have been eliminated or placed in the budget as an expenditure. This has increased the efficiency of tax policy and administration, and introduced transparency into the system. The cost of exemptions is now explicit. The categories of exemptions that remain cover bilateral, UN, multilateral agreements and protocols of a reciprocal nature. Technical Note 3: Adjusting Trade Policy for Export Growth  ADJUSTING TRADE POLICY FOR EXPORT GROWTH1 1. Uganda's external trade regime has been liberalized gradually since the late-1980s both in the context of the Preferential Trade Area for Eastern and Southern African States (PTA) and 2 unilaterally. Apart from changes in trade policy instruments -- such as reduction in tariffs and removal of quantitative restrictions -- shifting from foreign exchange rationing to market-based allocation has also been trade liberalizing, terminating implicit subsidy for imported capital and intermediate goods in favor of nontradable investment goods and consumer imports. Trade liberalization has occurred as part of broader liberalization of the economy, notably that of interest rates, prices and marketing and export of agricultural commodities, and subsequently with privatization of public enterprises and many institutional reforms. Table 1: Effective Protection in Selected Industries (%) Industry Domestic sales wrt Domestic sales wrt Export sales in world market regional market world market Soft drinks 75 75 -19 Brewing 167 167 -24 Fish processing 111 96 -9 Grain milling 282 186 -26 Horticulture 24 24 -21 Sugar 34 11 -17 Tobacco products 266 266 -11 Clothing 220 99 -38 Leather 61 16 -2 Paints 181 130 -5 Plastic goods 35 16 -11 Misc. manufacturing 53 29 -24 Unweighted average 93 64 -16 Source: World Bank, 1996 2. Despite liberalization, the incentive regime is still biased against exports. Despite lower tariffs, low value added in domestic production at world prices and the remaining tariff differences between inputs and final products continue to produce high effective rates of protection for goods sold in the domestic market. As Table 1 shows, the effective rate of protection in extra-regional competition was, on average, over 90 percent in 1994, while that for The note was prepared by Ritva Reinikka (World Bank) as part of the background work for the Third Structural Adjustment Credit (SAC III) for Uganda. The author worked in close collaboration with John Short (consultant) during the missions. Data processing was carried out by Mimi Klutstein-Meyer (World Bank). 2 The World Bank (1996); Kasekende et al. (1994); UNDP-World Bank (1990) 14 3. Conceptually, in the case of imports where there is domestic production, import duties are the only source of anti-export bias as excise taxes and VAT are levied both on imports and domestically produced goods. If there are exceptions to this rule so that lower rates are applied to domestically produced goods, or they are exempted, while imports are being taxed, then also other import taxes apart from duties become protective. 4. When taxes are levied for revenue purposes on imports which do not compete with domestic production but are used as inputs, production costs go up. Unlike many import- substituting industries, exporters as price-takers in the world market cannot pass on the additional cost created by import taxes. Therefore, also non-protective import taxes contribute to a hidden anti-export bias. Finally, from the demand point of view, all import taxes raise the relative price of importables vis-A-vis exportable and nontradable goods, reducing the demand for imports, and hence reducing the demand for and the value of foreign exchange earned by exporters. 5. The focus of this note is on trade policy. Trade reform in Uganda is considered as an unfinished agenda, hence the paper suggests further adjustments to tariffs and regional trade preferences to remove the bias against exports based on Uganda's inherent comparative advantage. It is important to note that good trade policy is only a necessary but not a sufficient condition for export growth. In order to become more competitive in the world markets the Government has to continue investing in infrastructure and education, while ensuring that sufficient incentives are in place for increased private sector participation and adoption of better technologies and managerial and marketing skills. 6. Given the poor performance of export incentive schemes in Sub-Saharan Africa, lowering of import tariffs is seen as the main policy instrument of promoting exports in Uganda.4 This would also reduce consumer prices on products that the poor majority of Ugandans use in consumption or as assets, such as bicycles and hoes. Given that the industrial sector represents a small share of total production, the cost of changing its structure will be smaller compared to what it would be later. Significant export expansion is indispensable to ease the balance of payments constraint, while diversification of the export base would make the economy less vulnerable to external shocks. But export growth will require both improved access to the regional market and opening of larger and more distant markets. 7. Sections A and B examine the present system and implementation of import taxes that contribute to the anti-export bias, and propose measures to harmonize and reduce tariffs. Given the importance of various import taxes in public revenue (18.8 percent of the cif value of imports, and over 50 percent of public revenue during the first half of 1995/96), it is necessary to ensure that the trade reform will not contribute to a sudden revenue loss which would undermine the credibility of the reform. The revenue implications of the proposed tariff changes are investigated in section C. Given the strong regional preference that characterizes the current Ugandan system of trade taxes, it is useful to discuss the extra-regional and regional trade regime separately. There are currently no export taxes. With 22 percent of the total, petroleum products are the largest single source of public revenue. The available Customs data include only a small fraction of imports of oil products; hence the data analysis in this paper covers mainly non-oil merchandise imports. Another omitted source of imports is smuggling. 15 A. EXTRA-REGIONAL TRADE POLICY 8. In order to reduce effective protection, the level and dispersion of import duties applied to extra-regional trade require further reduction. With a few exceptions, there are four non-zero duty rates in extra-regional trade, varying from 5 to 30 percent. Although the nominal rates do not appear very high, effective protection, as demonstrated in Table 1 above, remains high, given the low value added in domestic production at world prices. Table 2 shows the composition of non-PTA trade by duty category. As can be seen, the share of imports in the 10-, 20- and 30- percent categories fell considerably in 1995/96 compared to the previous years (from 89 to 58 percent), while the zero-rated and the 5-percent tariff category for raw materials, introduced in 1995/96, had a corresponding increase in their share (from 11 to 42 percent). 9. The largest increase during the first half of 1995/96 was in zero-rated capital goods (from 11 to 25 percent). Previously, licensed investors were exempted from duties on all capital good imports under the Investment Code, but these exemptions were replaced by zero rate in the Tariff Code in 1995/96. Initially, zero-rating was to be non-exclusive (all importers would be face the same rate) but restrictive in terms of goods to be included. However, pressures from investors who lost their privileges under the Investment Code resulted in a considerable expansion of the definition of capital goods. Table 2: Non-PTA Trade: Composition of Imports by Duty Category, 1994-95 As Share of Total ciflmport Value (%) 10% 5% 10% 20% 30% Total Exempted 1994 Jan-June 10 - 36 31 23 100 49 July-Dec 10 - 36 31 23 100 49 1995 Jan-June 11 - 39 28 22 100 52 July-Dec 25 17 23 16 19 100 19 Source: Calculations from Customs (URA) data 10. In order to see whether the observed shift in the composition of imports can be attributed to the policy change, we apply the previous year's tariff rates to the 1995/96 imports (July- December). The resulting composition confirms that the shift was entirely policy-induced. In other words, importers were able to declare most of their previously exempted capital goods at zero rate. If there was any misuse of the exemptions under the Investment Code, it was only transferred under another heading. 11. To determine whether the increase in capital goods is excessive, we classify Ugandan imports by the UN Broad Economic Categories (Table 3). As 15 percent are actually capital goods, the extent of zero-rating for capital goods seems excessive. However, the net effect of the new policy on exemptions was positive as only 8 percent of the 5-percent duty category for raw materials was exempted. 16 Table 3: Non-PTA Trade: Imports by Broad Economic Category, 1994-95 As Share of Total cif Imports Value (%) 1994 1995 Jan-June July-Dec Jan-June July-Dec Capital goods 12 14 17 15 Vehicles 20 15 17 18 Intermediate goods 48 49 46 47 Consumer goods 20 22 20 20 Total 100 100 100 100 Source: Calculations from Customs (URA) data 12. Efficiency in duty collection depends on the Customs administration and the extent to which exemptions are being granted. Given wide-spread exemptions until recently and inefficiencies in tax collection, the effective (actually collected) average tariff has varied between 7.8 and 8.9 percent of the cif value of imports during 1994-95. Three observations stand out from Table 4. First, the collection efficiency was only little over 50 percent prior to 1995/96. Second, duty collection did not keep up with the import surge resulting from the coffee boom and collection efficiency fell by 5 percentage points during the latter half of 1994/95. However, part of the fall can be attributed to increased imports of capital goods. During the first half of 1995/96 collection efficiency improved substantially (from 49 to 75 percent) thanks to zero-rating a quarter of imports and subjecting imports to a new 5-percent band. Third, collection efficiency tends to be higher in lower tariff bands. Now that exemptions of raw materials have largely been curtailed, this relationship seems to hold. Table 4: Non-PTA Trade: Collection Efficiency and Effective Average Tariffs, 1994-95 In Percent 5% 10% 20% 30% Overall Effective Efficiency Average Tariffu 1994 Jan-June - 36 68 48 52 8.8 July to Dec - 39 69 47 54 8.9 1995 Jan to June - 37 65 42 49 7.8 July-Dec 96 100 81 53 75 8.9 Source: Calculations from Customs (URA) data. I/Calculated as ratio of total tariffs collected and, total cif value of imports. 13. To make the incentive regime more favorable to the production of exports, Government should, in the interim, consider lowering the highest rate and consolidating duties into two non-zero rates, say, 20 and 10 percent. As final goods would have 20 percent tariff and intermediate goods and raw materials 10 percent, this option would entail some cascading of protection related to the stage of 17 related to the stage of production. Capital goods should continue to be zero-rated but in order to prevent misuse of zero-rating, their share should not be expanded on an ad hoc basis beyond the conventional classification of such goods. The proposed tariff system would not only lower effective protection, but with removal of a large number of exemptions, it will make the system of trade taxes more transparent and easier to administer. The Government has expressed interest in imposing a low uniform tariff (12 percent, for example) on all extra-regional imports as its medium-term goal. When implemented after rather than prior to a positive trade shock (the coffee boom), trade liberalization stands a much better chance of being perceived credible by the private sector. B. REGIONAL TRADE POLICY 14. In PTA-trade which is over 30 percent of total imports, mainly from Kenya, the overall level of tariffs is low but their dispersion is wide. In 1995/96, for example, there were 22 non- zero tariff bands solely for PTA trade, which unnecessarily complicates customs administration. Although a few prohibitively high PTA tariffs exist for protective purposes, over 80 percent of the total cif value of PTA trade has a tariff of 10 percent or less (Table 5). During the first half of 1995/96, as much as 74 percent of regional trade had a 6-percent tariff or below. Table 5: PTA Trade: Composition of Imports by Duty Category, 1994-95 As Share of Total cif Import Value (%) 0% 2-4% 5-6% 7-8% 9-10% 11-15% >15% 1994 Jan-June 29 29 11 1 11 15 4 July-Dec 12 37 14 1 17 15 4 1995 Jan-June 14 35 15 1 16 13 6 July-Dec 17 38 19 1 12 11 2 Source: Calculations from Customs data, URA 15. The composition of regional trade is different from Uganda's trade with the rest of the world: intermediate goods are about three quarters of total imports and the rest is mostly consumer goods (Table 6). 6 In this regime, exporters would not be able to obtain their imported inputs at world prices. A duty drawback system, if working adequately, could achieve this objective. 18 Table 6: PTA Trade: Imports by Broad Economic Category, 1994-95 As Share of Total cif Imports Value (%) 1994 1994 1995 1995 Jan-June July-Dec Jan-June July-Dec Capital goods 2 5 4 3 Vehicles 2 5 4 3 Intermediate goods 78 71 74 78 Consumer goods 18 19 18 16 Total 100 100 100 100 Source: Calculations from Customs (URA) data 16. Collection efficiency of PTA tariffs is high (Table 7). This can be expected as rates overall are low and hence there is less incentive to evade taxes or seek tariff exemptions. But this may also reflect tax evasion if goods are being excessively declared as regional imports to take advantage of a low tariff. In the coffee boom year 1994/95 there was a substantial fall in collection efficiency from 90 to 81 percent. 17. In order to harmonize the PTA/COMESA duty rates Government replaced the 22 duty bands by five non-zero bands in July 1996 (2,4,5,6 and 12 percent). In line of the 80 percent regional preference, announced in the 1996/97 budget, and taking into account the above proposal of 0-10-20 percent rates for extra-regional trade, regional trade should carry two positive rates, 2 and 4 percent. In some instances this structure could lead to modest negative protection, as inputs from outside the region would carry a duty of 10 percent. Table 7: PTA Trade: Collection Efficiency and Effective Average Tariff 1994-95 As Share of Total cif Import Value (%) 2-4% 5-6% 7-8% 9-10% 11-15% >15% Overall/ AEDZ 1994 Jan-June 90 91 103 82 110 43 90 5.2 July-Dec 84 81 103 75 92 59 81 5.5 1995 Jan-June 92 88 78 78 84 69 82 5.5 July-Dec 119 98 100 81 88 62 93 5.0 Source: Calculations from Customs data, URA I/ The overall efficiency rate takes into account all tariff bands. Efficiency rates over 100 percent are likely to reflect mid-year changes in the Tariff Code. 2/ Average Effective Duty calculated as the ratio of tariffs collected and total cif value of imports. 18. To protect domestic industries from outside competition, particularly from Kenya, Government introduced a surtax of 10 percent on about 50 items in July 1996. This measure is 19 supposed to replace the previous ad hoc excise duties and tax waivers for domestic producers.7 As long as excise duties remain an option to provide ad hoc protection, firms will continue to seek to get them. If granted they aggravate the anti-export bias and reduce competitiveness (even if they are now included in ministerial votes). Also their vetting has an opportunity cost in terms of scarce staff resources available for tax policy being spent on exemption applications. Therefore, as a matter of policy, protective excise duties should be discontinued in their entirety as soon as possible. 19. Given that Uganda already allows highly preferential access to most products originating from the PTA region to its markets, there should be a strong incentive to seek similar treatment for Ugandan exports, which are mostly agricultural products (90 percent of Uganda's PTA exports). The Government could therefore adopt a more pro-active role in its trade policy by initiating negotiations based on reciprocity within the region, particularly with Kenya, to promote regional trade liberalization of agricultural products. This would also have a positive impact on rural incomes including the poorest households as one third of the agricultural income of the two lowest expenditure quartiles is derived from maize and beans which are the main export items regionally. The 'bargaining' approach to trade policy may not bring immediate results, but since significant potential for further export expansion exists in the region, it could in the longer term result in a more binding commitment to liberalization. C. REVENUE IMPLICATIONS OF THE PROPOSED TARIFF CHANGES 20. Cross-country evidence shows that initial conditions with regard to tax structure and the level of fiscal deficit matter for successful trade liberalization. The high dependence on import taxes for revenue and the emphasis of reducing fiscal deficit typically constrain the reduction in the tariff level. In many developing countries trade taxes have also been used as an endogenous policy measure, either to prevent run-down of reserves in a less flexible exchange rate regime, or to prevent fiscal deterioration due to an unexpected revenue shortage (The World Bank, 1992). This of course can compromise the credibility of liberalization with the private sector. First, protective excise duties have been imposed on selected imports from the PTA countries as countervailing duties in response to higher protective tariffs in the neighboring countries. These imports are wheat flour, matches, sacks and bags, raw cane sugar, bicycles and bicycle parts, soap products and wheat flour. Second, a number of firms have been granted waivers from taxes on domestic production, while these are imposed on imports. Locally produced furniture, for example, bears a lower sales tax than furniture imports. Radios, tv sets and some steel products have been added to this list. 20 Box 3: CROSS-COUNTRY EXPERIENCE IN TRADE LIBERALIZATION AND FISCAL ADJUSTMENT A World Bank cross-country study on trade policy reforms and their sequencing examines two groups of countries: more intensive adjusters - Ghana, Indonesia, Mexico, Morocco and Turkey - and less intensive adjusters - Cote d'Ivoire, Colombia, Jamaica and Pakistan. The former group reduced both quantitative restrictions and tariffs, while the latter failed to reduce the tariff level. The difference in the extent of tariff liberalization is explained by the degree of dependence on import taxes for revenue. The higher dependence on trade taxes for revenue constrained the ability to reduce the import tariff level consistent with fiscal stabilization. One of the conclusions is that a domestic tax reform that reduces excessive dependence on trade taxes for revenue can successfully complement tariff liberalization. The less intensive adjusters also started with a higher level of fiscal deficit compared to the more intensively adjusting countries. Both groups improved their fiscal balance partly by converting non-tariff barriers into tariffs and hence transferring private rents to the government. Finally, the more successful adjusters pursued a more active exchange rate policy than the other group. Source: The World Bank, 1992 21 When comparing import duties (as percentage of total imports), Ugandan average tariff rates at present are at the levels of the more intensive adjusters as defined in the above World Bank study. Only Indonesia and Mexico had a lower average tariff than Uganda today. Furthermore, Uganda has maintained fiscal stability since 1992 by essentially running a cash budget, financed by tax revenue, grants and foreign concessional borrowing (but without any domestic borrowing). The potential revenue loss due to further liberalization is also mitigated by two factors. First, Government is a net seller of foreign exchange. When import taxes are reduced, the demand for imports increases, and the Shilling depreciates improving the Government's position. Second, lower tax rates result in better compliance and encourage the growth of the formal sector which allows the tax base to expand. 22. As shown in Table 8, consolidation of non-PTA tariffs into two rates would reduce public revenue slightly, assuming current collection efficiency and composition of trade (reduced by 4 percent). If all discretionary exemptions were removed, the proposed reform would be more or less revenue-neutral. However, it can be expected that the composition of trade would change as a result of trade liberalization which alters relative prices, but such changes and their revenue implications are difficult to forecast. Imports that now have a tariff rate above 20 percent are likely to increase, hence potentially increasing public revenue, but imports in the lower tariff brackets may shrink considerably if reduction in protection causes firms to produce less output and hence import less intermediate inputs. This is one of the risks of the reform. 21 Table 8: Non-PTA Trade: Revenue Implications of Proposed Tariff Harmonization In U Sh Millions Current 0% 5% 10% 20% 30% 60% Total Imports 83,351.7 54,433.2 74,044.4 50,303.0 60,361.3 30.3 322,523.9 Duty 656.6 2,620.8 7,399.2 8191.3 9,672.4 4.7 28,545.0 Efficiency (%) 96.3 99.9 81.4 53.4 25.8 74.5 Proposal 0% 10% 10% 20% 20% 20% Total Duty; Current 0.0 5,241.6 7,399.2 8,191.3 6,448.3 1.6 27,281.9 efficiency Change 100.0 0.0 0.0 -33.3 -66.7 -4.4 Duty; Improved efficiency 0.0 5,171.2 6,664.0 8,551.5 8,450.6 4.2 28,841.5 Change (%) 97.3 -9.9 4.4 -12.6 -9.7 1.0 Uniform rate 12% 12% 12% 12% 12% 12% Total Duty; Current efficiency 9,631.6 6,289.9 8,879.0 4,914.8 3,869.0 0.9 33,585.2 Change (%) 140.0 20.0 -40.0 -60.0 -80.0 17.7 Duty; Improved efficiency 9,502.1 6,205.4 7,996.8 5,130.9 5,070.3 2.5 33,908.1 Change (%) 136.8 8.1 -37.4 -47.6 -45.8 18.8 Source: Calculations from Customs (URA) data. Note: Current values and rates based on July-December 1995 data. 23. The uniform rate of 12 percent would be revenue enhancing (increased by 19 percent), given the present pattern of imports. This scenario assumes that the uniform tariff will be imposed on capital goods as well. However, it may turn out to be difficult to tax investment, given that there is not likely to be domestic production of capital goods. As capital goods are a large share of imports, their zero-rating in the 12-percent uniform tariff scenario would reduce revenue substantially (by 16 percent).8 24. In PTA trade (Table 9), assuming the current composition of imports, revenue would fall substantially (by 59 percent). However, in absolute terms PTA duties are only a small proportion of total revenue. In the interim, the 10 percent surtax (which is expected to remain until the other countries reach Uganda's tariff level) would compensate part of the revenue loss. It is assumed here that taxation of petroleum products (which are not included in the import data provided by URA and hence these calculations) remains at the present level. The high taxation of petroleum aggravates the anti-export bias by raising costs. An operational duty drawback based on input coefficients could help exporters to overcome the bias created by high taxation on petroleum. 22 Table 9: PTA Trade: Revenue Implications of Proposed Tariff Harmonization In U Sh Millions Current 0 % 2-3 % 4-5 % 6-9 % 10-15 % 15 % Total Imports 17,201.6 38,397.3 19,658.6 5,512.9 15,594.8 2,042.9 102,408.2 Duty 39.9 1,376.8 955.3 777.5 1,668.0 339.7 5,157.2 Efficiency(%) 120.6 98.3 95.1 81.8 61.9 93.6 Proposal 0% 2% 2% 4% 4% 4% 4% Duty; Current 0.0 755.1 386.6 362.0 510.0 50.6 2,9064.2 efficiency Change (%) -45.2 -59.5 -53.4 -69.4 -85.1 -60.0 Duty; Improved 0.0 755.1 386.6 372.9 530.2 69.5 2,114.3 efficiency Change (%) -45.2 -59.5 -52.0 -68.2 -79.6 -59.0 Source: Calculations from Customs (URA) data. Note: Current values and rates based on July-December 1995 data. Collection efficiency exceeding 100% reflects mid-year changes in HS Code. Technical Note 4: Estimates of Effective Protection and Manufacturing Efficiency  ESTIMATES OF EFFECTIVE PROTECTION AND MANUFACTURING EFFICIENCY' Introduction 1. This paper sets out to examine the incidence of effective protection and manufacturing efficieny in Uganda. From the basic data on imports by PTA and non-PTA and value added, effective protection rates under different broad tariff scenarios have been estimated. Finally, there is a discussion of how the application of tariffs and exemptions have impacted on two sectors that are crucial to the rural sector and by implication the poor. The Data 2. Data were collected at the enterprise level by using a questionnaire supported by company visits during September to December 1994. A summary of the methodology is included as an annex. 