Tanzania - The challenge of reforms : growth, incomes and welfare (Vol. 1 of 3) : Main report

Report No. 14982-TA Tanzania The Challenge of Reforms: Growth, Incomes and Welfare May :31, 1996 ( ()LI1) I I v ( )I )( 'I )1 I I( \l1 I ) 1 j I( JR II AIr i iA I( i i) p i Document of the World Bank  Covernment Fiscal Ycar FY94=.July 1. 1993 to June 30. 1994 Currency Equivalents Currency Unit: Tanzania Shilling (T Sh) Bureau Selling Rate: US$1.00 =T Sh 572 (March, 1996) Bureau Buyinlg Rate: US$1.00 - T Sli 552 (March. 1996) ACRONYMS AND ABBREVIATIONS ADIS Agricultural D)ivcrsilication and Intensificationi Survey AIDS Acquired Immune Deficiency Syndrome ATC Air Tanizania C'orporation BAWATA Women's Council ol Tanzania BOT Bank of T'anzania CRDB Cooperative and Rural D)evelopment Bank CSPD Child Survival, Protection, and Development Program DHS [)emographic and Ilealth Survey EPZ Export Processing Zone ERB Economic Research Bureau ERP Economic Recovery Program ESRF Economic and Social Research Foundation FAO Food and Agriculture Organizatioln FERP First Fconomic Recovery Plrogramii (Zanzibar) FGT Foster, Girecr. TIhorbeclke F Y Fiscal Year (GlDP Gross D)omestic Product GNP (iross National Product GTZ Giermiiani Technical Assistance Agency HC'TI lHighi Court of Tanzania HIV Human Immun.o-deficiency Virus HRD Humani Resource Development IDA Internationial Development Association IDA Iron Deficiency Anemia (in C'hapter 4) IDD Iodine Deficiency Disorder IFM Institute of Finance Management IEPRI Interinationial Food Policy Research Institute IL1O International Labor Organization IRT Individualization, Registration, and Titling LFS Labor Force Survey lIMA Law of Marriage Act LRC Law Refiorm Coilmimiission of Tanzania  MOF Ministry of Finance MP Member of Parliament NBC National Bank of Commerce NCPI National Consumer Price Index NGO Non-Governmental Organization NIGP National Income Generating Program NISS National Informal Sector Survey NLP National Land Policy NMC National Milling Corporation OECD Organization for Economic Cooperation and Development PAYE Pay As You Earn PC Planning Commission PEM Protein Energy Malnutrition PLC Presidential Land Commission PMO Prime Minister's Office PPA Participatory Poverty Assessment PSLE Primary School Leaving Exam SADCC Southern African Development Coordination Conference SERP Second Economic Recovery Program (Zanzibar) TACOSODE Tanzania Council for Social Development TANESCO Tanzania Electric Supply Company TANGO Tanzanian Association of Non-Governmental Organizations TAZARA Tanzania Zambia Railway Authority TFNC Tanzania Food and Nutrition Center TRA Tanzania Revenue Authority TRC Tanzania Railways Corporation UDSM University of Dar Es Salaam UNDP United Nations Development Program UNICEF United Nations International Children's Emergency Fund UNU United Nations University UPE Universal Primary Education URT United Republic of Tanzania US$ U.S. Dollar VAD Vitamin A Deficiency VAT Value Added Tax WHO World Health Organization ZCT Zanzibar Commission for Tourism ZILAP Zanzibar Integrated Land Use Plan ZIPA Zanzibar Investment Promotion Agency ZSTC Zanzibar State Trading Corporation ZSTCDA Zanzibar Stone Town Conservation and Development Authority   COUNTRY DATA - TANZANIA Area: Population, incl. Zanzibar (1994) Density (1994) 945200km2 28.8 millions 32.8 per km2 POPULATION CHARACTERISTICS* HEALTH* Crude birth rate (per 1000): 46 Population per physician (1992) 21496 Crude death rate (per 1000): 15 Population per hospital bed: 981 Infant mortality (per 1000 live births): 115 INCOME DISTRIBUTION* HOUSEHOLD INCOME* % of national income, highest quintile: 41.6 Share of top 20% of household: 45.4 % of national incomc, lowest quintile: 22.3 Share of bottom 20% of household: 6.8 ACCESS TO SAFE WATER (between 1988-93) ACCESS TO ELECTRICITY* % of population - urban 75.0 % of household - urban: 35 % of population - rural 46.4 % of household - rural: I NUTRITION (1994) EDUCATION* Caloric intake per person: 2206 Adult illiteracy rate (% of population) 23 Primary school enrollment - (% of 68 relevant age group) GNP PER CAPITA IN 1994: US$140 GROSS DOMESTIC PRODUCT (1994) mln. US$ _ GDP at market prices 3576 100.0 Gross domestic investment 1098 30.7 Consumption 3479 97.3 Exports of goods and NFS 944 26.4 Imports of goods and NFS 1944 54.3 OUTPUT at factor cost (1994) Value Added mln. US$ % Agriculture 1860 57.1 Industrv 544 16.7 Services 853 26.2 Total 3257 100.0 GOVERNMENT FINANCE Central Government bln. TSh % GDP FY95 FY90 FY93 FY94 FY95 Current receipts 331.2 14.1 12.9 15.0 14.5 Current expenditure 431.0 15.5 19.0 18.4 18.9 Current balance -99.8 -1.4 -6.1 -3.4 -4.4 Development expenditure 164.0 3.3 5.1 4.6 7.2 * 1993 estimates, unless indicated othenvise.  COUNTRY DATA - TANZANIA MONEY, CREDIT & PRICES (end-period) FY86 FY88 FY89 FY90 FY91 FY92 FY93 FY94 FY95 Money supply (billion TSh) 38.7 62.8 86.2 125.3 159.1 223.5 321.5 469.1 647.3 Bank credit to public sector (billion TSh) .. .. 108.3 131.4 140.3 128.4 193.8 239.7 305.4 Bank credit to private sector (billion TSh) .. .. 21.6 26.5 40.7 53.2 78.3 170.8 168.8 Money as % ofGDP 29.7 22.0 18.1 18.7 18.8 21.6 25.4 29.1 28.3 Consumer price index (FY87=100) 77.0 129.7 167.3 204.5 269.3 336.0 416.0 541.7 725.8 Annual percentage changes in: Consumer price 32.4 29.7 29.0 22.2 31.7 24.8 23.8 30.2 34.0 Bank credit to public sector .. . . 21.3 6.8 -8.5 51.0 23.7 27.4 Bank credit to private sector .. .. .. 22.7 53.7 30.7 47.2 118.0 -1.2 BALANCE OF PAYMENTS (million US$) MERCHANDISE EXPORTS (1994) FY86 FY92 FY93 FY94 FY95" mln. US$ % Share Exports of goods andNFS 425.7 572.8 617.2 854.7 1039.8 Coffee 115.4 22.2 ImportsofgoodsandNFS 1105.2 1701.0 1850.7 1913.1 1990.9 Tea 39.5 7.6 Resourcebalance -679.5 -1128.2 -1233.5 -1058.4 -951.1 Cotton 105.1 20.2 Tobacco 20.6 4.0 Netfactorincome -105.1 -194.5 -177.9 -153.8 -139.4 Cashewnuts 51.2 9.9 Net current transfers 241.7 456.4 463.2 450.0 436.6 Total 519.3 100.0 Balance of Current Account -542.9 -866.3 -948.2 -762.2 -653.9 Official grants 278.3 569.0 582.1 459.0 376.0 EXTERNAL DEBT as of Dec. 31, 1994 mln. US$ Net MLT borrowing -0.8 146.1 -167.0 -154.7 -57.2 Public & publicly guaranteed' 6444.4 Disbursements 201.5 288.6 189.4 206.6 265.1 Non-guaranteed private debt 12.0 Amortization 202.3 142.5 356.4 361.3 322.3 Other capital (net), -80.5 41.2 41.3 272.2 96.1 including errors & omissions DEBT SERVICE RATIO, FY95 Overall Balance -345.9 -110.0 -491.8 -185.7 -239.0 Public & publicly guaranteed 45.9 Non-guaranteed private debt Gross reserves (end year) .. 384.9 294.6 306.3 255.1 RATE OF EXCHANGE (annual average) IBRD/IDA LENDING as of Sept. 30, 1995 mrln. US$ FY86 FY92 FY93 FY94 FY95 Outstanding & disbursed 2716.2 TSh per US$ 1.00 18.6 250.5 340.1 477.6 536.4 Undisbursed 816.0 US$ perTSh 00.00 5.36 0.40 0.29 0.21 0.19 Outstanding incl. undisbursed 3009.0 1 / Preliminary estimates. 2/ including use of Fund credit, but excluding arrears.  TANZANIA TE CHALLENGE OF REFORM: GROWTH, INCOMES AND WELFARE Volume I: Main Report CONTENTS Preface ........................................................ i Executive Summary ........................................................ v Part I: Reforms and Economic Recovery 1. ECONOMIC REFORMS AND RECOVERY ..1..................................... A. OVERVIEW OF ECONOMIC REFORMS AND PERFORMANCE . .1....................... B. REFORMING THE INCENTIVE SYSTEM ........................................................ 3 Price and Market Reforms ........................................................ 3 Exchange Rate Reform ........................................................ 4 Trade and Industrial Policy Reforms ........................................................ 4 Export Incentives ........................................................ 5 C. STRUCTURAL REFORMS ........................................................ 5 Financial Sector Reforms ........................................................ 5 Monetary Policy and Inflation ........................................................ 8 Parastatal Reforms ........................................................ 9 Reform of Public Administration ........................................................ 10 D. ECONOMIC RECOVERY AND RECENT ECONOMIC DEVELOPMENTS .. 12 Recovery of Agriculture ........................................................ 13 Min ing and Quarrying ........................................................ 17 Tourism ........................................................ 18 The Informal Sector ........................................................ 18 Export Recovery ........................................................ 20 E. THE PERFORMANCE OF INVESTMENTS ........................................................ 21 Performance of Investments ........................................................ 21 Issues in Investment ........................................................ 23 F. BASIC MACROECONOMIC BALANCES ........................................................ 24 Recent Performance ........................................................ 24 Resource Balance ........................................................ 26 Savings and Investment ........................................................ 26 Resource Mobilization ........................................................ 27 Balance of Payments ........................................................ 28 External Environment ........................................................ 30  2. FISCAL AND PUBLIC RESOURCE MANAGEMENT ............................................. 31 A. OVERVIEW AND RECENT DEVELOPMENTS ........................................................ 31 Overview ........................................................ 31 Recent Fiscal Perforn ance ........................................................ 32 SECTION I. EXPENDITURE MANAGEMENT ...................................... 36 B. EXPENDITURE PERFORMANCE ........................................................ 36 Recurrent Spending ........................................................ 36 Recurrent Savings ........................................................ 38 Development Expenditures ........................................................ 39 Public Administration Expenditures ........................................................ 39 C. DEBT SERVICE ........................................................ 41 D. SECTOR EXPENDITURES ........................................................ 43 E. WHO Is BENEFITING FROM GOVERNMENT SPENDING? ...................................... 46 Education Expenditures ........................................................ 46 Health Expenditures ........................................................ 47 Spending on Water ........................................................ 48 F. ANALYSIS AND DISCUSSION ........................................................ 49 G. GOVERNANCE ........................................................ 50 SECTION II. REVENUE MOBILIZATION ........................................................ 52 H. LOCAL FUNDS AND TAX REFORM .......................................... 52 Introduction ........................................................ 52 Tax Structure ........................................................ 53 1. TAX REFORM ........................................................ 54 Tax Administration and Compliance ........................................................ 54 Customs Duties ........................................................ 55 Sales and Excise Taxes ........................................................ 56 Income Tax ........................................................ 56 Preparations for the Value Added Tax (VAT) ........................................... 58 Part II: Growth, Incomes and Household Welfare 3. INCOMES, INEQUALITY AND POVERTY ........................................................ 59 A. OVERVIEW ........................................................ 59 B. INCOMES AND EXPENDITURES ........................................................ 60 Inequality ........................................................ 62 C. POVERTY UPDATE ........................................................ 65 Definition of the Poverty Line ........................................................ 65 Rural Poverty Between 1983 and 1991 ..................................................... 65 Recent Evidence on the Incidence of Poverty ........................................... 66 Spatial Mapping of the Poor ........................................................ 68 Regional Mapping of the Poor in 1993 ...................................................... 69  D. THE HUMAN DEVELOPMENT DIMENSION OF POVERTY ..................................... 71 Access to Water ............................................................. 71 Literacy ............................................................. 72 Infant Mortality ............................................................. 73 Distance to Major Centers and Access to Amenities ................................. 73 Tanzania and Neighbors ............................................................. 74 Conclusion ............................................................. 74 E. POVERTY, INEQUALITY AND GROWTH: OUTLOOK FOR THE FUTURE ................. 75 Poverty and Growth ............................................................. 76 Poverty and Inequality ............................................................. 76 Sectoral Growth and Poverty ............................................................. 77 F. CHARACTERISTICS OF POVERTY ............................................................. 78 Demographic Characteristics of the Tanzanian Household ....................... 78 Pattern of Expenditure ............................................................. 79 Sources of Income ............................................................. 80 Diversification of Income ............................................................. 81 G. WHAT FACTORS UNDERLIE POVERTY? ............................................................. 81 Education ............................................................. 81 Infrastructure and Market Integration ........................................................ 83 Education and Infrastructure ............................................................. 85 Supply of Inputs ............................................................. 85 Savings and Credit ............................................................. 86 Distribution of Assets ............................................................. 87 4. FOOD SECURITY AND NUTRITION .................................................................... 91 A. OVERVIEW .................................................................... 91 B. FOOD SECURITY .................................................................... 91 C. NUTRITION .................................................................... 93 Protein Energy Malnutrition .................................................................... 94 Micronutrient Deficiencies .................................................................... 95 D. FOOD SECURITY AND MALNUTRITION .............................................................. 95 E. CAUSES OF MALNUTRITION .................................................................... 96 Proximate Causes .................................................................... 96 Household Factors .................................................................... 97 Food Consumption and Energy Intakes ..................................................... 98 Community Level Factors .................................................................... 98 National Level Factors .................................................................... 99 F. CONSTRAINTS TO FURTHER MARKET INTEGRATION ........................................ 101 G. NUTRITION POLICY AND INTERVENTIONS ....................................................... 102 Policy Interventions on Micronutrients ................................................... 102  5. THE STATUS OF WOMEN ............................................... 105 A. OVERVIEW .............................................. 105 B. LEGAL STATUS .............................................. 105 Possession of Property .............................................. 106 Legal Illiteracy .............................................. 106 C. POLITICS, ADVOCACY AND ORGANIZATION ............................................. 107 D. ECONOMIC AND SOCIAL STATUS ............................................ 107 Female-headed Households .............................................. 107 Intra-howsehold Inequality ...............................................111 6. LAND USE AND LAND REFORM ISSUES .............................................. 113 A. ENVIRONMENT AND POVERTY .............................................. 113 Overview and Issues .............................................. 113 Environmental Stress and its Results .............................................. 115 Improving Environmental Quality .............................................. 119 B. LAND TENURE AND LAND POLICY .............................................. 120 Overview .............................................. 120 Evolution of Land Policy .............................................. 121 Recent Developments in Land Policy .............................................. 122 Part III-A: Employment and Targeted Programs 7. EMPLOYMENT AND WELFARE .............................................. 125 A. OVERVIEW .............................................. 125 B. LABOR FORCE AND EMPLOYMENT .............................................. 125 The Profile of the Labor Force .............................................. 125 Recent Trends in Employment: 1978-88 .............................................. 126 Current Employment Situation by Key Sectors ....................................... 127 Types of Employment and Poverty .............................................. 128 C. UNEMPLOYMENT AND UNDEREMPLOYMENT .............................................. 129 Unemployment Rate .............................................. 129 Underemployment .............................................. 131 Household Unemployment and Poverty .............................................. 131 D. FUTURE EMPLOYMENT PROSPECTS .............................................. 132 Outlook for Employment Growth .............................................. 132 Employment Growth Potential by Sector .............................................. 132 E. POLICY CONSIDERATIONS BY SECTOR .............................................. 134 Agriculture .............................................. 134 Informal Sector .............................................. 135 Formal Sector .............................................. 135 Supply-side Issues .............................................. 136  8. TARGETED PROGRAMS AND SAFETY NETS .................................. 137 A. OVERVIEW .............................................................. 137 Traditional Safety Nets .............................................................. 138 Formal Safety Nets .............................................................. 138 B. LABOR-INTENSIVE INFRASTRUCTURE PROGRAMS ........................................... 139 Lessons From Previous Attempts ............................................................ 139 Future Potential .............................................................. 140 C. CREDIT-BASED SELF-EMPLOYMENT PROGRAMS FOR WOMEN ........................ 141 Performance .............................................................. 141 D. URBAN HOUSING FOR Low-rNcoME FAMILIES ................................................ 142 E. PROGRAMS FOR REHABILITATION OF AIDS ORPHANS .................................... 142 F. NGOS AND IMPLEMENTATION OF TARGETED PROGRAMS ................................ 143 Part III-B: Outlook on Growth and Welfare 9. OUTLOOK ON GROWTH AND WELFARE ..................................... 145 A. INTRODUCTION AND OVERVIEW .............................................................. 145 B. THE CHALLENGE OF GROWTH AND POVERTY REDUCTION .............................. 147 C. FISCAL AND PUBLIC RESOURCE MANAGEMENT .............................................. 149 Fiscal Restructuring .............................................................. 149 The Role of Effective Government .......................................................... 153 Public Institutions .............................................................. 154 D. EXTERNAL ASSISTANCE .............................................................. 155 E. AGRICULTURAL AND RURAL DEVELOPMENT ................................................... 156 F. COMMERCIALIZATION AND PRIVATE SECTOR DEVELOPMENT ......................... 158 Other Industries .............................................................. 159 G. INFRASTRUCTURE CONSTRAINTS .......................................... 160 Energy .............................................................. 160 Transport .............................................................. 161 H. HUMAN RESOURCE DEVELOPMENT .............................................................. 163 Tables Table 1.1: Growth of Selected Macro-Indicators, FY76-94 ..............................................3 Table 1.2: Distribution of Commercial Banks Domestic Lending ...................................7 Table 1.3: Money, Credit and Inflation ...............................................................8 Table 1.4: Direct and Indirect Subsidies for Parastatals ....................................................9 Table 1.5: GDP and Real Growth at Constant 1976 Prices ............................................. 12 Table 1.6: Selected Agricultural Production Indices ....................................................... 16 Table 1.7 Non-traditional Exports Share in Total Exports, 1980-93 .............................. 20 Table 1.8: Composition of Fixed Capital Formation ....................................................... 22 Table 1.9: Percent of GDP in Current Prices .............................................................. 25 Table 1.10: Balance of Payments .............................................................. 29  Table 2.1: Summary of Central Government Operations ................................................ 34 Table 2.2: Economic Classification of Recurrent Expenditures, FY89-FY94 ................ 37 Table 2.3: Development Expenditure ........................................................... 39 Table 2.4: Debt Service ........................................................... 42 Table 2.5: Sector Expenditure ........................................................... 44 Table 2.6: Distribution of Water Subsidies, Rural-Urban Differential ............................ 48 Table 2.7: Central Government Tax Revenue, FY90-94 ................................................. 54 Table 3.1: Some Household Surveys Available in Tanzania ........................................... 61 Table 3.2: Gini Coefficient for the Distribution of Income in Tanzania, 1969-93 .......... 64 Table 3.3: Household's Perception of Changes in Inequality, 1985-1995 ...................... 64 Table 3.4: Poverty Measures by Location and Poverty Line, 1993 ................................. 67 Table 3.5: Evolution of Poverty, 1983-1995 ........................................................... 67 Table 3.6: Changes in Rural Literacy Rates ........................................................... 73 Table 3.7: Evolution in Access to Services, 1976-93 ...................................................... 74 Table 3.8: Evolution of Social Indicators ........................................................... 75 Table 3.9: Probability of Low Income, 1993 ........................................................... 78 Table 3.10: Expenditures Patterns (Cash and Kind) by Household Type, 1993 ... 79 Table 3.11: Sources of Household Income (1993) ........................................................... 80 Table 3.12: Education and Rural Household Welfare ...................................................... 82 Table 3.13: Net Enrollment Rate by Age Group, by Location and Welfare Quintile, 1993 ........................................................... 83 Table 3.14: Road Quality and Household Welfare ........................................................... 84 Table 3.15: Access to Land and Size of Holding ....................................... 88 Table 3.16: Livestock Ownership and Distribution, 1993 ................................................ 89 Table 4.1: Evolution of Calorie Intake ........................................................... 92 Table 4.2: Prevalence of Nutritional Deficiencies .......................................................... 93 Table 4.3: Indicators of Malnutrition in Children Under Five ........................................ 94 Table 5.1: Features of Male and Female Heads of Household ..................................... 109 Table 5.2: Mean Expenditure by Household per Adult, 1993 ...................................... 109 Table 5.3: Rural Poverty Status by Head of Household in 1993 and 1995 .................. 110 Table 5.4: Percentage of Individuals Over 14 Who Own Small Assets, 1993 ............. 111 Table 5.5: Women's Enrollment as Percentage of Total, 1990 .................................... 111 Table 6.1: Breakdown of Deforestation ........................................................... 117 Table 7.1: Profile of the Labor Force 15-64 Years Old (1991) ..................................... 126 Table 7.2: Employment Distribution By Status ........................................................... 127 Table 7.3: Poverty and Employment Status/Sectoral Employment ............................... 129 Table 7.4: Unemployment Rates by Sex, Region and Age ............................................ 130 Table 7.5: Possible Employment and Underemployment Patterns in the Future for Working Age (15-64 years Old) Work Force ..................................... 133 Table 8.1: Tanzania: Construction of Unpaved Feeder Road with Gravel Surface: A Comparison of Equipment-intensive with Labor-intensive Methods: 1994-95 ........................................................... 140 Table 8.2: Credit for Productive Activities for Women: Loans Advanced and Operating Costs, 1991-94 ........................................................... 142 Table 9.1A: Outlook for Growth and Poverty ........................................ 146 Table 9.1B: Sector Growth Scenarios .............................................. 146 Table 9.2A: Low Case Scenario Selected Indicators ................................... 150 Table 9.2B: Base Case Scenario Selected Indicator ................................... 151  Figures Figure 1.1: Real Exchange Rate Index ........................................................ 4 Figure 1.2: GDP Growth Per Capita ........................................................ 13 Figure 1.3: Manufacturing Growth Rate ......................................................... 17 Figure 1.4: Export Growth, Volume ......................................................... 20 Figure 1.5: Incremental Capital to Output Ratio ........................................................ 22 Figure 1.6: Savings and Investment ........................................................ 26 Figure 1.7: Current Account Balance ......................................................... 28 Figure 2.1: Government Expenditure ......................................................... 35 Figure 2.2: Government Financing ........................................................ 35 Figure 2.3: Recurrent Revenue & Expenditure ........................................................ 36 Figure 2.4: Recurrent Savings ......................................................... 38 Figure 2.5: Administration and Debt Service ........................................................ 40 Figure 2.6: Social and Economic Services ........................................................ 40 Figure 2.7: Distribution of Education Subsidies per Welfare Quintile ............................ 46 Figure 2.8: Distribution of Health Subsidies per Welfare Quintile ................................. 48 Figure 3.1: Expenditure per Adult Equivalent, 1993 ....................................................... 62 Figure 3.2: The Distribution of Income by Quintile, 1993 .............................................. 63 Figure 3.3: Distribution of Poverty by Location, 1993 .................................................... 68 Figure 3.4: Distribution of Population by Welfare Level, 1993 ..................................... 68 Figure 3.5: Distribution of Welfare Index ......................................................... 70 Figure 3.6: Distribution of Population with Access to Safe Water .................................. 71 Figure 3.7: Infant Mortality Indicator ......................................................... 73 Figure 3.8: Poverty and Growth ......................................................... 76 Figure 3.9: Poverty and Inequality ......................................................... 77 Figure 3.10: Family Size and Poverty, 1993 ......................................................... 78 Figure 3.11: Agriculture as Income Source, 1993 ........................................................ 81 Figure 3.12: Percentage of Population Older than 21 Completing Primary or Secondary School, 1993 ........................................................ 82 Figure 3.13: Income and Distance to Markets (1993) .83 Figure 4.1: Food and Energy as Percentage of Requirements (1984/85-1989/90). 92 Figure 4.2: Official Purchases of Food Grains (1983/84-1991/92) .100 Figure 6.1: Land Use in Tanzania .116 Text Boxes Box 3.1: Access to Water ......................................... 72 Box 3.2: Agricultural Markets ......................................... 86 Box 4.1: Coping Mechanisms for Food Shortages ........................................ 