3. The survey results are presented in Table 1 at the 4 digit ISIC level in the tables rather than at the company/product level and at the individual HS 6 digit code. Most companies produce more than one product that is likely to fall with this 4 digit ISIC. These products would come under many different HS codes and indeed may be in different chapters of the Customs Book. The analysis in the text refers to the main product of the survey companies. The final scenario is, however, presented by main product grouping. 4. Sixty-one companies returned usable questionnaires with seventeen in the primary and food products sector, twenty-six in the intermediate products sector and eighteen in the final goods sector. These survey companies employed 6,029 people. Output was sold primarily to the domestic market. The only exports were fish (rest of world), leather, soap and hoes (PTA). 5. The survey shows that materials inputs were in the main sourced domestically or when imported these came from the rest of the world and from the PTA. Effective Protection Rates and Domestic Resource Cost Ratios 6. The EPR analysis shows that moderate levels of protection are the rule with some exceptions: fish (exported, and consequently negatively protected), one animal feed producer and one paper packaging producer (do not get duty exempted inputs as competitors do and are negatively protected). Some products have negative value added at world prices that indicates that the manufacturing process destroys value in Uganda rather that adds it: chickens, ketchup, animal feeds, foam blocks, concrete blocks (2), steel goods (1), steel wool (1), matches, garments (2), plastic shoes, leather shoes and TV assembly (1). 7. The DRC analysis presents an encouraging picture in terms of the potential for efficient production: * There are ten companies that indicate negative valued added at world prices. These would be expected to fail under greater competition from imports that a CET would This note was prepared by Mr. John Short. The results shown here are part of wider study carried out by REPIM in association with IMANI Development on the impact of a common external tariff (CET) for the Common Market of Southern and Eastern Africa (COMESA). 25 generate. (Chickens, animal feeds, foam, concrete blocks (2) steel goods (1), matches, garments (2) and TV assembly (1).) However the existence of sample companies in some of these sectors exhibiting economic efficiency would suggest that a Ugandan company should be able to capture their market and thus expand without the consequent employment loss. * Forty companies have a DRC less than one at using the DRC3 measure. These would be expected to expand both in the domestic market and the export markets. (Fish proc, meat proc, maize, flour (5), tea (2), cooking oil, chalk, paint (1), paper packaging (1), timber, plywood, bricks (2), batteries, steel goods (4), wire mesh, steel wool (1), hoes, weaving, cigarettes, soap, paper tissues, garments (1), blankets, shoes (2), aluminium products, TV assembly (1), furniture (2) and mattresses.) * Of the remaining eleven companies some protection would ensure that they might be financially viable. They might survive depending on the extent of competition from PTA imports under low tariffs and provided they could improve efficiency. (Plastic shoes, leather shoes, steel wool, concrete blocks (1), printing, leather, paint (1), paper packaging, industrial chemicals, animal feeds, ketchup, and Pharmaceutical, Paint, Metal Goods, Yarn/Fabric, Soap, Clothing, and Wooden Goods.) As some of these products are in the expansion category, the possibility of greater efficiency is demonstrated already. Indeed, leather is exported which indicates some competitiveness in a particular sub-sector (semi-processed). However, if some of these companies failed, the existence of sample companies in some of these sectors exhibiting economic efficiency would suggest that a Ugandan company should be able to capture their market and thus expand without the consequent employment loss 8. In Table 2, EPR estimates are presented for two scenarios with the following assumptions: Scenario 1 All competing imported outputs are from non-PTA sources and inputs are obtained duty free following the duty remission granted in 1994. (EPRrow) Scenario 2 All competing imported outputs are from PTA sources and inputs are obtained duty free following the duty remission granted in 1994. (EPRp,) 9. These give an upper and lower boundary of the effective protection that companies face depending on the mix of competing imports from PTA and non-PTA sources. 10. The estimates in Table 2 indicate that effective protection under each scenario is positive. The range of EPR in both of these scenarios and the actual EPR is as follows with the number of companies with EPRs in the broad bands: EPR <0 1>25 25>50 50>100 100> -VA Actual 3 12 21 7 4 14 Scen 1 0 5 17 14 11 14 Scen 2 0 32 11 4 0 14 While EPRs are the lowest in Scenario 2 relative to the other two cases, EPR is still sufficiently high to encourage domestic production. 26 Table 1: Uganda - EPR AND DRC Estimates ISIC EPR DRC1 DRC2 DRC3 Primary Products And Food Maize 17 0.72 0.68 0.58 3144 -262 1.47 0.67 0.68 3111 30 0.24 0.21 0.16 3112 10 0.13 0.12 0.13 3111 -VA -VA -VA -VA 3116 34 0.65 0.53 0.62 3116 42 0.14 0.10 0.12 3116 28 0.33 0.32 0.32 3116 69 0.38 0.38 0.30 3116 43 0.37 0.36 0.28 3121 31 0.36 0.35 0.28 3121 32 0.36 0.33 0.33 3115 25 0.23 0.22 0.22 3121 -VA -VA -VA 2.62 3122 -VA -VA -VA -VA 3122 -96 3.74 3.03 1.55 Intermediate Products 3511 45 2.23 1.74 1.82 3521 117 3.11 3.11 1.91 3521 44 0.43 0.37 0.37 3419 23 2.28 2.04 1.66 3419 -49 0.44 0.42 0.34 3560 -VA -VA -VA -VA 3231 95 4.95 4.59 1.92 3311 21 0.42 0.35 0.37 3311 31 0.17 0.16 0.16 3420 4 2.61 2.50 1.48 3691 32 0.10 0.10 0.10 3691 10 0.35 0.33 0.34 3691 63 2.8 2.35 1.92 3699 -VA -VA -VA -VA 3699 -VA -VA -VA -VA 3699 12 0.30 0.23 0.27 3830 136 3.60 1.99 0.96 3813 50 0.23 0.23 0.20 3813 -VA -VA -VA -VA 3813 31 0.24 0.23 0.21 3813 41 0.21 0.19 0.20 3813 90 1.21 1.17 0.83 3813 7 1.20 0.85 0.40 3819 -VA -VA -VA 1.86 3819 29 0.35 0.32 0.32 3811 12 0.43 0.40 0.38 3211 4 1.46 0.46 0.45 27 Table 1: Uganda - EPR and DRC Estimates ISIC EPR DRC1 DRC2 DRC3 Final Products 3529 113 0.65 0.49 0.58 3699 -VA -VA -VA -VA 3523 78 0.43 0.38 0.29 3419 46 1.49 1.22 0.98 3220 66 0.71 0.60 0.45 3220 -VA -VA -VA -VA 3220 -VA -VA -VA -VA 3212 3 1.80 1.02 0.66 3240 -VA -VA 2.19 2.44 3240 -VA -VA -VA 1.86 3240 123 0.71 0.55 0.37 3240 41 0.23 0.18 0.21 3811 21 0.14 0.14 0.13 3830 -VA -VA -VA -VA 3830 29 0.33 0.33 0.29 3320 43 0.26 0.25 0.23 3320 36 0.22 0.19 0.20 3320 48 0.19 0.15 0.18 28 Table 2: Uganda - EPR Estimates ISIC EPRow EPR Primary Products And Food Maize 31 9 3144 252 76 3111 31 28 3112 17 5 3111 -VA -VA 3116 39 12 3116 62 19 3116 45 14 3116 81 24 3116 60 18 3121 36 11 3121 50 10 3115 46 14 3121 -VA -VA 3122 -VA -VA 3122 118 36 Intermediate Products 3511 95 28 3521 153 46 3521 58 17 3419 28 9 3419 39 12 3560 -VA -VA 3231 179 54 3311 70 21 3311 42 13 3420 49 15 3691 37 11 3691 12 4 3691 112 34 3699 -VA -VA 3699 -VA -VA 3699 17 5 3830 206 62 3813 57 17 3813 -VA -VA 3813 42 13 3813 37 11 3813 107 32 3813 152 45 3819 -VA -VA 3819 43 13 3811 17 5 3211 37 11 29 Table 2: Uganda - EPR Estimates ISIC EPR,. EPRta FINAL PRODUCTS 3529 136 41 3699 -VA -VA 3523 154 46 3419 82 24 3220 99 30 3220 -VA -VA 3220 -VA -VA 3212 95 29 3240 -VA -VA 3240 -VA -VA 3240 183 55 3240 62 19 3811 22 7 3830 -VA -VA 3830 57 17 3320 70 21 3320 49 15 3320 57 17 30 ANNEX Methodology The effective rate of protection is designed to capture the protection accorded to value added in production rather than to the finished product. It is defined as the difference between value added per unit of output in domestic prices and value added in world prices, expressed as a percentage of the latter. The effective rate of protection (ERP) is equal to Vad-Vaw ERP= ------- .100 Vaw where Vad represents value added at domestic prices and Vaw represents value added at world prices. The result can be positive or negative depending on the sizes of the two measures of value added. However, Vaw itself could be negative which indicates that domestic production is so inefficient that it destroys value. The level of Effective Protection that is measured in this report is at the company level taking account of all sales whether to the domestic market or exported. This shows how much a company is protected rather that the amount of protection given to the product or groups of products that it produces and is sold to the domestic market. The level of protection that a company that exports all of part of its output will be less than for a company that sells all of its production in the domestic market. There are no tariffs on exports. Therefore, where a company in the survey is an exporter, the actual protection of the company's production for the domestic market will be underestimated. The DRC is a measure of how worthwhile it is to use domestic factors of production (Vad) in adding further value to raw material or semi-manufactured inputs to earn foreign exchange. If the value of the factors of production used (Vad) is greater than the value of the additional foreign exchange earned or saved (Vaw) then it is not worthwhile to use them in this way. If the converse is true, the particular production process should be encouraged. This is derived from the identity for a unit of output: Economic Benefit (EB) = Price (P) less Intermediate Inputs(I) less Domestic Resources used (Vad) EB = P-I-Vad as P-I= Vaw EB = Vaw - Vad Therefore if Vad > Vaw, an economic loss is made (EB is negative); but no reference has to be made to the size of economic benefit. Thus, a company's DRC is represented most simply by the ratio of total domestic value added valued at opportunity cost to the net foreign exchange saved or earned. Thus: domestic value added per unit of product DRC= --------------------------------------------------------- world price per unit- foreign exchange cost per unit 31 The numerator represents domestic value added (Vad) and the denominator value added at world prices (Vaw) as it is the residual of revenue and tradable costs valued at border prices in the same currency unit. If the ratio is less than one it is efficient, if greater than one it is inefficient in its use of domestic resources. What the DRC measure indicates is whether a production process is an efficient user of domestic resources. Companies and sectors that could survive in the new environment are identified by using the DRC technique of analysis. Companies that are likely to be viable are identified on current performance. Efficiency gains based on the potential elimination of constraints that would lead to increased capacity utilisation and better utilisation of inputs in terms of output generated have been simulated. Domestic Resource Cost has been estimated for each company under three conditions. The first (DRC1) measures DRC at the operating capacity and at the actual levels of depreciation used. This is the actual short-run DRC. Actual depreciation values (book values) were taken for each of the capital categories and adjusted to border price equivalent for tradables with non-tradables decomposed into tradables and non-tradeables. This measure provides an estimate using some element of capital consumption. A more desirable estimate would be obtained by measuring the capital consumption where the year's capital consumption was based on the economic life of the replacement cost of the capital equipment used. This would be a long-run measure. However, even this has difficulties as the use of replacement costs, while a much better estimate of capital consumption, would not be compatible with the other inputs used. As some of the companies may be old and using old machines the input-output relationship reported for these machines will not be the same for new machines that is implied by replacement cost. Replacement investment would be of a different vintage and, therefore would incorporate operating efficiency which existing machines do not. The second (DRC2) is a very short-run measure that does not include any element of capital consumption. This avoids any distorting influence from unreliable measures of capital consumption, but in itself is deficient as it does not incorporate capital. Nevertheless, for many companies this measure is appropriate as their capital equipment is so old that its life should be extinguished and already consumed. Whatever depreciation is being used is merely for tax purposes. However, both the DRC1 and DRC2 measures provide an indication of the extent of efficiency in the absence of a concrete and definitive data set. Domestic resource cost analysis is an "art" rather than a science. The third measure (DRC3) is one that makes an adjustment for capital utilisation and potential operating efficiency gains. These could be achieved by better access to finance, better management, more appropriate marketing function, more timely spare parts, etc. This recognises that the companies may not have been operating under anything approaching optimum conditions. As a result the DRC under the first two assumptions may understate their potential if some of these problems were removed. This third DRC simulates each company operating at some level of capacity that is likely to be achievable given its potential market. It does not use installed full- capacity operations that are not realistic as each company is unlikely to be able to sell output at that level. Even if it could, the technical condition of the company's equipment would not allow it to operate at installed capacity. What is important is the process of assessing whether companies that would benefit from restructuring and have the potential to succeed, but whose potential is not established because of a variety of factors. Whether or not these hindering factors can be removed, can only be addressed by studying the company in greater detail. The DRC3 measure is used for this purpose: by simulating the effect of the removal of these factors, some indication of the potential of the company is established based on current performance. Technical Note 5: A Proposal for an Effective Duty Drawback System  A PROPOSAL FOR AN EFFECTIVE DUTY DRAWBACK SYSTEM Introduction 1. Anti-export bias exists in Uganda from a range of sources. These include the additional costs that exporters face both as producers and consumers from the imposition of import duties, as well as the greater profitability of selling to a protected domestic market than to a highly competitive export market. 2. To facilitate export growth, a simple duty drawback needs to be be implemented. The system would be designed to reimburse exporters for the additional costs on inputs that the tariff system imposes. In the interim the already established system has been revitalised, but retaining the rigities which previouly rendered it inoperatable. The MFEP has recently established a working group to consider alternative schemes including a proposal based on input coefficients payable by tax credit.. The working group is expected to complete its deliberations soon, and a workable scheme, including its administration, would be introduced by September 1996. 3. Another important export promotion measure involves streamlining export documentation. Given the considerable improvements in the data collection of the Department of Customs with the installation of ASYCUDA, the Export Certificate has become redundant. This is to be discontinued. Further, given the changes in the Exchange Control Act and that the Customs Declaration Form contains all necessary information, the Bank of Uganda declaration form (known as CD3) will be abolished. For agriculture products, the need for a Phytosanitary Certificate could be left to the importer, as such inspection is often carried out at the port of entry anyway. This note outlines the mechanics of the duty drawback that is being considered for implemenation by September 1996. The Development Of Procedures 4. Although duty drawback has been, in theory, available to exporters in Uganda, it rarely functions in practice. One problem was that a cash payment had been envisaged for drawback purposes and in a climate of government revenue shortages, cash was rarely reimbursed and never in real terms when inflation and the lengthy wait for reimbursement are taken into account. More recently money has been set aside to make payments, but the verification system is still cumbersome and time consuming. A workable system also has to use as little time as possible of customs officers as well as avoiding the regular pitfalls of the bureaucratic process. 5. For these reasons a system has been recommended that sacrifices complete accuracy in computing the amount of drawback to be granted to exporters in favour of greater simplicity. Instead of requiring proof of purchase and duty/tax paid for imported and domestic inputs which would then be refunded on proof of export, the system would thoroughly evaluate this once. The percentage of export value represented by duty, once established, would then become the "coefficient" for the export of that product until a revision is deemed necessary. This coefficient This note was prepared by Mr. John Short, REPIM, Northumberland, UK. 33 would then be used to calculate the duty drawback due to the exporter on any future shipment of the export product in question. Customs officers need do little more than a short calculation for each shipment. 6. The inclusion of tax/duty drawback on domestic inputs is crucial to the system for without it the domestic suppliers of inputs would find it harder to compete with foreign suppliers whose prices are not affected by tax in Uganda. Methodology For Calculating The Coefficient 7. In order to apply for a Duty Drawback coefficient, companies would be required to complete an application form which enables customs to break down the value of an export consignment into: (a) Value of imported inputs (b) Value of duty/tax on imported inputs (c) Value of domestic inputs (d) Value of duty/tax on domestic inputs (e) Value of wages, water, electricity (f) Value of exports. 8. A simple calculation is then made of the percentage of the consignment's value represented by duties. This coefficient would then be applied to that product from that particular company whenever it has been exported. Given that prices and duties paid on inputs do not vary greatly year by year (in most cases, especially in the manufacturing sector) then this coefficient can remain in operation for up to a few years without a revision of the coefficient being necessary. 9. If a product is exported by more than one company then the coefficient should be the same for all companies if possible. For example, if a number of companies export toilet soap then the same coefficient should be used for all exports of toilet soap. In cases of significant differences in quality then it is possible that coefficients may vary. Calculation Of Coefficients 10. Coefficients have been calculated for three exported products: cured tobacco, fresh fish and frozen fish to illustrate the methodology. The importance of duty drawback in encouraging exports can be demonstrated as it establishes the extent of duties in the value of exports. 11. The coefficient is based on material inputs both imported and domestic sourced and the export value of the good. Capital costs could be included in the calculations for duty drawback but the reality in Uganda mitigates against this as few capital goods have any book value due to age, and those which have a book value have been purchased under the Investment Code which suspends duty on capital equipment. As well, exports are often based on marginal cost pricing which would excluded a capital cost element. 34 DUTY DRAWBACK ASSESSMENT COST OF PRODUCTION FOR EXPORT GOODS TOBACCO IMPORTED INPUTS (including fuel) Shilling Million Description Total Cost Duty Paid Tariff Code Tariff Rate Non- (net of sales (%) PTA tax) 1 Fertilisers 372.5 0 3102.10 0 0 2 Iron Sheets 92.3 7.3 7210.41 10 3 3 Pybuthrin 32.7 0 3508.10 0 0 4 Seed Bed Packs 90.2 0 3404.90 0 0 5 Thermometers 5.6 0.4 9025.11 20 2 6 Carton Boxes 29.0 0 4819.10 10 7 Carton Labels 5.0 0 4821.10 30 TOTAL 594.6 7.7 Packaging material exempted due to Second Schedule provisions. DOMESTIC SOURCED INPUTS (including fuel) Shillings M. Description Cost (net of Tariff Code Tariff Rate Implied sales tax) Duty 1 Empty Drums 2.8 7310.10 20 0.47 2 Hoes 3.5 8210.30 10 0.32 3 Fuel 139.0 100 69.5 4 Buckets 2.3 7324.90 20 0.72 5 Sprayers 61.5 8424.81 10 5.59 6 Jute Twine 4.2 5303.10 10 0.38 7 Watering Cans 22.0 3917.22 30 5.08 8 Polythene Tubing 14.0 3917.22 10 1.27 9 Hessian Cloth 36.0 5303.10 10 3.27 10 Hoop Iron 2.4 7313.00 10 0.20 TOTAL 287.7 86.8 For fuel the cost is inclusive of all taxes. A 100% inclusive tax figure is used for illustrative purposes. 35 IMPLIED DUTY =(t/(100+t))*(COST) where t is the tariff rate OTHER INPUTS Million Shillings Wages And Salaries 442.0 (Including Benefits) Electricity 1 84.1 Water J Transport 650.0 Total 1176.1 EXPORT SALES VALUE Description Million Shillings Cured Tobacco 2058.4 (The actual figure for export sales was not provided by the company, and has been requested. Discussions with the company indicated that the export of tobacco has not been profitable so the total of all costs have been taken as the export value. CALCULATION OF DUTY DRAWBACK Million Shillings Total Cost Duty Paid Implied Duty Imported Inputs 594.6 7.7 xxxxxxxx Domestic Goods 287.6 xxxxxx 86.8 Other Costs 1176.1 xxxxxx xxxxxxxx Total 2058.3 7.7 86.8 Duty drawback coefficient calculation ((duty paid + implied duty)/ total export value ((7.7 + 86.8)/ 2053.3) = .046 This coefficient would be applied to the declared exported value on the customs export form. A tax credit of UgSh 94.5 million would be issued to the exporting company. 36 DUTY DRAWBACK ASSESSMENT COST OF PRODUCTION FOR EXPORT GOODS FRESH FISH (Per Kilo) IMPORTED INPUTS (including fuel) Shillings Description Total Cost Duty Paid Tariff Code Tariff Rate % (net of sales tax) (excluding duty) 1. Packaging 184 0 4819.10 10 TOTAL 184 0 Packaging material exempted due to Second Schedule provisions. DOMESTIC SOURCED INPUTS (including fuel) Shillings Description Cost Tariff Code Tariff Rate Implied Duty (net of sales tax) 1. Fish 1840 0302.69 30 0 2. Fuel 46 100 23 TOTAL 1886 23 For fuel the cost is inclusive of all taxes. A 100% inclusive tax figure is used for illustrative purposes. Basic raw materials are not included as part of duty calculation. IMPLIED DUTY =(t/(100+t))*(COST) where t is the tariff rate OTHER INPUTS Shillings Wages And Salaries (Including Benefits) Electricity, Water - Transport Total 230 37 EXPORT SALES VALUE Description Shillings Fresh Fish 2622 CALCULATION OF DUTY DRAWBACK Million Shillings Total Cost Duty Paid Implied Duty Imported Inputs 184 0 xxxxxxxx Domestic Goods 1886 xxxxxx 23 Other Costs 230 xxxxxx xxxxxxxx Total 2300 23 Duty drawback coefficient calculation ((duty paid + implied duty)/ export value 23/2622 = .0088 This coefficient would be applied to the declared exported value on the customs export form 38 DUTY DRAWBACK ASSESSMENT COST OF PRODUCTION FOR EXPORT GOODS FROZEN FISH (Per Kilo) IMPORTED INPUTS (including fuel) Shillings Description Total Cost Duty Paid Tariff Code Tariff Rate (net of sales tax) (excluding duty) 1. Packaging 92 0 4819.10 10% TOTAL 92 0 Packaging material exempted due to Second Schedule provisions. DOMESTIC SOURCED INPUTS (including fuel) Shillings Description Cost Tariff Code Tariff Rate Implied Duty (net of sales tax) 1. Fish 1840 2. Fuel 46 100 23 TOTAL 1886 23 For fuel the cost is inclusive of all taxes. A 100% inclusive tax figure is used for illustrative purposes. Basic raw materials are not included as part of duty calculation. IMPLIED DUTY =(t/(100+t))*(COST) where t is the tariff rate OTHER INPUTS Shillings Wages And Salaries (Including Benefits) Electricity I Water J Transport Total 230 39 EXPORT SALES VALUE Description Shillings Fresh Fish 2300 CALCULATION OF DUTY DRAWBACK Million Shillings Total Cost Duty Paid Implied Duty Imported Inputs 92 0 xxxxxxxx Domestic Goods 1886 xxxxxx 23 Other Costs 230 xxxxxx xxxxxxxx Total 2208 23 Duty drawback coefficient calculation ((duty paid + implied duty)/ export value 23/2300 = .01 This coefficient would be applied to the declared exported value on the customs export form Technical Note 6: Issues in Local Taxation  ISSUES IN LOCAL TAXATION Introduction 1. The decentralization of governance may be taken to mean the strengthening of an autonomous local government system, or it may be taken to mean a deconcentration of line ministries to plan and deliver services at the regional level. In the former case, accountability of local officials is to the local voters and in the latter case where the local governments are spending agents of the center, accountability is largely to the center. 2. It seems clear from the Preamble and from chapter eleven of the Ugandan Constitution, that local governments with some degree of fiscal autonomy are to be the basis for decentralization in Uganda. The Constitution requires popular election of local councils, assigns responsibility to local governments for many important expenditure functions, and defines a set of local government revenue sources. The spirit of the decentralization thrust of the Constitution is that the objective is to move government decisions closer to the people. 3. The details of the fiscal decentralization program have not yet been worked out, and particularly little attention has been paid to the revenue side of the program. This note offers some recommendations about methods that might be used for increased local government revenue mobilization. Two topics are covered. The first is the statement of a set of guiding principles that might be used in structuring a local revenue system for Uganda, and the second is a set of more specific recommendations for changes to the structure of grants and taxes. Principles In Designing A Local Revenue System 4. Decisions about fiscal reform, such as the structuring of a local government revenue system, ought to be guided by what the government wants to accomplish with the reform. Moreover, such decisions ought to be taken in a context of comprehensive reform, rather than as a tax-by-tax set of adjustments. It would seem an ideal time in Uganda to develop a revenue policy for local its local governments, and to design a timetable for its implementation. 5. From the Constitution, the Local Government Law, various position papers on decentralization, and the direction of economic policy, it would seem that the GOU wants to achieve the following with its decentralization program: * Move fiscal decisions (the size and the composition of government expenditures and financing) closer to the people. Involve more of the population in the process of governance by giving them a say in spending and taxing decisions. * Make local officials more accountable to their constituents and therefore increase the efficiency of local public services. * Increase the overall rate of revenue mobilization, and therefore provide enhanced government services This note was prepared by Professor Roy Bahl, Georgia State University, Atlanta, GA. 42 * Build capacity for governance at the local level. 6. To achieve these goals with a decentralization policy, the government faces difficult choices. On the revenue side they must decide what taxing instruments to give to local governments, what autonomy to give local governments to alter the rate and base of these taxes, what revenue sharing arrangements should exist between the center and the districts, and whether all local governments should be given the same revenue raising powers in the new system. Such questions can be better answered if the government develops a set of principles to guide its fiscal decisions. The following is one possibility for such a set of principles. 7. First, the basis for successful decentralization is popularly elected local officials, and locally appointed chief officers of the local government. If local officials are appointed, they will be accountable to the central government and not to the local population. 8. Second, local governments must have some power to set tax rates. Without local taxation, the link between government officials who preside over the delivery of services and voters who judge the adequacy of these services, is weak. Local voters must see its officials as being responsible for both providing the services which they value, and levying the taxes which pay for a significant part of these services. 9. Third, local taxes should be thought of as general user charges for locally delivered services. This means that the primary purpose of local taxes should be to recover the costs of locally provided services, and that the burden of local taxes should be borne in the area where the benefits from the services financed by those taxes are received. The latter means that local taxes should not be easily exported to other regions. The former implies that local taxes in Uganda should not be structured with a primary goal of redistributing income -- other taxes in the system are more suitable for that purpose. 10. Fourth, in so far as possible, local taxes should be relatively simple in structure and administrable at a relatively low cost. An exception to this general rule is that in Uganda, an important purpose of local taxation may be to involve more of the population in the system of governance, hence a relatively high administrative cost might be tolerated to accomplish this goal. 11. Fifth, local taxes should be designed to enhance overall revenue mobilization and not to compete with the central government for the same tax base. The goal should be to combine the comparative advantages of the central and local governments: the locals have a better familiarity with the local tax base and especially with the hard-to-tax self employed sector, and the center has a stronger overall tax administration capability and access to better information on large taxpayers. A combined central-local tax administration effort could raise total revenues. 12. Sixth, the grant system should be compatible with the goals of accountability and increased revenue mobilization. If a local government knows that its revenue shortfall will be made up with a deficit grant, it will have little incentive to push up local revenues by increasing tax rates or by pursing a more aggressive tax administration. Conversely, if a local government recognizes that its increased revenue effort will not be penalized with reduced grants, it will exert a maximum revenue effort. 13. Finally, one system for local government revenue raising powers will not fit all local governments in the nation. Urban local governments have a much greater capacity to implement local taxes than do rural areas, and so should be given a greater opportunity to become more 43 fiscally self-sufficient. For now, and likely for the long term, rural local governments will be more reliant on transfers from the center. Such a policy may serve another purpose, i.e., higher local taxes in the more urbanized areas may properly reflect the higher marginal cost of urban service provision. Opportunities For Increased Local Revenues 14. The Constitution (Article 191) gives local governments the power to raise revenue from "rents, rates, royalties, stamp duties, personal graduated tax, cess fees on registration and licensing, and any other fees and taxes that Parliament may prescribe". Neither the Constitution nor the Local Government Act give great detail, however, and it is not clear what autonomy the local governments have in setting tax rates and rates of user charges. 15. The recommendations offered here are based on an assumption that it is desired to increase local government revenues and to give local governments some discretion in setting the level of their taxes. Graduated Personal Tax 16. The graduated personal tax (GPT) is arguably the major source of locally raised revenue. Because audited financial reports for local governments are not available, it is not possible to produce firm documentation of this. However, an ODA survey of 10 municipal and 10 district councils for 1994 shows that GPT accounts for between 18 and 40 percent of own source revenues for the municipalities, and 61 and 85 percent for the districts. Though it is an extremely important source of financing for local governments, its revenue yield is not great: about U Sh 4000 per capita in Kampala City in 1994, and less than U Sh 1000 per capita in several of the districts included in the ODA survey. 17. Problems and Strengths. In fact, the GPT is an income tax. Its legal base is income. In practice, depending on the district and the subject of taxation, GPT might be described as an income tax, a wealth tax or a poll tax. The tax has 36 rate brackets, with a graduated (specific) rate up to a maximum payment of U Sh 80,000 at an annual income level of U Sh 820,000. For incomes below U Sh 820,000, the GPT is progressive. Above an income of U Sh 820,000 it is highly regressive. Since U Sh 820,000 is about half the lower threshold of the central government income tax on individuals, the tax is quite regressive over much of the income scale. However, the amounts involved (in per capita terms from $US 4 in Kampala to less than $US I in rural districts) do not suggest that the overall distribution of income is markedly effected by this structure. 18. The GPT is assessed and collected as a presumptive tax on almost all tax subjects in the rural areas, and as a presumptive tax from the self-employed in the urban areas. This is a very expensive form of tax administration, and a very subjective approach to assessing tax liability. The subjectivity and arbitrariness of presumptive assessments usually brings unfairness with it, and this is another critique of the GPT. Most believe that the informal sector is under taxed relative to the formal sector. In the urban areas (primarily Kampala), GPT is collected from formal sector employees as a withholding tax, at a relatively low cost. About 60 percent of the revenue collected in Kampala is from payroll deductions. There are no good nationwide estimates of collection costs for the GPT, but anecdotal estimates suggest that the costs are high. 44 19. Despite these shortcomings of the GPT, there are reasons why it might well serve the GOU objectives for decentralization. * It has a very broad coverage, hence it serves the purpose of involving a maximum number of people in the process of governance and taxpaying. Many would argue that this is an first important step toward the successful decentralization of governance. * It more or less fits the profile of a general user charge, that is, the burden of the tax falls within the benefit zone of locally provided services. * It is accepted by the population and a mechanism is in place to assess and collect the tax. * It is administratively difficult, but no more so than would be an expanded property tax. Moreover, the PAYE portion of the GPT in Kampala is collected at quite low cost. 20. The advantages of the GPT would seem to outweigh its disadvantages, especially in light of the alternatives available. Should the government decide to continue with the GPT as the cornerstone of local government finance, and the view here is that it should, then substantial revision in this tax is necessary. 21. Proposed Structural Reform. Many have suggested a reform of the GPT over the years. To simplify its structure, to reduce its regressivity, and to merge it with the income tax are the most often mentioned changes. These proposals have been rejected, usually for political reasons. With the new constitution and the decentralization initiative, and the functioning of a LGFC to provide guidance, the time would once again seem right to consider reform of the GPT. 22. The approach recommended is to convert the GPT to a general user charge for local public services. The fundamental change in the rate structure recommended here can result in a tax that is more fair, more revenue responsive to income growth and inflation, and simpler. Three structural revisions are proposed. (a) The present 36 rate brackets should be converted to a flat, ad valorem rate. (b) A maximum shilling payment could be imposed (as at present), but this should be set at no less than the amount that would be paid by a person earning the threshold level of the government income tax (gross income of about U Sh 1.5 million). (c) The local governments should be given some authority to set the rate of tax (within specified limits), as well as the ceiling payment. The central government could set limits on maximum and minimum rates, and on the ceiling, but the local authorities ought to be given substantial discretion within these limits. 23. These three reforms would accomplish several goals that would seem consistent with the government's decentralization strategy. (a) The revenue yield of the GPT could be significantly increased and could provide a basis for a larger local government expenditure budget. The rate and the 45 maximum payment could be chosen at a level that would generate significantly more revenue. If decentralization is ultimately to succeed in Uganda, local governments will have to provide significant amounts of services to the population. (b) Some natural growth in revenues could be built into the system with the ad valorem rate, and local councils would not have to wait on the central government to make discretionary changes in the structure. (c) Local governments would have some control over the level of revenues raised, and would become more accountable to the voters for the expenditure of this money. (d) Kampala City could be given authority to tax itself more heavily. This fits a notion that the tax rate ought to be larger in the largest urban areas to reflect the higher cost of service provision, and it fits the idea of an equalizing component to a grant structure that would require higher income local councils to become more self-sufficient. (e) The overall regressivity of the system could be lessened, if the local authority chose a higher ceiling and a lower ad valorem rate. However, one might take the view that this is more in the nature of a user charge than a tax, that other taxes in the system are better able to carry out the redistributive function, and that the magnitudes of GPT are such that major equity effects are not likely. (f) The replacement of 36 rate bands with a single ad valorem rate would be less subjective, future adjustments would be less arbitrary, and the GPT would become more understandable to taxpayers. (g) The administration of this structure should be no more difficult than at present. Collection schedules could be developed in terms of shillings to be paid at each income level, just as at present. In fact, administration could be enhanced because the revenue rewards for better capturing those in the informal sector would be significantly greater. 24. There also would be problems, and possibly substantial resistance to this program. First, while the regressivity of the overall system would be lessened, the distribution of tax burdens would become less progressive over the very low income range (U Sh 2000 to U Sh 840,000. Though this might be justified on grounds that GPT is really a user charge, a better route would be to sweep much of the very low income population out of the tax net by exemption or simply charge them a flat, nominal amount. The revenue loss to poor districts might be compensated with an equalization grant. 25. A second problem is political. The population in Uganda, as that in most countries, will resist increased taxes, and so such a reform will be unpopular. The poor will see themselves as being exploited, and the higher income will resist increases on grounds that they do not receive any services for this payment. Public reaction has gone as far as tax riots in the past. A popular argument is that people are not willing to pay local taxes because they do not receive value for money in terms of local services. But service provision is constrained by limited resources. Where does one start? Moreover, it has been three years since GPT has been adjusted, and it is 46 likely that some increase will soon be made in any case. The same sort of public reaction will be expected to this increase. 26. Third, and the biggest problem, is that all districts will not be effected the same by this proposal. First, those with significant numbers of taxpayers in the highest income brackets (above U Sh 820,000 will benefit most, and those with very few taxpayers above this range could actually suffer a revenue reduction, depending on the tax rate chosen. The way to resolve this problem is with equalization grants as are provided for in the Constitution. 27. Choosing a Rate Structure. In theory, there are many combinations of a rate level and a tax ceiling that would yield any given amount of revenue. Unfortunately, it is not possible to simulate the revenue and tax burden effects of alternative GPT structures. To do this would require data on the distribution of taxpayers at each income level. The government does not presently have such data at hand. The necessary data to carry out such analysis do exist at the subcounty and the municipal level, and it would be possible for the MLG to organize a collection program and to re-establish a regular reporting program. 28. Administrative Reforms. A second set of reforms has to do with streamlining the administration of GPT. Here one must consider the two methods of collection of GPT: (a) directly from taxpayers, and (b) as a withholdings tax that is paid by employers to the local government. Relatively little can be done about the method used for those not employed in the informal sector - face-to-face visits and presumptive assessments based on valuation manuals are required. 29. With respect to the PAYE component, however, cooperation between URA and at least the Kampala City Council can yield benefits to both. For example, the KCC coverage of the informal sector for GPT is much broader than that of the URA for income tax. Even though the taxable value as determined by KCC might not be directly useful to URA in determining income tax liability, the KCC tax rolls provide information that would assist URA in identifying informal sector members who should be on the central government income tax roll. The taxation of the self-employed in Uganda, as in most developing countries, largely excludes the self employed from the tax base. If presumptive assessment information would assist URA in expanding the coverage of the income tax, then increased revenue mobilization would result at relatively little administration cost. 30. The KCC could also benefit from cooperation. The ad valorem GPT could be collected along with the income tax, and turned over to the Local Government. The KCC would also benefit from audits undertaken by URA, an activity that the KCC is ill equipped to carry out. 31. Implementation. It may not be possible for GOU to move from the present system to one such as that recommended here in one step. And the timing may be particularly bad in that this reform may coincide with local elections or the tenure of newly elected local officials, and would provide a ready platform for anti tax campaigns. Moreover, this program would be coming on the heels of the newly introduced VAT. Another problme is that the system of equalizing grants, which will need to be introduced hand-in-hand with the new GPT system, has not yet been implemented. 32. As an alternative, the government might introduce an intermediate reform that goes some distance toward this program, with the goal of fully introducing the system within the next three years. Such an intermediate program could include the following: 47 (a) Cut the number of rate brackets below U Sh 820,000 from 36 to 5. This is in line with the proposal made by ODA. Cover the very poor with a flat charge. (b) Add three new rate brackets, New Tax Bracket (in thousands of U Sh) Income Tax 821 - 1000 U Sh 90 1001 - 1500 U Sh 100 above 1500 U Sh 150 or 1% of taxable income, whichever is greater, up to an absolute maximum of U Sh 200 (c) Collections from all taxpayers covered by the government income tax would be made by the URA and turned over to the local government. A cooperative agreement between the URA and the local governments would provide for an exchange of information. (d) Introduce equalizing grants to compensate poor districts for the revenue loss from the new structure, taking care not to reward slack effort. 33. This intermediate program would move the GPT in the direction of a flat tax and would lessen the shock at the time of full introduction, say in three years. It would improve the elasticity and lessen the regressivity of the current system, and it should be no harder to administer. It could lead to significant increases in revenues for Kampala City, a conservative estimate for the rate schedule proposed above is between a 10 and 15 percent increase, if no increased rate of compliance is assumed. 34. There would be problems with this intermediate system. The wider rate brackets would create notches that would cause marginal tax rates to jump by great amounts as persons moved into a higher class, thereby increasing the incentive to evade; the higher tax rates would further penalize those already in the system in favor of those who presently avoid all income taxes; and the distribution of tax burdens would become regressive after the ceiling level of taxable income was reached. The latter two problems will characterize any reform that is likely to be made. Rental Income tax 35. The central and the local governments both levy a tax on the rental income received. The central government taxes 80 percent of the income earned by individuals from rental properties. After allowing the standard deduction of U Sh 1.5 million, rental income is taxed at a rate of 20 percent. Virtually all of the revenue (abut U Sh 3 million) is collected in Kampala. At the same time, the local councils collect a property tax assessed against the rental value of properties. The basic tax rate is 10 percent. Hence the same owner is subject to separate taxes on essentially the same base. 48 36. The central government individual income tax on rents is a good candidate for a local tax, in particular, for an urban government tax. Three arguments might be made in support of this proposal. First, essentially the same base is being taxed twice, but different sets of deductions are allowed, and different rates are applied to these adjusted bases. The overall tax effect on residential property investment is as a consequence unknown, and certainly not planned. Second, there are economies of scale in cooperative if not joint administration. Third, the rental income tax fits the maxims for a good local tax -- its burdens are locally borne and property owners do benefit from local public services, arguably in proportion to the letting value of their property. Fourth, a sharing of the rental income tax would increase local government revenues, primarily in Kampala. 37. The rental income tax on individuals could be converted to a shared tax with local governments, each receiving 50 percent of the collections. Responsibility for collection and audit would remain with URA, but a cooperative program of information exchange would be put in place. 