94 Box 6.1: Perceptions of Soil Fertility ........................................ 116 Box 6.2: The Time Costs of Depleted Fuelwood Stocks ........................................ 118 Bibliography Map IBRD 27941   PREFACE The Country Economic Memorandum/Poverty Assessment (CEM/PA) is the result of a collaborative effort involving the Government and the major stakeholders of Tanzania. A joint team comprising members of the Government, World Bank, other bilateral and multilateral donors, and local research organizations worked together to prepare the report. The theme and topics covered and the process for preparing the report evolved from initial meetings held in Dar es Salaam in December 1994. The CEM/PA aims at deepening the understanding of economic growth and poverty in Tanzania. The report examines the progress that has been made in economic policy reforms in recent years, and it analyzes the impact of the reforms on economic growth, on overall macroeconomic performance and on the welfare of households. It identifies the factors which have had impact on improving household living conditions. It concludes with a discussion of the main challenges that require priority attention in the effort to accelerate broad-based economic growth and to achieve a fast reduction in poverty in Tanzania. Volume I is the main report for the Mainland, and Volume 2 covers appendices and statistical annexes. Zanzibar is covered in an additional volume. Volume 1 is presented in three parts. Part I reviews the status and impact of the economic policy reform program initiated in the mid-1980s. Chapter I discusses both the progress made in implementing economic reforms and the overall macroeconomic performance during the reform period. Chapter 2 focuses on fiscal management; it investigates both the basic factors underlying the persistent macroeconomic instability and the options for restoring a stable fiscal and financial environment for economic development. Part II analyzes the living conditions of the population. Chapter 3 uses household income and expenditure surveys to analyze the evolution of rural household living conditions. It examines the principal factors that contributed to the reduction of poverty. Chapter 4 deals with nutrition and food security, with emphasis on vulnerable groups, including women and children. Chapter 5 studies the status of women, especially their economic well being, access to social services, and legal and property rights. Chapter 6 examines the scope for sustainable development, focusing on land use, environmental quality, security of land tenure and land reform. Part III examines the prospects for the future. Chapter 7 analyses the profile of the labor force, unemployment, and employment prospects. Chapter 8 reviews the performance of targeted programs and safety nets. It identifies selected, potential income generating activities, which could be used to provide rural employment and enhance the living conditions of targeted lowest income groups. Chapter 9 presents a brief synthesis of the overall report, with emphasis on the prospects for accelerated and broad-based economic growth and fast reduction in poverty. It identifies some major development challenges that require priority attention in order to make faster progress in improving household welfare. The joint team that organized the preparation of the report was chaired by Dr. J. Kipokola, currently Principal Secretary (PS) of the Ministry of Energy and Minerals, and assisted by Mr. R. Mlolwa, currently Economic Advisor in the Office of the President. Other members were Mr. P. M. Lyimo, currently Commissioner for the Budget in the Treasury;  ii Preface Hon. M. A. Kigoda, currently Minister of Industries and Trade; Mr. J. Zayumba, Planning Commission (PC) and Mr. N. Mboyi (PC). Ambassador F. Kazaura, formerly PS of the PC was the supervisor. Mr. S. Odunga, currently PS in the Ministry of Education and Culture, organized a team of Ministry of Finance staff, which discussed sections of the report concerning the budget. Mr. C. Kimei, Director of Research, Bank of Tanzania ( BOT), was a resource person and commented on sections of the report. The members of the team from the World Bank comprised A. Agbonyitor (task manager), Y. Abe and A. Lee (AF2CO); L. Ferreira (responsible for Poverty Assessment) and L. Goodhart, consultant, (AF2PH); and K. Subbarao (ESP). Mr. K. Havnevik (economist, sponsored by SIDA) joined the main mission. The report draws on background papers, which were reviewed by the joint team and also presented for discussion in a workshop. Most of the background papers were prepared by local consultants. They include Professor S. Wangwe and Dr. S. Haji, ESRF, (Growth Potential); Mr. J.K. Kabyemera, FAO, (Agricultural Development); Hon. M.A. Kigoda (Alternative Measures of Poverty. The State and the Voluntary Sector); Dr. M. Mmuya, UDSM, (The Political System and Poverty); Dr. H.P.B. Moshi, UDSM, (Reforms anld Economic Performance); Dr. N. Osoro, USDM (Tax Reform); Ms. M.K. Rwebangira, HCT, (The Status of Women and Poverty); Messrs. F.C. Schechambo and G.L.K. Jambiya, UDSM, (Environment); Dr. J. Semboja, UDSM (The Economy of Zanzibar); and Mr. L. Shirima, IFM, (Participatory Assessment Study of the Voluntary Sector in Poverty Alleviation. An overview Paper on Safety Nets). Other background papers were prepared by: Y. Abe (The Status of Employment and Unemployment. Investments); L. Ferreira and L. Goodhart (Incomes, Inequality and Poverty. Nutrition and Food Securijty. Who benefits from Government Expenditures?); K. Subbarao (Targeted Programs ond Safety Nets); D. Narayan (Participatory Poverty Assessment); C. Kanda (NGOs); A. Agbonyitor (Fiscal and Public Resource Management) and K. Havnevik (Urgent Land Tenure Issues). The Central Bureau of Statistics, and the statistics divisions of BOT and the PC provided valuable support. A. Lee provided statistical support and prepared the statistical annexes. The workshop to discuss the background papers was held in Arusha on May 14-20, 1995, with the theme, "Socio-economic Growth and Poverty Alleviation in Tanzania". It was organized by the Government of Tanzania in collaboration with the World Bank, with support from the Governments of Japan and the Netherlands, as well as ODA, SIDA and DANIDA. The workshop was opened by Hon. Lt. J.M. Kikwete, former Minister of Finance. The workshop provided an opportunity for the representatives of all the major stakeholders to participate in discussing issues relating to economic growth and welfare in Tanzania. Several individuals assisted the joint team in monitoring the presentation and discussion of the workshop papers, especially L. Shirima, Consultant, C. Davies 0 l, {! Tf.1; E. Malangalila (AF2TA) and M. Van Vliet (The Netherlands Embassy). More than 30 NGOs participated in the Workshop, and about 70 took part in a participatory assessment of the NGO sector. Information on household welfare was obtained  iii Preface through several household surveys, as well through a Participatory Poverty Assessment that surveyed more than 1,000 households in 80 rural villages. ODA supported a Participatory Poverty Assessment, the production of a video and a consultant. The Government of Netherlands supported the consultation workshop at Arusha and the draft report discussion workshop at White Sands. The Govemment of Denmark supported a study of Safety Nets. SIDA managed two of the background papers for the report (Political Systems and Poverty. Women and Poverty) and supported a third (Urgent Land Tenure Issues). Throughout the process the donor community was kept informed of the status of the work through a monthly newsletter. The report was prepared under the general supervision of Messrs. M. Carter, Chief, and G. Gebhart, Acting Chief (AF2CO); J. van Lutsenburg Maas, Chief (AF2PH); F. X. Colaco and J. Adams, Directors (AF2). The Lead Economist at the inception of the work was Mr. P. Miovic; Mr. S. Yusuf, Lead Economist, supervised the completion of the report. Mr. M. Konishi (former Resident Representative) and R. Brigish (current Resident Representative in Tanzania) supported the mission's work. The peer reviewers were E. Ablo (ESP), 1. Bannon (LA2CO), B. Milanovic (PRDTE), F. Iqbal (EA3CO), and S. Shetty (EA I CO). The team wishes to acknowledge the helpful comments and support received from participants of workshops in Arusha and White Sands, members of the Tanzania country team, and from interested staff in the Bank and in the donor community, especially H. Altvall, E. Korsgren and K. Havnevik (SIDA); P. Evans and D. Peddley (ODA); P. Shaw and J.W. van Holst Pellekaan (AFTHR); W. Shaw (AF2CO); J. Coates and P. Dewees (AF2AE); G. Swamy (AF2PH) and J. Maweni (AF2EI). S. Hadler (AF2TA) guided the Zanzibar study. A. Folimer (AF2PH) provided editorial support. Z. Ahmed (AF2CO) assisted in proofreading the final report. K. Rivera, L. Canamaso and R. Covington (AF2CO) provided secretarial and production support. L. Tsoflias provided invaluable assistance in organizing the workshop. Administrative and logistic support was also provided by H. Taddese and L. Cuadro- Meliotes (AF2PH), M. Lynch (AF2CO), and by C. Makoye, E. Sakaya, M. Mwakangale, M. Sabai, G. Minja and L. James at the Resident Mission, Dar es Salaam. A draft of the report was discussed with the Government, private sector individuals and donors in March 1996. Discussions included a one day workshop for Govermment officials. The workshop was chaired by R. E. Mariki, PS Planning; it was opened by M. Lumbanga, Chief Secretary. The mission appreciates the cooperation received from the Tanzanian authorities, local consultants and other stakeholders.   EXECUTIVE SUMMARY A. ECONOMIC PROGRESS AND WELFARE 1. For about a decade now, Tanzania has pursued a program to reform the regime of restrictive state controls which, together with a series of external shocks, had led to economic stagnation and a sharp deterioration of living conditions between the 1 970s and early 1980s. The economic reforms have now been accompanied by political reforms which ushered in a multi-party democracy in November, 1995. The last Country Economic Memorandum (CEM) prepared in 1991 discussed in detail the thrust of the reform program starting in the mid- 1980s and the progress that had been made, especially in liberalizing trade and agricultural markets. It also emphasized the need for structural reforms for the financial sector, the parastatals, and the system of public administration. 2. This report examines the progress made in the adjustment effort since the 1991 CEM. It assesses policy and overall macroeconomic performance. It uses household surveys to examine the impact of the reforms on household expenditures and poverty. Finally, it identifies the main challenges that require priority attention in order to advance the reform effort and to promote faster economic growth and poverty reduction. 3. Overall, Tanzania has made progress in phasing out many of the remaining restrictions. For example, controls on agricultural input and output markets and prices were abolished, and farmers' incentives were restored. Restrictions on traditional exports and on the retention of export receipts were removed. Quantitative import controls were lifted, except for selected items related to national security and public health. Taxes were simplified; for example, the number of sales tax rates dropped from seven to two in the 1990s. Official exchange rate controls were phased out: the two-tier exchange rate system introduced earlier was replaced by a unified rate system with auctions in 1993; in turn, the auction was replaced by a market-determined system in 1994. Interest rate controls and restrictions on private banking were lifted. Investment policies were liberalized, with steady progress in the parastatal reform program initiated in 1993. Of the 300 odd parastatals in existence initially, 75 sales agreements were signed as of December, 1995, and some 51 non-performing assets were liquidated. Agreements with employees remain critical for accelerating the pace of the parastatal reforms. Much remains to be done to sustain the reform process in areas where limited progress has been made, especially in fiscal management and macroeconomic stability, the state-owned banks, the non-financial parastatal sector and public administration.  ii Executive Summary 4. GDP growth at factor cost averaged about 4 percent, with continuing growth in per capita income in the 1990s; the 4 percent growth rate compares favorably with the median growth rate of 2.6 percent for 128 countries during 1986-94. The reforms, supported by the rehabilitation of major roads, helped agriculture to expand at an average annual rate of about 5.4 percent during 1991-94. This compares favorably with average annual growth under 3 percent during the decade prior to the start of reforms, and is more than twice the average for African countries. The abolition of state monopolies and deregulation of investments, commerce, external trade and agricultural marketing have fostered informal sector growth. The informal sector, which is not covered by the official data, expanded faster than official GDP. The share of the urban informal sector in total GDP rose from 10.3 to 14.5 percent between 1985 and 1991. The declining trend of agricultural exports was reversed; real export growth averaged about 4-5 percent during 1990-94, and improved access to imports ended severe shortages and hardships to consumers. 5. The economic growth improved household living conditions. Household income surveys show that the incidence of rural poverty has declined from about 65 percent in 1983 to 51 percent in 1991. More recent household expenditure surveys show mixed results, though poverty was broadly the same in 1995 as in 1991. Widespread low income levels persist, however, because of the deep decline in per capita incomes during the period of strict controls and economic stagnation. The quality of land, cash and export crop production, education and infrastructure are among the strongest contributors to household income and welfare. Educational and infrastructure investments are more effective in increasing household welfare when implemented jointly rather than separately. Access to national and international markets, especially the marketing of export crops, is an important factor in raising the welfare of agricultural households. The regions and the households that lie beyond the network of market institutions and which do not grow cash crops are most likely to be poor. The cash and export crop connection to rural household incomes underlines the role of macroeconomic, exchange rate and other policies that make exports competitive. 6. Considerable efforts have been made to reverse the stagnation and decline of the social indicators. These indicators slipped with the decline of the economy and collapse of revenue in the mid- 1 970s, and remain poor. International experience supports the view that there are lags in the response of social indicators. Current consumption tends to lead social indicators, which depend in part on the built-up stock of physical and human capital. The distribution of social sector subsidies has been inequitable. The bottom 20 percent of the population receives 14.5 percent of total subsidies, while the top 20 percent receives 28.6 percent. The most equitable programs are primary education and primary health services; the least equitable are expenditures on university education, hospitals and water.  Executive Summary iii B. CHALLENGES BEYOND LIBERALIZATION 7. About half of the population is estimated to be living at very low levels of welfare. While this represents a substantial improvement over performance in the 1 970s and 1980s, the current economic growth rate will not provide dramatic improvement in living conditions. Using 50 percent poverty incidence as a base line, assuming (a) the current 4 percent economic growth rate and 2.8 population growth rate, and (b) no change in income distribution, an estimated 45 percent of the population will be living in poverty by the turn of the century, and 35-40 percent will be in a similar position a decade from now. Any deterioration in equity would worsen poverty. Even if economic growth were 8 percent, a child born in Tanzania today would have an average income equivalent to approximately US$440 (estimated per capita income at the current exchange rate) on reaching the age of 15 years. By the turn of the century, about a third of the population would still live in poverty. This underlines both the magnitude of the challenge facing Tanzania and the importance of economic growth to poverty reduction progress. 8. Tanzania has the land resource to achieve faster and diversified agricultural growth and to raise welfare. There is substantial potential for tourism, mining and manufacturing. It is suitably located to provide shipping and transit facilities to neighboring landlocked economies in competition with Mozambique and South Africa. But the country faces serious challenges. Agricultural growth is susceptible to drought conditions as well as to various constraints. Inadequate infrastructure, including roads and water resources, lack of secure and timely supply of inputs during the planting season, and lack of access to financial services and to markets are major concerns in agricultural households. Much of the potential in other sectors, such as tourism, mining and manufacturing, is underutilized. Most of these sectors were controlled by inefficient parastatals until recently. To realize the growth potential, there is an urgent need to give greater attention to removing limiting factors, especially those concerning inadequate economic management; weak delivery of basic infrastructure and social services; and shortage of skills. 9. Economic management has slipped and remains a serious concern. Revenues fell by 2.3 percentage points of GDP even though total expenditures rose by 7.1 percentage points of GDP during FY92-FY95. Even including aid, the fiscal balance deteriorated sharply from a small surplus in FY92 to an estimated deficit of 8.9 percent of GDP in FY95; the situation remained worrisome during the first half of FY96. It is becoming increasingly difficult to finance basic public services from domestic revenues; as a result, aid dependence has increased. The immediate causes of the difficulties are weak tax administration and lax expenditure control. These reflect a basic underlying problem of ineffective public administration. Debt overhang is heavy and problematic. Total external debt is equivalent to over 200 percent of GDP, and debt service would absorb about half of export earnings under existing terms. The upsurge of fiscal deficits is worsening domestic debt. Interest payment on domestic debt tripled to an estimated T Sh 57.7 billion during FY94-FY95. Inflation accelerated from about 22 percent to over 30 percent in 1992-94, instead of the target one digit, and is now about 27 percent.  iv Executive Summary 10. Public service delivery needs to be improved. In particular, non-wage recurrent costs of most programs are under-funded, affecting the operation and maintenance of services, including extension service, water, and other infrastructure and social services. The lack of local counterpart funds makes it difficult to expand development services to new communities. Public administration is considered ineffective for various reasons, including shortage of high level skills, low pay, overlapping controls and over-staffing at lower level skills. 11. The banking system remains weak, even with the removal of interest rate controls, abolition of state monopoly of banking, and the growth of private banks. The bad loans of the dominant state-owned National Bank of Commerce (NBC), for example, culminated in losses of over US$180 million in FY94. Cash requirements to cover weak and failing banks (NBC, Meridien BIAO, Tanzania Housing Bank and the People's Bank of Zanzibar) were about half of the total broad money growth in FY95. Bad debtors, including parastatals and cooperatives, remain a serious problem to all state-owned financial institutions. The combination of weak banks and inadequate fiscal management is crowding out the private sector's access to credit, which fell by 1.2 percent in FY95. 12. Human resource development requires greater attention to provide the managerial capacity and skills needed to harness natural resources. Over a half of the labor force is under 30 years and mainly unskilled; few have post-primary education or training. While the literacy rate is favorable at about 80 percent, functional literacy is considered to be lower. Social service delivery is weak; about 63 percent of households surveyed in 1993 said that the availability of drugs in health facilities was poor or very poor. School outcomes in the public sector have declined. Gross public secondary school enrollment is about 5-8 percent. While the lifting of restrictions on private schools have raised secondary school enrollments, access to lower income groups remains very limited. C. TOWARDS A REFORM AND DEVELOPMENT AGENDA Priority Issues 13. A priority reform agenda is essential to consolidate the gains made and to advance the process. As in many other countries, reforms in Tanzania have made progress where measures are simple in administrative terms; for example, in liberalization and abolition of controls. Progress has been slower where reforms require the creation of management infrastructure, administrative capacity and restructuring of institutions. The situation warrants focusing on selected priorities to avoid over-stretching capacity. 14. The most serious and urgent issue is revenue and expenditure reforms to restore a stable macroeconomic environment for growth and for the orderly financing of critical development programs. The restructuring of public administration is an inescapable element of fiscal reforms. Improvement in tax administration is critical for (i) closing various sources of tax evasion, such as leakages through Zanzibar, exemptions and transit trade, (ii) reviewing income taxes to cover various allowances which are currently  Executive Summary v excluded and (iii) broadening the tax base to new payers, including small scale activities. Effective tax reform could raise revenue by up to 1.1 percent of GDP a year, on average, during the next four years. Fiscal measures will need to limit total expenditure growth while redirecting allocations towards priority areas, such as social services, research and extension services in agriculture, and rural infrastructure. This will involve monitoring each sector budget ceiling to curb expenditure overruns. Recent steps taken to expand private sector role also have priority. Other reforms would stress restructuring public administration; identifying critical and specific tasks; and reforming institutions, personnel policies and incentives to perform those tasks more effectively. Fiscal reforms will need to be supported by a debt management strategy, which will limit domestic borrowing and reschedule eligible external debt on terms favorable to Tanzania. 15. Reorganizing the financial institutions, especially NBC, to mobilize and allocate resources efficiently will be critical for eliminating the losses, periodic capital infusion, and the subsidy burden on the budget. The implementation of on-going measures to limit lending to non-performing creditors, and to downsize and stabilize NBC is an important first phase of measures to make NBC self-sustaining. More work is needed to design options for further reform, including privatization and joint ventures, and to design an approach to developing effective rural financial service institutions. 16. Human resource development would emphasize: (i) allocating more resources for drinking water, basic health, and primary and secondary education as the fiscal situation improves; (ii) continuing to promote private sector role in social service delivery; (iii) assigning a greater role for local level agencies in delivering services, as they build capacity; and (iv) implementing targeted programs, such as education of girls from low income households. Reforming the education budget and eliminating its fragmentation among so many agencies would reduce over-staffing and rationalize resource use. 17. "Pro-poor" growth policies would give top priority to rural and agricultural development. Agricultural productivity and growth would require various policies, including: (i) maintaining liberalized agricultural markets, (ii) providing access to improved technology for small farmers through research and extension, (iii) expanding access to farm land by making unused state farm lands available to small farmers, (iv) building rural roads to provide access to inputs and help evacuate farm surpluses to markets, and (v) extending financial services. Selected labor-based rural infrastructure projects would help to relax the infrastructure constraints while providing off-season job opportunities for lower income households. This could build on experiments under way to train small contractors to execute small scale labor-intensive rural projects. 18. Mining, tourism, manufacturing and infrastructure will build on progress made in deregulating investments and privatizing parastatals. Continued implementation of parastatal divestiture is a priority. A transparent process and compensation to affected employees will be needed to accelerate divestiture. Other steps include (i) selective joint ventures in extractive industries based on guarantees against expropriation and repatriation of profits, (ii) support services for export development, including market  vi Executive Summary information and feasible incentive duty drawback schemes, and (iii) emphasis on commercialization of utilities and infrastructure activities. Recent Reform Efforts 19. The new Government has clearly expressed its intention to improve fiscal management and to continue the structural reform program. The Government has made a start by adopting a mini-budget for the second half of the fiscal year ending in June 1996, supported by an informal program. The main fiscal and financial benchmarks of the program include improvement in revenues, on an annual basis, by 1.7 percent of GDP; reduction in expenditures and in the recurrent deficit to 2.7 percent of GDP from a projected 4.4 percent of GDP without the program; and deceleration of money growth from 30 to 25 percent and of inflation from about 27 to 22 percent. Measures to reduce expenditures would focus on limiting parastatal subsidies and civil service hiring. Tighter financial policies are expected to reduce commercial bank liquidity by adjusting the reserve requirements. The program has structural benchmarks focused on key areas, including tax administration, civil service reform, banking and expenditure control. The rest of the summary will focus on the key reforms. D. MOBILIZING DOMESTIC RESOURCES Mobilizing Revenue 20. Recent Actions. Revenues are not responsive to income growth, and widespread tax loopholes undercut collections. The revenue to GDP ratio of 14.5 percent in FY95 is lower now than it was three years ago; the effective revenue to GDP ratio is estimated at 8-10 percent, based on appropriately adjusted national income accounts. Among the major steps taken recently to address these issues: (i) gross instances of tax evasion have been investigated, and legal action is being taken against offenders; and (ii) a Revenue Authority Act was passed in FY95 to establish an autonomous institution for tax administration. Customs duties have been simplified significantly since the mid-1980s from over 20 rates and a maximum rate of about 200 percent to just four rates and a maximum rate of 50 percent in FY95. The FY96 budget committed the Government to eliminate duty exemptions for capital goods imports and to apply a duty of 5 percent. A start has been made to computerize customs, and pre-shipment agencies were hired to assist with assessment, collection and documentation of customs duties. Over 100 bonded warehouses, which were viewed as channels for tax evasion, were closed down. However, to date, these measures succeeded in raising nominal revenue, but the revenue to GDP ratio stagnated. The lessons of past tax reform efforts suggest: (i) a need to improve services to enlist tax compliance; (ii) the importance of an autonomous revenue authority to enforce compliance; (iii) that tax reduction will not enhance revenues unless it is accompanied by effective administration to broaden the tax base; and (iv) that lower tax rates have to be accompanied by the elimination of exemptions. Implementing a reform program along these lines is critical.  