38. There could be advantages in tax administration to both parties under such a sharing arrangement. Jointly, they could identify a tax base that neither could do separately. (a) The number of individuals registered as liable for the URA administered income tax on rents could be significantly increased. This is because the KCC now maintains a roll of properties, owners, and rental value estimates. It is conceivable that this roll is more extensive than the roll of rental value properties now in place at URA. (b) The KCC information on rateable value would give URA a cross check on the level of declared rental income, and could increase the tax base. Likewise the level of declared rent income and the result of URA audits could give the KCC additional information that might expand its tax base. 39. The revenue yield for the KCC could be significantly increased by such an arrangement. Based on 1994/95 data, about U Sh 2.8 billion in rental income tax was collected nationwide, and about 90 percent of this came from the Kampala collection districts. With a 50 percent sharing rate, the KCC would receive about 1.2 billion in additional revenues, which would be equivalent to about 10 percent of the total amount raised from their own tax sources, about 50 percent of the amount raised from the GPT, and about three fourths the amount raised from the property tax. 40. On the negative side, this could also be viewed as a revenue loss to the central government. In 1994/95 the rent income tax accounted for 2.5 percent of central government revenues. Further, it might be possible to reduce grants to KCC as a compensation for this increased local revenue. Certainly the shared local tax is more of an unconditional transfer than the present grant system, and hence this is consistent with the decentralization thrust of the government. Moreover, if the cooperative administration has the effect of expanding the tax base, the hit on central government revenues will be less and the gain to local government revenues will be greater. Property Taxation 41. The property tax is a natural source of local government revenue, and it meets many of the criteria for a good local tax in Uganda. It is already in the system and accepted by the population, it can be revenue productive, it can be progressive since property ownership is 49 concentrated in the hands of higher income persons, and its burdens fall within the local area where it is administered. Local governments have a natural comparative advantage in property tax administration, because of their familiarity with land ownership and land use in their jurisdiction. 42. The property tax is an emerging source of revenue in Uganda. The experience is mixed. All district councils may levy the tax, but at present, none have chosen to do so. The updated land valuation schedules are not in place and the skilled personnel to carry out this valuation work is limited. It would not seem likely that the property tax will become a mainstay of local financing for the rural councils in the near future. 43. The story is only somewhat more promising for the urban councils. Many use the tax sparingly or not at all. However, the larger urban councils do rely on the property tax. It accounts for about 16 percent of total own source revenue in KCC and for nearly 30 percent in Jinja. The valuation of properties is nearing completion in Kampala, and collection is now contracted out. However, even in Kampala, where one would expect the property tax to have its best success as a revenue source, there are questions: (a) At present, the KCC has only one qualified valuer. Even with contract valuation a possibility, maintaining a valuation roll with updates every fifth year will be a formidable task. It is estimated that at least six valuers are needed. (b) The property tax system is not yet computerized and the record keeping requirements for an effective property tax are quite substantial. There is need to merge the records on ownership and deeds to keep a current file for billing purposes, and at least on larger properties, new construction must be tracked. (c) Payment and delinquency records must be computerized if enforcement is to be properly supported. The success with program such as the cooperative arrangement with the URA suggested above will depend on the accuracy and the completeness of these records. 44. The revenue potential for the property tax is substantial. KCC estimates revenues for 1996/97 of about U Sh 2.7 million. However, the collection efficiency appears to have been running at about 50 percent historically, and the base would appear to be undervalued. This suggests that actual property tax revenues will yield about the same amount in Kampala as do business licences and market charges. 45. Our view is that the property tax can became an important source of revenue for the urban councils but not for the districts, at least not in the near future. There are several reasons why rural local governments might not be well advised to pursue property taxation. One is that the administrative costs of valuation and preparing a roll are prohibitive relative to the revenue that might be raised, and will take time. Another is that the district councils are not yet equipped to maintain a property roll and carry out a collection function. Finally, and perhaps most important, the GPT as it is assessed in the rural areas is in fact a tax on property values, capitalized by a schedule provided by the central government. It is in fact, a much cheaper form of administering the property tax in the rural areas. To impose ratings on top of this would be to tax much of the same base again. 46. We recommend that the property tax be strengthened as a major source of urban local government revenue. On this point we are in agreement with the ODA study which calls for "... 50 remedial action to re-establish property tax/rates as a key urban tax revenue source." The development of an action plan to achieve this goal, however, requires further study. Grants and Expenditure Discretion 47. The major decentralization initiative undertaken by the government thus far is the creation of a grant system. The Constitution calls for (a) conditional grants, (b) Block Grants, and (c) equalization grants. The former were to give an incentive to honor central government expenditure priorities. The block grants were to give local governments some discretion in the allocation of funds among alternative uses. The equalization grants were to help guarantee nationally-set minimum service levels in the districts where the gap between needs and resources is greatest. 48. Initially the emphasis was on the block grants, and by now all 39 districts have been brought into the block grant system. However, the program seems to have strayed some from the original intentions. The labor component of the block grant is about 75 percent of the total, and this is now allocated among jurisdictions on a basis of the number of approved positions existing within the functions to be transferred. Hence local governments have discretion only as regards the expenditures for the non wage component (which is allocated by formula) and even in this case there is a movement to introduce some earmarking. 49. There is no plan to introduce the equalization grants soon. Apparently, there is no agreement on how to measure the level of needs called for in the constitution, and work in this area has not proceeded. Moreover, it is unclear whether the staff responsibility for this work rests with the MLG or the LGFC. 50. The Bank recognizes that decentralization is a government policy, and has no position other than the governments position on what is good policy in this area. However, if the GOU is going to hold its course on the decentralization program outlined in the Constitution, then the following would seem a reasonable set of next steps: (a) Determine the responsibility for designing the intergovernmental system as between the MLG and the LGFC. (b) Develop the indicators to be used in measuring needs for the equalization grant program, and develop a policy paper for implementation of that program. (c) Come to a clear decision about whether block grants are to be a discretionary revenue source or local governments, or whether local governments are to become sending agents with the center setting the policy. Then redesign the grant program to match the objective. (d) Provide a training program for those Ministry officials who are to guide and monitor the intergovernmental program in Uganda. The success of decentralization will ultimately depend on the ability of the central government to guide it. (e) Develop a readily usable census of local government finance and taxation, which all can draw on. This is an essential ingredient to successful decentralization. Technical Note 7: Outcome-Oriented Budgetary Process  OUTCOME-ORIENTED BUDGETARY PROCESS A. INTRODUCTION 1. Public services in Uganda are financed through: (i) the recurrent budget of central government ministries and institutions; (ii) the budgets of local governments; and (iii) the development budget. For 1996/97, the central government recurrent and development budgets 2 total Ush 1,281 billion , of which 32% is allocated to the recurrent budget of central government ministries and agencies, 14% as transfers to district councils3 and 53% to projects in the development budget. 2. The three-way division of the budget makes it difficult to determine resource allocations for key public services, particularly as a substantial share of recurrent financing is provided through the development budget. One consequence is that there is little assessment of the effectiveness and impact of sector expenditure programmes. This problem is accentuated by budget practices that concentrate on the detailed budgeting of inputs or expenditures rather than on the level of services to be provided. 3. This Note sets out the main elements of a budgetary reform programme that aims to make the budget process more strategic with a stronger "outcome-orientation" and which builds on the successful government initiative in introducing a medium-term expenditure framework (MTEF). It begins by outlining the budget cycle and key features of the budgetary process (Section B), and goes on to identify an agenda for budgetary reform (Section C). The final section sets out an action plan for budgetary reform under the SAC III programme (Section D). A separate Technical Note considers requirements for strengthening expenditure planning and management under decentralisation. B. THE BUDGETARY PROCESS The Budget Cycle 4. The fiscal year in Uganda runs from July to June. Traditionally the budget cycle starts with the budget call circular issued in January/February. Ministries then prepare their Estimates submissions which are reviewed by the Ministry of Finance (MOF) during April/May. The draft Estimates are endorsed by Cabinet and submitted to Parliament with the Budget in June. During the 1980s and early 1990s, the budget process faced considerable uncertainties over government revenues. This and its one-year time horizon meant there was little scope for addressing This note was prepared by Mr. Andrew Bird, Mokoro Ltd., Oxford, UK, in collaboration with Ms. Cecilia Hermansson, Ministry of Finance, Sweden and Mr. Carlos Montes, London, UK. 2 Excluding recurrent consolidated fund services. Source: draft Estimates submitted to Parliament. District councils also raise local revenues primarily from graduated personal tax and property taxes. In some districts local revenues may finance up to one third of the district council budget. 52 strategic issues on the composition of the expenditure programme, even though the need for major restructuring was widely recognised. 5. In order to provide a strategic basis for expenditure planning and budgeting, MOF has since 1993 prepared a Budget Framework Paper (BFP) for submission to Cabinet at the outset of budget preparation. The initial BFPs were concerned with establishing a realistic macroeconomic framework for the budget and setting the aggregate resource ceilings. In 1995, the BFP was expanded to include a three-year medium-term expenditure framework (MTEF) that addressed the strategic shifts in public expenditure required to meet policy objectives (Box 1). The MTEF identifies three-year resource ceilings for recurrent expenditures in each sector, broken down between central and local government functions. These ceilings provide the framework for subsequent preparation of the annual budget. Box 1: The Medium-Term Expenditure Framework (MTEF) The MTEF comprises three main elements. * a proiection of the o%erall go%ernment budget for the three-%ear period. broken don betmeen- (i go%ernment recurrent expenditure: (iii de'elopment expenditure, and (iii net lending and in' estment. * a projection of ministerial recurrent expenditures broken dow%n by sector. ministerial Volte and decentralised programmes for the three->ear period: and * a projection of sectoral financing under the de%elopment budget A ke% stage in preparation of the NITEF is the sectoral programme res that take place between the MOF-Tl1PED and the line ministries. The obiectie of the retiews is to facilitate impro%ed public serice delivern b- * ensuring that budgetar% resources are directed towards the highest priorr programmes: and * reducing allocations to programmes that. ii) are no longer a priorir\ in terms of sectoral policies and strategies: (ii although potentially important cannot be financed meaningfully gi\en budgetarx constraints, and lit fund ser ices w%hich could be adequately pro% ided %%ithout recourse to budgetar, financing. For each sector, a separate review meeting is chaired by MOF/MPED and involves the line ministry, other Vote-holders in the sector, the Ministr of Local Go'ermnent and the Ninistr\ of Public Senice The meeting briefly re% ie"s each of the main expenditure programmes ikithin a sector. current levels ot fundin- and the extent to wkhich the programme justifies a greater or lesser share of the sector budget. Based on the conclusions of this meeting. indicatie resource ceilings for each sector for the coming three-,ear period are included in the NiTEF. The introduction of the MTEF has faced a number of difficulties including: (i un%%illingness of stakeholders to consider rationalisation of their programmes. ii the limited detail in the sector re%ie% submis%ions required from line ministries: tiii) dela\ s experienced in finalising the BFP and ecninL Cabinet agreement to the resource ceilings. and ( in difficulties in holding ministn budgets to the resource ceilings Ne%ertheles,. the exercise has been successfully used by MOF to facilitate the reallocation of resources to"ards decentralised programmes. and has also pronided a stronger framework for the Expenditure Department to e%aluate ministerial budget submissions. Central Government Recurrent Budget 6. The format of the recurrent budget dates from the 1970s. Ministries and some autonomous institutions are each assigned a separate Vote in the budget which is divided 53 '4 between a number of "programmes" . Below the programme level, expenditures are allocated to individual items, each of which has a unique four digit code. The first two digits of the item 5 code reflect the broad category of expenditures and the last two digits the item of expenditures The classification used is fairly broad and expenditure allocations are grouped under a relatively 6 small number of items . 7. Line ministries typically adopt an incremental approach in preparing their recurrent budgets. Requests are based on the previous year's allocation plus a growth factor reflecting the perceived level of underfunding. Often budget requests are 2-3 three times previous allocations (even where MOF has advised that no real increase in funding can be expected). Evaluation of budget requests by MOF involves arguing over individual line item allocations, rather than the wider programme allocations. 8. Wagebill costs are treated differently from other recurrent items. These are determined by the Ministry of Public Service (MPS) and MOF, based on approved establishments and the current payroll, and there is minimal involvement of line ministries. District Budgets 9. Since 1993/94 almost one third of the recurrent budget has been decentralised to the districts7 which are now responsible for the delivery of sectoral programmes except for "national functions" (e.g. foreign policy, defence and immigration, tertiary education, main roads, and the national referral hospital). Central government funding for district level services is channelled through separate Votes for each district and involves a combination of unconditional grants (which districts are free to allocate as they wish), delegated funding for secondary education and 8 hospital services, and conditional grants linked to sectoral policy objectives . The local government legislation also provides for equalisation grants which are intended to address imbalances in service provision between districts, but which have yet to be introduced. District councils prepare an annual budget which allocates resources between the different sectoral programmes. These follow a separate programme and item classification. The recurrent budget of a typical line ministry would be distributed between 10-15 programmes. There are seven broad categories of expenditure in the recurrent budget: 10 - Employee Costs; 20 - Administration Costs; 30 - Supplies and Services; 40 - Transport and Plant Costs; 50 - Property Costs; 60 - Transfer Payments to Other Agencies or Persons; and 70 - Other Expenditures. Examples of four digit budget codes are: 1010 - Staff Salaries; 1040 - Travelling and Transport of Persons (Inland); and 4010 - Operation and Maintenance of Vehicles (Administration). 6 For example, the recurrent budget Vote of the Ministry of Agriculture is divided between 20 expenditure items. The decentralisation was phased in over a three year period covering 13 districts in 1993/94, 27 districts in 1994/95 and all 39 districts in the 1995/96. 8 Conditional grants were introduced in the 1996/97 budget and are initially providing additional funding to support the introduction of universal primary education and increased funding for district road maintenance. 54 Development Budget 10. The development budget is broken down into a number of "Budget Heads" corresponding to the Votes in the recurrent budget. Under each Budget Head is a series of projects (each separately coded) which are defined to match the donor-funded projects through which around 90% of the financing for the development budget is provided. There is no division of projects into sub-sectoral programmes. Allocations for each project follow an item classification with three main categories (Fixed Assets, Payments to Personnel, and Non-Wage Goods and Services) under which are a total of 19 items. 11. The development budget forms the first year of the three-year Public Investment Programme (PIP). Before projects are included in the PIP they are first approved by the joint MOF/MPED "Development Committee" which reviews the justification of the project and whether it merits further preparation. Inclusion in the development budget follows further detailed appraisal and the securing of external financing. Domestic financing is primarily to cover the counterpart costs of externally financed projects and few projects are fully financed 9 from domestic sources . Because the development budget reflects approved funding allocations for projects, it tends to be more realistic and better linked to project objectives and planned outcomes than is the case for the recurrent budget. C. THE BUDGETARY REFORM AGENDA 12. Significant improvements have been made to the budget process recent years, particularly with the introduction of the MTEF. Nevertheless, there remain areas in which present practice is inadequate to ensure effective public expenditure planning and management. Box 2 identifies some of the main issues in relation to the recurrent and development budget and their implications for budgetary reform. The remainder of the section then discusses each of the main features in greater detail. 9 In recent years the most important locally financed projects have been the Mityana-Fort Portal Road, the Rural Feeder Roads Programme and the Entandikwa Credit Scheme all of which have been funded from the Coffee Stabilisation Fund. Box 2: The Agenda for Budgetary Reform Feature Recurrent Budget (RB) Development Budget (DB) Effect on Expenditure Management Implications for Budgetary Reform Outcome - Budgeting practice incremental and * Budgeting linked to appraisal and & Lack of explicit link to service standards => Link sector policies and strategies to Orientation focused on the financing of inputs. inclusion of projects into the PIP. lcads to inefficient resource use, public expenditure priorities. * Little attention given to the outcomes of & Objectives, outcomes and outputs a Little incentive to rationalise progr- Clarifyi programme objectives, activities the public expenditure programme. considered during project appraisal. ammes to reflect budget constraints, and outcome indicators. O t Link resource allocations to service delivery. Sectoral a Sector budget divided between ministry, * Broader sectora) perspectives relained in e Annual budget does not indicate overall Emphasise integrated sectoral exp- Perspectives autonomous bodies and districts. three-year PIP. budget allocations for each sector. enditure planning role of MTEF. * Line ministries uncertain over wider * DB contains a substantial share of re- * Recurrent financing from DB not taken =~Emphasise role of line ministries in sectoral expenditure planning role. current financing, into account in setting RB allocations, coordinating sector programmes. * Variations in adequacy of recurrent => Line ministries should determine ser- financing between and within sectors, vice standards and funding norms. District Focus e Budget decentralised - spatial distrib- 9 Single centralised budget - district level a Risk of resource ceilings in MTEF not -Strengthen consultation between min- ution of expenditure is a specific concern allocations not identified. adequately reflecting districte'oncemns. istries and districts in MTEF process. L of the budgeting process. * Differences in spatial distribution of DB Disaggregate DB between central and l * Minimal district involvement in MTEF expenditure contribute to variations in district allocations. exercise. service delivery standards. => Inform districts of likely DB3 funding prior to district budget preparation. Programme * Budget for each Vote divided between *No programme classification used in DB. * Weak links between sectoral strategies = Rationalise "programmes" to reflect Classification programmes. and "programmes" in the budget. main areas of service delivery. * Programmes do not always reflect the * Impossible to determine overall resource -> Ensure linkage between programmes in strategic areas for public expenditure. allocations at programme level. ministry and district budgets. * No clear statement of programme obj- - Links between central and local gov- Introduce same programme classific- ectives and intended outcomes against enent programmes not always clear. ation to RB and DB and disaggregate which to monitor performance. d_districts,three-year_PIPbudgetallocationsforeachsector.nDB projects by programme.F. Iterisation * Number of budget items have been Different iternisation used from rB.- *Impossible to determine total recurrent - Clearly identi capital, recurrent, and rationalised. nIternisation does not explicitly distin- spending on service delivery. TA elements of DB projects. * Different item classifications used in guish between capital, technical Vu In medium term introduce common central and local govemsment budgets. assistance and recurrent costs, item classification for RB and DB. Implementation t During budget implementation funds Insufficient information available on * Information on actual Db expenditure f Limit ministries' ability to vire funds and Monitoring often reallocated towards overheads and actual expenditures. available only at project level, not at item from service delivery to overheads away from service delivery. o Projects account against donor financing level. > Introduce dual coding system for DB so * Little monitoring of programme agreement not DB itemisation. * Virtually no information available on that expenditures can be accounted for performance. * Implementation monitoring and project programme performance and against budget itemrisation. * "Checks and balances" on budget imp- evaluation tends to be donor-lcd. effectiveness. o Introduce programme performance lementation funds place excessive o Programme managers ton involved in the component to MTEF exercise. demands on programme managers detail of budget management. g Eliminate unnecessary controls on without improving accountability. comesIagain programme managers. 