Executive Summary vii 21. Task Ahead. Currently, there are competing pressures on the Government to use tax instruments to: (i) provide fiscal incentives for investments; (ii) support stabilization policies; and (iii) generate revenues for rural-biased development projects. For the immediate term, the authorities are emphasizing using fiscal instruments for collecting revenues and restoring fiscal stability. Realizing that the Revenue Authority requires more legal backing to be effective, the Government is working to amend the Revenue Act to strengthen the Revenue Authority in enforcing tax compliance, including limits on tax exemptions. Other steps are to: (i) adopt new personnel policies, including pay reform and qualification standards, to raise staff performance; (ii) update tax payer records and register new tax payers; (iii) introduce tax payer identification; (iv) review interest rate penalty for default; (v) establish a tax court; and (vi) enforce revenue laws in Zanzibar as consistent with a single customs union. The Customs Department also needs greater capacity to document and manage expanded transactions in the liberalized environment. It requires skilled staff and equipment to interdict illegal trade, incentive pay and personnel policies to motivate staff and limit corruption, and updated operation procedures dating back to the East African common market era. In the area of sales tax, the main concerns are to simplify and unify the rates for the same products in preparation for a VAT; convert specific and ad valorem rates on excises to a single ad valorem rate to simplify tax administration; and rationalize the differential tax rates between goods and services. Income tax reforms are also being formulated to expand the personal income tax base to include all allowances and fringe benefits, and phase out income-tax-based fiscal incentives, including tax holidays. Finally, more of the small scale and informal sectors could be brought into the tax net through a simple presumptive tax. 22. The work under way to expand the tax base through a consumption-based tax or a VAT requires thorough preparation. Implementing a VAT requires strengthening the system of tax administration, eliminating ambiguities in the existing taxes and unifying tax rates for the same product, establishing requirements for business record keeping and improving the accounting system. It is also necessary to train tax collectors, tax payers and accountants, and to disseminate information on the process of the VAT. Other potential sources of revenues include deregulating tariffs and instituting commercial measures to generate surpluses from (i) parastatals, such as in petroleum, electricity, ports and shipping; and (ii) natural resources, such as forestry, hunting and tourism. Financial Reforms and Savings 23. Recent Actions. Financial institutions are critical for mobilizing domestic resources to support private sector development. Actions have been taken to abolish restrictions on private banks and on interest rates. The Bank of Tanzania has been restructured to conduct monetary policy more effectively in a market system. But the banking system remains very weak, reflecting weak management and a legacy of directed credit and financing of non-performing enterprises. While reforms are being pursued to protect depositors of NBC and to stem the losses, the problems are massive. NBC's Board of Directors was reconstituted in early 1995, and a new management team has been  viii Executive Summary given a mandate to make the bank self-sustaining. A major reform program is under way, including closure of over 30 branches and 14 regional offices, and reduction of staff. 24. Task Ahead. The commercialization of NBC is a major step to make it break even and to compete eventually as a private bank. In addition, more work is being planned on supporting the growth of the whole financial system. Beyond the reform of NBC, work is needed to design and extend financial services to the rural areas to expand the base for financial savings, and to develop lending instruments for small scale businesses. The recent failure of a private bank has also raised concerns and points to a need to reexamine entry and exit requirements and prudential regulations, and to develop personnel and technical capacity in bank supervision to ensure the integrity of the small banks. Currently, because of the growing fiscal deficit, treasury bills have become the most dominant instrument for public borrowing. Containing the fiscal imbalance and the deficit financing are most relevant for the growth of other financial instruments. Eventually a successful banking system requires a broad pool of entrepreneurs who will generate returns to pay back on loans. Support for expanding the entrepreneurial base by training and business services is essential. Equally important is to maintain a hospitable environment to encourage reinvestments and savings in local currency. E. PUBLIC ADMINISTRATION AND EXPENDITURE REFORMS Public Administration 25. Recent Actions. Following the progress made in liberalization, reform of public administration has become critical to redefine the role of the Government, improve fiscal performance and enhance the effectiveness of the civil service in managing the activities that will remain in the public domain in the foreseeable future. The efficiency review of major ministries has been done; capacity requirements and redundancies have been identified. The initial phase of the reforms involved retrenching about 50,000 redundant staff, including unknown workers on the payroll. While the retrenchment reduced the total government staff size to about 310,000 and limited the expansion of the public sector, there has been no noticeable fiscal impact, because of increases in the minimum pay and rehiring. The pay increases in FY96 raised the wage bill and resulted in minimal allocations for non-wage costs in major departments. Staff reduction is costly, and the savings generated can be sustained only if those retrenched are not rehired or replaced. 26. Task Ahead. The main issue in the next phase of reforms is to redress the ineffectiveness of the administrative system as an integral part of reforming the system. This means building capacity, providing incentive pay and controlling wage costs and overall expenditures. The effort requires shedding old functions, departments, agencies and personnel, which are no longer relevant, and developing new capacity and appropriate incentives to motivate effective performance. The Government is undertaking a pay reform, which will involve incorporating most of 36 allowances into the basic pay to make the system transparent. Providing incentive pay requires further retrenchment to contain the overall size of the wage bill. Moreover, economic growth  Executive Summary ix and poverty reduction have to provide a context for the restructuring effort in order to ensure that measures necessary for rural and agricultural development and for the delivery of infrastructure and social services will be carried out. To this end, it is desirable to identify priority tasks clearly, and equip the relevant institutions (revenue authority, customs department, expenditure management and control, agricultural extension) to perform those tasks. Last, experience suggests that the reform of public administration is a process, rather than a once-and-for-all activity, and it has to be followed through over time to get results. Also, because of its political nature, it has to be embraced at the highest levels of political authority. Expenditure Priorities 27. Recent Actions. Actions were proposed in the past to impose expenditure discipline with limited success. Recent efforts involved monitoring of sector budget targets; limiting local fund expenditures to revenue collections; applying the financial codes to stop expenditure overruns; and instructing banks to reject checks not backed by adequate funds. Efforts to formulate and implement a core program succeeded in merging projects, but this did not impact much on overall spending. The widespread under-financing of non-wage recurrent activities suggests limited scope for further cuts in that area, without crippling basic public services. 28. Task Ahead. The main challenges are to (i) continue limiting subsidies for commercial parastatals; (ii) enforce accountability for observing sector budget limits at the highest level of political authority; (iii) continue with private delivery of services as well as delivery by NGOs and local communities to the extent feasible; and (iv) emphasize selected development activities, while phasing out others. With aid support and as revenues improve, activities which require more resources, in terms of non-wage recurrent spending, are primary health care, primary and secondary education, and agriculture. 29. In the case of the social sectors, the emphasis is on using public resources efficiently and enlisting the role of the private sector to mobilize more resources. Proposed reforms include shifting spending priorities to target low income groups and women; enhancing the role of the private sector and local level agencies and communities; and providing options for choice among users. There is also scope to make spending more efficient. Examples include emphasizing more day secondary schools rather than expensive boarding schools, and reforming the education budget to eliminate the fragmentation of allocation among too many agencies and to reduce over-staffing. For the health sector, the main options are to (i) shift sector resources in favor of preventive care and major killer diseases, such as maternity and child health, malaria, TB, and water-borne diseases, and (ii) continue to encourage private sector delivery of services, while focusing the Government on the rural sector. For water, simulations show that user fees for private connections could be raised and the current subsidy halved. This would generate an estimated T Sh 7 billion (US$12 million) for expanding access. This would also improve the equity of the water program substantially. The results of the  x Executive Summary simulation show that the highest quintile would then receive 34.9 percent of the water subsidy, and the lowest 20 percent of the population would receive 12.7 percent, compared to the current allocation of 41 percent and 11 percent respectively. 30. Reforms in agriculture would need to emphasize more support for research and extension; water resource management; supportive rural road infrastructure; improved management of parks and game reserves; and development of rural financial markets. Rural roads are particularly critical for overcoming the isolation of farmers from markets by reducing transportation costs and motivating private purchasing agents. On the other hand, a reduced government role is warranted in other areas: (i) imposing limits on new government projects in various activities, including dairy, general livestock, cashew nuts, sugar, area development and cotton, among others; (ii) implementing divestiture of state farms and other sector commercial parastatals; (iii) phasing out government operations in agricultural input markets, especially agricultural chemicals; and (iv) limiting public role in the cooperatives to foster viability under competitive conditions. F. DEBT SUSTAINABILITY 31. While revenue and expenditure reforms must be the centerpiece of efforts to contain the fiscal difficulties, this alone will not be sufficient. Tanzania's extemal debt is very large, and rescheduling on favorable terms is needed. About 50 percent of the debt is bilateral; about 44 percent is multilateral, and the rest is private. Debt service is difficult to manage under existing circumstance, resulting in a heavy burden and accumulation of arrears. The recent analysis of debt sustainability suggests that debt rescheduling will be helpful, provided it is part of an overall fiscal reform package. The analysis of debt sustainability is based on: (a) strong performance in the exports of goods and services, with average real growth of goods exports of 4 percent and real growth in services of 8 percent; (b) average import growth of about 4 percent; (c) rescheduling of eligible external debt on favorable terms, involving 67 percent reduction in the stock of eligible debt as well as reduction of 85 percent in commercial debt through discounted debt buyback; (d) financing of new requirements on concessional terms; and (e) limits on domestic debt with net domestic borrowing under one percent of GDP. 32. The results of the analysis suggest that debt rescheduling on favorable terms, together with a fiscal reform program, will improve liquidity and debt service. The debt service ratio to exports will fall, reaching under 20 percent. However, the debt stock burden will be heavy in the medium term. The debt stock indicator (represented by the ratio of present value of debt to exports) will exceed the threshold level of 200 percent at the end of the decade, and Tanzania will require exceptional consideration to overcome the medium term burden. The fiscal burden will improve, though it will remain problematic. The share of external debt service from revenues will fall from over 40 percent in FY96 to about 20 percent by the end of the decade. Domestic debt service will similarly fall from about 13 percent of revenue to about 10 percent.  Executive Summary xi G. BROAD-BASED ECONOMIC GROWTH Rural and Agricultural Development 33. Recent Performance. Agriculture accounts for about half of the GDP and is a source of employment for 80 percent of the population. Over the past five years, the removal of controls on agricultural production and marketing and rehabilitation of road transport infrastructure, together with favorable weather conditions, helped to achieve an average annual agricultural growth rate of over 5 percent. However, much of the potential in the sector remains to be harnessed fully. Agricultural exports are lower now than they were in the late 1960s and 1970s, and agricultural productivity is lower than the LDC average. Improved access to technology, access to markets, water resource development and application of modem inputs would help raise productivity and incomes. 34. Task Ahead. Reforms must now move beyond liberalization to the next stage of addressing structural constraints. The recent Agricultural Sector Memorandum discussed the sector development issues in detail. It is envisaged that technology will be the main instrument for raising productivity, based on extension support, improved tools and other inputs. The other main inputs and options include the following: * Emphasis on smallholder agriculture, especially cash and export crop production. The types of crops vary by agro-ecological zone and by locality. Smallholder expansion will broaden the base for growth, in addition to the enclaves of large farms which exist for tea, sisal and, to a lesser extent, coffee. * Addressing rural infrastructure and human resource constraints, especially road transport, water, and post-primary training and education. In Tanzania, investments in infrastructure and education have been found to be more productive if implemented jointly rather than separately. The Second Integrated Roads Project has developed a priority list of rural roads. * Adopting labor-intensive methods in constructing suitable rural road projects would foster rural development, while providing off-farm employment outside the peak farming season. This will build on the pilot program being conducted to train small contractors in labor-intensive construction methods. This effort would require institutional support, appropriate ways to finance and monitor the program, and contractual arrangements to enlist participation. The option of intermediate means of transport such as bicycles, and hand and animal drawn carts, needs to be explored in areas with low traffic density and high construction costs. * Expanding access to inputs requires diversifying the sources of supply. Steps have been taken to remove controls on input markets, and to eradicate rent- seeking. However, periodic monitoring at the local level would identify and  xii Executive Summary address the causes of seasonal supply shortages regarding fertilizer, veterinary drugs, farm implements, improved seeds and fuel. While government support for extension services to small farmers is important, the private sector should be encouraged, especially to serve the large commercial farms. Land Policy Reform 35. Recent Policies. Given the role of land as a major productive asset, security of land tenure is critical for agricultural development. A National Land Policy (NLP) was adopted by the Cabinet in March, 1995, and the necessary legislation is now being drafted. Among other things, the NLP: (i) vests all land "in the President as trustee on behalf of all citizens," (ii) recognizes both customary and statutory rights of occupancy as equal in law, (iii) states that it is the Village Council which is to administer village lands, and (iv) provides that women be entitled to acquire land in their own right. 36. Task Ahead. First, a lack of administrative capacity and infornation, including land survey data to implement NLP, is a major concern. A process which recognizes existing customary rights, and is based on local participation and management will most likely help to mitigate the information gaps. Second, there are about 3.5 million small farmers in Tanzania, farming about 4.1 million hectares, and some additional 2.0 million hectares allocated among 730 farms owned and operated primarily by parastatals. In sharp contrast with the smallholder area, only 30 percent of parastatal land is cropped. In regions where land is relatively scarce and the Government owns over 80 percent of area under large farms, redistribution of land to smallholder farmers is likely to have a positive impact in reducing poverty. The Government can promote small farmer activities by designing criteria for returning parastatal farms to small farmers, including women. Commercialization and Private Sector Development 37. Recent Actions. Potential growth sectors, such as manufacturing, tourism, mining, shipping, aviation and other infrastructure, were dominated by inefficient parastatals in the recent past. This resulted in large scale investments, but low growth rates. A start has already been made in deregulating investments and in implementing a program for parastatal reforms. Over 100 parastatals have been sold or liquidated. The main tools which are being used for private sector development include: (a) selling parastatals, liquidating the unviable ones, and forming joint ventures with private management control and state minority participation; (b) deregulating cumbersome procedures regarding private investments; and (c) promoting export development, with emphasis on manufactured exports, gold and gemstones, traditional agricultural exports, forestry and marine exports, and tourism. 38. Manufacturing. The manufacturing sector accounts for about 8 percent of GDP and grew by about 3.5 percent average per annum in recent years. Growth activities include food processing, beverages, tobacco, footwear, electrical equipment, batteries and  Executive Summary xiii chemicals. However, inefficient parastatals persist with loss making units in textiles, steel, saw milling cement and paper. The next stage reforms need to redress the structural constraints which continue to hinder manufacturing and exports. Replacing existing aged machinery and equipment requires emphasis on banking reforms and viable financial institutions to support new investments, raw material purchases and export credit. Export policies being developed need to be followed through with implementation. These policies focus on: (i) reforming institutions to administer incentives and services including business registration and customs processes, (ii) developing infrastructure with emphasis on power and telecommunication; (iii) sponsoring alliances between domestic and foreign investors through market information and supportive laws for joint ventures, and cooperating with neighboring countries to facilitate intra-regional trade; and (iv) fostering private sector development, including accelerated reform of manufacturing parastatals. Limits on tax exemptions and tightening of customs management of import taxes would rationalize industrial incentives and provide a level field to compete with imports. 39. Mining. Mining has potential as a major source of revenue. While official exports amount to about 2 percent of GDP, smuggling is believed to be much higher than reported output. The state monopoly control of mining has been abolished; but the Government needs to follow through with the implementation of proposed reforms to phase out the role of the state as a direct investor and manager of enterprises. Guidelines and legal codes to control environmental problems, including strip mining, need to be provided and enforced. Other important issues are to equip the customs department to enforce reporting on the export of precious minerals; expedite licensing procedures for small labor intensive miners; provide marketing services for small-scale operations and support them with appropriate technology; and formulate an incentive scheme to support skill development in mining technology and related areas. 40. Tourism. Tourism has been dominated by the Tanzania Tourist Corporation and its 13 subsidiary parastatals. The privatization of most commercial tourist enterprises, such as hotels and tour operators, has resulted in a marked increase in capacity (number of beds, chartered aircraft), especially in the northern circuit, and in the quality of services offered. There are no reliable figures on tourist receipts and estimates, but raising earnings through volume and high paying tourists would require greater investment in infrastructure and improvement in services, such as transport, hotels and telecommunication. Wild life, parks, sites and beaches are environmentally sensitive; and they require environmental guidelines and monitoring, including greater transparency in the public management of hunting. 41. Infrastructure. There is substantial potential for growth in infrastructure, especially electricity, shipping, railways, aviation, telecommunications and road transport. Current policies call for greater private sector role through direct investments, privatization, and suitable contracting arrangements with existing parastatals. Power, telecommunications and transport sectors, including air travel, are open for private participation. The Tanzania Electricity Supply Company, for example, is to contract out  xiv Executive Summary the functions of the in-house construction unit. In shipping, there is scope to expand service to neighboring landlocked countries, but the competitiveness of Tanzania's seaports, especially Dar es Salaam, have been of much concern, given the substantial trade and traffic flows that are developing through South Africa and Mozambique. Measures to improve shipping services include the rehabilitation of Tanga port to improve the yard for stacking containers and repairs to the quay; providing more operating vessels, reducing high operational costs, and upgrading navigational equipment. Proposals to commercialize the Tanzania-Zambia railway and shipping need to be followed through with implementation. Commercialization of the major international airports and the establishment of an independent regulatory body for air systems are in progress. Internal air transport pricing needs to be fully liberalized to support the commercialization effort. H. HUMAN RESOURCE DEVELOPMENT 42. Recent Performance. As noted earlier, investment in human resource assets began to erode when the economy declined and revenues collapsed in the late 1970s and the early 1980s. There were restrictions on private social services until recently, so the private sector could not compensate for the public sector decline. Expenditure on the social services began to recover in the 1990s, but this is yet to be reflected in the social indicators, which normally lag current consumption. The literacy rate is high at about 80 percent, though functional literacy is lower. Most social indicators have stagnated. 43. Task Ahead. The Government has proposed a Social Sector Strategy to arrest the deterioration in services. The main elements of the strategy, which is being pilot-tested include the following. (i) Increase budget allocations for the social sectors as revenues improve; emphasize basic education and basic health care; target expenditures on the lowest income groups, including women and girls; and provide greater budget allocation to support non-wage recurrent cost inputs. (ii) Decentralize authority for social service delivery and give more responsibility for planning, budgeting and managing programs at the local level. This will require improvement in management capacity at the local level. (iii) Promote private participation in the provision of social services. The recent liberalization of the primary school sector, the earlier liberalization of secondary school ownership, and the re-legalization of private medical practice are all examples of government commitment to this strategy. 44. An important issue is the imbalance in enrollments between primary education and secondary education. Tanzania needs post-primary education to develop the skills  Executive Summary xv the economy needs to grow. The public secondary school gross enrollment of about 5-8 percent is the lowest in the world, though the overall secondary enrollment has increased with recent expansion of the private sector. While the growth of private secondary education is good for those able to pay, access for low income groups is critical for achieving generational changes in income and poverty. This situation suggests a need to reconsider the overall approach to providing all levels of education. Secondary education needs to use resources more efficiently than is currently the case. It also needs more resources to expand access and quality. Such improvement is necessary to provide a base for improving primary school instruction and outcomes. 45. Serious consideration needs to be given to implementing proposals to improve efficient resource use in the education sector to: (i) reform the education sector budgetary process to eliminate the fragmentation and over-manning of programs; (ii) shift provision of secondary education from boarding schools to day schools, which are less costly and would be more accessible; (iii) review and adjust inputs into education to emphasize supplies and other non-wage activities; (iv) rationalize training programs spread out among ministries, phase out duplicative activities and consolidate programs to reduce costs and make training relevant for the job market; (v) involve private sector employers in training programs to improve market relevance, and explore the use of incentives to motivate firm level training; and (v) initiate debate on the financing of higher education to build a consensus for reviewing the subsidy policy. I. EMPLOYMENT AND TARGETED PROGRAMS Employment 46. An estimated 0.4 to 0.6 million new job seekers would enter the job market annually, and employment growth will have to exceed 3.5 percent a year in order for the unemployment rate (10.7 percent) to decline. With 80 percent of the working age population engaged in agriculture, most people will continue to earn their livelihood on rural farms in the foreseeable future. Enhancing agricultural growth from subsistence to cash crop production will enable farmers to hire both farm and off-farm labor. It will also stem urban unemployment, which is fed in part by rural to urban migration. While the informal sector has potential to create jobs, most of these will be single operators. The sector needs institutional support to obtain space for operating, basic entrepreneurial training, financial services and lending tools for small operators. Making a single institution responsible for ensuring the coordination of programs would probably be more effective than multiple uncoordinated activities spread across the various government departments. The vast majority of the work force is unskilled and would require training to gain jobs likely to come up in manufacturing, mining and tourism.  xvi Executive Summary Rural Works Programs 47. Targeted labor-intensive infrastructure construction and improvement of rural roads, irrigation projects, and water supply projects present potential to enhance broad- based growth, while providing additional employment opportunities and income for unskilled rural labor. Pilot projects executed by labor-intensive contractors have shown that such methods are technically and financially feasible, especially in road improvement and maintenance, and in rural road construction. The Integrated Roads Project has started with setting aside 50 km of roads in some selected regions for labor-based contractors. A program is also being implemented to train small contractors in labor-intensive methods. 48. Expanding the program effectively in the future would require institutional management, financial support and appropriate design. Whenever feasible the projects should: (i) ensure the adoption of self-targeted approaches; and (ii) be located in regions where the depth of poverty is highest, provided that the strategy is consistent with public expenditure priorities, the labor-based methods are technically and economically feasible, and the projects are integrated with the local development plans. To self-target the project towards the very poor, the wage should be set at a level below the reservation wage of the non-poor. Also, these programs should not coincide with peak agricultural seasons, when labor shortages tend to occur. Women and Children 49. Although the gender gap in Tanzania is narrow in regard to literacy and primary education completion rates, women are still at a disadvantage in other areas. Girls face obstacles in completing their secondary education and are far less likely than boys to continue on to upper secondary school. Based on perceived high social returns to girls' education, the Government will pre-test a pilot -project which will assign a block of scholarships to girls from low income households. 50. With the rapid spread of AIDS, there has been a dramatic increase in orphaned children-estimated to be 600,000-stretching the traditional safety net to its limit. Presently there are 32 children's homes in the country with a total capacity of care of 3,000 children. However it is expensive to provide for children in this way, and it may have negative effects on the psychological development of the children. Other types of care arrangements such as foster care and adoption, though contrary to traditional practices, need to be explored. It also seems more effective to assist financially the families who adopt the affected children, rather than set up children's homes which reach only a few. J. EXTERNAL ASSISTANCE 51. Access to external aid has been important to finance imported inputs and public services and to aid economic growth. Weak capacity in public administration and the shortage of local counterpart funds for project implementation have limited project aid  Executive Summary xvii absorption. The financing problem intensified in both FY95 and FY96, and reduced the pace of implementation of development projects. External aid is not the only constraint on public investment. Project aid utilization is conditioned by complementary domestic inputs, especially implementation and local funding capacity to absorb the aid pipeline. In view of the weak capacity to support projects, non-project aid has become more and more important. But, if public administration, revenue performance and overall absorptive capacity is improved, project aid utilization and access would most likely rise, as non-project aid is reduced. 52. In the meantime, local currency resources generated from non-project aid are very critical for bridging the recurrent cost financing gap to ensure the continued delivery of critical social services. Also, donor support will be critical for implementing reform in public resource management, especially for phasing out low priority projects. This is inevitable for re-channeling support to the top priority activities. In addition, new donor support will have to emphasize the priority activities, especially in the social sectors, and rural infrastructure. Finally, even with the best effort, creditors provided relief for debt service in the past, and further action in that regard would relieve the debt service burden.   Part I Reforms and Economic Recovery  I  1. ECONOMIC REFORMS AND RECOVERY A. OVERVIEW OF ECONOMIC REFORMS AND PERFORMANCE . I Tanzania is a low income country with an estimated official per capita income of about US$100, though household surveys show a value closer to US$200.1 From the mnid-1960s to the mid-1980s, following the Arusha Declaration, the country pursued policies for economic growth and development based on restrictions and controls, and direct state investments in commercial concerns in all sectors of the economy. The last Country Economic Memorandum discussed in detail the evolution of these restrictive policies, their impact on economic development, and the eventual initiation of reforns.2 Initially, the country made progress in economic and social development. GDP growth averaged 4.7 percent up to the mid-1970s. Primary school enrollments tripled. Access to health services improved, with gains in life expectancy and infant mortality. But the impact of the inefficient parastatals and the restrictive policy regime, together with a series of external shocks, including severe drought, the war with Uganda and oil price hikes, led to serious macroeconomic imbalances. Severe shortages occurred. The economy stagnated; per capita income declined, and living standards plunged. 1.2 For nearly a decade now starting from about the mid-1980s, the country has embarked on a reform program to liberalize virtually all aspects of the economy to phase out the pervasive controls and to reverse the economic decline. The Government's initial reforms were fornulated under the National Economic Survival Program (1981-82) and the Structural Adjustment Program (1982-85). These efforts were strengthened by an Economic Recovery Program (ERP) in 1986 and the Economic and Social Action Program in 1989; these latter programs received significant donor support. This chapter will assess the impact of the main reform measures on economic performance. Also, it will identify options for further reforms to sustain the progress that has been made. The main objectives of the reforms were to: (a) increase domestic production of food and exports; For various reasons, the official estimates of the national income accounts grossly distort the level of economic activity. Work is currently under way to update the national income accounts as discussed in the Appendix to this Chapter in Volume 2 of this report. 2 ' Tan-ania: Towards Sustainable Development in the 1990s". World Bank Report No. 9352-TA, 1991.  2 Chapter I (b) restore efficiency in the mobilization and the utilization of domestic resources; (c) reduce the rate of domestic inflation from about 30 percent by at least half; (d) restore internal and external balances by pursuing prudent fiscal, monetary and exchange rate policies; (e) rehabilitate the physical infrastructure, in particular transport and comrmunications to support directly productive activities; (f) revamp the industrial sector using appropriate trade and industrial polices; and (g) rehabilitate and maintain social services. 1.3 Tanzania has made significant progress in liberalizing its economy. The incentive system has been restructured, and the restrictive and pervasive state control regime now belongs to the past. Controls on agricultural markets have been effectively abolished, and industrial investments have been deregulated. Restrictions on trade, foreign exchange rates, and private investments, interest rates and private banking have been lifted. Parastatal reforms are being implemented. However, macroeconomic management is grossly inadequate, and not much progress has been made in other critical areas, such as the banking system, the public expenditure program and in public administration. Macroeconomic instability and inflationary pressure remain serious and recurring concerns. The financial system is extremely weak. The public expenditure program is over-extended, and social services have yet to recover. Progress has been slower where reforms require the creation of management infrastructure, administrative review, monitoring and restructuring of institutions. 1.4 Economic performance over the reform period has been mixed (Table 1.1). Economic growth recovered and exceeded the population growth rate, reversing the declining trend of per capita income. Household expenditures improved significantly and the incidence of poverty declined (see Chapter 3). Exports recovered also, growing at over 4.5 percent average in real terms. External assistance and foreign earnings from export recovery allowed imports to recover also. But the fiscal deficit and inflationary pressure have persisted. The current account balance has remained weak, though it showed some gains lately with the coffee boom. While investments recovered, this has been financed mainly with external assistance, as domestic savings declined. The rest of the chapter will examine policy and economic performance in detail.  Chapter 1 3 Table 1.1: Selected Macro-indicators, FY76-94/4 FY94 FY76-80 FY81-85 FY86-90 FY91-93 Estimate GDP growth at factor cost 2.0 1.8 3.9 4.5 3.6 GDP per capita, at factor cost -1.5 -2.2 0.6 1.4 0.7 Exports (growth) -5.7 -3.3 4.4 4.9 12.8 Imports (growth) 2.1 -6.9 2.3 -0.3 22.6 Current AC deficit /GDP \1 -- 8.8 15.7 22.0 22.6 Fiscal Deficit/GDP -- 10.6 7.6 5.4 10.2 Total Expenditure/GDP -- 29.1 23.2 20.2 23.0 Revenue/GDP -- 18.3 15.8 15.2 15.0 Tenns of trade (rate of change) -5.7 -7.9 -5.0 -4.5 -1.9 Domestic inflation rate 16.5 29.7 28.6 22.0 36.0 GDI/GDP /2 24.7 21.6 27.7 32.8 31.3 DS/GDP/3 -- 13.2 7.6 4.7 3.3 1 At current price. 2 GDI standsfor Gross Domestic Investment. 3 DS standsfor Domestic Savings. 4 FY isfiscal year, for example FY93 = 1992/93. Source: Tanzanian authorities and staffestimates. B. REFORMING THE INCENTIVE SYSTEM Price and Market Reforms 1.5 Initial reforms were focused on reviving agriculture to lead the economic recovery. Producer prices in agriculture were freed between FY91 and FY93; most of the controls, including restrictions on the marketing and transportation of food grains, were abolished in 1989. The marketing of non-traditional export crops was deconfined and extended to traditional crops, such as coffee, tobacco, cashew and cotton in 1990. Private traders started buying cashew nuts during the 1991/1992 marketing season, and private marketing of major traditional export crops was officially allowed in 1993. The list of items subject to price controls shrank from 400 in the early 1980s to petroleum products and electricity. Restrictive exchange rate policies which were biased against agriculture and agricultural exports were relaxed. Input subsidies, for example, for fertilizer, which caused severe shortages and encouraged rent-seeking, were phased out. However, discriminatory regulations and practices which give advantages to the public sector and to the state-sponsored cooperatives remain to be rationalized. In other areas, the main trunk and regional roads were rehabilitated to provide better access to markets. However, access to rural transport infrastructure remains limited.  4 Chapter 1 Exchange Rate Reform 1.6 The system of foreign exchange controls Figure 1.1: Real Effective Exchange Rate Index with about 800 percent premium in the parallel exchange rate market in the mid-1980s was ad- 250 dressed with effective 200 actions. There were a series of devaluations to 150 make the exchange rate 100 competitive, and foreign 50 exchange bureaus were introduced. The nominal 0o I I I I ! I i E 0 CN A rate to the US dollar rose o o ON co Oc N CD from T Sh 33 in 1986 to T Sh 298 in 1992. By mid-1992, the huge exchange rate distortions of the mid- 1980s had been almost eliminated. Subsequently, official exchange rate controls were abolished. The two-tier exchange rate system was unified in August, 1993, following the introduction of an auction system earlier with participation of the foreign exchange bureaus. Under this arrangement, the official exchange rate was set on the basis of the exchange rate prevailing in the foreign exchange auction. The exchange system was rationalized further in June, 1994, when the auction was replaced by an interbank market for foreign exchange. The official rate is now based on the interbank rate. The exchange rate reform and trade liberalization have been accompanied by export recovery. 1.7 Figure 1.1 shows the pattern of the real exchange rate, indicating a trend of substantial depreciation since the mid-1980s. The rate has been more or less stable during 1993; but it began to indicate appreciation in real terms after 1993 for various reasons, including a slippage in macro-fiscal management, excessive deficit financing and upsurge of inflation. Trade and Industrial Policy Reforms 1.8 Restrictive trade and industrial policies were also liberalized. Permission to import from "own funds", which accounts for a third of imports, was given earlier in 1984. Subsequently, quantitative import restrictions were replaced with a system of open general license. The tariff system was rationalized. Prior to June 1988, there were 20 tariff rates, and the maximum tariff was 200 percent. But these have been reduced to four tariff rates by 1990, and the maximum tariff rate fell to 50 percent by July 1994. The tariff reforms helped to narrow and rationalize the tariff structure. Until 1988, the sales tax structure had over 26 rates and a maximum sales tax rate of 300 percent. The number  Chapter 1 5 of sales tax rates was reduced to seven, and then to only two ad valorem rates of 25 and 30 percent in 1995. Further measures to eliminate discretionary tariff and sales tax exemptions were introduced in the FY96 budget, but implementation remains to be seen. 1.9 The exchange rate and trade policy reforms noted above were integral to the overall effort to foster efficient industrial investments. The main initial objectives were to: (i) emphasize rehabilitation over new investments; (ii) direct more resources towards potentially viable export-oriented activities; (iii) promote greater private investments; and (iv) achieve a better balance between industry and other sectors, such as transport and agriculture. Also, greater importance was to be attached to agriculture rather than industry as the engine of growth, and the public sector was to be downsized to give the private sector a greater role in industrial expansion. Accordingly, a program for public enterprise reforn, including divestiture and liquidation, was launched. Export Incentives 1.10 Impediments to exports were removed, including (i) abolishing surrender requirements on foreign exchange receipts from non-traditional exports, (ii) eliminating licensing and registration requirements for exports, (iii) reducing the number of export items subject to permit and (iv) allowing private participation in traditional exports. A duty drawback scheme was introduced in 1986, but is yet to function well despite efforts to streamline it. An export retention scheme was introduced to allow exporters to retain varying degrees of export receipts; this has been improved progressively. Export retention for traditional exports and non-traditional exports is now 100 percent of receipts. Also, restrictions on the current account regarding international payments and transfers were eliminated. Export volume rose by a total of over 50 percent during 1986- 94, though the impact of export receipts on the external accounts was offset by a deterioration in the terms of trade. The revival of traditional and non-traditional exports coincided with the depreciation of the exchange rate as well with the measures which were introduced to deregulate agricultural exports. C. STRUCTURAL REFORMS Financial Sector Reforms 1.11 Prior to 1993, the banking system consisted entirely of insolvent and inefficient government owned banks. The largest, the National Bank of Commerce (NBC), accounted for roughly 90 percent of commercial bank deposits and the Cooperative and Rural Development Bank accounted for approximately 5 percent. The remainder of the commercial banking sector consisted of the Tanzania Housing Bank3 and the People's Bank of Zanzibar. In 1991, the Government issued a policy statement on financial sector Tanzania Housing Bank ceased operations in August, 1995.  6 Chapter I reform which acknowledged the perverse impact on the sector caused by a lack of competition and government interference. In an important reversal of policy, the Banking and Financial Institutions Act was enacted to allow for private banking. Also, interest rates were liberalized; banks were allowed to set lending rates below an announced maximum and to set deposit rates freely, subject to the 12-month deposit rate being positive in real terms. NBC was recapitalized as part of a restructuring program which entailed branch rationalization and the transfer of some of NBC's non-performing assets to a specialized and autonomous loan recovery agency. 1.12 Efforts made to strengthen the banking sector were only partially successful. Two private foreign-owned banks, Meridian BIAO4 and Standard Chartered, began operations in Tanzania in 1993, and another five banks have been licensed. However, the new banks have targeted relatively few niche customers and have been reluctant to expand beyond Dar es Salaam. Hence, competition in the sector is still very limited. NBC's restructuring was not successful. The recapitalization was in the form of 11 percent government bonds, but the interest payments were never provided in full. NBC continued to suffer from high overhead costs, weak internal controls and the absence of a loan collection program. The estimated provision for bad loans is T Sh 100 billion of which the parastatals account for about 50 percent. The situation culminated in an FY94 loss of over T Sh 100 billion (US$186 million). This loss was not incurred entirely in one year but is a reflection of the losses of previous years and the inaccuracy of NBC's audited financial statements for those years. Finally, the Cooperative and Rural Development Bank remains financially weak even after its privatization. 1.13 Currently, more reforms are being pursued to protect depositors of NBC, to stem the losses experienced in the past, and to foster a more effective system of financial intermediation. NBC's Board of Directors was reconstituted in early 1995, and a new management team has been installed and given a mandate to make the bank self- sustaining. A major reform program is under way, including closure of 34 branches and 14 regional offices, and retrenchment of close to 50 percent of the staff. The business plan being implemented to effect this change will be strictly supervised by the Bank of Tanzania. The commercialization of NBC is expected to result in its eventual privatization. 1.14 In addition to restructuring and improving NBC's performance, one cannot overstate the importance of giving top priority to redressing the continuing weaknesses of the overall financial system to develop institutional support for mobilizing domestic resources and channeling them for development. For example, the recent failure of a private bank and the problems of the public banks signal the urgency of reviewing prudential regulations and strengthening the supervision of banks to ensure observance of such regulations. Bank supervision has been weak, and technical support for efforts being made by the Bank of Tanzania (BOT) to improve the situation will create a healthy In 1995, Meridian BIAO failed and was purchased by Standard Bank of South Africa.  Chapter 1 7 Table 1.2: Distribution of Commercial Banks' Domestic Lending (in percentage) 1980-85 1986-90 1991-94/' Average Share of Total Production 13.0 24.0 31.8 Agricultural production 4.6 7.2 8.8 Mining and manufacturing 8.1 16.4 21.3 Tourism 0.2 0.4 1.7 Trade 77.7 68.3 55.0 Marketing of agricultural produce 60.9 42.5 29.4 Export of agricultural produce 3.9 0.9 2.9 Trade in capital goods 0.4 0.4 0.7 Trade in non-capital goods 12.6 24.6 22.0 Other 9.4 7.7 13.2 Building and consLuucting 1.9 1.4 2.2 Transport sector 2.0 2.1 3.1 Other!2 5.4 4.3 7.9 Total 100.0 100.0 100.0 Memo Item: Average Annual Change Total 19.1 54.1 19.6 I As of June 30, 1994. 2 Includes public administration andfinancial institutions. Source. The Bank of Tanzania, Economic Bulletin, various issues. environment for private banking to grow. A second issue is to maintain an open banking system, but to review entry requirements to ensure the integrity and capacity of banking institutions. This is necessary to enlist trust and confidence in the system, especially in the case of the very small banks. A third issue relates to financial services for the rural sector. As shown in Table 1.2, the share of loans for the productive sectors more than doubled, with most of the increase going to manufacturing, mining and agriculture. However, the average annual growth of loans declined from 54 percent in 1986-90 to 19.6 percent during 1991-94, reflecting the necessary limits put on lending to non- performing borrowers in order to protect depositors. While the official data indicate that lending for agriculture increased, such increase would most likely have gone to large, rather than small farmers, since the banks generally have a minimum loan amount of T Sh 1 million and other requirements, which small farmers are not able to provide. In  8 Chapter I Table 1.3: Money, Credit, and Inflation (annual percentage changes) 1987 1988 1989 1990 1991 1992 1993 1994 1995 Broad money 20.2 34.9 37.3 45.4 26.9 40.5 43.9 45.9 38.2 Domestic assets, net 25.4 46.8 47.7 28.7 6.4 22.5 63.3 18.1 32.1 Domestic credit 27.5 52.3 34.6 21.5 14.7 0.3 78.9 26.4 15.1 Public sector -- -- -- 21.3 6.8 -8.5 79.7 3.9 26.0 Private sector -- -- -- 22.7 53.7 30.7 76.9 81.4 -1.2 National inflation rate 29.9 29.7 29.0 22.1 22.3 21.1 23.2 36.0 28.0 Source: Bank of Tanzania database. addition, the evidence from the Participatory Poverty Assessment (Chapter 3) indicates lack of access to financial services in the rural areas. This suggests a need for further work to explore options for extending access to rural financial services, and to develop appropriate lending instruments for small operators, including farmers. Monetary Policy and Inflation 1.15 As shown in Table 1.3, the inflation rate remained high, at about 30 percent and above target; it picked up in 1994. This reflects complex factors, especially (i) the persistent deficit financing and (ii) high money growth, due to lax loan policies of NBC during 1986-90 and the effects of the coffee boom on foreign exchange receipts in the 1990s. Fiscal management was unsteady, as central government borrowing from the banking system rose sharply in FY93 after a decline in the preceding year. This was accompanied by expanded credit to other public entities, especially the cooperatives. The inflation pattern was also susceptible to the abolition of price controls, and the drought condition in FY94 which affected food production. Food accounts for about 64 percent of the weight in the consumer price index. The government borrowing and inflation were accompanied by increases in the nominal interest rates, and the weighted treasury bill rate reached 66-71 percent in December 1994 rising from under 30 percent in the preceding period. 1.16 The weak state of the banks undercut the effectiveness of policies aimed at curbing financial instability and inflation. For example, the infusion of new capital into the banks is a major source of government borrowing [para 1.12]. Also, the reserve ratio was raised from 3 percent in July 1993 to 15 percent in January 1995 to tighten money growth, with no impact. The BOT was restructured in December 1993, and treasury bill auctions were introduced to improve monetary management in a liberalized financial environment. However, as revenues slipped and the fiscal deficit rose, the treasury bill instrument which was originally designed to mop up excess liquidity became instead more of an instrument for government borrowing. While the recent passage of the BOT Act (1995) to increase its autonomy and strengthen its role in monetary management is important, tighter fiscal control and reform of the weak banks remain key to restoring  Chapter 1 9 macroeconomic stability. Fiscal prudence is critical to reduce government borrowing and the high treasury bill rate, and to enable other financial instruments to develop. Parastatal Reforms 1.17 In the past, there were huge investments in some 300 odd parastatals, which dominated all sectors, including mining, tourism, transport, commerce, finance and insurance, hotels and other services. Most of the parastatals were inefficient.5 Output losses from technical inefficiency in the manufacturing parastatals were estimated at 6 about 50 percent average during 1968-88, equivalent to losses in GDP growth in the range of 1.0 to 1.5 percentage points a year. Severe balance of payments crisis and shortages of imported raw materials and spare parts led to low capacity utilization and negative returns.7 Estimates of total financial losses of the parastatals vary; but the lowest indicate losses of about 3-5 percentage of GDP per annum in 1988-92, excluding various costs from monopoly pricing, preferential access to credit and tax exemptions. 1.18 Until recently, the parastatals have operated under a Table 1.4: Direct and Indirect Subsidies for soft budget constraint, and have Parastatals received various explicit and (T Sh billion) implicit subsidies (Table 1.4). FY91 FY92 FY93 FY94 These include: (a) direct ________________________ subventions from the treasury From Budget /a 13.4 11.3 35.6 17.0 subventions)from the Custom Exemptions 15.0 na na na b) tax exemptions, (c) bad loans Counterpart Fund from commercial banks, and (d) Arrears /b na 104.6 89.6 70.1 default on payments arising from Notes: commodity import schemes. The a Including crop authorities and excluding strategic grain reserves. b Data from Income Tax Department. parastatals account for the greater Source: Budget Data and Income Tax Department. part of the default on local currency funds generated from commodity import schemes, and have an accumulated arrears of T Sh 70 billion, about 4 percent of GDP in December 1994.8 The parastatals account for over half on the non- performing assets (over T Sh 100 billion) of NBC. There are only a few exceptional cases of parastatal efficiency relative to the private sector in the textile industry. See A. Mbelle and T. Stemer, EastAfr ican Economic Review, 1991, pp. 31-38. 6 L. Rustayisire, "An Empirical Analysis of Public Enterprises' Performance in Tanzania's Manufacturing Sector and Some Suggestions for Improvement", East African Economic Review, 1991, pp. 13-30. See B.J. Ndulu, "Investment, Output Growth and Capacity Utilization in an African Economy: The Case of Manufacturing Sector in Tanzania", Eastem Africa Economic Review, Vol. 2, No. 2, 1986, pp. 14-29. The largest debtors include the Coffee Marketing Board, Tanzanian Oxygen Ltd., and TFC Fertilizer Company.  10 Chapter 1 1.19 In 1993, the Government initiated a parastatal sector reform program with emphasis on introducing greater competition, liquidating uneconomic units and divesting others. Also, there were measures to enforce a hard budget constraint. The reform program is making progress. Of the 300 odd parastatals in existence initially, 75 sales agreements were signed as of December, 1995; about 104 non-performing assets were identified for liquidation, and 51 of these have wound down. Studies have been conducted on another 111 units to determine the appropriate strategy for reform. 1.20 Because the implementation of parastatal restructuring began late (1993), it is too early to assess the impact of actions taken on the productivity of investments. However, the imposition of hard budget constraint is having some effect. For example, commercial bank credit to the parastatal sector, including the cooperatives and marketing boards, has declined by 23 percent to T Sh 92.8 billion during FY93-FY95. Similarly, direct budgetary subsidies fell to less than T Sh 5 billion in FY95, less than a third of the previous years; and tax exemptions are being restricted. 1.21 While many loss making parastatals persist, their losses appear to be mitigated by the gains of the surplus units. The most recent survey of parastatal performance shows 41 units with total losses of T Sh 24.3 billion. Three enterprises (Tabora Textiles, Southern Paper Mill, and National Urban Water Authority) account for 59 percent of the losses. There were 43 profitable units with surpluses totaling T Sh 13.9 billion. Overall, the estimated net losses would be under 2 percent of GDP. Faster progress in the reform effort requires reaching agreements on compensating affected workers; seeking suitable foreign private investors to overcome the limitation of the local market in buying the enterprises; continuing to enlist public support by providing information on the receipts and on the disposition of earnings. Reform of Public Administration 1.22 The need to reform the system of public administration to make it function in the liberalized system is now widely recognized. Among the main issues: (a) The effectiveness of the civil service has declined. This is attributed to: (i) the thin layer and wrong placing of the highly skilled civil service; (ii) absorption of unskilled cadres and political functionaries into the civil service, which affected quality and professionalism; and (iii) outmoded and highly centralized organizational structure and practices.9 (b) The civil service is now at 310,000, following the retrenchment of about 50,000, including undocumented staff at a cost of US$34.7 million. The retrenchment has helped to contain the growth of government staff size. 9 For a recent analysis of the administrative competence of GOT, see E. Bukuku, "Economic and Social Reforms in Tanzania" in Development Challenges and Strategies for Tanzania: An Agenda for the 21st Century. 1994. Dar es Salaam University Press, pp. 25-38.  Chapter 1 11 However, there is still overstaffing in certain cadres and productivity is low. The structure is overloaded with overlapping functions and duplication of responsibilities among agencies and between local and central government authorities. (c) While the civil servants are not over-paid, it absorbs a large share of the budget because revenues are low, and there is over-staffing in some cadres. Also, the compensation system is complex; there are 36 nontaxable allowances not reflected in the basic pay, and the system is perceived to be arbitrary and lacking equity and transparency. Reform is costly, and the recent retrenchment was offset by a 75 percent increase in the minimum wage and some rehiring. As a result, the budget showed no improvement. 1.23 The main elements of the reform are to: (i) reduce the role of the Government by redefining and focusing its functions, downsizing, eliminating redundant staff and reforming personnel policies; (ii) reform the pay system to make it transparent and incentive-oriented; (iii) restructure key departments and functions to make them effective; (iv) build new capacity; and (v) decentralize government functions to give greater authority and responsibility at the local level. The immediate challenge is to downsize the Government while building capacity through pay reform, skill development, and reorganizing the institutional and administrative structure. Providing incentive pay requires further retrenchment to contain the overall size of the wage bill. 1.24 The pay increases required'0 to correct distortions in the pay structure and provide enhancement for the different levels would be in the range of 67 to 141.7 percent."I The extra revenues required to meet these increases would be equivalent to 3.3 percentage of GDP. Thus, a scaled down pay reform is being implemented to consolidate most non- incidental allowances into the basic pay. The minimum pay was raised by 75 percent to T Sh 17,500 per month in July 1995. In March 1996, the Government initiated efforts to reform the pay system and consolidate most of the allowances into the basic pay in order to control the total wage bill. A new pay structure will be introduced with separate salary scales for Government, operational and protective service workers and rationalize the overall grade at an initial cost of T Sh 168 billion. Most non-incidental allowances (for skill, duty, drivers, workshop, ration, electricity, transport, telephone, etc.) will be incorporated into the basic pay; no new incidental and non-incidental allowances will be introduced; lunch and sitting allowances are to be abolished; and rent and housing allowances are to be reviewed. As part of the pay reform, a prograrn for early retirement and retrenchment will be introduced. An estimated 20,000 will leave the service; the total size of government staff will not exceed 283,000 in FY97, and the target for new 10 This assumes a benchmark of T Sh 10,000 per month for each scenario, and the alternative levels of decompression of the pay structure are: (i) 1:10 for scenario one; (ii) 1:12 for scenario two; and (iii) 1: 16 for scenario three. Based on an estimated initial wage bill of about T Sh 50 billion.  