56 Outcome Orientation 13. The links between government policy objectives and strategies, budgetary resource allocations and standards of public services are currently poorly defined. The resulting lack of any significant "outcome orientation" in the budgeting process has contributed to inefficiencies in the public expenditure programme (Box 3). Box 3: Consequences of the Absence of an Outcome-Orientation to the Budgeting Process Some of the consequences of an inadequate "outcome-orientation" in the budgeting process include: Inadequate linkage between government polic% objecti%es and strategies and the allocation of public sector resources. Line ministries and local governments ha%e little incentive to rationalise their expenditure programmes in the face of se%ere budgetary constraints As a result most government programmes have remained se%erel% underfunded, implying severe inefficiencies in public ser ice delikern and resource use. Budetary resources are "retained" to meet administrative and overhead expenses rather than being made available to meet costs of service deli%er-. In the most extreme cases, allocations for operational programmes have been reduced in order to meet the costs of o%erseas travel for senior officials A recent "tracking stud>" found that 640o of the non-%age elements of district primar% education budgets was retained at the discretion of the district education officer and only 360o was made available to school budgets''. Resources reaching ser ice deliverN points are often not reflected in ser ice standards. For example, the tracking studN estimated that 69(o of drugs reaching primar health facilities "leaked" into the private market. Without a clear statement of programme objecti%es and deliver targets. there is little basis against Ahich to monitor and evaluate service deliken. 14. The introduction of a stronger outcome-orientation to the budgeting process will involve three main sets of measures: * Strengthening sectoral policy and strategy development: so that resource allocations are better linked to government policy objectives. To be effective, the process of strategy development needs to recognise the extent of budgetary constraints and to prioritise supporting expenditure programme proposals within these constraints. * Improving programme definition: involving: (i) a clear statement of the objectives for each sub-sectoral expenditure programme; (ii) a description of the activities to be funded under the programme and the standards of service delivery that can realistically be achieved; and (iii) the identification of achievement indicators both for outcomes relating to programme objectives, and for outputs relating to the implementation of activities under the programme. 10 "Tracking of Public Expenditure on Primary Education and Primary Health Care", Consultants' Report, Economic Policy Research Centre, Kampala, June 1996. 57 Linking resource allocation to service delivery: throuph greater use of budgetary norms that link resource allocations to service delivery outputs . The use of such budgetary norms should be backed up with studies by line ministries to determine standards of service 12 provision that can attained within a given level of funding 15. The Action Plan for Poverty Eradication will constitute a critical input for sector strategy development and the redefinition of sectoral expenditure programmes. Sectoral Perspectives 16. In recent years there has been a significant loss of sectoral focus in the annual budget as sector budgets have become fragmented between line ministries, semi-autonomous institutions and districts. The problem has been accentuated by the funding of substantial allocations of recurrent costs from the development budget. The resulting lack of transparency in sectoral resource allocations has contributed to considerable variations in the adequacy of financing of public services between and within sectors. At the institutional level, line ministries are often uncertain of their role, under decentralisation, in coordinating the expenditure programme for their sectors and in setting and monitoring standards for decentralised services. 17. Accepting that the annual budget no longer provides a framework for determining sectoral resource allocations, implies that more prominence will need to be given to the MTEF as the instrument for addressing sectoral issues in the public expenditure programme. This will require: * Greater emphasis on the role of the MTEF in setting overall sectoral resource ceilings at programme level. This would be facilitated by the introduction of "integrated sector programming" that analysed together expenditure allocations from the recurrent and development budgets and from the central and local government budgets. * Clarification of the role of line ministries in the planning and coordination of sectoral expenditure programmes. In particular, ministries will need to develop their roles in providing technical advice and backstopping to districts. * A stronger input from line ministries in the MTEF exercise requiring the preparation of more detailed sector review submissions to MOF/MPED. There may also be scope for the MOF to enhance consultation by for example reviewing MTEF resource ceilings with line ministries before these are submitted to Cabinet. For example, budgets for training institutions could in future be developed on the basis of "contracts" for training a certain number of students at a standard cost per student Similarly, primary health care services could be budgeted on the basis of per capita funding allocations that reflect age distribution and disease incidence among the local population. 12 Under decentralisation, line ministries are expected to determine "minimum standards of service provision" to be provided by district councils. There is a danger of such standards being set without adequate reference to budgetary constraints. The concept of "attainable service levels" realistically linked to budgetary norms is more relevant. 58 District Focus 18. Decentralisation has provided a spatial dimension to the budget that was previously lacking. However, it is still not possible to identify total budgetary resource allocations for each district since the development budget does not net separate out funding for district level project activities (Box 4). Box 4: Decentralisation of the Development Budget The practicalities of decentralisation of the development budget are currenlN bein investicated bN consultants funded b% the Institutional Capacit- Building Project. There are tmto distinct issues that will ha%e to be addressed In the budgeting of de%elopment prolecis there is an urgent need to identifk district leel allocations so that o%erall budgetary allocaions to each district are transparent With regard to decentralisation of implementation arrangements and financial procedures. proiect implementation units should be gi%en guidelines on managing district-le%el components of their prcljects. The first of these issues is central to the SAC III budgetarn reforms. 19. In order to improve the "district focus" of the budget, it will be necessary to: * Strengthen consultation between line ministries and districts in determining sectoral and sub- sectoral resource envelopes. * Separate central government and district level expenditures in the development budget. This could be accommodated through the creation of new budget items identifying transfers to districts. The development budget should in future show aggregate allocations to the districts. * Ensure that districts are consulted and informed in advance over planned resource allocations from projects in the development budget so that these can be taken into account in the preparation of their budgets. This might be expected to be the responsibility of the project implementation unit. Programme Classification 20. A stronger strategic focus to the expenditure programme implies significant redefinition of the functional programme-level classification used in the budget. The present structure of "programmes" in the recurrent budget tend to reflect narrow-administrative units rather than strategic areas of service delivery. The aim should be for each programme in the budget to reflect sectoral priorities and be comprehensive in covering both central and local government responsibilities, and recurrent and development budget financing. This will require: * Making the programme classification in the budget more strategic, reflecting broad areas of public service delivery, rather than narrow departmental boundaries. In most sectors, the budget should be divided between no more than 6-8 core programmes. * Better linkage between programmes at the national level and those at the district level in order to ensure greater transparency and consistency with sectoral objectives. 59 Introducing the same programme classification into the development budget as in the recurrent budget so that total funding allocations for each programme can be determined. For projects with a broad cross-sectoral focus this will require disaggregation between two or more programmes (in effect increasing the number of "projects" in the development budget). 21. It will be important that the reclassification of "programmes" in the budget is linked to meaningful restructuring of sectoral expenditure programmes and the post-decentralisation reorganisation of ministries. This suggests that the reclassification should be undertaken sequentially, sector-by-sector, rather than as a single exercise involving all sectors simultaneously. Itemisation 22. A strength of the current itemisation used in the budget is that expenditure allocations are divided between a relatively small number of budget items. This prevents the budget becoming excessively detailed and allows managers some flexibility in the management of their programmes. However, the usefulness of the classification for budgetary planning and analysis is undermined by the use of different item categories in the recurrent and development budgets. 23. The immediate priority should be to ensure a clear distinction between capital, salary and wage, non-wage recurrent and technical assistance costs in the development budget in order to allow meaningful aggregation of programme allocations across the different budgets. In the medium-term, a common itemisation should be adopted for both the recurrent and development budgets as a step towards full integration of the central government budget. Implementation and Monitoring 24. Weaknesses in budget implementation and monitoring procedures result in resources not reaching service delivery points and in little information being available on service delivery performance. Areas in which reform and strengthening of existing procedures are required include: * Protection of the integrity of the budget by limiting the scope for switching funds between programmes in the budget. * Ensuring that more timely dissemination of information on donor funded expenditures in development. There is also a need to introduce a dual coding system for development budget expenditures so that they can be accounted for against the itemisation in the budget as well as against the categories in the financing agreement. * Eliminating procedures and controls which duplicate or place unnecessary demands on programme managers. In particular there is a need to streamline the complex system of 13 checks and balances in the utilisation of budgetary funds * Introducing a programme performance component to the MTEF sectoral review exercise. 13 Examples of inappropriate or excessive controls include: (i) the involvement of the Auditor General in verifying supplies of goods and services before payment is authorised, posing a potential conflict with his ex- post audit role; and (ii) the requirement for a final verification by Treasury before vouchers are passed to the Accountant General for payment. 60 D. ACTION PLAN FOR BUDGETARY REFORM 25. The approach to budgetary reform under the SAC II programme will build on existing initiatives, particularly the MTEF, and take account of capacity limitations. There are three main components to the action plan: * Strengthening the strategic focus of the expenditure planning and budgeting cycle, by developing the role of the MTEF and the BFP in reviewing and setting sectoral and programme level resource allocations. * Improvements to budget presentation and budget procedures that make the budgeting process more strategic, increase transparency across the budget, and improve the efficiency and operation of budget processes. * Improvements to the content of the budgeting process through the development of integrated sectoral programmes, initially in the agriculture, health and education sectors. 26. The reform process will need to be coordinated by the Director of Budget in the MOF supported by the Budget Adviser who is to be appointed under bilateral technical assistance. Implementation of the reforms will involve primarily the Expenditure Department in MOF on issues relating to budgeting procedures and MPED on issues relating to the sectoral review component of the MTEF. Component 1: Strengthening the Strategic Focus of the Budgeting Cycle 27. Improvements to the strategic-level planning of public expenditures will require the strengthening and embedding of the two-stage budget cycle that has been introduced with the preparation of the MTEF. In so doing it will be important to allow for greater involvement of the districts in the planning of sectoral expenditure programmes. Box 5 shows how the expenditure planning and budgeting cycle might in future operate. Box 5: Sectoral Expenditure Planning and Budgeting Cycle Aug Sep Oct Nov Dec Jan Feb Mar Apr May Jun " Review on-going programmes. *Line ministries and centi at Line o Statement of objectives, acfv government institutions prepare ifies Stamnt oftobjeicts detatled budget submissions Ministries consistent with programme- Minstres" Determine fundino priorties. level resource ceilings. * Indicative 3-year resourceA ceilings for each programme.. * Financing to be provided through :. ... conditional grants. Budge Preliminary Call Sectoral Ras- Cicular ource Ceilings. Central U Lpdate inacio fiame- * Final budget negotiations. BFP Sectoral * Budget Estimates ROsourcO Guidelines sectoral * Finalise expenditure Techn"ica finalised, approved by Resouce Gudelies Prgrammng :Review ceitings in BFPPaian. Issued MConsuolaenns nMeBFng Consultations Cabinet and submitted to Ministries . BFP presented to and Padiament. approved by Cabinet. Peiisary Districts informed of c@ins for block and Block Grants conditional Urants. * Review programme performance and outcomes * Resource ceilings given to line Districts * Determine outline financing re- isti budt quirements at programme level. L * Line depamensepare draft di*tudgt presete * Calculate financing gap. dct Counil fo * Discuss requirements for cond- * Review of draft budget approl f itional grants with line ministries. submissions. Oudget Framowork Paper - Medium-Term Expenditure Framework Preparation of the Annual Budget * Outcome Orientation * Institution Based " Stratqgic - fnulti-year, sectoral perspective, programme focus. * Detailed - item focused within pre-determined programme ceilings * ritegrated - recurrent/development, central government/dishict 62 28. The budgetary cycle will be closely linked to the harmonisation of central and local government 14 fiscal years . This will have two main implications for the budgetary process. First, information will be available on actual district budget allocations for the current financial year in time for the sectoral programme review exercise which precedes the preparation of the MTEF. Second, the MTEF rather than the annual budget will become the mechanism through which central government grants to districts are determined. Once the MTEF resource ceilings are approved, preparation of central government and district budgets will take place concurrently. Preparation of the MTEF 29. Developing the MTEF as the strategic budgeting exercise, implies that it should be undertaken in greater detail than at present. The key requirements are that MTEF should be: * Sectoral and programme based, covering both central and local government responsibilities. * Comprehensive, covering both the recurrent and development budgets and both wage and non-wage elements of the budget.. * Multi-annual, projecting trends in expenditure allocations between and within sectors consistent with public expenditure priorities. 30. The starting point for preparation of the MTEF should be the allocations included in the previous BFP, adjusted for subsequent changes included in the annual budget and developments in the macroeconomic framework. Thereafter, a stronger input into the programme review exercise will be required from line ministries in: (i) providing a clear statement of the objectives, activities and outcome indicators for each of the main programmes within their sectors; (ii) reviewing the performance of on- going programmes and the extent to which planned outcomes are being achieved; (iii) determining financing priorities within the sector consistent with overall sectoral goals and strategies; (iv) setting indicative 3-year resource ceilings for each programme, identifying the split between central and local government budgets; and (v) identifying financing to be provided to the districts through conditional grants and eventually equalisation grants. 31. Districts will need to undertake a parallel exercise in reviewing their own programmes, collaborating with line ministries, in order to determine for each sector: (i) outline financing requirements at the programme level; (ii) the extent to which these could be met from the block grant; and (iii) the financing gap (to be covered by conditional and equalisation grants). The results of these reviews should be fed into the MTEF programme proposals presented by the line ministries. 32. Based on the line ministry submissions and subsequent sectoral review meetings, three-year MTEF resource ceilings should in future be determined for each programme, divided between central and local government allocations (Box 6). Currently the local government fiscal year runs from October to September. Under the Local Government Bill presented to Parliament in October 1996, the local government fiscal year will be harmonised with that of central government. 63 Box 6: Political Endorsement of the MTEF Sectoral Resource Ceilings Both the 1905 96 and 1996'97 BFPs ha%e faced considerable delaNs in receiving Cabinet approval This has undermined the usefulness of the MITEF since resource ceilings have not been aailable when ministries have started preparing their budget proposals. A contributor- factor has been the length and comple.irn of the BFP document. In seeking to make [he BFP process more manareable consideration should be given to a phased presentation of the different sections of the BFP to Cabinet, %%ith the MTEF sectoral re%-ie%% being one such component. This would ha%e the advaniage of allow ing for more detailed discussion of sectoral priorities and resource allocations. Preparation of the Annual Budget 33. One consequence of the strengthening of the MTEF component of the budget cycle is that the annual budget should in future be concerned primarily with the "technical efficiency" of public expenditures by focusing on determining an appropriate input mix within programmes, rather than on deriving overall programme resource allocations. This will imply changes in the "tactics" used by line ministries in preparing their budgets and by MOF in evaluating budget submissions. Line ministries will be required to adhere to the programme-level resource ceilings previously sanctioned in the MTEF and it would be made more difficult to switch funds between programmes. MOF, in reviewing budget submissions, would be primarily concerned with ensuring an appropriate distribution of funding between budget items rather than identifying items on which allocations could be reduced. Component 2: Budget Presentation and Budget Procedures 34. The previous section (Section C) identified a number of ways in which budget presentation and budget procedures need to be improved. However, major reform of budget procedures will involve considerable costs. For this reason a cautious approach should be adopted that (i) responds to demands generated through implementation of the strategic budgetary reform measures; and (ii) takes account of experiences in other SSA countries. A key task of the Budget Adviser in MOF should be to identify the requirements for modemisation of budget procedures and to develop a phased programme for the design and implementation of new procedures. Budget Classification 35. The main area in which budget presentation will need to be improved relate to the adoption of a common programme classification to be applied to both the recurrent and development budgets linked to the harmonisation of the item classification. Implementation would need to be linked to the "integrated sector programming" exercise in the three pilot sectors, so that the reforms to the sector expenditure programme generated by this exercise could be reflected in the annual budget. The main requirement will be for modifications to the structure the development budget to incorporate a programme classification and the identification of resource allocations to districts. 64 36. A longer-term aim should be the full integration of the recurrent and development budgets into a unified budget. However, this will need to be implemented in a way that did not cause the budget to become unwieldy (Box 7). Budget Discipline 37. A stronger outcome-focus in the budget will also necessitate better budget discipline in order to allow programme managers greater certainly in the implementation of their programmes. Specifically, this will require: (i) stricter controls over supplementaries which result in ad hoc cutbacks elsewhere in the budget; and (ii) restrictions on the viring of funds between programmes which undermines the integrity of the budget process. Improvements in budget discipline should also permit MOF to progressively move away from the monthly releases system thereby increasing the predictability of funding. Flexibility in Determining Resource Mix 38. Increasing the accountability of programme managers for budget outcomes requires that they should be given greater flexibility in managing inputs within their programmes. Procedures for shifting funding between budget items within overall programme funding allocations should be simplified. Greater flexibility should also be allowed in adjusting staff establishments (subject to compliance with government establishment and payroll policies) in order to: (i) reflect changed skill demands; and (ii) shift financing from wagebill to operations and maintenance expenditure items. Box 7: Integration of Recurrent and Development Budgets The separation of recurrent and development budgets dates from the Independence period when Uganda undertook major investment in setting up the infrastructure, public services and institutions of a modem state. The development budget was primarily an investment budget adding to the stock of capital and generating downstream operations and maintenance iO&M) costs to be met from the recurrent budget Over the %ears the de%elopment budget has become primarill an "'aid" budget containing a substantial share of recurrent financing The recurrent expenditure element of the de%elopment budget in 1995 96 %%as estimated at USh 156 billion in 19)5 Q6 representing around one third of total non-w%age recurrent allocations. Furthermore. man\ capital projects are no%% financing rehabilitation of existing infrastructure rather than incremental imestment. The present separation makes it difficult to analse overall recurrent expenditure allocations and leads to seere imbalances in recurrent financine bet%een programmes and sectors. The case for merging recurrent and de%elopment budgets is strong and conceptuall% the tak is simple. Capital and recurrent costs could be identified separatelN in the itemisation of the budget, and donor and domestic funding reflected in different columns within the budget. However, in practice. integration of a large number of projects could substantialli increase the complexit\ of a unified budget. An interim measure would therefore be to introduce a common programme and iem classification for each sector to be applied in both the recurrent and de\elopment budgets. This would allo% total resource allocations against each programme to be determined and reflected in the budget summar) tables Computerisation of Budget Processes 39. Scope exists for more systematic computerisation to support: (i) budget preparation in line ministries and MOF; (ii) budgetary analysis; and (iii) budget reporting and monitoring. 