12 Chapter 1 employment in FY98 will depend on progress with the organizational and efficiency reviews and the retirement program. Other proposed actions relate to the review of the funding of public institutions. 1.25 Much more remains to be accomplished in three major areas, namely: (i) in restructuring and focusing the functions of key departments; (ii) building needed capacity, and (iii) downsizing and laying ground work for a more effective system and also reducing the fiscal burden. Recently, an organizational review of major departments was completed. Ten departments/functional areas are to become autonomous self- financing agencies. Other preparations are under way to: (i) adopt a new personnel control system to guide hiring, promotion, training and capacity building, (ii) improve budget planning, payroll and personnel management, and (iii) revise the role of local government, and its financing. Experience in other countries suggests that civil service reform is a continuous process, and it requires sustained commitment and support from the highest level of political authority in order to make progress. D. ECONOMIC RECOVERY AND RECENT ECONOMIC DEVELOPMENTS Table 1.5: GDP and Real Growth at Constant 1976 Prices (in T Sh million and in percent) GDP at 1976 Constant Prices Average Growth Rate in T Sh Million (in Percent) 1979 1986 1990 1994 1980-85 1986-90 1991-94 Agriculture 9066 11557 14055 17083 3.0 5.2 5.1 MiningandQuarrying 200 154 180 262 -1.5 1.5 11.7 Manufacturing 2821 1991 2338 2669 -4.9 2.5 3.5 Tourism /I 2839 3098 3575 4048 2.3 2.8 3.2 Other Non-agriculture 7813 8410 9756 11060 0.4 3.0 3.3 GDP at factor cost 22739 25210 29904 35122 1.1 3.9 4.1 GDP per capita, at factor cost(TSh) 1297 1150 1200 1251 -2.0 0.6 1.1 I Including trade, hotels, and restaurants. Source: Bureau of Statistics. 1.26 The implementation of the reforms led to economic recovery. Table 1.5 shows the sector growth pattern before and during the reform period. The real GDP growth rate averaged about 4 percent between the mid- 1 980s and the 1 990s, which is more than three times the average growth rate obtained during the five-year period preceding the initiation of reforms, and it compares favorably with a median growth rate of 2.6 percent for 128 countries during FY86-94. The economic recovery was led by agriculture, which accounts for about 60 percent of GDP and grew at about 5 percent. Industry, which is about 1 1-12 percent of GDP, grew at an average 5-6 percent. For 1994-95, estimated real GDP Growth is about 3.6 percent, due to widespread drought conditions which reduced hydropower supply as well as agricultural and industrial production.  Chapter 1 13 1.27 As shown in Figure Figure 1.2: GDPGrowth perCapita 1.2, the GDP growth (percent) performance during the reform period exceeded the population growth rate and helped to reverse the decline of per capita 2 incomes from about minus A 2 percentage points in 1976-85 to a positive _2 growth position since the -4 V mid-1980s. The rise in per -6 -6 capita income is consistent 00 with the general - X X, improvement in living conditions. As discussed in detail in Chapter 3, the recovery of per capita income is confirmed by the growth of household expenditures and a decline in the incidence of rural poverty from 65 percent in 1983 to 51 percent in 1991, and about the same in 1995. However, because of the deep decline in per capita income and sharp erosion of living conditions prior to the initiation of reforns, widespread low incomes persist even as the economy recovers, and a very large segment of the population lives at low levels of welfare. Based on the official data, real per capita income in 1994 was about 96 percent that of 1979, and the per capita income in 1986 was 87 percent that of 1979 (Table 1.5). Low levels of welfare persist even considering the informal sector. Recovery of Agriculture 1 .28 A detailed analysis of the impact of reforms on the agricultural sector has been provided in the recent (1994) Tanzania Agricultural Sector Memorandum. The main reforms which led to agricultural recovery include exchange rate adjustment, removal of marketing and pricing controls, freeing of external trade and payments arrangements, and rehabilitation of transport infrastructure to facilitate access to markets. The favorable policy regime and improvement in producer prices promoted sector recovery, especially for high valued cash crops and export crops, such as coffee, tea, cashew and cotton, which show positive supply price response. 2 Other factors contributing to growth 12 See, for example, R. Mshomba, "Price Elasticity of Supply of Tanzania's Major Export Crops" Eastern Africa Economic Review, June 1989, pp. 9-23. Also, see N.E. Luvanga and F.M. Musonda "Export Promotion through Diversification: Prospectsfor Non-Traditional Exports". Mimeo. Dar es Salaam, 1992.  14 Chapter 1 include the availability of imported incentive goods;'3 improved supply of inputs, such as farm implements; expanded acreage;'4 and a cessation of severe drought conditions. 1.29 While precise estimates of real growth are difficult, evidence on sector recovery is extensive, and it confirms official findings which indicate real average growth in the range of about 5 percent during 1986-94, compared to about 2 percent during 1966-85; this is more than twice the average for African countries. Growth performance varied among subsectors.'5 Crop production, about 65 percent of agricultural GDP, maintained its long term growth rate of about 2 to 2.5 percent. 16 The production of maize, the main staple, grew at about 5 percent average during the reform period; and maize imports fell, leading to national sufficiency and cross-border sales since FY87, though this remains susceptible to weather conditions. The production of paddy improved rapidly. The growth rate of pulses was significant at over 5 percent. Finally, export crops (coffee, tea, cotton and cashew) showed significant recovery, reversing the decline in the 1970s and early 1 980s. The efficiency of agricultural investments also improved. For example, the aggregate rate of return on sector investment was higher in FY86-9 1, compared to 1981- 85. 17 1.30 The agricultural recovery played a very significant role in increasing household expenditure and reducing the incidence of rural poverty. The most important source of income in mainland Tanzania is agriculture. It accounts for 96 percent of the income of the bottom 20 percent of the population and 62.4 percent of the top 20 percent. The impact of the agricultural recovery on rural incomes varied. Generally, cash crop producers, areas close to urban markets and those connected to transport infrastructure gained most.18 Variations in the gain reflect differences in access to inputs, proximity to markets and infrastructure, and in resource endowments.'9 Generally, the relatively well- off areas include those with the most intensive agricultural systems, especially the: (i) coffee, maize and legume systems of the Southern Highlands, and (ii) coffee, banana and 13 See D. Bevan, P. Collier and J.W. Gunning, "Peasants and Governments - An Economic Analysis" Clarendon Press, Oxford, 1989, pp. 153-222. 14 See K.J. Havnevik, "Tanzania, The Limits to Developmentfrom Above". Nordiska Afrikainstitutet, Sweden, 1993; pp. 295-314. 15 See Tanzania Agriculture Sector Memorandum, op. cit. 16 Food crops account for 55 percent of agricultural GDP; livestock account for 30 percent; cashcrops account for 8 percent; and fishing, forestry and hunting account for the rest. 17 This is based on simple estimates of the ratios of agricultural GDP to agricultural investments. For example, the inverse of the ICOR averaged 0.6 in 1981-85 and 1.8 in 1986-91, the latest period for which data are available. The corresponding estimates for public sector investments in the sector were -0.5 and 1.0. J. Lugalla, "Poverty and Adjustment in Tanzania: Grappling with Poverty Issues during Adjustment Period", URT, Planning Commission, 1993, pp. 20-31. 9 Tanzania: Agricultural Diversification and Intensification Study, 1992.  Chapter 1 15 dairy systems of the North. On the other hand, the relatively poorer regions include the: (i) agro-pastoralist zone of the semi-arid central plains and (ii) the cashew and cassava areas of the South. However, there are variations even within localities. 1.31 While the progress is encouraging, much of Tanzania's agricultural potential remains to be harnessed. Output levels are below those of the early 1970s for several crops. For example, total export of cashew nuts was over 100, 000 metric tons during 1972-74, compared to less than 60,000 in 1994, even though this is much higher than in the early 1980s. Most of the farming is low intensity, with minimal use of inputs that would enhance productivity. The potential for high-value horticultural activities remains to be tapped; and agricultural productivity is low. Maize yield, for example, is about a third of world average. The main constraints relate to the need to access more productive technology, including tools and extension service support; supply of inputs, including chemical fertilizers, fuel and financial services; and provision of rural infrastructure, including transport and water resources. The need to relax official constraints on grain exports to neighboring countries has now been recognized.  Table 1.6: Selected Agricultural Production Indices 1980/81 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93 1993/94 1994/95p Food Crops Maize 100 145.2 122.0 131.7 132.0 121.1 126.8 121.0 124.0 117.4 139.5 Paddy 100 119.5 146.2 223.8 219.5 210.6 116.1 112.7 183.4 175.7 206.9 Wheat 100 108.2 79.1 83.1 89.8 117.0 92.5 70.7 86.2 65.2 82.9 Pulses 100 159.1 92.6 139.6 141.9 141.5 156.3 114.8 149.5 142.4 139.2 Cash Crops Coffee 100 96.2 75.6 83.2 104.4 96.9 101.8 105.6 87.2 89.6 99.3 Cotton 100 88.7 134.6 129.3 120.8 74.3 86.4 134.9 121.3 133.2 130.5 Tea 100 88.1 80.1 90.3 90.9 109.1 102.8 104.4 121.0 127.3 140.9 Cashewnuts 100 35.2 30.6 45.2 35.7 31.7 61.1 74.4 85.7 86.1 88.9 Tobacco 100 84.5 111.5 87.2 78.4 75.0 79.7 111.1 157.4 162.0 123.6 Sisal 100 39.8 36.6 40.2 40.4 39.2 40.8 43.6 29.5 30.1 30.9 Pyrethrum 100 51.9 44.4 51.9 48.1 59.3 63.0 81.5 100.0 103.7 88.9 P Preliminary estimate. Source: Ministry ofAgriculture, MDB, and Food Security Unit.  Chapter 1 17 Recovery of Industry 1.32 The immediate Figure 1.3: Manufacturing Growth Rate period prior to reforms (percent) was one of deindus- trialization. Manu- facturing output growth declined by 4.9 percent 15 average in FY80-FY85 (see Figure 1.3), 10 reflecting foreign ex- change shortages which 5 constrained access to 0 imported inputs and the co _ N4 weaknesses of the import -5 substitution industriali- * - zation strategy based on inefficient parastatals. -15 1.33 With the reforms, industrial activity has expanded at an annual average rate of about 4 percent between FY87 and FY94, the latest data available. Estimated manufacturing growth fell in FY95 as drought conditions reduced power supply, causing power shortages. Overall manufacturing output expanded as access to foreign exchange improved and trade and exchange rate reforms reduced the prevailing anti-export bias. The recovery reflected (i) improved capacity utilization averaging about 50 percent currently and higher than levels of 4 to 20 percent in the early 1 980s, (ii) growth of agricultural incomes and growth of total consumption at over 4 percent of GDP average, and (iii) much more favorable trade and exchange rate regime, inducing exports. Some industries, such as tobacco, newer textile firms, chemicals, beverages, light industries and plastics expanded. Others, such as older textile firms and garments, continued to limp and operate below capacity. Manufacturing industries remain constrained by power shortages, poor management, a weak financial system and inadequate financial support to acquire spare parts and raw materials, and uncompetitive costs. Mining and Quarrying 1.34 The growth of mining exceeded 20 percent in 1990, and declined to about 6 percent average in the last three years, probably reflecting illegal exports rather than output decline. Officially, mining accounts for less than 2 percent of GDP, though the  18 Chapter 1 actual potential for expansion is enormous.20 While small scale mining is growing, large scale mining has been in the hands of the State Mining Corporation; performance has been limited by inadequate management, poor physical infrastructure and poor technology for exploring, extracting and processing mineral resources. The recent abolition of government monopoly over mineral production and trade are in the right direction. Expanded private sector role, joint ventures, and the adoption of a mineral policy to provide production and environmental guidelines should help the sector to expand. The expansion of this sector would help to generate revenues with which to finance critical social and infrastructure programs essential for improving living conditions. Tourism 1.35 Tourism is growing though much of the potential is under-utilized. The number of tourists visiting Tanzania increased from about 50,000 in 1983 to about 250,000 in 1994. Earnings from tourism are believed to have increased, though no reliable estimates are available. Earnings are a fraction of Kenya, which is said to have less potential tourist resource base than Tanzania. Inefficient parastatals, especially the Tanzania Tourist Corporation and its 13 subsidiary companies, including hotels and tour operators, dominated tourism until recently. While private role is now expanding, other factors, such as service staff, internal air transport, good roads, power, hotels and communication facilities are required to make tourist services reliable. In the short term, there is scope to expand tourist activities in the northern circuit, especially Kilimanjaro, Ngorongoro and Serengeti. However, careful management of sites will be critical to contain the pressure in these areas. Options for expansion include development of facilities to attract high- paying tourists and improvement in infrastructure and services in other potential areas to spread out benefits. Monitoring environmental quality will make it possible to ensure sustainable development of tourism. The Informal Sector 1.36 Official surveys define the infornal sector (excluding agriculture, which is already in the GDP) mainly, but not solely, in terms of documentation of transactions and number of employees. The existence of the informal sector in Tanzania is not due to 20 Tanzania has abundant mineral resources, including reserves of iron ore (85 million tons), coal (324 mt), magnesite (4.5 mt), gypsum (2.6 mt), as well as commercial quantities of nickel, gold, diamond, rubies and tanzanites. There is, however, lack of consistent data on mineral production. Most of the production is said to be smuggled out through illegal flight from the mining areas. Estimates of production vary in the range of US$200-400 million a year. 21 The infornal sector was defined to include commercial (i.e., non-subsistence) urban agriculture, commercial fishing in the rural and urban areas, and private enterprises with five or less employees, excluding professional enterprises (for example, lawyers, doctors, and micro firms with high skills and high technology). See "Tanzania The Informal Sector", Planning Commission. URT, 1991, pp. 1-1 to 1-15.  Chapter 1 19 the presence of controls, since the policy regime has been significantly liberalized; it is mainly due to inadequate statistical coverage and documentation. Also, the informal sector is largely outside the tax net. There is widespread evidence that significant growth is occurring, which is not covered by the official national accounts.22 It has become much more difficult to collect statistical data, with the decline of the parastatals which used to be the official data sources. With the expansion of the private sector, production and sales data collected by existing parastatals understate overall activity. The abolition of most state monopolies, deregulation of investment, removal of restrictions on agricultural marketing, and the growth of imports made possible by trade reforms have fostered the growth of both the rural and urban informal activities. Much of this is not routinely documented, and occasional surveys are not integrated with national accounts. The National Informal Sector Survey estimated the size of the informal sector at 32 percent of official GDP in 1991.23 The share of the urban informal sector in total GDP rose from 10.3 to 14.5 percent between 1985 and 1991. About 51 percent of the informal sector value added is in services, mainly trade, restaurant and hotel activities. Manufacturing, including tailoring and domestic textile and garment industries, constitute the second largest activity, accounting for 16.1 percent of the value added and 24 percent of total establishments in the informal sector. The informal sector has, in the process, emerged as a growing source of jobs, employing over 2 million people and accounting for 24 about 20 percent of total employment. This has enabled households to cope with job losses in the formal sector. 1.37 Given that average monthly incomes in the informal sector are higher than the wages for low skill government employees, strengthening the informal sector is essential for reducing poverty. But no clear policies for supporting the sector exist. This is indicated by recent measures taken to restrict informal traders in some urban centers. A new informal sector policy is under preparation. Support for the informal sector through policies and provision of infrastructure and location would help the sector to expand and also provide a basis for bringing it into the tax net. T.L. Maliyamkono and M.S.D. Bagachwa, "The Second Economy of Tanzania". James Cury, Heinemann, 1990. 23 Other estimates put the share of the informal sector at about 30-39 percent of official nominal GDP in the 1980s and at between 46-49 percent during 1989-90. However the differences in the estimated figures reflect differences in the definition as well as methodology of the estimates. See M.S.D. Bagachwa, "Estimates of Informal Parallel and Black Market Activities in Tanzania". Paper presented at the International Conference on Development Challenges and Strategies for Tanzania: October, 1993. 24 The Labor Force Survey of 1991 shows about I million employed in the informal sector. The National Informal Sector Survey of the same year shows a larger number of over 2 million.  20 Chapter 1 Export Recovery 1.38 Figure 1.4 shows the recovery of exports. Real Figure 1.4: Export Growth, Volume export growth averaged 4-5 (percent) percent during FY86-FY93 following the negative growth during the preceding period. Estimated export growth 25 exceeded 5 percent in FY94- 20 FY95. The official estimates 15 understate actual performance; 10 over US$400 million a year reported as private transfers are believed to be unrecorded exports. The main exports are -15 agricultural products, particu- -20 larly coffee, tea, tobacco, sisal and cotton. Non-traditional exports include manufactured goods, marine and forestry products. T he main manufactured exports are food products, tobacco, and selected garments. The export growth is attributed to various factors, including: (i) import liberalization and access to inputs to enhance production, (ii) improvements in the internal terms of trade for exportable goods, due to reduction in the taxation of exports and deregulation of export marketing, and (iii) the devaluation of the shilling, which raised the profitability of exports. 1.39 Financial sector reforms are essential for export Table 1.7: Non- financing, especially for new operators. Other constraints traditional Exports Share which need to be relaxed relate to infrastructure; for in Total Exports, examnple, power, telecommunication and transport 1980-93 infrastructure; bureaucratic bottlenecks and cumbersome Non procedures. With basic liberalization policies completed, YearaNon the institutional and infrastructure constraints need to be Exports Share addressed more vigorously. Constraints regarding single in Total clearance points, customs hours, and dissemination of 1980 29.6 informnation about export procedures need to be relaxed; 1984 22.2 the proposal to reorganize the Customs Department needs 1985 24.0 to be pursued with urgency to expand its operational 1986 21.0 capacity, improve the quality of the personnel and provide 1988 35.5 more resources to modernize record processing. In 1989 40.8 addition, existing export facilities not administered 1990 33.1 effectively, mainly because of weak institutions, need to 1991 28.0 be reformed. Some of these include the Duty Draw Back 1992 30.6 Scheme and other programs administered by the Board of 1993  Chapter 1 21 External trade to provide market information to exporters. As the fiscal situation improves, the Duty Draw Back Scheme needs to be supported with an operational budget and it should also be monitored to ensure that beneficiaries are actually exporters. E. THE PERFORMANCE OF INVESTMENTS Performance of Investments 1.40 The trend of investments shows a marked increase since the reform started. However, precise estimates are not easy because of weaknesses in the national accounts, including the collection of investment data. Officially, fixed capital formation as percentage of GDP (at factor prices) increased from 22.2 percent during 1979-85 to about 30 percent annual average during 1986-94. This official investment ratio partly reflects the underestimation of national income statistics. Both the output and investment of the informal sector are yet to be recorded in the national accounts. Adjusted for the underestimation, fixed capital fornation would indicate an increase probably from about 12-14 percent of GDP in the pre-reform period to about 18-20 percent of GDP subsequently. 1.41 The share of' the private sector in fixed capital formation increased from an average of 61.8 percent in 1980-85 to 66.2 percent in 1986-92 (Table 1.8). Investments in construction, transport and communication increased notably. The slight official decline in manufacturing investment probably reflects under-estimation since lending to the sector increased (Table 1.2). The decrease in agricultural sector investment shown by the official data is also misleading, as it may reflect inadequate official reporting on farm level activities. The shift from state farms to more private farming makes it more difficult to record farm investments. Most recent indicators of agricultural sector performance show significant investments.25 Area cultivated expanded during FY87- FY91 at an annual average of 5.1 percent for maize, the main staple, and 7.4 percent for paddy. Investment in irrigated area expanded at the rate of 2.3 percent during 1985-93. Also, investment in livestock (about 25 percent of agricultural GDP) rose during 1985-94 at an average annual rate of 0.7 percent for cattle, and about 2.5 to 4 percent for ruminants. 1.42 The available data suggest that efficiency of investment improved to some extent. Returns to investment, indicated by incremental capital-output ratios (ICOR), have not only increased but also become less volatile as shown in Figure 1.5. Similarly, in terms of the GDP growth rate as a ratio of investment, the efficiency has more than doubled between 1981-85 and 1991-94. Yet, the returns to investment in Tanzania remain lower 25 See "Indicators of Agriculzitural Sector Performance and Environmental Status in Eastern Africa", Eastern Africa Department. World Bank, 1995.  22 Chapter I Table 1.8: Composition of Fixed Capital Formation (in percentage) 1980-85 1986-91 By investor type Public 38.2 33.8 Private 61.8 66.2 100.0 100.0 By economic sector Agriculture 8.7 3.3 Service 46.5 41.8 Industry 44.8 54.9 100.0 100.0 By key subsector Manufacturing 24.2 23.6 Construction 10.4 20.3 Transport & communication 24.1 30.5 Source. National Accounts of Tanzania, 1976-1994, Bureau of Statistics, Tanzania. Figure 1.5: Incremental Capital to Output Ratio (at constant prices) 50 40 30 20 ' a 0 0l ''l -t w - -------- -30 -40 _ICOR+I - C ICOR+3  Chapter 1 23 in comparison with other Sub-Saharan African countries. The official GDP growth rate as a ratio of GDI/GDP has increased from 5.1 percent in 1980-85 to 13.0 percent in 1991- 94. For the 1986-90 period, this ratio is estimated to be 35.0 in Ghana, 22.5 for Kenya, and 59.2 in Uganda. The lower investment efficiency of Tanzania is attributed to under- reporting of incomes and inefficient parastatals. Under-reporting of incomes, even by the formal sector, is attributed to high tax rates. For example, import tariffs were up to 200 percent in the 1980s; resident companies were taxed 35 percent, and non-residents 40 percent in FY95. The state dominated investments, which were inefficient, resulting in under-utilization of capacity. Average capacity utilization is about 50 percent and as low as 20 percent in manufacturing. Uncompetitive public investments in manufacturing persisted until parastatal reforms began in 1993, and parastatals accounted for over a third of manufacturing investments during 1990-92. Most of the capital stock is in need of repair. Public goods investments are also under-utilized because of inadequate funding for operation and maintenance costs. Recent public investments went mainly for rehabilitation rather than expanding the capital base; thus, net investment is palpably low. Issues in Investment 1.43 Firm Level Efficiency. There are some indications of improvement in efficiency reflected in the recovery of exports and in the slow but noticeable export diversification arising from the growth of non-traditional exports, including more manufactured goods exports. Most of the wage jobs created since the mid-I 980s came from the private formnal and informal sectors, in contrast to the 1960s and the 1970s when the state was an employer of last resort, and public administration was a steadily growing activity. While agriculture still absorbs most of the work force, employment in the formal private sector more than doubled from 0.15 million in 1984 to 0.3 million in 1991; private manufacturing employment expanded by 7-9 percent a year during the 1980s. 1.44 While this process of transition is encouraging, it is slow. Consequently, the industrial sector as a whole shows only a slow recovery from the de-industrialization of the early 1980s.26 Firms with foreign licenses and foreign assistance contracts, which facilitate the acquisition of technology, are few in Tanzania. Existing business support services such as the Small Industry Development Organization (SIDO) are not known or used by most firms.27 Thus, the efforts to remove firm-level constraints by developing infrastructure, providing market information and encouraging the acquisition of entrepreneurial skills and training need to be increased. 1.45 Investment Incentives and Other Issues. Some specific measures were introduced to promote private investments. For example, the National Investment Promotion Act was passed in March 1990 (and revised in 1994) to establish the 26 The share of the industrial sector in GDP decreased from over 15 percent in 1970s to about 8 percent by the mid- 1980s. Since the reform, the share has increased gradually to the current level of about 13 percent. 27 Center for International Business Research, Helsinki School of Economics. "Development and Growvth of Industrial Enterprises in Tanzania, 1994. "  24 Chapter I Investment Promotion Center (IPC). The original Act provides investment incentives, identifies priority areas for investment,28 delineates activities reserved for parastatals, and it provides safeguards to protect private investors against expropriation and to guide arbitration and transfer of profits. In addition, the Government spelt out details of investment incentives. Some of the major incentives provided initially include exemption from corporate income tax for the first five years, exemption from withholding tax on dividends and exemptions from import duties and sales taxes on machinery equipment, spare parts and material supplies. Investment incentives have come under scrutiny recently mainly because of gross abuses, such as tax evasion, which have contributed to significant revenue losses. Thus, measures have been proposed to revise the role of the IPC to emphasize business promotion rather than the administration of incentives. It is also proposed to phase out the tax holidays and other tax exemptions. There is a perception that redressing the deteriorating macroeconomic situation should receive top priority relative to the provision of fiscal exemptions which erode the revenue base. F. BASIC MACROECONOMIC BALANCES Recent Performance 1.46 Macroeconomic performance improved during the second half of the 1980s, but it has slipped since FY93. Recent performance has been grossly inadequate with instability and widening imbalances. The fiscal deficit before grants was 10.4 percent of GDP in FY93-FY94; it was estimated at 13.5 percent of GDP in FY95, compared to 2-3 percent during FY91-FY92.29 Domestic borrowing escalated to 3.9 percent of GDP in FY95, higher than 2.5 percent in FY94 and 3.7 percent in FY93. Inflation is about twice the target at about 27 percent presently, compared to under 24 percent in FY92-93 and over 30 percent in FY94. In addition, the external balance is basically weak, despite recent improvement with the coffee boom. The current account deficit rose from 7.8 to over 20 percent of GDP during FY86-FY94, while savings declined. The resource imbalances reflect deep underlying causes, especially inadequate domestic resource mobilization, an ineffective public administration, an over-extended public expenditure program and a difficult debt service burden. External resource inflows have been important for the improved economic growth discussed earlier. 1.47 The macroeconomic situation has implications for living conditions, and is reflected in recent household surveys using participatory techniques. For example, the survey results showed that only 35 percent of the very poor indicate that they are better off now than they were four years ago, compared to 51 percent who felt they are better off now than they were 10 years ago. On the other hand, 53 percent felt they are worse off The priority areas include agriculture, manufacturing, petroleum, mining, transport, high technology, and tourism. 29 The national GDP figures were adjusted from calendar year to fiscal year basis to correspond to the fiscal accounts.  it Table 1.9: PERCENT OF GDP IN CURRENT PRICES Average FY81 FY86 FY88 FY90 FY91 FY92 FY93 FY94 FY81-85 FY86-90 FY91-94 GDPatmarketprices 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 Resource Balance -10.1 -13.0 -23.1 -22.1 -25.8 -29.1 -29.2 -28.0 -8.5 -20.1 -28.0 GrossDomesticExpenditures 110.1 113.0 123.1 122.1 125.8 129.1 129.2 128.0 108.5 120.1 128.0 Consumption 80.9 93.7 94.8 85.7 91.2 97.8 96.8 96.7 86.8 92.4 95.6 Private 68.6 77.5 84.1 78.4 83.8 90.5 88.6 88.4 72.9 81.1 87.8 Public 12.3 16.2 10.7 7.3 7.4 7.4 8.2 8.3 13.9 11.3 7.8 Investment(GDIYI 29.2 19.3 28.3 36.4 34.6 31.3 32.4 31.3 21.6 27.7 32.4 Private/2 17.8 12.9 16.0 25.8 25.0 21.6 22.8 22.7 13.2 17.5 23.0 Public 11.4 6.5 12.4 10.6 9.6 9.7 9.6 8.6 8.4 10.2 9.4 Memo Itmes: Domestic savings/GDP 19.1 6.3 5.2 14.3 8.8 2.2 3.2 3.3 13.2 7.6 4.4 National savings/GDP/3 21.0 12.7 18.5 26.1 22.7 19.2 20.7 20.1 15.0 18.7 20.7 National savings/GDP/4 18.7 11.5 11.3 16.7 15.1 10.3 7.0 8.8 12.8 11.9 10.3 Currentaccountdeficit/GDP/3 10.5 7.8 17.0 19.7 19.5 21.0 25.4 22.6 8.8 15.7 22.1 Overall budget deficit/GDP /5 13.3 6.4 10.5 5.3 3.2 2.2 10.6 10.2 10.6 7.6 6.6 1/ sharp increasc in investment starting from 1987 atributes to higher investment cost in manufacturing, electricity and water supply, construction, and trnsport and comnunication due to the appreciation of foreign exchange. 2/ includes stock chwnges. 3/ officiallyreported. 4/ adjusted to be consistent with national accounts and external account. 5/ checks cleared and before granta, and the ratios are in fical year (FY) ending June 30.  26 Chapter 1 now than they were four years ago, compared 40 percent who indicated they were worse off 10 years ago. The serious macroeconomic imbalances and the inflationary pressure have to receive greater attention in the next stage of reforms in order to arrest the erosion of incomes, and to sustain the economic growth and the progress that has been made in reducing poverty. Resource Balance 1.48 A most serious aspect of the resource situation is the widening resource gap. As shown in Table 1.9, the overall resource balance more than doubled from 8.5 percent of GDP to 28 percent between FY80-FY85 and FY91-FY94, indicating growing aid dependence. While investmnent increased, its efficiency is low for reasons discussed earlier [para 1.43]. The increased investment occurred, increased, even as domestic savings declined from 13.2 percent of GDP in FY80-FY85 to 4.4 percent of GDP in FY9 1 -FY94. The reform and economic recovery led to consumption growth, following a decade of stagnation and severe shortages of consumer goods. Consumption rose by 4.5 percentage points of GDP between the early 1 980s and the early 1 990s, reflecting mainly the growth of private consumption. Public consumption appears to have declined, though various claims on the budget, including those for public administration, debt service and parastatal losses are substantial. Savings and Investment 1.49 Figure 1.6 reflects the expanding Figure 1.6: Savings and Investment savings-investment (percent of GDP) gap, which is financed mainly by external resource inflows, including private and official transfers. 40 Increases in net 35 official development 30 assistance from an 25- Invest average of about 10 12 -|saving| percent of GDP in 10 1980-85 to about 20 5 percent of GDP in o 1986-94 allowed co c X co X co 0 0 Tanzania to expand U. investment in spite of the sharp devaluation, which raised the costs of capital goods. However, the low productivity of investments, especially the poor performance of the financial and non-financial parastatals, which was noted earlier,  Chapter 1 27 accounts in part for fragile economic growth, low incomes and low levels of domestic resource mobilization.30 1.50 Savings performance has also been poor, though the fact that official private savings are estimated as residuals warrant caution in interpreting the data. Large amounts of private transfers (US$485 million in FY94) reported in the external accounts are believed to consist mainly of unrecorded exports, suggesting under-estimation of savings by about 6 percent of GDP.3' Even with these caveats there is a serious resource problem as indicated by growing aid dependency. While there are no reliable estimates of private savings, such savings are not likely to exceed the African average. The officially estimated gross domestic savings to GDP ratio of 4.4 percent in FY91 -FY94 is a far cry from the African average of about 17 percent during about the same period. Even for the earlier 1981-90 period, Tanzania's savings ratio was 10.6 percent of GDP, compared to 33.2 for East Asia, 31.9 for Southeast Asia and 12.6 for Sub-Saharan Africa. Resource Mobilization 1.51 Experience in the East Asian countries suggest the kinds of "pro-savings policies" that may help Tanzania to improve its resource mobilization.32 These policies have to be implemented as a package and pursued steadily to be mutually reinforcing. One option is to pursue policies that will curb inflation and raise real interest rates so as to provide positive financial returns. In East Asia, for example, interest on savings averaged a positive 3-6 percent, compared to minus 13.5 percent in Tanzania during 1980-94. The adoption of prudent fiscal and financial policies to restore macroeconomic instability and to contain inflation will be central to pro-savings policies; for example, the nominal interest rate on savings has risen from 10 to over 25 percent since 1986, but this has been offset by inflation. It is also important to: (i) restructure the financial institutions, especially the dominant state-owned NBC, to contain its huge losses and (ii) nurture the growth of a viable banking system by, for example, strengthening bank supervision and prudential regulations. Providing financial services for the rural areas might also help expand the base for financial savings. Third is to reform the tax system to mobilize revenue and to restrain expenditure to eliminate recurrent dissavings of about 4 percent of GDP; however; this will affect private savings. A fourth option is to reduce the drain of parastatals by accelerating reforms to liquidate non-performing units and expand private sector management control. Fifth is: (i) to expand entrepreneurial base by sponsoring training and business support services, including market information, and (ii) to maintain a hospitable environment for established businesses to save in local currency rather than abroad. Last, and most important perhaps, is to emphasize policies that promote robust rural income growth to expand the base for rural savings. 30 Agrawal, N., Z. Ahmed, M. Mered, and R. Nord, "Structural Adjustment, Economic Performance, and Aid Dependency in Tanzania. " World Bank Policy Research Working Paper No. 1204, 1993. 31 M. Gavin, 1994. "Saving and Investment in the Tanzanian Economic Reform; p. 38. Gavin's estimate of savings underestimation from unrecorded exports was 10 percent of GDP in 1994, which would be equivalent to about 6-7 percent based on revised national accounts of August, 1995. 32 p. Harrold, et al.,"Practical Lessons from East Asia in Industrial and Trade Policies." World Bank, 1995.  28 Chapter 1 Balance of Payments 1.52 There were some favorable deve- lopments in externalFigure 1.7: Current Account Balance lopments in extemal ~~(percent of GDP) receipts. For example, there was a steady increase in net private transfers from US$242 million in FY86 to US$450 million by (n L - c FY94. These private - 5 > > transfers are presumed -10 to be earnings from -15 CA: Grants undocumented exports, - CA Ex Grants and they have become -20 important after -25 introduction of the Own Funds Imports -30 Scheme, which lifted restrictions on private importers. In addition, receipts from services improved. Nevertheless, the external financing situation is weak as shown in Figure 1.7. Even though export volume has grown faster than imports, the current account deficit (without grants) in proportion to GDP has widened from 8.8 percent of GDP in FY81-FY85 to over 20 percent during FY91-FY94, with some improvement in FY94 as export prices rose. The major factors underlying the deterioration of the current accounts are (i) a sharp rise in interest payments on external debt and (ii) a rise in import prices and deterioration in the terns of trade. (See Table 1 .1.) 1.53 Capital inflows and private transfers enabled official reserves to rise. Total official reserves stood at about nine weeks of imports at the end of FY95, compared to under three weeks in FY86. Direct foreign investments began to rise with the liberalization of private investments. Grants more than doubled from the level of the mid-1980s, reaching a peak of US$459 million in FY94 (Table 1.10). Long term net loan disbursements were negative in FY93-FY94, reflecting the external debt problem; but total net capital inflows, excluding private transfers, were estimated at about US$700 million in FY94. Overall official capital inflows are believed to be understated by about US$150-200 million because of offshore payments for consultants, contractors and turnkey projects.  Chapter 1 29 Table 1.10: BALANCE OF PAYMENTS (in millions of USS) FY81 FY86 FY88 FY90 FY91 FY92 FY93 FY94 Trade of goods, net/I -654.5 -708.0 -842.0 -956.7 -987.7 -1030.3 -1060.3 -949.9 Exports, fob 537.2 316.6 338.0 389.3 393.6 414.1 411.4 485.9 Imports, cif 1191.7 1024.6 1180.0 1346.0 1381.3 1444.4 1471.7 1435.8 Factor income services, net 35.0 -76.6 -268.5 -267.8 -237.6 -292.4 -351.0 -262.3 Private Transfers, net 22.3 241.7 519.0 460.6 407.7 456.4 463.2 450.0 Current Account Balance (before -597.2 -542.9 -591.5 -763.9 -817.6 -866.3 -948.1 -762.2 official transfers) Official Transfers (net) 243.7 278.3 489.0 530.0 548.6 569.0 582.1 459.0 Current Account Balance (after -353.5 -264.6 -102.5 -233.9 -269.0 -297.3 -366.0 -303.2 official transfers) Capital Account, net 254.8 -81.3 -145.8 23.0 89.5 187.3 -125.7 117.5 Direct foreign investment .. .. .. .. 10.0 15.0 61.7 63.0 Net M&LT borrowing 162.0 -0.8 -0.5 37.2 60.7 146.1 -167.0 -154.7 Disbursement 162.0 201.5 189.5 224.1 211.4 288.6 189.4 206.6 Amortization .. 202.3 190.0 186.9 150.7 142.5 356.4 361.3 Other capital inflows and 92.8 -80.5 -145.3 -14.2 18.8 26.2 -20.4 209.2 errors and omissions Overall Balance -98.7 -345.9 -248.3 -210.9 -179.5 -110.0 -491.7 -185.7 Financing: 98.7 345.9 248.3 210.9 179.5 110.0 491.7 185.6 Bank of Tanzania/2 14.6 -3.7 -60.8 -56.7 -116.2 -117.0 131.2 9.7 Grossreserve 28.7 -16.8 -94.4 -66.0 -85.7 -175.5 90.3 -11.7 Net use of IMF credit -14.1 13.1 33.6 9.3 -30.5 58.5 40.9 -10.2 Other .. .. .. .. .. .. .. 31.6 Commercial banks .. .. . .. 30.4 -16.1 -13.6 -116.7 Counterpart to monetization .. .. .. .. 22.8 30.9 5.5 0.0 Arrears 84.1 -4.9 259.8 -28.2 88.6 -121.9 30.1 -30.1 Debt rescheduled .. 354.5 49.3 295.8 153.9 333.9 338.5 322.8 Memo item: Gross reserve level .. .. .. 124.3 209.4 384.9 294.6 306.3 Current Account Balance as % of GDP Before Official Transfers -10.5 -7.8 -17.0 -19.7 -19.5 -21.0 -25.4 -22.6 After Official Transfers -6.2 -3.8 -3.0 -6.0 -6.4 -7.2 -9.8 -9.0 I/ Merchandise trade only. 2/ The negative sign means increase. Source: The Tanzanian authorities, and the Fund staff estimates.  30 Chapter 1 External Enviroument 1.54 Tanzania ratified the Final Act of the Uruguay Round and became a member of the World Trade Organization in early 1994. The agreement poses challenges and also opportunities for orderly trading, and for reduced tensions and trade restrictions. Tanzania's share of world exports is small, and the external environment does not constrain export growth. However, the economy is susceptible to commodity price fluctuations and to adverse terms of trade. As the recent coffee price boom subsides, the export environment seems uncertain. Existing reviews suggest likelihood of over-supply and decline in the price of tea, a tightening of stocks and price increases for cotton, a softening for tobacco, and uncertainty for coffee during the next 2-3 years.33 On the import side, sharp increases in the price of oil would undermine the external accounts, given the dependence on oil which accounts for about 35 percent of export earnings; however, while oil prices may inch up, sharp increases which would disrupt orderly external financing are not considered likely. 1.55 Tanzania's location also provides challenges and advantages within the Southern and Eastern African Region. To promote intra-regional trade, the East African Economic Union was revived at a meeting of the Heads of State of Tanzania, Kenya and Uganda in March, 1996. Tanzania is a member of the Southern African Development Community (SADC) and a party to the tariff reductions agreement signed in February, 1995 to promote intra-community trade. However, this effort duplicates the Preferential Trade Area of the Common Market of Eastern and Southern Africa (COMESA). Improved competitiveness would enable the country to expand regional trade and also provide rail, shipping and road transport services for neighboring landlocked countries. The port of Dar es Salaam already provides freight services to Zambia, Zaire, Malawi, Burundi, Rwanda and Uganda. Improved efficiency would be critical for retaining and expanding market share as peace in South Africa makes that country a competitor in the industry. On the regional political front, the country has managed to contain the refugee problem created by instability in neighboring countries, with external support. 1.56 Given the expanded use of aid, Tanzania is vulnerable to declines in external assistance. Even if improved economic management results in the resumption of balance of payments support, sustainable growth and development would require greater domestic resource mobilization. 33 See "Commodity Markets and the Developing Countries". A World Bank Quarterly, May 1995; pp. 33-37.  2. FISCAL AND PUBLIC RESOURCE MANAGEMENT A. OVERVIEW AND RECENT DEVELOPMENTS Overview 2.1 This chapter focuses on fiscal performance. The progress made in economic reforms so far is being undermined by severe fiscal and financial difficulties. There is serious shortage of revenue, inadequate financing for critical development activities, and an unstable and widening fiscal situation. Addressing these problems is central to a credible strategy for economic growth and poverty reduction. Measures are needed to (i) contain excessive deficit financing and inflation which erode real incomes, (ii) restore a stable macroeconomic environment for private sector expansion and growth and (iii) limit overall expenditure growth while redirecting expenditures towards selected activities, especially human resources and basic infrastructure needed for growth. The key issues involve: * Greater revenue mobilization to eliminate persistent public dissavings and to finance critical development programs: among other things, this requires actions focused on restructuring a weak system of tax administration and a narrow tax base, which is not responsive to income growth. Revenue is low at about 8-10 percent of GDP, if based on realistic national income estimates. * Curbing expenditure over-commitment: rising expenditure claims have contributed to a persistent fiscal deficit, and a fragile and unstable fiscal system. Recurrent expenditures alone exceeded revenues by 6.1 percent of GDP in FY93, and by 4.4 percent in FY95. Debt service imposes a heavy burden and steady financing for development expenditure has become very difficult. Inadequate enforcement of fiscal discipline and laxity in public accountability have been concerns, as government investigations have revealed cases of questionable conduct in revenue administration. Efforts being made to address the problem by enforcing financial codes and spending limits need to be strengthened. * Achieving balance in allocating public resources with emphasis on education, health, agriculture and rural infrastructure. These activities have been under- funded and the asset base eroded. Public administration is weak, though it has been the fastest growing government activity until the recent reforms, including retrenchment, were initiated. 2.2 The weak fiscal situation cannot support expenditure expansion. The challenge, therefore, is to (i) raise revenues and tighten aggregate spending, (ii) reduce government spending in selected areas, (iii) allow private delivery of services to the extent feasible  32 Chapter 2 and (iv) restructure the expenditure program to provide adequate financing for critical development activities. The restructuring of public administration, together with aggressive efforts in revenue mobilization and debt relief, will be key to improving the fiscal situation and channeling resources to the productive services. The sectors which require the most support are education, agriculture services and rural infrastructure. Others which are also under-funded are drinking water and primary health care. 2.3 In addition, measures to overcome the difficult fiscal challenges facing Tanzania would need to be implemented as a package comprising: (i) reorganization of the system of public administration to build needed capacity, provide incentive pay and limit the size of Governrnent; (ii) an aggressive domestic resource mobilization effort; (iii) enforcement of fiscal accountability and adherence to financial rules and budgetary ceilings; and (iv) external debt relief. The strategy for debt management already formulated will need to be implemented, including limiting exposure to commercial debt, rescheduling eligible debt on very favorable terms, and. limiting domestic borrowing. Recent Fiscal Performance 2.4 Fiscal performance went through various phases during the reform period. During the early phase of the reforms, the budgetary deficit narrowed significantly. There was a decline in the fiscal deficit, due mainly to a sharp drop in expenditures in relation to GDP from 29.1 percent in FY81-85 to 23.3 percent in FY86-90;34 domestic borrowing declined relative to GDP by more than half. However, the fiscal situation remained fragile, with an upsurge of deficits and wide fluctuations during FY91 -FY95. 2.5 There was a setback in mid-1992, which persisted to the first half of 1994. The Government had attempted to improve revenues by lowering the tax rates and broadening the base. But the measures were not enforced effectively enough to offset revenues lost by the tax simplification. Revenues dropped by 3.9 percent of GDP in FY93; import taxes declined by 1.2 percentage points of GDP and domestic sales and excise taxes by 2 percentage points of GDP. Also, expenditures rose by 5.2 percent of GDP, reflecting mainly the expenditures carried over from the previous period. Economic services, especially roads and bridges, accounted for most of the increase, rising from 2.9 percent of GDP to over 7 percent during FY92-FY93. The fiscal deficit deteriorated sharply from 2.2 percent to 10.6 percent of GDP between FY92 and FY93. 2.6 For a while starting in mid-FY95, the Government again intensified its efforts to bring the fiscal situation in line. A mid-year budget was adopted to raise revenues and enforce tighter spending controls. Tax exemptions appeared to have been curbed. Restrictions on check issues by line ministries were proposed, and banks were given authority to reject checks in excess of funding limits. However, performance remained below expectation due to weak tax administration and overrun on recurrent expenditures 34 The fiscal deficit is defined in terms of checks cleared before grants.  Chapter 2 33 from various sources, including wage payments, clearing of arrears to the electricity corporation, retrenchment costs of the civil service reform and higher than expected interest payments on domestic debt. 2.7 The weak fiscal performnance persisted during the first half of FY96. Revenues are now estimated at T Sh 439 billion in FY96, compared to the original budget estimate of T Sh 456 billion. Recurrent expenditures are expected to exceed the budgetary target by T Sh 119 billion, about 4 percent of GDP, reflecting excess spending on the elections, larger than expected interest payments on domestic debt and lack of control over the wage bill. 2.8 The new government, which assumed power in November 1995, has clearly expressed its intention to improve fiscal performance by increasing revenue and maintaining better control over expenditures. The Government has made a start by adopting a mini-budget, and it has renewed discussions with the Bank and the Fund in the various areas, especially by adopting an informal program for the second half of FY96. The main fiscal targets of the informal program include improvement in revenues by 1.7 percent of GDP; reduction in expenditures and in the recurrent deficit to 2.7 percent of GDP from a projected 4.4 percent of GDP without the program; and deceleration of money growth from 30 to 25 percent and of inflation from about 27 to 22 percent between January and June 1996. The outcome of this effort remains to be seen. 2.9 Table 2.1 and Figures 2.1 and 2.2 show the evolution of fiscal performance. Table 2.1 shows the fiscal deficit, which expanded from 2.2 to 13.5 percent of GDP during FY92-FY95. Figure 2.1 summarizes the profile of government expenditures since the 1980s.35 Figure 2.2 shows the main sources of financing. Revenue showed no steady recovery in FY94-FY95. Access to aid improved; there was a pressing need to complete development projects under implementation. Increases in expenditures, together with the unsteady revenue pattern, made it difficult to contain deficit financing. Domestic borrowing rose from minus 3.1 percent of GDP in FY92 to 3.9 percent of GDP in FY95. 35 Table 2.1 and Figure 2.1 understate actual resource and expenditure flows, mainly because of off- budget items, including technical assistance, turn-key projects and activities financed directly by donors outside the budget. Technical assistance aid alone is estimated at about US$200 million average per year.  Table 2.1: SUMMARY OF CENTRAL GOVERNMENT OPERATIONS (in millions of Tanzanian shillings) j FV81 FV86 FY88 FY90 FV91 FY92 FV93 FV94 FV95' FV9%2 Average FVSI-85 FY86-90 FY91-94 Total Revenue 8742 20831 46431 94655 133238 173566 164110 242444 331200 439258 12666 52961 178340 Total expenditure and net lending 14371 32373 73368 126242 151863 194891 305059 370910 595000 613465 19772 74684 255681 Recurrent expenditure 9998 26912 51896 103866 135375 162286 240349 296221 431000 518873 15220 59188 208558 Developmentexpenditureand 4373 5461 21472 22376 16488 32605 64710 74689 164000 94592 4552 15497 47123 part of net lending Overall deficit (checks issued) -5629 -11542 -26937 -31587 -18625 -21325 -140949 -128466 -263800 -174207 -7106 -21723 -77341 Adjustment to cash and other -612 3205 -2982 4129 -8750 -1870 6628 -36753 -45500 -18000 219 -1918 -10186 items (net) Overall deficit (checks-cleared) -6241 -8337 -29919 -35716 -27375 -23195 -134321 -165219 -309200 -192207 -6887 -23641 -87528 Financing: 6241 8337 29919 35716 27375 23195 134321 165219 309200 192207 6887 23642 87528 External grants . .. 15909 27664 22876 32796 58313 76909 105600 84373 .. 18990 47724 Foreign Financing, net 2702 2222 7318 1489 8858 22676 29436 47754 113600 -40181 2415 2037 27181 Domestic Borrowing 3538 6115 6692 6563 -4359 -32277 46572 40556 90000 148014 4472 6413 12623 As % of GDP at market prices Total Revenue 18.6 16.0 16.3 14.1 15.7 16.8 12.9 15.0 14.5 14.7 18.3 15.8 15.1 Totalexpenditureandnetlending 30.6 24.8 25.7 18.8 17.9 18.9 24.1 23.0 26.0 20.5 29.1 23.2 21.0 Recurrentexpenditure 21.3 20.7 18.2 15.5 16.0 15.7 19.0 18.4 18.9 17.3 22.2 18.1 17.3 Developmentexpenditureand 9.3 4.2 7.5 3.3 1.9 3.2 5.1 4.6 7.2 3.2 6.9 5.1 3.7 part of net lending Adjustment to cash and other -1.3 2.5 -1.0 -0.6 -1.0 -0.2 0.5 -2.3 -2.0 -0.6 0.2 -0.2 -0.7 items (net) Overalldeficit(checks-cleared) 13.3 6.4 10.5 5.3 3.2 2.2 10.6 10.2 13.5 6.4 10.6 7.6 6.6 Financing: 13.3 6.4 10.5 5.3 3.2 2.2 10.6 10.2 13.5 6.4 10.6 7.6 6.6 External grants5 .. .. 5.6 4.1 2.7 3.2 4.6 4.8 4.6 2.8 . 5.1 /1 3.8 Foreign financing, net 5.8 1.7 2.6 0.2 1.0 2.2 2.3 3.0 5.0 -1.3 3.8 1.0 2.1 Domestic borrowing 7.5 4.7 2.3 1.0 -0.5 -3.1 3.7 2.5 3.9 4.9 6.8 2.5 0.6 1/ Likely outturn. 2/ Mid-year revised budget with new measures. 3/ For the period FY87-90. 4/ For the purpose of calculating the ratios, the GDP figures were adjusted to fiscal year basis to correspond to fiscal accounts. c 5/ No specific breakdown for foreign grants prior to FY87. Source: The Tanzanian authorities.  Chapter 2 35 Figure 2.1: Government Expenditure (percent of GDP) 35 30 25 20 T ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~4Total Exp 20 -~~~~~~~~~~~~~~~~Rec Exp 15 -~-Dev Exp 10 5 0 _ N c7 LO C)D r-c 0) C) O 4 _ NLO L ;> > :; L L 2 Figure 2.2: Government Financing (percent of GDP) 20- 15- 10 +-4-- Revenue -4+-Foreign Aid 5 - Domestic Borrowing 0 - co I 0 I 0 I IC- co S LL L L UE L L 'Los L L  36 Chapter 2 SECTION I. EXPENDITURE MANAGEMENT B. EXPENDITURE PERFORMANCE 2.10 Excessive govern- ment expansion and over- Figure 2.3: Recurrent Revenue and Expenditure commitment underlie the (percent of GDP) fiscal and public resource manage-ment problem facing Tanzania. The expansion was carried over from the socialist 30 era, and it has persisted 25 even with reforms. 20 _ Exp During FY91-95, for 15 RevcEnu example, expenditures 10 rose by an annual average 5 equivalent to 2.0 0 l l l __l l l l _ l l _ _ percentage points of c9 GDP, compared to minus ' 0.3 percentage points of GDP for revenues. Recurrent Spending 2.11 One aspect of the problem is that revenues are not adequate to cover budgeted recurrent cost items fully. The gap between recurrent expenditures and revenue widened in recent years as shown in Figure 2.3. The additional revenue that would have been required to finance budgeted recurrent costs fully was equivalent to 6.1 percent of GDP in FY93 and 3.1 percent of GDP during FY94-FY95. Moreover, recurrent cost financing needs are much more than those actually presented in the budget. For example, estimated true recurrent cost for FY96 is T Sh 747 billion, while proposed financing is T Sh 488 billion, a shortfall of 35 percent. In addition, reforms aimed at expanding the delivery of social services will require more recurrent cost financing as the social sectors require greater recurrent funds than other sectors.36 36 For example, in the FY95 budget proposed under the rolling plan for FY95-FY97, the ratio of capital expenditures to recurrent expenditures were 1:3.5 for human resource development; 1:0.45 for agriculture, and 1:0.23 for infrastructure.  Chapter 2 37 Table 2.2: Economic Classification of Recurrent Expenditures, FY89-FY94/1 FY89 FY90 FY91 FY92 FY93 FY94 Composition of Recurrent Expenditure: Percent /2 Wages and Salaries 25.8 25.8 22.5 23.3 23.9 24.9 Interest Payments 18.8 16.1 14.8 14.7 13.1 10.5 Domestic (8.3) (7.4) (6.4) (6.5) (7.2) (10.4) Foreign (10.5) (8.7) (8.4) (8.2) (5.9) (8.5) Other Goods and Services /3 38.9 34.6 45.1 43.2 42.5 43.1 Transfer Payments /4 16.5 23.5 17.6 18.8 20.5 21.5 Crop Authorities /5 (1-9) (3.4) (1.7) (2.2) (2.4) (2.0) Parastatals (2.6) (4.3) (5.2) (1.0) (5.7) (5.4) Other Domestic (10.8) (14.3) (9.8) (15.0) (11.6) (13.6) Foreign (1.2) (1.5) (0.9) (0.7) (0.8) (0.5) Recurrent Expenditure 100.0 100.0 100.0 100.0 100.0 100.0 Memorandum Items Military/6 12.0 10.4 12.1 8.9 8.7 Wages and Salaries /7 28.0 28.3 22.8 21.6 34.9 31.2 1 Checks-issued basis, partly estimated. 2 Defined as Recurrent Expenditure plus Defense Development Expenditure. 3 Determined as residual. 4 To parastatals, universities and research institutions. 5 Including cost of Strategic Grain Reserve. 6 Estimated military percent of total. 7 Percent of revenues. Sources: Data provided by the Tanzania authorities and IMF staff estimates. 2.12 Allocations of recurrent costs need to be rationalized. Expenditures on wages and salaries are excessive relative to non-wage expenditures, but the problem has been obscured by the way expenditures are classified. As shown by the official statistics in Table 2.2, wages and salaries accounted for 25.8 percent of recurrent expenditures in FY89 and 25 percent in FY94. On the other hand, purchases of other goods and services accounted for about 40 percent of recurrent expenditures in FY89-FY94. This is misleading. The official figures on wages and salaries obscure various allowances and non-monetary rewards which are part of employee compensation, but are not appropriately classified; they are included in "Other Services and Goods". There are close to 40 such allowances, and the total estimate average about 50 percent of basic pay,  38 Chapter 2 with a range of about 20 to 380 percent.37 To illustrate, an adjustment of the official wage and salary figures based on the low end figure of 20 percent underestimation of total compensation would raise the total compensation costs close to 40 percent of recurrent expenditures in FY94, which is high by developing country standards. 38 2.13 It is important to note that the large share of estimates devoted to wages, salaries and allowances does not necessarily mean that public officials are over-paid. The main issues are that: (i) the compensation system lacks transparency and undermines the urgency of expenditure reforms, and (ii) that revenues are too low to support employee compensation adequately. Other major components of recurrent costs which expanded in recent years are transfer payments and interest on public debt. As shown in Table 2.2, transfer payments are dominated by transfers to parastatals and transfers to various institutions, including universities and local councils. Recurrent Savings 2.14 Public recurrent savings, that is, the Figure 2.4: Recurrent Savings Government's recurrent (percent of GDP) savings from revenues, have been negative since the start of the 1 980s except in FY92. This 2 pattern (Figure 2.4) l indicates that public 0 . . . . . . . recurrent savings have not 21 been improving even as the -2 economy is recovering. 