65 Budgetary Monitoring and Analysis 40. Within both MOF and line ministries more attention will need to be given to budget monitoring and analysis. Monitoring will have to take place not only at the institutional level, but also at the programme level, bringing together central and local government expenditures from both recurrent and development budgets. A key requirement will therefore be to bring budget information together from different sources and make it available to different users. A greater emphasis will need to be given to output and outcome monitoring linked to the establishment of the proposed "value for money" audit unit in the MOF. Component 3: Integrated Sector Programming In Three Pilot Sectors 41. The integrated sectoral programming exercise will aim to introduce a stronger programme focus to expenditure planning and budgeting in the agriculture, health and education sectors. It will link resource allocations to programme objectives and activities and provide a basis for monitoring programme outcomes and outputs. The exercise will be linked to preparation of the 1998/99-2000/01 MTEF and will commence in the first quarter of 1997. 42. The critical activities to be supported under the SAC III programme are: (i) sector strategy development; (ii) institutional and programme restructuring; (iii) development of integrated sector budgets; and (iv) preparation of sector proposals for the MTEF. The results will then be reflected in the preparation of the MTEF and the 1998/99 annual budget. Sector Strategy Development 43. The definition and refinement of sector objectives and strategies is an on-going part of the policy responsibility of a line ministry. Nevertheless, there is a need to bring the process to some conclusion so that integrated sector programmes can be developed. In each of the three pilot sectors, strategy exercises are under way or planned and should be sufficiently advanced by the second quarter of 1997 to provide an adequate basis for the restructuring of ministry programmes. Institutional and Programme Restructuring 44. Sector programmes and institutions will need to be restructured to reflect the revised sector strategies and institutional roles post-decentralisation. This will have to involve not only the line ministries, but also consultation with district councils, MPED, MOF and MPS. Technical assistance to support the preparation of institutional and programme restructuring proposals in the pilot sectors will be available under the results-oriented management (ROM) initiative funded under the Bank's Institutional Capacity Building Project (Box 8). 66 Box 8: The Results-Oriented Management (RON) Initialie The "results-oriented management" iRON) initiatie is beLng iniroduced bN the NIPS and should pro%ide man\ of the -buildine blocks" for the introduction of outcome-oriented budgering processes The details of the RON initiative are till being defined. but %II probablN in%ol\e defining the mission statements and objectives of ministries; 0 defining the key result areas against which achievement of these objectives can be measured; 0 setting specific performance targets in each of the key result areas; 0 preparing costed action plans for meeting these performance targets; 0 basing the annual budget negotiations between ministries and MOF/MPED on this costed action plan; 0 negotiating and agreeing a "performance contract" with the PS specifying the mission, objectives, key result areas and performance targets for the ministry and the obligations of Government to provide the resources required to meet these objectives; 0 delegating increased powers to the PSs in areas such as finance and staffing to enable them to fulfil their performance contracts; and 0 periodically assessing the performance of ministries against the terms of the performance contract. It will be essential that the ROM proposals are developed jointly with the proposed budgetary reforms. The initial consultancy to support the introduction of ROM will take place during the first half of 1997 and should be used to assist ministries in redefining their programmes and in determining appropriate outcome indicators. Development of Integrated Sectoral Budgets 45. Once the new programmes have been defined, it will be necessary to restructure sector budgets consistent with these programmes. This will involve: (i) distributing recurrent, development and district budget ollocations between programmes; and (ii) dividing development budget allocations between central and local government functions. The exercise would be carried out on the 1997/98 central and local government budgets in order to provide a reference point for subsequent preparation of the MTEF proposals. Preparation of Sector Proposals for the MTEF 46. Preparation of the sector proposals for the 1998/99-2000/01 MTEF will need to involve a more thorough review of programmes than in previous MTEFs and increased consultation with districts on issues of programme performance and future funding requirements. Preliminary resource ceilings should be provided by MOF so that line ministries can develop realistic programme financing proposals for their sectors for the three-year period. Implementation 47. Box 9 proposes a timetable for the integrated sector programming exercise. A description of the current status of the expenditure planning process in each of the sectors is given in Annex A. The exercise will need to be closely coordinated by MOF/MPED to ensure that common approaches are adopted across sectors and that systemic issues relating to budget presentation and procedures are addressed. Within the line ministries the planning units should have lead responsibility for the 67 exercise. It is likely that some technical assistance will be required to support the background analysis required for the development of the integrated sector budgets and in the preparation of the sector proposals for the 1998/99-2000/01 MTEF. This might involve up to 3-4 months of consultancy in each of the pilot sectors, although there might be scope for utilising existing TA staff attached to the planning units. In this respect a concern is the lack of capacity in the ministerial planning units. All three ministries have key posts in the planning units which they are unable to fill due to the current freeze on recruitment into the civil service.  Box 9: Action Plan - Integrated Sectoral Expenditure Programming in Three Pilot Sectors A. Development of the Integrated Sector Programme Activity Timing Tasks for the Ministries of Health, Education and Agriculture Related Actions Sector Strategy Development On-Going 0 Define sector objectives consistent with government policy statements. 0 Coordinate with MPEDJ on inter-sectoral 0 Review existing activities and on-going government initiatives, policy issues and with MOF on likely levels of S Determine sector strategies distinguishing between policy and expenditure funding. measures. Assess expenditure programme implications for both capital and 0 Link sector strategies with the Poverty Action recurrent costs. Plan. 0 Identify programme management requirements. 0 Announce national priorities and sectoral strategies. Institutional and Programme Jan-Jun 1997 0 Identify objectives and key activities for each programme and the roles of the line 0 Link to ROM exercise under MPS. Closely Restructuring ministry, government agencies and districts. coordinate with MPED and MOe. 0 Develop ministerial restructuring proposals. 0 MPS to prepare ministry restructuring 0 Agree on ministerial restructuring measures to be implemented with the 1998/99 proposals and seek Cabinet approval for budget. implementation with the 1998/99 Budget. Development of Integrated Sector July-Sep 1997 0 Revise budget programme classification for introduction with 1998/99 budget 0 Agree revised budget programme classification Budget (reflecting ministerial restructuring and strategic sector programmes). with MO. 0 Prepare short statements of objectives and activities for each budget programme and 0 MOF/MPED to provide guidelines and identify macro-level outcome monitoring indicators technical assistance for preparing the 0 Using the 1997/98 budget, develop an "integrated sector budget" showing recurrent "integrated sector budget". and development allocations against each budget programme, divided between line ministry, other central government agencies and districts. 0 Determine (where possible) simple budgetary norms for service delivery to inform the preparation of the MTEF and budget. Preparation of Sector Proposals for Sep-Oct 1997 0 Using 1997/98 "integrated sector budget" as a reference point, prepare a budget 0 MOF/MPED to issue guidelines for 1998/99- the M(EF programme proposal for the MTEF. For each budget programme this will involve: 2000/0 MTEF which should include 0 reviewing its relative priority in the sector and assessing adequacy of present indicative sectoral resource ceilings (based on financing levels; updated outer years of previous MTEF * consulting with districts on reviewing programme performance and exercise). prioritising expenditure allocations; 0 MOF/MPED to liaise closely with line 0 determining relative priority of capital and recurrent allocations; ministries in pilot sectors during preparation of S developing 3-year resource allocations consistent with indicative sector MTEF programme proposals. ceilings for recurrent and development budgets; and determining the allocations between central government and districts, and between recurrent and development expenditures. 0 Submit programme proposals for the MTEF to MOF. Box 9 (cont.): Action Plan - Integrated Sectoral Expenditure Programming in Three Pilot Sectors (2/2) B. Preparation of the MTEF and Annual Budget Activity Timing Tasks for ministries of Health, Education and Agriculture Related Actions Finalisation of the MTEF Nov-Jan 1998 0 Discuss sectoral programme proposals of pilot ministries at sector review meeting. 0 MOF/MPED organise sectoral review meeting 0 Determine the level of, and guidelines for the use of conditional grants. with MPS and MLG. 0 MOF/MPED and MLG determine indicative sector programme aflocations for districts, dividing allocations between unconditional, conditional and equalisation grants. 0 MOF finalises MTEF. 0 Budget programme level ceilings discussed at Cabinet meeting on BFP. Preparation of 1998/99 Budget Feb-Apr 1998 0 Prepare recurrent and development budget submissions on the basis of new budget 0 MOF circulates MTEF analysis to line programmes and in line with MTEF sectoral and programme ceilings. ministries and issues budget call circular. O Provide technical advice and backstopping to districts in budget preparation. 0 MLG provides details of central government grants to districts. Finalisation of 1998/99 Budget May-Jun 1998 0 Final budget negotiations with MOF. 0 MOF finalises budget and presents to Parliament. 0 District administrations finalise and present their budgets to District Council. 0 MOF updates BFP to incorporate final budget figures as part of "Background to the Budget". 70 Annex A PROGRAMMING AND BUDGETING IN AGRICULTURE, HEALTH AND EDUCATION Introduction 1. This annex outlines the current status of programme development and budgeting in the three pilot sectors and identifies the implications for the integrated sector programme exercise to be supported under the SAC III programme. The main findings are that the process of sector strategy development, institutional reform and programme restructuring is most advanced in health. In the other sectors, line ministries are facing considerable difficulties in adjusting to their changed roles following decentralisation. The greater emphasis on the policy and programme monitoring roles of ministries implies considerably greater demands on ministerial planning units (MPU). However, in health and education these units are poorly staffed, while almost 50% of posts in the larger MPU in the Ministry of Agriculture are currently vacant. Sector Policy and Strategy Development 2. In all three sectors policy and strategy exercises are either under way or due to commence shortly. * In hath, a medium-term planning cycle was established with the Health Plan for the 1993/94-1995/96 period. This was subsequently extended for a further two years, and work on a new five year planning cycle is due to begin in early 1997. The Health Plan has a clear strategic focus emphasising: (i) increased resource mobilisation, particularly to primary health services; (ii) reallocation of resources to the most effective public health interventions; (iii) effective implementation of decentralisation in the sector; (iv) restoring functional capacity and improving efficiency of essential government services; (v) facilitating a greater role for NGOs, the private sector and communities; and (vi) capacity building. * In agriculture, policy roles have been confused by the institutional fragmentation that formerly existed. Until 1992 there were three line ministries serving the sector (agriculture, animal industries, and cooperative and marketing) with the Agricultural Secretariat in the Bank of Uganda having a cross-sectoral policy role. Subsequent rationalisation has given a stronger policy role to the Ministry of Agriculture, Animal Industries and Fisheries (MAAIF), although programme support for agricultural marketing falls under the Ministry of Trade and Industry. A sector strategy exercise has been initiated with the preparation in mid 1996 of a policy paper "Modernisation of Agriculture - The Way Forward 1996-2000". This sets out an extensive list of initiatives to be supported but with limited prioritisation or strategic focus. A more comprehensive strategy exercise is to be initiated in early 1997 under the Agricultural Sector Management Project (ASMP). * In education, current sector policies and strategies are based on the 1989 Education White Paper and subsequent policy initiatives, particularly universal primary education. A sector strategy exercise is currently being initiated by the planning unit 71 in the Ministry of Education and Sport (MOES) and will address sector-wide policy issues and the role of MOES within a decentralised education system. In view of staffing limitations in the MOES planning unit, additional support for this exercise may be required. Institutional and Programme Restructuring 3. Linkage between institutional structures and service delivery programmes appears to be strongest in health, with the MOH having a relatively simple structure (only 7 programmes in its recurrent budget) with clearly defined links between departments in the ministry and the districts. An exception is primary health services for which there is no clear departmental responsibility in MOH and instead services are provided through a number of project initiatives. In agriculture, the institutional and budget structure tends to be more fragmented and links between central government and district programmes are less transparent. In education, the proliferation of semi-autonomous institutions with separate Votes has undermined the sector policy and coordination role of MOES. 4. The need for further institutional restructuring following decentralisation is recognised in all three sectors. This will involve considerable rationalisation of line ministry functions to reflect their substantially reduced operational responsibilities and the relatively greater importance to be given to policy and programme monitoring functions: * MAAIF organised a ministerial retreat in mid 1996 which made initial proposals for restructuring of the ministry. The Ministry of Public Service (MPS) has been requested to review and develop further these proposals. * In MOES, proposals exist for the reorganisation of the ministry into two Directorates for general education and higher education. * MOH is engaging a consultant to prepare proposals for reorganisation of the ministry. 5. All three line ministries are included in the first phase of the implementation of the results-oriented management (ROM) initiative being introduced by MPS. The starting point of the ROM exercise will be a review of line ministry functions and structures. The consultants appointed to initiate the ROM exercise are expected to begin work by January 1997. Programming and Budgeting 6. In all three sectors the development budget is an important source of recurrent financing. In 1995/96, the proportion of non-wage recurrent expenditures from the development budget was 80% in agriculture, 54% in health and 25% in education. Despite this dependence on the development budget, there is remarkably little coordination in the preparation of the recurrent and development budgets in line ministries, with the development budget prepared by the MPU and the recurrent budget by the ministerial finance and accounts unit. * In education, there appears to be relatively weak ministry oversight of the externally financed component of the development budget (which is seen as the responsibility of donor supported project implementation units) with the MPU primarily concerned with the budgeting and management of the domestic funded component of the development budget. 72 * In agriculture, the dependence on the development budget for recurrent financing has resulted in fragmentation of service delivery functions and uncertainties over the continuity of funding levels. It has also had implications for the decentralisation process with district agricultural services being almost totally dependent on projects for meeting operations and maintenance costs. MAAIF has become increasingly concerned about the extent to which the sector programme is being driven by donor interests. Under the ASMP it is proposed to develop a "sector investment programme" to replace the present reliance on funding from stand-alone projects. In order to integrate it with the sector programming approach under SAC 1II, the scope of those elements concerned with public sector service and infrastructure should be broadened to include recurrent expenditures. * It is in health that the greatest progress has been made towards a more integrated approach to expenditure planning and budgeting. This has been reflected in the relatively thorough analysis submitted for the MTEF/BFP sector review meetings and also in the efforts to take account of development budget funding in the district work plans. It also appears that donor support is better coordinated in health than in some other sectors. 7. The MTEF exercise introduced in the 1994/95 and 1995/96 BFPs has proved most useful in the health sector. MOH has found that the resource envelope provided by the MTEF has facilitated prioritisation within the expenditure programme. However, with particular reference to primary health services, the MOH felt that the MTEF needed to be backed up by measures to ensure that resources "released" from central government programmes for reallocation to district-level services actually resulted in increased allocations to the targeted district-level services. Decentralisation 8. MAAIF and MOES have faced difficulties in adapting to their changed roles under decentralisation. Although operational responsibilities have been transferred to districts, the technical backstopping and monitoring roles of these ministries have yet to be effectively operationalised. The MOES planning unit has recently prepared a paper on the ministry's role and functions in relation to decentralised programmes and a consultant is being hired to develop a system for monitoring district-level programmes. It has also started on a programme of visits to districts to discuss performance monitoring and budgeting norms. In agriculture, discussions have begun with districts on developing local-level programme planning capacities and on identifying critical funding gaps. In 1997, the ASMP will support the development of district- level planning and implementation capacities. This will involve the preparation of guidelines for district agricultural plans followed by assistance with district plan preparation. 9. Health provides an example of how the line ministry/district relationship can work under decentralisation. With resources from the District Health Services Project, the Ministry has provided considerable support to districts with the preparation of health plans and work programmes. These have incorporated both decentralised financing and the resources available from development projects. Efforts have been made to link resource allocations to outputs and to programme objectives and outcome indicators. Backstopping and monitoring have been improved through the introduction of a programme of supervision visits, whereby each district is visited by a team from the MOH at least twice in a year. The initial focus under the DHSP has 73 been on a number of pilot districts, and the programme is currently being extended to all districts. Institutional Issues 10. The greater emphasis on strategic sectoral programming within a two-stage expenditure planning and budgeting cycle will have the following implications for institutional responsibilities: * MPUs should be given lead responsibility within line ministries for the MTEF exercise since this should: 0 be policy and strategy driven and concerned with programme objectives and outcomes as well as issues of prioritisation between programmes; 0 cover the whole sector (not just the line ministry budget) including semi- autonomous institutions (e.g. NARO, Makerere University, and Mulago Hospital) and decentralised programmes; 0 cover both recurrent and development budgets; and 0 be multi-annual, covering the three year MTEF period. * Finance and accounts units would be responsible for the subsequent preparation of the annual budget for the line ministries. This would be primarily concerned with achieving an appropriate input mix within the overall resource ceilings established under the MTEF. There is also a strong case for transferring responsibility for the preparation of the development budget from the MPU to the finance and accounts unit. 11. The greater emphasis on the policy and programming role of line ministries will place added demands on MPUs which in the case of health and education are inadequately staffed. Although MAAIF has a planning unit with an establishment of 30, there are currently 14 vacant posts and considerable reorganisation of its workload will be required. An added problem is the recent deterioration in incentives for middle-level professional and managerial staff, following the phasing out of donor-funded staff allowances. In case of MAAIF, it was felt that the inadequacy of incentives considerably reduced the motivation of staff in the MPU to take on new tasks and responsibilities. 12. In all three sectors sources of technical assistance are available to support the development of the policy and programming function both from Bank funded projects in the sectors (DHSP in health, ASMP in agriculture, and PETD in education) and from bilateral sources as in case of the assistance being provided by British ODA to the health planning unit. Technical Note 8: Improving the Management and Efficiency of Public Expenditures Consistent with the Decentralization Program  IMPROVING THE MANAGEMENT AND EFFICIENCY OF PUBLIC EXPENDITURES CONSISTENT WITH THE DECENTRALIZATION PROGRAMi Introduction 1. The budgetary system in Uganda as elsewhere should be geared towards achieving three interrelated objectives: aggregate fiscal discipline, or the control of aggregate spending and deficits; strategic prioritization of spending among competing sectors and programs; and given the level and composition of spending, technical efficiency in the use of budgeted resources. Measures introduced since 1992 -- particularly, strict expenditure control through monthly cash management -- have resulted in considerable progress on aggregate fiscal discipline. In addition, the government has taken steps to improve expenditure prioritization by instituting a medium- term expenditure plan, with increased funding to Priority Program Areas in key development sectors. It has also sought to improve technical efficiency by reducing the size of an over-staffed civil service by half and raising civil service wages. At the same time, the government has launched a far-reaching program which decentralizes decision-making and service delivery for a large number of functions to district governments. Building on recent initiatives, there is substantial scope for strengthening institutional arrangements to enhance efficiency and effectiveness in the allocation and use of public expenditures. During the discussion of the SAC III program, the government agreed to reform its expenditure management program in order to achieve these desired improvements. The essence of this reform is to institute an outcome- oriented sector programming process which integrate recurrent and development expenditures and is consistent with the decentralization program, and to ensure greater flexibility for line agencies and districts to deliver budgeted outputs and outcomes. The new process would place strong emphasis on transparency and accountability in the design and implementation of programs and ensuring that actual expenditures are in line with outcomes. Implementation of this reform will involve almost all ministries and agencies of the government, both central and local, and will take several phases and iterations, over some time, to complete. This note addresses the issues and processes of the reform that are related to its implication for the decentralization program. Consistency of the Budget Process with the Decentralization Program 2. There are several significant aspects in which the present budgetary process is not consistent with the far-reaching decentralization program that the government has already implemented. 3. First, a large amount of budgetary resources are being allocated to central government ministries for functions that have been decentralized. The division of responsibilities in the Constitution specifies that the central government will largely be responsible for policy formulation and monitoring, maintaining implementation responsibilities only for national public This note was prepared by Mr. James Ford (World Bank). 75 goods such as defense and trunk roads. However, the staffing and administrative expenses of several central government ministries such as agriculture and education are inconsistent with this assignment of responsibilities. 4. To address this problem, the BFP for 1998/99 must begin by identifying the residual pool of resources after allocations for block grants and statutory payments (e.g., interest and debt repayments) have been made. This residual pool could well be quite substantial (e.g., around 50 of the budget). The BFP would need to allocate this residual pool among central government functions and conditional or equalization grants for decentralized functions. 5. Second, following from the above, substantial resources will be transferred from the central to district governments through conditional and equalization grants, but the modalities governing these grants need to be developed and implemented. It is through these grants that the contractual arrangements for policy implementation are to be specified. More fundamentally, these are the principal instruments available to the central government in a decentralized structure for ensuring the achievement of national priorities. At the same time, districts need to have indications of the amount of resources potentially available to them and the rules under which thesc grants may be accessed, including the monitoring and reporting systems. The mechanism through which sector ministries will effect the flow and management of resources for sector policy implementation is through the conditional grants established inder the new Constitution. Similarly, the critical means of effecting policy reduction policy through improvement of services to the poor, is through the likewise mandated equalization grants. The programs and objectives that the central government would wish to support should be clearly delineated, and districts would need to submit their program proposals in concert with these objectives in order to gain access to the funds. Further, access to budget allocations will be linked to the preparation and submission of operating plans based on consultation and agreement with relevant communities, and fund releases to reporting on expenditures. For equalization grants, the potential allocation will be based upon the deficits in services to the poor relative to national standards. Access to those resources also will be dependent on the preparation and submission of satisfactory programs. Background analyses and the preparation of service standards by sector ministries will require more time, but should be ready for implementation by FY 1998/99. 6. The remaining sections of this note outlines the agreed framework for the preparation of plans to put into practice the expenditure management reform measures related to the implementation of the GOU decentralization program. Preparation of Integrated Sector Programs by Pilot Ministries 7. The Ministries of Agriculture, Education and Health have been selected to initiate the preparation of sector programs and budgets which incorporate appropriate outcome-oriented systems and procedures and appropriate roles, responsibilities and authority for central and district governments and agencies. The substantive character of the sectors and the roles of the central and district Governments differ by sector. In addition, state of preparedness of the Ministries to implement programs in a decentralized context also vary. As a result, each Ministry's program will be different in its details. Nevertheless, a consistent framework needs to be utilized for the preparation of individual plans that would incorporate the Decentralization Program. This framework is outlined in a series of six interrelated questions below. Preparing 76 answers to these questions will assist the Ministries in clarifying for themselves, the requirements and steps to implement the reform. I. How Much Resources Should Go Into the Pools for Conditional and Equalization Grants? 8. The 1998/99 budget would need to identify the pool of resources remaining after statutory payments and allocations for block grants are made, and to allocate this pool among central government functions and conditional and equalization grants for decentralized functions. In practice, it is anticipated that allocations would be made sectorally. The Ministry of Finance (MOF) would provide indicative allocations of total resources - combining recurrent and development - based on macro conditions, strategic commitments, priority areas, recent experience and policy imperatives within priorities. It is proposed that the MOF specify a maximum wage component and the line Ministries be allowed to plan for and reduce actual wage expenditure below that ceiling without losing resources; Ministries also would be allowed to trade off positions within their establishments as long as they stayed within their ceilings. MOF also may retain a pool of resources during the preliminary BFP planning stage, from which the Ministries would be able to compete for incremental resources based on their output targets and expenditure proposals. 9. The line Ministries would be responsible for determining allocations between wage, non- wage and counterpart funds for donor financing, as well as the pools for conditional and equalization grants within their sectors. Because of the emphasis on poverty reduction, overall allocations to the equalization pools may require interministerial and donor consultations. Overall pools will need to be divided into separate pools for each program within the sector strategy. As noted below, the Ministries need to incorporate consultations and bottom-up, as well as top-down planning in determining the size of pools. II. What Should Be the Rules and Conditions by Which the Grants Would Be Distributed? 10. Unlike block grants, which are designed to support local autonomy and are thus constitutionally within the sole authority of districts to determine their use, conditional grants and equalization grants are intended to reflect national policies and priorities. In effect these latter grants establish contractual relationships between the central government Ministries and the districts which elect to receive them. The contractual conditions would need to include three key elements: performance targets, input/output relationships and instruments and mechanisms. 11. In the context of an outcome-oriented system, performance targets are essential to establishing clearly what are the expected results that district-level agencies would be held responsible for under a particular program. Input/output relationships determine the resources that are anticipated to be required in order to achieve the agreed results. Specifically linking inputs and outputs has several other potentially important contributions which are discussed under sections (Monitoring and Adjustments) below. 12. In agreeing resources, it is plausible that districts may be asked to contribute from their own sources, as well, although there are obvious limitations to reasonable requests. This issue may best be dealt with by the districts preparing their own plans and proposals which include attention to meeting national priorities. Except where there are major externalities, there is 77 considerable overlap between national and local priorities. Primary education and primary health care, rural roads and increased agricultural production are of substantial local as well as national concern. Given the opportunity to express their preferences, local communities would be anticipated to ask that considerable resources be devoted to improving and expanding these services. The central government would be able to meet national targets by expanding the resources available to the districts through the conditional grants. 13. Sector ministries, however, should make independent assessments of the circumstances of a local government that are relevant to particular national programs. A district may be particularly suited for growing a crop with excellent export markets; conversely, a district may be unusually susceptible to a certain type of infestation. These considerations should inform the ministries proposed geographical allocations. 14. It also is possible that sector ministries would make proposals for local governments to undertake programs that a district or city may not be anticipated to include in its plans, but which are of national importance. This may include programs with externalities that are not controllable with the local jurisdiction and/or require the collaboration of several local governments to implement successfully. A malaria or Tsetse fly control program, or a watershed management program are examples of programs with these characteristics. Consideration of, and planning for these programs would be facilitated by early communications with local governments. 15. Instruments and mechanisms set out the broad processes and standards to be followed in the planning and management of the program. It is important that these instruments and mechanisms are carefully designed to give clarity to mutual expectations between the Ministries and the district agencies about their roles and responsibilities without becoming means of micro- management by the Ministries. Line Ministries should develop and disseminate to local governments budgt models which include the key factors and their standard costs that go towards the achievement of program outputs and outcomes. Such models convey the Ministries technical expertise and accumulated experience appropriately and effectively in guiding and assisting local governments in their program planning and preparation. This also would facilitate aggregation and comparison of districts' proposals. III. How Will Decentralized Programs Supported by Conditional and Equalization Grants Be Monitored? 16. Monitoring is a crucial requirement for a successful output oriented, decentralized program. In this context, several layers and interrelated types of monitoring programs are required. At least five levels are involved, and among the most important is at the level of the clients. This is an appropriate place to start because feedback from clients provides the ultimate test of success or otherwise, as well as directions for improvement. Districts need to establish systems appropriate to their managerial responsibilities; the Ministries must assure themselves that contractual outputs and outcomes are achieved and maintained, and inputs used appropriately; the Auditor General has constitutional and legal responsibilities for reporting on the use of public funds; and the Ministry of Local Government must supervise and guide the development of the decentralized system. 17. Monitoring programs must be accompanied by a program of systematic and timely publication. Making the public, and stakeholders in particular, fully aware of the performance of programs and program elements is a key element in maintaining and improving performance. A 78 corollary requirement is that the media and forms of publication should be readily understandable and accessible to the intended audience. IV When Monitoring Indicates the Need for Improvements in Decentralized Programs, How Should these Programs Be Adjusted? 18. Although the reform builds on a number of ongoing activities of the Government, it will necessarily involve much that is new. It is expected, therefore, that as programs are monitored, problems will be revealed. Success will only be built over time, through refinements and adjustments. Some issues may be anticipated and planned for. Among these are the relationships between input and outputs, many of which will be clearly identified for the first time. Districts and Ministries will need to identify quickly where planning assumptions were incorrect and make necessary adjustments both within and between programs. The SAC III discussions have emphasized the importance of allowing Ministries the necessary flexibility to make these adjustments. Ministries would also need to be assured that the 3-year horizon of the medium- term Budget Framework Paper will allow them a predictable resource pool within which to work towards improvements. V. Who Will Need To Be Involved In, and Responsible For Various Stages and Components in the Planning and Implementation? 19. Ministries have shown differing capacities to respond to the realities of decentralization. The reform program requires that a working consensus is created among decision-makers, analysts and implementers within the Ministries. That consensus must be reflected in the preparation of the pilot programs. VI When Will the Various Elements of the Plan be Prepared and Implemented? 20. A timetable has been established under SAC III that reflects agreement on when the elements of the plan for the implementation of the reform program will be ready. This timetable has been distributed separately, but broadly indicates a six month preparatory period beginning in January, 1997, during which initial one-time major adjustments in role, functions and structures of organizations, and broad integration of recurrent and development budgets would take place. Beginning with the new Financial Year, the new BFP and Budgetary procedure would be initiated, and their results incorporated in the 1998/99 budget. Preparing the Pilot Ministries' Programs 21. The Ministries will prepare programs which are based on the substantive characteristics and conditions of their sectors and their roles and priorities within them. 22. The Ministry ofAgriculture will build on discussions already begun with the districts on bottom-up planning of programs and activities. These discussions have revealed resource gaps in proposed activities that may be filled appropriately by the use of conditional grants. 23. The Ministry of Education has prepared a document outlining its strategy for implementation of decentralization. It has started on a series of visits to districts to discuss performance target and input output relationships, and to obtain feedback from district education 79 officials on problems and issues at that level. While they are not yet at the stage of formulating district based, outcome oriented work programs, the introduction of the Universal Primary Education (UPE) policy has provided an opportunity and resources to introduce a condition grant in support of this policy in January 1997. In preparation for this, the Ministry of Education proposes to utilize a budget model that incorporates the significant choices that need to be made in determining the pace and annual cost of UPE and the allocation of resources between factors influencing education performance. 24. The Ministry of Health has made considerable progress in initiating district-based health plans and programs. It also will seek to consolidate this progress by working from locally- prepared work plans to build input-output relationships and utilize existing examples as pilots for other districts. In addition, the Ministry will undertake a new initiative which will focus on the preparation of central-district partnerships for public health programs, beginning with a national mosquito control program. The Ministry intends to have the parameters for an output-oriented budget in a decentralized framework by December, 1996. The Role of the Districts 25. Districts need to reexamine their roles, functions, structure and processes in light of their expanded role under the Constitution and the new Local Government Law (forthcoming). The expenditure management reform process provides the vital framework for this reexamination. Issues in Restructuring Local Governments Operations 26. Districts now have dual roles in serving their constituents and being the executing agencies for national programs. Fulfilling these roles will require that local governments become more effective and efficient managers of their functions as service providers. Critical to achieving these objectives of improved efficiency and effectiveness, is districts recognizing that responsibility as service providers do not require them to be service deliverers. While special circumstances or the absence of viable alternatives may force them to adopt the service delivery role, local governments should look for, and carefully examine alternative delivery mechanisms. Districts are already doing so in primary heath care, where they partner with NGOs who deliver the service with financial and other support from districts. Some districts have contracted out portions of their functions including the operations of markets and collections of certain fees and taxes. This movement of the part of districts to create and extend working relationships with the private sector is in keeping with international trends in both developing and industrialized countries. Experience in this regard has indicated that capacity and resources may be brought more rapidly to the delivery of public services, and utilized more efficiently, in well designed relationships with private partners. Districts should note, however, that the most important feature of designing these relationships is that competition either in the delivery of the service or for the rights to deliver the service for a specified period, must be the basis of the design. 27. When local governments determine how best functions may be performed and services delivered under their prevailing circumstances, they will need to develop proposals for institutional and organization structures and relationships to implement these decisions. Where the local government will also deliver services directly, the requirements for the agencies to do so need to be defined. Where responsibilities are to be contracted, the regulatory framework needs to be established. In all cases the monitoring systems and procedures need to be specified. 80 28. Local governments also will need to analyze the relationship between services outputs and outcomes and recurrent and development expenditures. An important aspect of the new expenditure management system is the integration of both budgets so that decisions on how to improve and/or expand services may be made with full appreciation of resource constraints and implications. In this manner, all the factors that go towards providing and delivering improved services may be considered together. Districts should establish, based on local experience, the relationships between various types and combination of expenditures and services outputs or outcomes as set of norms to guide their planning. As a consequence, when districts establish or agree to performance targets for services, they will be able to quickly assess the resource implications of their decisions. Budget models provided by the sector ministries, calibrated for local conditions, would be useful tools for this purpose. 29. Finally, in preparing for their new roles, local governments should recognize where support from central ministries in technical guidance, systems design, training and other capacity building and enhancement measures would be helpful. These measures should become an important part of the agenda for consultation with the Ministry of Local Government and the sector ministries. District and City Level Sector and Program Planning 30. Local governments already have a well defined planning and budgetary process for the functions assigned under the original decentralization law of 1993. Much of this process will remain valid under the new expenditure management program and the new roles and responsibilities of local governments. The important changes will be: earlier initiation of the process during the financial year; planning in response to national priorities and targets; establishment of district absorptive capacity and performance targets; detailed consultations with communities and stakeholders on local and national priorities and programs; publications of program plans, targets and outcomes; consultations and negotiations with sector ministries; and, performance contracting with sector ministries for conditional grants (and eventually, equalization grants). 31. An early start in the process will result in an almost continuous planning and budgeting process at the local government level. This extended cycle is necessary for three important reasons: to accommodate the change in the budget year to coincide with the central government's; the increased number of preparatory steps; and, to allow for consultation and negotiations with sector ministries. 32. As local governments will be responsible for the implementation of programs to achieve national priorities, they will need to prepare for this by identifying and planning for the activities that will need to be carried out in their jurisdictions in order to fulfill their responsibilities. Clearly, the endowments of particular districts will determine the extent of their role in achieving specific national targets. Consequently, it will be important that every local government examine their individual role and potential and establish as well, their capacity to play their role to the fullest extent. Having understood their potential and capacities, districts and cities must set their individual performance targets with regard to national priority programs. Local governments have responsibilities to their own communities as well, and to ensure that these responsibilities also are met, local governments must examine local priorities and programs in a manner similar to the national priorities exercise. However, as noted in the discussion of the sector ministries role, there will be considerable overlap of local and national priorities, with the result that a 81 significant part of this preparation will be consideration of how addition resources would allow program outputs and outcomes to be maximized within the capacity constraints of the local authority. An output of this preparatory work will be the local government's estimates of the priority program targets it would be able to meet from its own resources; the additional targets it could achieve with specified levels of conditional grant resources; and any capacity building requirements that would be needed to enable it to achieve its targets. 33. To ensure that its proposals respond to community priorities, local governments must maintain a process of consultations and confirmation with local communities. Previous experience and stakeholder inputs would have informed the process of initial prioritization and planning. However, before finalizing proposals to be presented to the sector ministries, local governments need to confirm with communities that their priorities are adequately represented. These community consultations would be facilitated by publication and dissemination of the previous year's accomplishments and providing communities with planning parameters such as service levels and pertinent unit costs of service provision and delivery. These consultations also would provide an appropriate opportunity to inform the communities about, and discuss the local government's role in national priority programs. 