4 The magnitude of the -5 dissavings would be much -6 higher if amortizations -7 were included. The _ dissavings on recurrent 9 spending were equivalent to 4.4 percent of GDP in See K. Kiragu and R. Mushi, "Tanzania: Monetization of Benefits and Pay Reform Proposals". URT and World Bank Report, August 1992. 38 The ratios for some comparator countries are Mauritius (32 percent, 1985-90); Kenya (30 percent, 1980-87); sub-Saharan Africa, excluding Nigeria (27 percent, 1980-87); Zambia (20 percent, 1980- 87). See Government Finance Statistics Yearbook, 1990, p. 72; UNDP/World Bank, African Economic and Financial data; 1990.  Chapter 2 39 FY95. One consequence of the dissavings is a severe shortage of local funds to provide the counterpart resources needed to support development projects. Development Expenditures 2.15 Development ex- penditures are also over- Table 2.3: Development Expenditures extended. With domestic (T Sh millions) resources in short supply, the development expen- FY92 FY93 FY94 FY95\ ditures program is totally dependent on aid, is over- Total 32604 64710 74688 49700 crowded with too many activities and lack Foreign 12003 49815 61020 38700 activities, and lacks adequate local counterpart Local 26601 14895 13668 11000 funds to implement pro- Local % 63.2 27.2 18.3 18 jects effectively. Local a Budget. counterpart fund Source: Rolling Plan and Forward Budget/or Tanzania. Planning Commission. contributions to deve- lopment projects declined steadily from 63.2 percent in FY92 to 18.0 percent in FY95 as shown in Table 2.3. Aid support for development increased in the early 1990s, but the aid could not be fully utilized because of the local fund shortage. In fact, because of the severe shortage of local funds in FY95, the local counterpart funds allocated to the development budget were frozen, and only a small portion was actually disbursed. The allocation in the development budget in FY95 was T Sh 121.9 billion; actual expenditure was only T Sh 49.7 billion. This resulted in a sharp decline in development expenditure from 4.6 percent of GDP in FY94 to 2.8 percent in FY95. It is important to note that the counterpart fund problem is more severe than Table 2.3 indicates mainly because of off- budget development expenditures. Public Administration Expenditures 2.16 Another major source of over-commitment is public administration, which was discussed in Chapter 1. Figure 2.5, based on official data, indicates the general pattern of expenditures for general public services, defense and debt service. Together they accounted for 48.2 percent of government spending in FY82 and 65.7 percent in FY92; and they also absorbed over 80 percent of revenues (excluding aid) during the early 1990s. General public services, including general administration, external affairs, and public order and safety, continued to expand in the 1980s as well as the 1990s, rising from under 4.9 percent of GDP in FY81-85 to 9 percent of GDP in FY90-94. The expanded claims by public administration have contributed to displacing the financing for other activities.  40 Chapter 2 Figure 2.5: Administration and Debt Service (percent of GDP) 9 8 l 7 8 - ?i>|Ge bAsdm.: 4 3 2 O I I t l l l l l l l l~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ I o co - co co co O co co go a) a)N L L L L L LL >U L> L> L Figure 2.6: Social and Economic Services (percent of GDP) 14 12 10 _ - Soc serv. -2U- Ec serv. 6 4 2 O g I I I co I I  Chapter 2 41 2.17 During the period of economic decline in the 1970s to the mid 1980s, Government adjusted to the economic decline and fall in revenues by protecting general public services, while allowing erosion of the productive base of the economy. This affected social services, economic services, and infrastructure. Figure 2.6 shows the pattern of resource claims by social and productive sectors. Expenditures on economic services, including infrastructure, declined much faster than the decline of overall economic activity. By 1990, allocations to economic services as a share of GDP were less than half the level in the early part of the 1980s.39 C. DEBT SERVICE 2.18 A significant third element of the budget over-commitment is the debt service burden. Recent estimates put Tanzania's total external debt stock at about US$6.8 billion, over 200 percent of GDP. Almost all the debt, about 99.8 percent, is public or publicly guaranteed debt. Over 90 percent of the long term debt is owed to official creditors; about 44 percent or so of the total is owed to multilateral institutions, and about 50 percent is bilateral. Private creditors account for the rest. Based on the originally contracted repayment schedule, external debt service ratio to the export of goods and 40 services averaged about 40 percent annually during FY90-FY94. However, the actual debt service ratio averaged less than 30 percent because of rescheduling and accumulation of arrears. The total arrears for FY96 are estimated at US$290 million, owed mainly to bilateral creditors. Even with the arrears, lower payments impose a heavy budgetary burden. Total debt service rose from 28 percent of revenues in FY86 to 35 percent in the 1990s. 2.19 Domestic debt has also become a serious concern. Estimated total domestic debt reached T Sh 276 billion at the end of FY95, up from T Sh 167 billion at the end of FY94. About 14.9 percent of this was in treasury bills and the rest stocks and bonds. Interest payments on domestic debt rose from T Sh 19.6 billion in FY94 to T Sh 57.7 billion in FY95. The overall increase in domestic debt reflects the fiscal deterioration over the last two years. With higher deficit financing, there was a sharp increase in the treasury bill rate, reaching a peak of 66-71 percent in November and December 1994. This raised interest payments on domestic debt service significantly. Also, the Government assumed responsibility for parastatal debt, including infusion of capital into the financial sector parastatals. 2.20 Under any realistic macroeconomic scenario, it is unlikely that Tanzania would be able to meet its existing scheduled debt service obligations while at the same time maintaining stable macroeconomic policies and providing orderly financing for critical 39 The figures are based on data available from the Planning Commission. They exclude other expenditures provided directly by donors or by the private sector; however, such expenditures are more significant for the social sectors than for economic service. 40 It was 38 percent in FY91, 34 percent in FY92, 50 percent in FY93, and 39 percent in FY94.  42 Chapter 2 Table 2.4: Debt Service (percent of GDP) FY86 FY87 FY88 FY89 FY90 FY91 FY92 FY93 Total Debt Service 4.5 6.6 7.2 7.0 5.1 5.4 5.5 5.3 Interest 2.6 3.7 3.7 3.1 2.5 2.4 2.3 2.5 Amortization 1.9 2.9 3.5 4.0 2.7 3.1 3.2 2.8 Domestic Debt 2.7 3.0 2.1 2.0 1.2 1.7 1.5 1.5 Interest 1.7 2.4 1.9 1.4 1.1 1.2 1.0 1.4 Amortization 1.0 0.6 0.1 0.7 0.1 0.6 0.4 0.2 Extemal Debt 1.8 3.6 5.1 5.0 3.9 3.7 4.1 3.7 Interest 0.9 1.3 1.7 1.7 1.3 1.2 1.3 1.1 Amortization 0.8 2.3 3.4 3.3 2.6 2.5 2.8 2.6 Memo Item Debt Service as % of 28.0 37.3 44.1 46.6 36.4 34.5 32.9 40.6 Revenues Source: Tanzanian authorities and staff estimates. development activities. In recognition of this, the Government has already formulated a debt management strategy to begin addressing the problem. The effort comprises (i) limiting exposure to non-concessional debt, (ii) instituting a debt buyback scheme for commercial debt and (iii) meeting debt obligations to prevent accumulation of arrears, while seeking rescheduling for eligible debt. 2.21 Tanzania requires exceptional debt relief measures to overcome its external debt service problem. An analysis of debt sustainability (Chapter 9) suggests that with strong fiscal and export performance, debt service would be favorable in the medium term, based on debt relief under Naples Terms and a discounted debt buyback to reduce commercial debt. This would involve a 67 percent reduction in the stock of official debt, plus 85 percent reduction in commercial debt through debt buyback and financing the gap on concessional terms. The trend of debt service ratio to exports will be favorable, falling to under 20 percent by the close of the decade. However, the debt stock indicator represented by the ratio of present value of debt to exports will remain high at over 200 percent, which will be difficult to manage in the medium term. Reducing this external debt overhang will require exceptional support. 2.22 In terms of the fiscal burden, the share of external debt service from revenues will fall from 45.3 percent in FY96 to 25.4 percent by the end of the decade. Domestic debt service will similarly fall from 13.8 percent of revenue in FY96 to 8 percent in FY2000. However, total debt service, including foreign and domestic debt, will be high at 31 percent of revenues by the end of the decade, suggesting the need for a tight domestic borrowing together with special consideration in external debt rescheduling. But  Chapter 2 43 consideration alone will not be sufficient. The feasibility of even this scenario requires strong improvement in economic management and strong export performance, revenue mobilization to finance the budgetary aspects of debt service, reduced domestic borrowing and measures to promote economic growth. Details of assumptions are discussed in Chapter 9. D. SECTOR EXPENDITURES 2.23 This section focuses on the sector expenditures. The sector data employed here exclude direct donor expenditures channeled outside the budget. This limitation and the general difficulties of reporting sector expenditures accurately warrant caution in interpreting specific numbers. Nevertheless, the general pattern of sector allocations indicates the priorities of public resource use. 2.24 The main conclusion is that the fiscal situation cannot support expansion under existing circumstances. The challenge, therefore, is to (i) reduce the role of the Governnent and focus it on selected activities, (ii) permit private delivery of services to the extent feasible and (iii) restructure the expenditure program to provide adequate financing for critical development activities. Rural infrastructure aimed at supporting agricultural and rural development would need to be the focus of government activity. Agricultural services, rural roads and education appear the least funded development programs. Others include drinking water and primary health care. 2.25 Table 2.5 shows the specific activities which declined the most, namely: (a) agriculture, including forestry and fishing, and (b) industry, comprising mining and manufacturing. The decline in spending on industry is consistent with the liberalization of industrial policies and parastatal reforms, which warrant less government role. Though agricultural growth recovered in recent years, the recovery was stimulated by deregulation, liberalization, and rehabilitation of roads more than direct government spending on delivering essential extension services. Allocations for social services and infrastructure began to recover in FY91 -FY94, following a decade of decline, though they are lower in proportion to GDP than in the early 1980s. Total expenditure on education, which is mainly recurrent, averaged 1.9 percent of GDP during FY91-FY94 and did not compare favorably with that of other developing countries.41 The sectoral expenditure share relative to GDP was lower in FY93 than in FY81. The inadequacy of expenditures on education has been reflected in the deterioration of school buildings and equipment, overcrowding of classrooms and shortage of supplies and materials. Primary school enrollment ratios have declined from over 90 percent in 1980 to about 69 percent currently; gross public secondary school enrollment is stagnant at about 5 percent. 41 The share of total government expenditure on education for Tanzania averaged 7.5 percent during FY91-FY94. While part of this may reflect inadequate reporting, there is evidence of erosion of education assets. In the World Development Report of 1994, the average shares of government expenditures on education for low income countries were 12.2 percent for selected Sub-Saharan African countries; 10.0 percent for Malawi; 20.1 percent for Kenya; 13.3 percent for Sierra Leone; and 21.9 percent for Lesotho.  44 Chapter 2 2.26 The share of total expenditures for health recovered in the 1990s (see Table 2.5). While the sector also experienced severe shortages of supplies and qualified staff, the expenditure shares compare favorably with those of other developing countries.42 In addition, donor expenditures on health, which are not channeled throufh the budget, amount to more than double the level of expenditures shown in the budget. 3 2.27 Expenditures on infrastructure also fell when the economy declined. In 1990, only 10 percent of trunk roads and 9 percent of regional roads were judged to be in good condition; over half of these roads were in poor condition, while others were in very poor condition. Subsequently, expenditures on roads and bridges, transportation and communication picked up significantly, with greater allocations for roads and railways. Expenditures on the road network went up with the start of the First Integrated Roads Project in FY90, mostly to rehabilitate assets which were deteriorating. The initial emphasis has been on areas with dense population and central markets, though some attention was paid to selected rural roads. By 1993, with the emphasis on road rehabilitation, 39 percent of trunk roads and 18 percent of regional roads have been recovered. The second phase of the Integrated Roads Project has developed a priority list of rural roads, though many are yet to be funded. 42 For health, the sector share from total govemment spending during FY9 1-FY94 was 6.2 percent for Tanzania. This compares favorably with the average of 5.6 percent for low income countries reported in the World Development Report of 1994. The shares of some comparator countries were 9.6 percent for Sierra Leone; 7.8 percent for Malawi; 6.6 percent for Madagascar; 5.4 percent for Kenya; and 11.5 percent for Lesotho. Moreover, govemment expenditure on health in Tanzania accounts for under 29.4 percent of spending for the sectors. Donor spending is 31.8 percent, and private sector spending is 38.8 percent. Data cited from "Regional Burden of Disease Study". World Bank. March, 1995.  Chapter 2 45 Table 2.5: Sector Expenditures (percent of GDP) FY81-85 FY86-90 FY91-94 Total 28.65 27.28 26.28 General Public Services 4.97 6.98 6.10 Defense 3.55 2.88 1.60 Social Sectors 6.07 3.48 4.05 Education 3.43 1.52 1.85 Health 1.58 1.14 1.50 Other Social 1.05 0.82 0.70 Economic Services 5.80 2.86 2.55 Agriculture!' 2.30 1.36 1.23 Industry/b 2.45 0.68 0.53 Other 1.05 0.82 0.80 Infrastructure 3.38 1.94 3.53 Powerf 1.08 0.50 0.83 Roadst 1.42 0.72 1.35 Transport/' 0.88 0.50 1.15 Other /f 4.78 8.52 7.30 Notes. a includes Fishing and Forestry. b includes Mining, Manufacturing, and Construction. c includes Water Resource Development and Electricity. d includes Bridges. e includes other Transport Development and Communication. f miscellaneous expenditures and debt service. Source: Based on datafrom the Planning Commission. 2.28 There is a need to address the problem of power shortages intensified by recent drought conditions, which reduced the capacity of the hydropower plants. The total installed capacity of 474 Mw is low, compared to the size of the economy; and the available capacity is even lower at 300 Mw. System losses are high at about 22 percent of net generation. There is over-staffing at low job levels with shortage of highly skilled staff at the Tanzania Electricity Supply Company (TANESCO). Customer arrears are high, as the billing system functions poorly. The state has monopoly control of procurement and refining of crude oil, though six private companies operate in the sector. 2.29 While various investments in power generation (e.g., Songo Songo gas project), transmission, and distribution are under way and more are planned to remedy the serious supply shortages, major reforms are required to ensure an efficient utilization of resources channeled to the sector. The major reform issues include: (i) commercialization of  46 Chapter 2 TANESCO, including divestiture of some of its functions (e.g., the in-house construction unit); (ii) adoption of economic pricing and implementation of tariff adjustments; and (iii) competitive marketing and pricing of petroleum products by phasing out monopoly controls. There is also a need to move towards eventual private management arrangements. E. WHO Is BENEFITING FROM GOVERNMENT SPENDING? 2.30 This section assesses the extent to which public expenditures on education, health, and water are equitable.44 The share of expenditures going to the lowest income groups is judged to be equitable if the share received by the group is larger than or equal to its proportion of the population. The results of the assessment show that the bottom 20 percent of the population receives 14.5 percent of total expenditures, while the top 20 percent receives 28.6 percent. The most equitable programs are primary education, health care centers and dispensaries. The least equitable expenditures are on university education, hospitals, and water. This finding suggests a need for reallocating resources within programs, for example, between curative and preventive care. Also, more action is required to have an appropriate balance in the allocation of resources to develop the quality of human resources needed for development. An example is to expand access to secondary education, which currently reaches only 5 percent of the eligible students. Education Expenditures 2.31 How well are education subsidies targeted? The two poorest quintiles receive 28 percent of the total subsidy for education, while the two richest quintiles receive 57 percent. The most equitable program reaching the lowest income groups is primary education. The per capita subsidy within each sub-sector shown in Figure 2.7, indicates that approximately 20 percent of primary school spending goes to each income group or quintile. For secondary schools, the better-off receive much more benefit than lower income households. Access to tertiary education is the most inequitable, since virtually all university students belong to the two highest income quintiles. In the secondary and tertiary programs, the two highest income groups receive 60.7 percent and 100 percent of the total benefits. Also, more is spent on urban than on rural areas. Allocation of estimated public sector subsidies to household members is based on their use of the publicly-provided service.  Chapter 2 47 2.32 The distribu- Figure 2.7: Distribution of Education Subsidies Per Welfare tion of education Quintile subsidies across different levels of schooling is 80 determined by 70 --- demographic and 60 - behavioral factors. In 50 -__.- Tanzania, as in many 8 40 __ opulation developing *- - - -~ -- OPnmAII Education other developing L 30-a PnmaryEducation a Secondary Education countries, lowest 20 *TeitiaryEdtion income families 20 1 ii | - generally have more i ii| 0 children. The average 1 2 3 4 5 number of children Welfare Quintile below the age of 18 is 4.1 for households in the bottom quintile and 2.4 for households in the top quintile. Enrollment rates among the bottom income families are much higher at the primary school level than at the secondary school level, but they are still lower than the rates for the well-off families. The greater number of children in poorer households means that they could receive a higher percentage of the education subsidy than wealthier households, but their lower enrollment rates would have the opposite effect. For secondary schools, the bottom income group has lower enrollment ratios and, therefore, receive lower subsidy than the highest income group. Health Expenditures 2.33 The distribution of health spending is relatively more equitable than that of education. Based on their usage of facilities, the two lowest quintiles (40 percent of the population) receive 18 and 20 percent respectively of the total health subsidy, while the 40 percent of the population in the two top income quintiles receive about 38 percent of the subsidy. Expenditures on health centers and dispensaries are the best targeted to the low income groups, with 45 percent of the total spending at health centers and dispensaries allocated to the bottom 40 percent of the population. By contrast, the poorest 40 percent receive only 25 percent of hospital subsidies, compared with the top 20 percent who receive 32 percent of total hospital subsidies. This results in the top quintile receiving an average of T Sh 1,815 per capita in health subsidy, compared with T Sh 1,460 for the lowest quintile. 2.34 The gap in benefits received between rural and urban populations is wide. First, there is a wide variation in the subsidy assigned to different types of facilities. While the unit cost per hospital is T Sh 3,057, the unit cost of the lower level health care facilities does not exceed T Sh 1,002. Second, there is a wide difference in the use of health facilities. In general, the poorest segment of the population in rural areas tend to use  48 Chapter 2 health centers and Figure 2.8: Distribution of Health Subsidies Per Welfare dispensaries more, Quintile while richer urban households make much greater use of hospitals. 35 These two factors result 30 in an inequality on the 25 - . HOp1i per capita health ID20 . l Dispensaries subsidy received by r 15 g Population each quintile in the rural 10 * Total Health and urban areas. 5 | 1 2 3 4 5 Welfare Quintile Spending on Water Table 2.6: Distribution of Water Subsidies: Rural-Urban Differential Quintile Poorest Second Third Fourth Richest Total Per Capita Subsidy (T Sh) All Tanzania 647 703 961 1379 2565 1,251 Rural 533 478 601 825 931 641 Urban 1,817 1,725 1.964 2,443 4.089 2,829 Incidence of Water subsidies All Tanzania .52 .56 .77 1.1 2.1 1 Rural .83 .75 .94 1.29 1.45 1 Urban .64 .61 .69 .86 1.45 1 Ratio of percent of total subsidies accruing to the quintile relative to the percenrt population in the quintile. In the perfect egalitarian situation all the cells would have ones. 2.35 The distribution of the water subsidy also has inequities (Table 2.6). Twenty percent of the population with the lowest income receive just 10 percent of the subsidy, while the 20 percent in the highest quintile receive 41 percent. These results are driven by the unequal distribution of water subsidies across income groups. Less than 2 percent of households in the lowest quintile receive the T Sh 29,700 estimated annual subsidy, while 63 percent in the highest quintile, and 20 percent in the fourth quintile receive the subsidy. The level of subsidy varies hugely between rural and urban residents. While on average a person living in a rural area receives T Sh 641 per year, the urban counterpart receives almost five times more. Further, a poor rural resident (in the lowest quintile) receives T Sh 533 per year, which is about 13 percent of the amount received by a well- off urban resident.  Chapter 2 49 2.36 The benefits are more equally distributed among the rural population. While the poorest 25 percent in the rural areas receive 21 percent of the subsidy, the top 13 percent receive 20 percent. In the urban areas, the poorest 8 percent receive just 4 percent of the subsidy, while the top 44 percent receive 54 percent. F. ANALYSIS AND DISCUSSION 2.37 Given the weak fiscal situation, the challenge is to: (i) raise revenues and tighten aggregate spending, (ii) reduce government spending in selected areas, (iii) permit private delivery of services to the extent feasible, and (iv) restructure the expenditure program to provide adequate financing for critical development activities. The restructuring of public administration, together with an aggressive effort in revenue mobilization and debt relief, will be key to improving the unstable fiscal situation and channeling resources to social and economic services. 2.38 For the social sectors, a strategy for improving access to service delivery has been formulated for testing on a pilot basis. The emphasis is on using public resources efficiently and enlisting the role of the private sector to bring in more resources. Measures include shifting spending priorities to target low income groups and women; emphasizing basic education; enhancing the role of the private sector and local level agencies and communities; and providing options for choice among users to the extent feasible. 2.39 Expenditure efficiency could be improved further by adjustments within programs. Examples include more day secondary schools rather than boarding schools which are more expensive; and more support for school materials and other non-wage inputs. Reforms in the university could request contribution from the well-off students while providing scholarships based on skill shortages and equity, including gender. Other measures for education reform include budgetary reform to eliminate the fragmentation of spending among too many agencies and reduce over-employment.45 2.40 For the health sector, the main options are to encourage private sector delivery of services and to focus the state on rural health service delivery, to allocate more resources in favor of preventive care, especially maternity and child health, and to focus on the major killers, such as malaria, TB, and water-borne and water-washed diseases. 2.41 For water, simulations show that user fees for private connections could be raised and the current subsidy halved. This would generate approximately T Sh 7,003 million (US$14 million) which could be spent on expanding access. This would also improve the equity of the water program substantially. The results of the simulation show that the highest quintile would then receive 34.9 percent of the water subsidy, and the lowest 20 5 P. Penrose, "Review of Public Expenditures in the Education Sector - The United Republic of Tanzania". July 1992.  50 Chapter 2 percent of the population would receive 12.7 percent, compared to the current allocation of 41 percent and 11 percent, respectively. 2.42 Reforms in agriculture would need to emphasize more support for research and extension; water resource management; supportive rural road infrastructure; improved management of parks and game reserves; and development of rural financial markets. On the other hand, a reduced government role is expected in other areas. These include: (i) imposing limits on new government projects in various activities, including dairy, general livestock, cashew nuts, sugar, area development and cotton, among others; (ii) accelerating divestiture of state farms and parastatal crop processing facilities; (iii) phasing out government operations in agricultural input markets, especially agricultural chemicals; and (iv) restructuring cooperatives, phasing out their preferential access to inputs and markets, and ensuring viability under competitive conditions. 2.43 In the case of road infrastructure, more needs to be done to reach rural areas with agricultural potential. Reforms could also foster employment in construction by adopting labor intensive methods for suitable projects. This, together with support for developing local small scale contractors, would help the evolution of local capacity in rural road development and maintenance. Other options include, mobilizing more resources for self-financing for the transport sector parastatals (ports, civil aviation, etc.), institutional changes to improve management and accounting practices, more private sector role in providing transport services, and deregulation of tariffs. G. GOVERNANCE 2.44 Overview. This section will focus on the issue of governance with emphasis on accountability. Tanzania is undergoing fundamental changes in the system of development management and governance. The shift from a statist economy to a more market-oriented system is part of this change. Another aspect of the change is the political reforms leading to a multi-party democracy. Multi-party elections which were held at the local level in 1994 went well; this was followed by presidential and parliamentary elections in late 1995. Overall, the process has been peaceful; though some of the election results are being contested. There is tension in Zanzibar arising from disagreement over the election results. The new government has revived the reform effort and is moving quickly to take action in the area of fiscal management as discussed earlier. Nevertheless, the transition is unsettling. The entire system is in a state of flux. The capacity to manage the difficult transition is very important for preserving a cohesive society and attaining a stable macroeconomic situation for growth and development. 2.45 Accountability. One of the more sensitive issues of governance which has emerged in recent years in Tanzania is public accountability. One aspect relates to overlapping responsibilities and inaction in enforcing agreed fiscal and financial codes. Recent efforts to contain expenditure growth did not succeed. Examples include: (i) laxity in enforcing the authority given to banks to reject checks which are not adequately covered by funds; (ii) delayed termination of retrenched workers after they have been  Chapter 2 51 compensated; and (iii) lapses in observing approved expenditure ceilings. Lack of progress in these areas contributed to the budget overrun. The expenditure float, for example, has grown from 3 to about 5 percent of GDP between FY87 and FY94. 