34. The preliminary program targets and planned outcomes that result from the planning and consultations should be published and disseminated as drafts to be discussed further with sector ministries. It is critical that information provided to communities about plans, programs, budgets and performance is presented in a manner that is pertinent to the audience and through media that is accessible to them. Consequently, communities need to understand in particular, the activities that will affect them most directly, and the context in which these activities will take place. By being informed, communities' feedback becomes an important part of an outcome monitoring system. 35. Having prepared plans and programs that have community and stakehoders endorsements, local governments are in an appropriate position to begin consultations with sector ministries. These consultations should clarify for the line ministries the local commitment to national priority programs and the requirements, including capacity building, to implement them. Districts and cities would understand the rationale of central government allocations available to the local program proposals; the performance targets and amounts available to meet them; technical guidance and design issues in detailing plans and implementing programs; the rules for accessing and continuing to utilize conditional grants; and, the monitoring and evaluation systems to be applied to the programs. 36. Local governments would be likely to need to modify their programs as a result of the consultations with the sector ministries. In instances where these modifications would mean significant departures from the programs and priorities agreed with local communities and stakeholders, the authorities need to discuss these changes with those affected and gain their support. The ministries, following agreement with the MOF, would finalize with the local governments, the size programs, including outputs and performance targets, which would be supported by conditional grants. These agreements and associated district or city programs would be published and disseminated in the manner of the prior drafts. 82 Budget Preparation by Local Governments 37. With agreements finalized on conditional grants, local governments would follow the final stages of their detailed budget preparation. This includes preparing detailed implementation plans and work programs, and activity plans with timetables and monitoring programs. In keeping with the principle and practice of publication of public plans and commitments in an accessible form, the budget would be translated into simple summaries including programs statements, by location of activities, stating what activities will occur, when and at what cost. Other media, including folk theater, also may be used for dissemination.  se 1 * *  REDUCING QUASI-FISCAL DEFICITS: PARASTATALS' 1. Background The first state-owned enterprises (SOEs) in Uganda were established after independence (1964), but the most significant growth of the SOE sector happened in the 1970s mainly as a result of nationalization or expropriation of major businesses. By 1993, the SOE sector comprised over 130 firms covering a diverse range of activities including trade and commerce, agro-production and processing, manufacturing, finance, insurance, and other services. These enterprises had a debt stock of USh. 968 billion (out of a national debt stock of USh. 3,500 billion) and an estimated employment of 78,000 (out of a workforce in formal employment of 275,000 including 164,000 civil servants). The sector was estimated to be operating at 25-30% capacity utilization and contributing to about 5% of GDP. 2. In 1993, the Ugandan Parliament approved the Public Enterprise Reform and Divestiture (PERD) Statute to provide a legal foundation to the PERD Program started by Government in 1992. In the first four years of the PERD program (1992-95), Government focused its attention on privatization, while laying the foundations for SOE Reform. The process of privatization started in 1992/93 but at a very slow pace. A review of the program in early 1995 resulted in targeting 85% of all SOEs to be privatized by the end of 1997 and in a substantial acceleration of the pace of privatization. As at the end of December 1995, 48 state-owned enterprises (including subsidiary units) had been divested. In SOE Reform, little progress was made until 1995. By 1996, however the Government had collected most of the necessary information on the performance of the SOE sector and had put systems in place to monitor most of the subsidies following to and from this sector. Progress still needs to made in enforcing financial controls on SOEs and in eliminating and/or regulating their monopolies. 3. SOEs' Impact on Economic Development. Most Ugandan state-owned enterprises are massively inefficient. They represent: (a) a very serious constraint to the development of the Ugandan private sector; and (b) a substantial drain on limited national resources. SOEs as a Constraint to PSD 4. According to a recent survey of over 250 Ugandan businesses, inefficient SOEs are a major constraint to the development of the domestic private sector. Overall, power breakdowns are a major obstacle to Ugandan businesses. Erratic supply has led many firms to purchase their own generators, substantially increasing both investment and operating costs. According to the Uganda Electricity Board (UEB), the state-owned monopolist in power generation and distribution, at present demand exceeds supply, and Uganda will need to make substantial investments to keep up with growing demand. Second to power problems are problems in telecommunications, another sector dominated by an inefficient SOE. A major project for rehabilitation of the telephone system in Kampala was completed in 1993, resulting in a marked improvement in the availability of telecommunication services. However, overall capacity is still very low, at 2 lines per 1000 people, compared with 7 in Kenya and 12 in Zimbabwe. The main problems of the Ugandan transport system lie in Uganda's dependence on Kenya and Tanzania for access to the sea (made more serious by the lack of an adequate route by road from I This note was prepared by Mr. Stefano Migliorisi, Milan, Italy. 84 Uganda to Tanzania), the poor maintenance of many roads, and the inefficient operations of the Uganda Railway Corporation (URC). 5. Overall, power breakdowns are a major obstacle and voltage fluctuations a moderate obstacle to Ugandan businesses. Firms overwhelmingly rely on power from the Uganda Electric Board, which ranks as firms' second least favorite public agency after the URA. Firms report high electricity costs and poor services as leading problems. Erratic supply had led 58 firms (24% of the main sample) in our sample to purchase their own generators. Continuous process manufacturing is virtually impossible without a generator. Generator costs various widely between enterprises, but the average initial cost was around $25,000 and annual costs average $10,000. Figure 1 - Infrastructural Constraints to Ugandan Entrepreneurs Power breakdown Voltage fluctuation u Telecommunications proble Quality of road Landlindustrial spac Water supp Garbage dispos Commercial truckin Waste water dispos Air freight servi Railway transpo Ports/Shippin r 1 1.5 2 2.5 3 3.5 4 4.5 5 Seveity (1N0 Obstacle 2mMinr, >.Madratem 4-Major) Table 1: Selected Energy Sector Performance Indicators for Uganda (1990-91) Indicator Description Value World Ranking Coverage Year (from bottom) Energy Consumption Kgoe 25 7 89 LDCs 1991 per Capita Capacity Factor % 55 9 107 LDCs 1990 __ __ __ 1__ _ _ (from top) I 1 _ _ _ System Losses % 40 2 44 LDCs 1990 Source: World Bank - IENIN Power Database. 85 6. While data are somewhat dated and the situation has often improved dramatically since 1990/91, it is interesting to note (Table 1) that Uganda has not only one of the lowest level of per-capita energy consumption in the world (less than half the Sub-Saharan Africa average2) but also one of the highest rates of capacity utilization indicating the existence of a serious power bottleneck if the projected rates of growth in GDP will materialize. According to UEB at present demand exceeds supply by 10 MW, a situation that should last till 1998, when the Owen Falls hydro plant will be fully rehabilitated. It is estimated that Uganda would need to invest US$ 100 million per year for the rest of the decade to keep up with growing demand, not considering the much needed investments in the rehabilitation and expansion of the transmission and distribution networks. Many business consumers operate stand-by diesel generating units, at a substantial cost as shown by the above survey results. Due to the land locked position of Uganda, petroleum products are also very expensive, and this increases the costs of operating stand-by generators. 7. Government has identified several locations for future development of mini-hydro power plants and licensed one of these sites to a private consortium. However, while production of power is liberalized, there is a need to review the UEB Act to allow for private distribution of power where UEB still enjoys a monopoly. With the current level of losses in the UEB grid, low payment ratios (only about 40% of all energy sent out in FY93 was actually paid for) and weak capital structure of UEB itself, the obligation of selling power to UEB instead of selling to final consumers is a serious deterrent to private initiatives in this field. Competition in distribution is at least as important as competition in production, and the two functions could be split even if UEB were to remain under state ownership. A management contract for UEB (or several management contracts for several "UEBs" if Government decided to break UEB into smaller units to be privately managed) could also help. Finally, there is a need to remove any regulatory function from UEB, like the granting of licenses to other generators or suppliers or setting their tariffs, functions that certainly do not encourage competition. 8. Second to power problems were telecommunications problems. A major project for rehabilitation of the telephone system in Kampala was completed in 1993, resulting in marked improvement in the availability of telecommunication services. However, overall capacity is still very low, at 1.9 lines per 1000 people, compared with 7.1 for Kenya and 11.9 for Zimbabwe. Firms in Arua reported tele-communications as a major or very severe obstacle, with an average score of 4.3 on a scale of I to 5. Jinja firms on average rated the problem at 3.4, Mbale/tororo at 2.8, Kampala/Entebbe at 2.6, and Mbarara at 1.9. Mbarara's telephone exchange was modernized in 1988, while Jinja struggles with a small, antiquated, system and Arua relies on a radio telephone link. Overall, 79.4% of firms responding said they had a telephone but 63.7% also said they did not have a sufficient number of lines. Of those that did not have sufficient lines 38.3% cited the length of time it takes to be connected, 31.4% high operating costs, 19.6% high connection fees, and 11.8% lack of phone lines. Of 98 firms providing an estimate, the average time they thought it would take to get a new line was 12 weeks. 2 0.03 tons of oil equivalent which compares to the average of 0.07 in Sub-Saharan Africa. 86 Table 2: Selected Telecommunications Sector Performance Indicators for Uganda (1992) Indicator Description Value World Ranking Sub-Saharan (from bottom) Africa Average Telephone Density Main lines per 100 0.15 3 1.47 population Waiting Period for Connection Years >10 13 5.3 Faults Cleared by % 14.8 2 - Next Working Day Local Calls % 44.8 6 Completion Rate Source: ITU's World Telecommunication Report - 1994. Data on 74 countries referring to 1992. 9. The main problem of the Ugandan transport system is not one of quality of infrastructure (although quality leaves to be desired) but a question of poor management and maintenance. Overall, Uganda's transport system consists of: (a) an extensive road network of about 29,000 km of which 8,000 km of main roads, 700 km of urban and 20,400 km of rural feeders roads; (b) a railways/wagon ferry system which consists of: (i) three rail wagon ferries operating on Lake Victoria and two Ugandan ferry terminals; and (ii) a modest domestic rail track network of three lines totaling 1,232 km; (c) an international airport in Entebbe and about 12 local airfields and airstrips across the country; and (d) some Lake Victoria water transport, and inland waterways which are out of use. 10. Given its land-locked position, roads are clearly vital to Uganda's competitiveness. The Uganda Railway Corporation (URC) plays a crucial role in Uganda's international transport, as it represents the only ground link to Dar es Salaam. According to USAID (1994), the cost of transporting goods from Mombasa to Kampala by road is US$ 75-100/tonne while the corresponding cost of rail transportation is US$ 42-58/tonne. It would be logical for the ferry system to carry the bulk of Uganda's goods from Dar es Salaam (Central Corridor; almost 1,600 km) and Mombasa (Northern Corridor; depending on the rout e between 1,200 and 1,300 km.). Ironically, railways rank as virtually no constraint to Ugandan firms, because the firms have no hope of ever using them. 11. Speed is of essence as substantial time is usually wasted to clear goods at the Dar es Salaam or Mombasa ports. Delays are due to a mix of bureaucracy, poor infrastructure and low productivity as shown by the fact that 50% of containers are verified in Eastern Africa, 10% in Asia and 2-3% in Europe, while port productivity in Eastern Africa is often only 30% of 19 countries including Uganda have a waiting period of over 10 years. 87 international norms for the same equipment. According to several traders, while a shipment from Europe to Mombassa by sea takes between three and four weeks, another three to four weeks are needed before they can take possession of the shipment after it arrives in Mombasa. It then takes 6 weeks for a consignment to move from Mombasa to Kampala. The total time between departure of a shipment from Europe and arrival in Kampala is therefore comprised between 3 and 3.5 months, three times what would be needed in ideal conditions. Table 3: Time Needed to Ship a Container from Europe to Kampala by Sea and Road/Rail Action Time Required Internationa (weeks) Comparisons Sea Transport 3-4 3-4 (from Europe to Mombasa) Clearance at Mombasa 3-4 5 days in Asia Transport from Mombasa to 6-7 3 days Kampala from London to Rome (same distance) Overall 12-15 4-5 Source: The Great Lakes Corridor Study. World Bank (1994) and own research. Table 4: Uganda Traffic by Route/Mode 1993: Volumes, Costs and Transit Time Estimated Transit Time Item Dry Cargo POL Total Cost 14 Cost 2 (Days ex ship (mil. (mil. (mi. % CIF % CIF arrival - dry tonnes) tonnes) tonnes) cargo) Imports Road via Kenya 0.29 0.15 0.44 16.5 9.3 39 Rail via Kenya 0.13 0.02 0.15 11.9 7.5 44 Rail via Tanzania 0.06 0.06 0.12 23.0 5.3 21 Imports Total 0.48 0.23 0.71 Exports Road via Kenya 0.10 0.00 0.10 8.1-10.6 9.0 42 Rail via Kenya 0.11 0.00 0.11 7.9-10.3 8.8 37 Rail via Tanzania 0.04 0.00 0.04 8.1 9.0 19 Exports Total 0.25 0.00 0.25 Grand Total 0.73 0.23 0.96 4 Costs of Transport to/from Kampala (Break Bulk): Direct (port changes, transport charges, and cost of agency fees) and Indirect (time in transit, reliability of service and loss and damage to the goods in transit). Cost of Transport to/from Kampala (Containerized): Direct (port charges, transport charges, and cost of agency fees) and indirect (time in transit, reliability of service and loss and damage to the goods in transit). 88 12. Given the low propensity of Ugandan firms to export, very few noted the pitfalls affecting Uganda in air transportation. While the refurbishing of the Entebbe airport is being successfully completed and the number of airlines operating here has increased dramatically, there is still a lack of cold storage facilities and cargo handling is deficient. The Entebbe airport, built in the 1970s, was designed to handle a maximum of 5,000 tonnes of cargo a year, while the current volume is over 20,000 tonnes annually, resulting in obvious congestion. SOEs as a Drain on the Budget 13. Inefficiencies in the SOE sector also have a significant fiscal impact, mostly in terms of forgone revenue. Experience worldwide indicates that SOE performance accounts for indirect and direct subsidies equivalent to 5-8% of GDP. In the case of Uganda, this amounted to USh 208 billion in 1994, or baout 8% of GDP. The subsidy was equal to five and half times Government recurrent expenditure on health (USh. 37.5 billion), more than twice the expenditure on education (USh.88.4 billion), and over 50% of the entire recurrent budget allocation (USh 375b) for 1994/95. Table 5 presents the subsidies to public enterprises since 1992 and projections through 1998 assuming no corrective measures are undertaken. As shown in the table, the fiscal implications of this "passive scenario" are massive. Table 5: Actual and Projected Subsidies to the SOE Sector in Uganda (1992-1998) Type of Subsidy *-Actual in USh hillions- +-Projections in JSh billions--+ (1993 prices) 1992 1993 1994 1995 1996 1997 1998 Direct Subsidies 9.4 15.2 19.4 20.1 22.0 18.0 20.0 linquit Support f 34.2 55.1 78.8 135.0 210.0 285.0 230.5 Finaminyi Teprms 16.9 27.3 57.0 62.7 69.0 75.9 83.5 Fiscal.Terns 28.4 46.8 18.0 15.1 18.0 11.5 11.0 Monopolh situations 8.8 20.2 35.3 37.4 15.0 8.0 4.0 Tou 97.7 164.6 208.5 270.3 334.0 398.0 349.0 % of CDP at:Mkt Prices 7.9% 14. Direct Subsidies. Direct subsidies accounted for 9% of total subsidies in 1994. They include direct cash transfers from Government, donor grants and equity from donor borrowings, and investment grants. These are by far the easiest to determine and control. They have remained roughly the same percentage of total subsidies in 1994 as in 1993. They are projected to maintain the same percentage peaking in 1996 when Government direct (local) transfers will be mainly for a few SOEs like National Enterprise Corporation, Kinyara Sugar Works, Uganda Air Cargo, Civil Aviation Authority, Uganda Posts & Telecommunications Corporation (NURP component) and Uganda Railways Corporation. The direct subsidies will start reducing in 1997 when many of the commercial SOEs will have been divested, thus requiring no subsidy at all, and the policies to ensure their reduction will start taking effect for SOEs yet under Government control. 15. Equity Support and Financing Terms. Equity support is mainly in the form of debt conversions to equity, take over of bad loans and delinking of subsidiaries. They relate to the way that foreign debt (the major source) has been invested in SOEs. Over 60% of this subsidy is attributed to debt conversions to equity in response to corporate emergencies for capitalization. These form by far the largest state subsidy accounting for 38% (USh. 78.8 billion). 89 16. Financing term subsidies are the next largest contributor to subsidies, accounting for 27% (USh. 57b). They derive from the fact that Government does not collect principal or interest, charges low on-lending rates and bears the foreign exchange risk cost, besides guaranteeing the loan to the SOE for no fee. Over 60% of this subsidy is the form of loan and current arrears, arising from the SOEs inability to repay loans, 17. Equity support as a percentage of total subsidies increased from 33% in 1993 to 38% in 1994. The big contributing factor was that in 1994, UEB received US$27m as capitalization through loan conversion. The projections for 1996-98 indicate a significant rise in the activity as a percentage of total subsidies as more loans to utilities will be turned into equity once the projects under implementation are completed and transferred to the SOEs, e.g., UEB from completion of Power III, NWSC from the new water works in up-country towns and Civil Aviation Authority (CAA) from the investment in the Airport through a grant of US$28.0 million from Denmark and a loan from Spain. 18. Fiscal Terms. These are subsidies which are a result of favorable tax treatment to SOEs and come as import duty and corporate tax exemptions, late payment of CTL and corporate tax, and late/underpayment of contributions to the National Social Security Fund (NSSF). The favorable tax treatment (corporation tax exemption) is currently granted to NWSC (by Statute) and Kakira & SCOUL (through investment licensing), while such subsidy is no longer available to other SOEs since 1991. Fiscal terms accounts for 9% of state duties. The largest contributors being duty exemptions and late payment of Corporate taxes combines at 63% of the fiscal terms, and 6% of total state subsidies. Their impact on the budget in terms of foregone cash inflows for Government is significant. 19. In addition a third fiscal subsidy is related to the National Social Security Fund (NSSF) payments. According to NSSF, the SOE sector owes over USh.800m in overdue payments. Some SOEs have collected the funds but failed to pass them on to NSSF while others register less employees than are eligible. Overall, the percentage of the losses due to fiscal terms reduced from 28% in 1993 to 9% in 1994 because of a reduction in duty exemptions and better collections of CTL. The projections are that this subsidy will fall in 1995 and remain stable thereafter. 20. Monopoly Situations. Some SOEs still continue to enjoy subsidies due to their unique monopoly positions in their respective sectors e.g., NWSC, UPTC, UEB and the Sugar companies. The annual total subsidies due to monopoly situations in 1994 are estimated at USh.35.3 billion compared to USh.20.2 billion in 1993, increasing in its contribution from 12.3% in 1993 to 17.7% in 1994. The subsidies due to monopoly situations will continue for sometime, especially for utilities, even if some of the activities of these corporations may be privatized and efficiency increased. The subsidies will peak in 1995 and start reducing as the telecom component of UPTC will have been hived off, and the monopoly broken up. The postal service is already facing competition. 21. Government Arrears. The aggregate Government arrears are estimated at USh. 30 billion in 1994 as compared to USh.55.2 billion in 1993. The reduction is attributed to a number of debt-swaps between Government and utilities over the two year period ending 1994. These arrears accumulate due to various factors, the major ones of which are highlighted below; 90 * Departmental accounting officers defer payment for services provided by public corporations, and instead settle more immediate needs or pay less flexible creditors. * Over time, budgetary provisions for these services have fallen far below the actual amounts charged, reinforcing the departments' reluctance to pay. * Budgetary provisions for utilities are diverted to other uses by the line Ministries. 22. In 1994, the arrears by Government to the SOE sector have significantly reduced since the practice of cross-debt settlement/deduction at source, although this may encourage its departments to accumulate arrears in the knowledge that all future arrears will in any case be offset. Government arrears are projected to decrease as Government is more strict now about its departments meeting their bills to the utilities. In addition to Government, the SOE sector has also accumulated arrears to the utilities estimated at USh.12 billion in 1994. The arrears situation in 1995 is projected to be more manageable as a mechanism for resolving these arrears has now been established. 23. Granting and Monitoring Subsidies. Previously, there has been difficulty in determining and monitoring subsidies in Government as responsibility was dispensed between many points, e.g. direct subsidies accounting for 9% of total subsidies were left to the sectoral ministry to administer. Now, however, this has been brought into the framework of the three- year Public Investment Plan which is continually updated during the budgetary process. Equity Support, mostly in the form of conversion of past funding into equity, which forms at least 38% of total subsidies have mainly been due to SOEs inability to repay or the need to replenish its net worth. These conversions were not accompanied by restructuring plans for remedying causes of SOE losses or amending on-lending agreements and in many ways SOEs may have regarded this as a grant. Now, however, a Debt Settlement Committee chaired by the Director of Budget and with members form the Ministry of Finance, the Attorney General, the Auditor General and the ICG, consider such conversion needs which are in turn to be discussed in the Ministry of Finance's Development Committee. I IMAGING Report No.: P 7097 UG Type: MDR