2.46 The other aspect of accountability is illustrated by recent episodes of tax exemptions and tax evasion. This issue cannot be separated from fiscal stability, growth and poverty reduction. First, as the findings of government investigations into revenue performance indicate, the lapses in fiscal performance were costly to the Treasury. The import revenue losses in 1994 alone could have covered the budget for the agricultural sector for three years.46 Second, it has consequences for equity and poverty reduction to the extent the fiscal laxity results in transfer of resources to the higher income groups as government investigations into identified cases of tax evasion have shown. The use of regressive inflationary taxation to finance the budget because of revenue loss also has adverse effects on equity. In addition, the shortage of local funds constrains the implementation of transfer programs which could benefit low income groups. 2.47 Lax accountability also undercuts the development effort, as low levels of local resource mobilization constrain the absorption of project aid and the expansion of services. For example, based on a local counterpart fund ratio of 25 percent, an extra revenue of T Sh 610 million (about US$1 million) channeled to the development budget would provide for leveraging an additional aid of US$4 million for capital formation. The officially estimated revenue loss from import taxes alone in FY94 was T Sh 70 billion. The less the capacity to leverage aid resources, the less the scope for investment in new projects necessary to make more inroads into poverty reduction. In addition, lax fiscal accountability contributes to public perception of erosion in the authority of the state. This perception is reinforced to the extent Government is unable to generate internal funds to cover its daily functions and to deliver basic services. 46 For example, the Presidential Statement on the issue indicated that lost revenue from import taxes, including exemptions, for the past financial year (January to September 1994) was T Sh 70 billion, about 4-5 percent of GDP. This is more than the allocations planned for agricultural development for the 3 year period FY94-FY96.  52 Chapter 2 SECTION II. REVENUE MOBILIZATION H. LOCAL FUNDS AND TAX REFORM Introduction 2.48 This section will examine tax issues. As has been noted, improvements in revenue mobilization (together with expenditure reforms) are necessary to finance critical social programs and economic services adequately while attaining a stable fiscal environment. Also, while Tanzania will continue to need substantial aid support, the ability to absorb the aid and to use it productively will depend on the availability of domestic resources. 2.49 Tanzania's revenue to GDP ratio is low. The current official average tax ratio is about 14 percent of GDP, but GDP is underestimated, so that the effective revenue ratio is about 8-10 percent, well below the average for Sub-Saharan African countries.47 Revenues are not responsive to income growth and inflation. Estimates of income tax elasticity are in the range of 0.7 to about 0.94.48 Among the various reasons for the low tax effort, the use of the tax system as an instrument for fiscal incentives has expanded at the expense of revenue collection. Also, until recently, payments of assessed import duties could be staggered, which contributed to the accumulation of arrears and affected revenue growth relative to inflation. There is a need to restructure the inherited institutions for tax administration, which used to handle transactions for a few state monopolies and now have to accommodate more clients in a market economy. The economy has large agricultural and informal sectors, which are yet to be brought into the tax net. 2.50 The Government has made repeated efforts, including the Presidential Commission of Inquiry into Taxation and Expenditure (1992), to reform the tax system and raise revenue. Despite efforts, revenue improvements have been far less than expected, and the official tax to GDP ratio is lower now than it was in FY92. The main culprits are failure to expand the tax base and weak tax administration. The initial success of the Government in generating a surplus of revenues over recurrent expenditures for the first time in FY92 (see Figure 2.3) led to a mis-perception that the path to macro-fiscal stability has been secured. Thus, in FY93, the Government began to emphasize the use of the tax system for providing fiscal incentives for investments. This Tanzania's effective tax to GDP ratio is higher than that of Uganda; however, Uganda has much lower Govermuent expenditure; deficit financing is close to nil, and inflation is less than 5 percent. Tanzania's official tax to GDP ratio averaged 14 percent in FY93-FY95, and is less than the average for LDCs and for Sub-Saharan Africa of about 18-20 percent. 48 Elasticity estimates vary by source of tax. Company tax, for example, is income elastic, while all the other taxes are inelastic. See N. Osoro, "Revenue Productivity Implications of Tax Reform in Tanzania", African Economic Research Consortium, September 1993.  Chapter 2 53 led to the proliferation of tax exemptions and lowering of tax rates. However, the Government was unable to strengthen tax administration, to expand the tax base and to enforce compliance to the lower tax rates; also, the exemptions were abused. In the event, the tax base was not expanded adequately enough to compensate for the revenues lost from the tax simplification and the exemptions. In FY93, the recurrent revenue surplus turned deficit, and this has persisted ever since. Tax Structure 2.51 Table 2.7 summarizes the relative importance of various taxes.49 The major tax revenue earners in FY95 were import duties and income taxes, which together accounted for about 60 percent of total taxes and 7.4 percent of GDP. The relative importance of income taxes declined from about 5.3 percent of GDP in FY81 to 3 percent in FY90, rising to 3.8 percent of GDP presently. Company taxes are the most important taxes among the income tax group, accounting for over 20 percent of total income taxes, followed by PAYE. In the 1980s, PAYE accounted for about 7 percent of total taxes compared to only about 4 percent in FY95. Import taxes are next in importance, accounting for about 29 percent of total taxes in FY95. These comprise customs duty (16 percent) and sales and excise taxes (13 percent). Domestic sales and excise taxes contribute about 26 percent to total tax revenue, equivalent to about 3.2 percent of GDP in FY95; this represents a decline from 4.7 percent of GDP in FY90. Erosion of domestic sales and excise taxes is a major reason for the lower tax to GDP ratio during FY90- FY95. Other taxes accounted for about 14 percent. In calculating the ratios, the GDP figures were adjusted to fiscal year basis to correspond with the fiscal accounts.  54 Chapter 2 Table 2.7: Central Government Tax Revenue, FY90-FY94 FY90 FY91 FY92 FY93 FY94 FY 95 (in millions of T Sh) Total Revenue 94655 133238 173566 164110 242444 331200 Tax Revenue 81,471 118,258 153,356 146,420 220,359 279,940 Import Taxes 22,960 31,504 38,617 3 1,637 56,651 81,318 Customs duty 11,930 17,321 21,103 16,288 28,404 46,723 Sales and excise tax on imports 11,030 14,183 17,514 15,349 28,247 34,595 Domestic sales and excise taxes 31,598 44,988 57,884 46,117 70,388 72,645 Income taxes 20,195 32,413 40,143 45,455 58,505 86,684 Othertaxes 6,718 9,353 16,712 23,211 34,815 39,293 (in percent of GDP) Total Revenue 14.1 15.7 16.8 12.9 15.0 14.5 Tax revenue 12.1 14.0 14.8 11.6 13.7 12.3 Import taxes 3.4 3.7 3.7 2.5 3.5 3.6 Customs duty 1.8 2.0 2.0 1.3 1.8 2.0 Sales and excise tax on imports 1.6 1.7 1.7 1.2 1.8 1.5 Domestic sales and excise taxes 4.7 5.3 5.6 3.6 4.4 3.2 Income taxes 3.0 3.8 3.9 3.6 3.6 3.8 Othertaxes 1.0 1.1 1.6 1.8 2.2 1.7 Source: Ministry of Finance database. 1. TAX REFORM 2.52 The objectives of tax reforms are to improve revenue yield by simplifying the system and broadening the base; to correct the effect of tax distortions on production and consumption by rationalizing fiscal incentives; to simplify tax administration; and to ensure equity among income groups. However, the principal objective is to generate revenues to finance critical development activities. Tax Administration and Compliance 2.53 The most urgent tax reform issue is to reorganize the system of tax administration which has lagged behind reforms taking place in the economy, and to address issues of skill requirements, motivation, and operational autonomy. Among the various steps taken recently are: (i) gross instances of tax evasion have been investigated, and concerned individuals have been identified for legal action;50 and (ii) a Revenue "Audit Report on Import Tax Evasion and Exemptions" URT, Office of the Controller General and Auditor General, May 1995.  Chapter 2 55 Authority Act was passed in FY95 to establish an autonomous and more professional institution for tax administration. Top priority reforms include: * Making the Revenue Authority operational by providing it with new guidelines, equipment, personnel policies and training, qualification standards, updated compensation and other incentives. This includes the need to reconstituting the Customs Department and the other key revenue departments. * Improving personnel policies and compensation for tax administrators. * Introducing a tax payer identification number system, and computerizing tax collection and auditing procedures at central, regional and local levels. * Updating tax records, raising interest penalty for tax default and strengthening legal sanctions against tax evasion. Customs Duties 2.54 Customs duties have been simplified significantly since the mid-1980s from over 20 rates and a maximum rate of over 100 percent to just four rates and a maximum rate of 50 percent in FY95. In FY95, the nominal duties consist of four non-zero rates: 5, 20, 30, and 50 percent. Serious leakages through tax evasion and discretionary exemptions have resulted in much lower effective duty rates.51 In 1989, total imports stood at T Sh 139.2 billion. Out of this value, exempted imports accounted for T Sh 51.0 billion, compared to T Sh 88.2 billion of fully taxed imports. Thus, 36.6 percent of total import value was exempted from taxes, excluding further losses from tax evasion and accumulation of arrears. 2.55 The Government intensified its efforts in FY94-FY95 to improve the performance of customs. For example, pre-shipment inspection agencies were hired to assist with assessment, collection and documentation of customs duties. Steps were taken to close down about 100 bonded warehouses, which were channels for tax evasion; and negotiations are to be undertaken to monitor Zanzibar's transit trade. Revenue performance improved in FY94 and FY95 in nominal terms, though they were below expectation relative to GDP. More reforms were adopted in the FY96 budget, including the elimination of duty exemption for capital goods imports and introduction of a duty of 5 percent on them, and (ii) strict monitoring of transit goods to curb tax evasion. Among the top priority reforms are to: 5! According to Section 7 of the 1976 Customs Tariff Act, the Minister of Finance has discretionary powers to exempt goods and individuals from full or partial payment of duty. The discretionary power covers all exemptions, including exemptions granted on a case by case basis or according to the Minister's assessment and desirability.  56 Chapter 2 * Enforce elimination of all discretionary exemptions, for example, on imports and profits; limit statutory exemptions, and rationalize duty rates. * Enforce the closure of undesirable bonded warehouses, impose restrictions on the use of others, and develop a framework to guide the operation of those which are needed. * Enforce agreed measures to monitor transit trade. * Restructure the Customs Department: update its facilities, equipment, pay and personnel policies and operational procedures to make it more effective. Sales and Excise Taxes 2.56 The sales tax has differential rates for services (5, 10 and 15 percent) and for goods (25 and 30 percent). Both imported and domestic goods bear the same rates for the most part; however, some goods bear specific rates. The tax base is narrow and focused mainly on manufactures. Excise duties cover petroleum products, alcoholic beverages, cigarettes, cosmetics and perfumes, and soft drinks. The top priority reform issues for sales and excise taxes are to: * Expand the sales tax base to include goods which are currently excluded, for example, building materials; and limit sales tax exemptions. * Simplify the sales tax and unify rates for the same products, such as beer, irrespective of origin; and convert specific and ad valorem rates on excises to a single ad valorem rate to simplify tax administration. * Rationalize the differential tax rates between goods and services. Income Tax 2.57 Corporate income tax is the single most important source of income tax. Until recently, resident companies were taxed at 35 percent and non-residents at 40 percent. Also, there have been various discrepancies in treating income from similar sources. For example, parastatals have preferential treatment in getting implicit subsidies through tax waivers. Corporations also had other provisions, including: (a) tax holidays; (b) an initial investment allowance of 20 percent partly to offset the original cost and depreciation allowances; and (c) privileges in carrying forward losses. 2.58 About 250,000 individuals, other than those under PAYE, are covered under the personal income tax. PAYE is easy to administer because it is deducted at source from salaries. Before the FY96 budget, the marginal income tax rates were between 5 and 30 percent, and the personal income tax structure had only six income tax brackets, with the  Chapter 2 57 top rate applied to taxable incomes in excess of T Sh 25,000 per month. The main exemptions were T Sh 150,000 on interest income from banks and from housing and transport allowances of civil servants, parastatal employees, and other specified groups. There are minimal tax credits for various purposes, including marriage and child relief. 2.59 Effective income tax rates have been regressive. According to Section 5 of the Income Tax Act of 1973, all incomes received from employment are subject to income tax. 1lowever, most fringe benefits and allowances received by the civil servants, parastatal staff, and private employees are not taxed in practice. Consequently, the income tax, excluding the fringe benefits, is regressive. These benefits include: transport, fuel, entertainment, responsibility and housing allowances; free electricity and water, and paid telephone. 2.60 Past income tax reforms focused on reducing rates for the upper income groups and broadening the base, while neglecting personal allowances and inflation adjustments, thereby negatively affecting the lowest income groups the most. Some income sources were taxed at different rates. Dividends, for example, were taxed at a withholding tax rate of 20 percent and capital gains at 20 to 30 percent. Poor record keeping makes it difficult to trace tax evaders. Among top priorities for income reform are the following: ; * Apply the same income tax to resident and non-resident corporations, and apply the same maximum income tax to both personal and corporate income taxes. * Expand the personal income tax base to cover all types of remuneration (including all allowance and fringe benefits). * Phase out income-tax-based fiscal incentives, including tax holidays. 2.61 Efforts were made in the FY96 budget to address some of these concerns and to reform the system. Some of the allowances are now included in the taxable income base and lower income groups have received more exemptions. Tax holidays are to be phased out. However, the number of income slots have been increased for tax purposes, which will make administration more complicated. 52 Estimates based on a sample from one parastatal show that the tax paid by a top cadre with base salary T Sh 42,150 is 21 percent; but if fringe benefits were included, the total eaming rose to T Sh 345,520, resulting in an effective tax payment of 2.6 percent. A middle level employee with a base salary of T Sh 32,270 paid a nominal tax of 18.6 percent on the tax base salary, but 2.4 percent on a total of T Sh 241,986, including fringe benefits. The low cadre, on the other hand, paid a nominal rate of 16.8 percent on a base salary of T Sh 12,280; but the actual income tax payment was 7.5 percent, based on a total of T Sh 27,760, including fringe benefits.  58 Chapter 2 Preparations for the Value Added Tax (VAT) 2.62 There has been an increasing acceptance of the principle of VAT in developing countries and Tanzania is no exception. The VAT is expected to expand the consumption tax base and consequently raise more revenue, and the Government is expected to make preparations to introduce it in 1996. Most countries which have introduced the VAT effectively have done so with thorough preparation, making changes in stages and adopting a limited VAT. Some of the important steps are to: (i) eliminate ambiguities in the existing taxes and unify rates for the same product, (ii) make adequate preparation for introducing the VAT by training accountants and tax officials, educating tax payers, and (iii) develop brochures and information centers to inform the public. With the apparent weak accounting system, lack of record keeping among individual businesses, and lack of culture of issuing and demanding receipts, the adoption of the VAT requires a lot of background work and preparation. Other steps are to introduce a credit system, or extend an existing credit system and unify it; and to base tax valuation on invoices. a  Part II Growth, Incomes and Household Welfare   3. INCOMES, INEQUALITY AND POVERTY A. OVERVIEW 3.1 The previous chapters examined progress made in implementing liberalization policies since the mid-1980s, and the impact of those policies on overall macroeconomic performance. Significant progress has been made in the abolition of controls on pricing, purchasing, marketing and export of agricultural inputs and outputs; liberalization of trade, exchange rate and industrial policies; and the rehabilitation of a substantial part of road transport infrastructure. This chapter will examine developments in household income, expenditures and welfare during the reform period; and it will discuss the lessons learned for future development policy in Tanzania. 3.2 An important finding is that household income and expenditures have grown, and the incidence of poverty has declined. About 65 percent of the rural population was living in poverty in 1983, compared to about 51 percent in 1991. More recent studies, including the 1993 household survey and the participatory poverty assessment conducted in 1995, confirm similar improvements in household welfare and incomes. However, not all households appear to have improved their living conditions. Rural poverty appears to be very fluid. Despite steady increases in rural consumption since 1983 and the general trend of declining poverty, the incidence of rural poverty based on expenditures increased between 1993 and 1995, resulting in a level only slightly lower than in 1991. Most of the poor are rural dwellers, though urban poverty is also a concern. Rural households account for 92 percent of the poor; they tend to have large families and are more likely to be illiterate and isolated from markets. 3.3 Literacy has increased; but infant mortality and life expectancy show little improvement, while primary school enrollment and access to safe water have declined. The social indicators started to slip with the decline of the economy and collapse of revenue in the mid-1970s to the early 1980s; there were restrictions on the private provision of social services until recently. Also, there are lags in the response of social indicators. Current consumption tends to lead social indicators, which depend in part on the built up stocks of physical and human capital. Sustained improvement in many social indicators requires reallocation of government expenditures for those objectives and expanded funding which is difficult in an environment of persistent fiscal crisis. 3.4 Generally, cash crop producers, households close to markets and those connected to transport infrastructure have gained more from the reform process. The relatively well-off areas include those with the most intensive agricultural systems, especially the (i) coffee, maize and legume systems of the Southern Highlands; and (ii) coffee, banana and dairy systems of the North. On the other hand, the relatively poorer regions include  60 Chapter 3 the (i) agro-pastoralist zone of the semi-arid central plains and (ii) the cashew and cassava areas of the North. 3.5 Furthermore, one of the most important features of the rural economy is the association between the impact of education and the impact of infrastructure on household income. The beneficial effect of proximity to roads on welfare increases with higher level of educational attainment. Thus, improvements in educational outcomes will raise households' welfare both directly and indirectly, through households' increased ability to take advantage of infrastructure. Improvements in educational attainment and programs of infrastructural improvements are distinctly more effective in impacting household welfare when done jointly than when implemented separately. Thus, it is recommended that the Government give priority to investing jointly in education and in rural infrastructures with emphasis on areas with growth potential. 3.6 It is found that growth in rural areas is four times more efficient than growth in towns, excluding Dar es Salaam, for reducing aggregate poverty; and that growth in Dar es Salaam is even less efficient than growth in other towns. This emphasizes the importance of rural development for the future of Tanzania; this requires increased productivity in agriculture. Given the importance of increasing productivity in agriculture to growth and poverty alleviation, evidence of constrained supply of agricultural inputs is troubling. It is thus fundamental to maintain the liberalization of agricultural input markets. B. INCOMES AND EXPENDITURES 3.7 This section examines household expenditures, income and inequality, which are major determinants of poverty. The analysis is based on four different household surveys. These are the Rural Household Survey conducted in September 1983, the Comell/ERB Survey conducted in 1991/1992, and the Human Resource Development (HRD) Survey conducted in October 1993. To complement the information from the household surveys, a Rural Participatory Poverty Assessment (PPA) Survey was conducted in March 1995.53 The PPA focused on assessing poverty as perceived by the households themselves. Table 3.1 summarizes the various household surveys based on income and expenditure data, and the scope of coverage in each case. 53 The rural household survey covered 498 households in four regions in the rural areas only. The Cornell/ERB Survey covered 477 rural and 570 urban households on the mainland only. The HRD Survey covered 5,184 households in all regions, including Zanzibar. The Participatory Poverty Assessment covered 768 households in the rural areas of mainland Tanzania. (See Appendix to Chapter 3 for methodological detail.) Unless otherwise stated all charts and figures are derived from the 1993 HRD Survey.  Chapter 3 61 Table 3.1: Some Household Surveys Available in Tanzania 1983 1991 1993 1995 Welfare Indicator Income Income Expenditures Expenditures Expenditures Scope of Survey Rural Tanzania All Tanzania All Tanzania Rural Tanzania 3.8 There is strong evidence from these household surveys indicating real per capita income and expenditure growth in Tanzania prior to 1995.54 For example, Young used the Heston and Summers data and OECD data sets to estimate that real per capita output grew by 2.3 percent per annum from 1960 to 1985.55 Sarris and Tinios conclude: "Comparing the average per capita total consumption expenditures in 1991 with those of 1976for each interval, it appears that within every interval in both rural and urban areas the 1991 per capita expenditures are higher than the average in 1976."56 The only exception to this finding is that expenditure was not found to increase for the lowest income interval.57 After a period of stagnation in the 1970s, rural income seemed to have grown sharply with the liberalization of agricultural markets and recovery of food, cash crop and export crop production. More recent household data indicate real rural per capita expenditure growth of 3.2 percent per year, and an overall (urban and rural) per capita expenditure growth of 5.5 percent per year between 1991 and 1993. Also, the 1995 PPA Survey indicates real rural per capita expenditure growth of 3.5 percent per year. These recent growth estimates are based on an inflation rate of 34.2 percent, which is higher than the official inflation rate and parallels broad money growth. These 54 D. Bevan, D. Collier, and J. W. Gunning, "Agriculture and the Policy Environment: Tanzania and Kenya", (Paris: OECD, Development Centre Studies, 1993); D. Booth, F. Lugangira, P. Masanja, A. Mvungi, R. Mwaipopo, J. Mwami, and A. Redmayne, "Social, Cultural, and Economic Change in Contemporary Tanzania: A People-Oriented Focus", Report to SIDA, 1993. See A. Young 1994, "Lessons from the East Asian NICs: A Contrarian View," European Economic Review, Vol. 38: 964-974; and R. Summers and A. Heston, 1991, "The Penn World Table (Mark 5): An Expanded Set of International Comparisons, 1950-1988", Quarterly Journal of Economics, May: 327-368. The exchange rate data compiled and published by Heston and Summers is based on a Purchasing Power Parity (PPP) valuation of international currencies and thus controls for differences in consumption baskets between countries. 56 A. Sarris and P. Tinios, "Consumption and Poverty in Tanzania in 1976 and 1991: A Comparison using Survey Data" (Cornell University Food and Nutrition Policy Program, 1994). This finding is similar to that of L. Ferreira, 1994, "Poverty and Inequality During Structural Adjustment in Tanzania", The World Bank Research Project, Income Distribution During Transition, Research Paper Series, No. 8, July, when comparing rural income in 1983 and 1991 that the poorest of the poor do not seem to have benefited from growth. However, as that author notes, the result may be driven by the small number of data points at the bottom of the distribution. 58 The inflation rate of 26.5 percent per year reported by the National Consumer Price Index (NCPI) indicates an annual growth in real rural per capita expenditure of 9.6 percent during 1991-93.  62 Chapter 3 growth rates based on household surveys Figure 3.1: Expenditure per Adult Equivalent, 1993 deviate from macro- economic growth estimates, which are much lower, mainly 450000 400000 because the former 350000 includes informal 300000 activities which are not 250000 captured by the official I 200000 statistics. 150000 100000 3.9 Despite recent 50000 growth, the overall 0 Poorest Betttr- All Rural Urban Dar es level of expenditure Off Sol am and consumption is still low. According to household survey estimates from 1993, national and rural per capita expenditures were equivalent to US$249 and IJS$193 respectively. Figure 3.1 displays expenditures per adult equivalent59 for different categories of Tanzanian households. It shows first the clear trend toward lower expenditure and welfare levels as one moves from Dar es Salaam to other urban areas, and then to the countryside. Second, it shows the large difference between expenditure levels of the richest and poorest quintiles.60 Inequality 3.10 Income distribution is quite uneven in Tanzania. Based on the 1993 survey, the average adult equivalent expenditure of the richest quintile was 6.24 times greater than that of the poorest quintile. Among rural households, the ratio was lower (5.25), increasing to 6.35 among urban households. Among the households in Dar es Salaam, the ratio between expenditure levels per adult equivalent in the top quintile and in the second quintile was 4.94. The fact that no individuals in the bottom quintile were found in Dar es Salaam reflects the normal difficulties of capturing survey data on the urban poor, who are homeless or have no fixed residence. An "adult equivalent" is a standardized measure which takes into account the different nutritional needs of people of different genders and ages, and economies of scale/size in the household. Exact estimates, both for expenditure per adult equivalent and for per capita expenditure, are reproduced in Appendix to Chapter 3. 60 Throughout the report the poorest refer to those belonging to the first quintile of the expenditure distribution, and the richest to those belonging to the fifth quintile of the expenditure distribution.  Chapter 3 63 3.11 In 1993, the wealthiest 20 Figure 3.2: The Distribution of Income by percent of Tanzanians (fifth quintile) Quintile, 1993 controlled 45.4 percent of total income, while the poorest 20 percent (first quintile) controlled only 6.8 percent (Figure 3.2). The inequality Poorest between rural and urban areas is also a 2nd Quintile matter of concern: while on average a rural Tanzanian spends the equivalent Better-off r of US$193, the counterpart living in Dar es Salaam spends on average US$587. 4th Quintile 3.12 Although people in the top income quintile are better off than those in the lowest quintile, they are still not in a position of wealth. The per capita income of those in the top 20 percent of the income distribution is equivalent to US$540, which is hardly a high income when measured against international or regional levels. For example, the per capita income for low and middle income Sub-Saharan African countries was US$520.6' 3.13 It is not easy to deternine precisely the pattern of inequality, since the estimates based on incomes and expenditures are not directly comparable across all years. The Gini coefficient for mainland Tanzania at various periods (Table 3.2) suggests that inequality may have deteriorated.62 Based on incomes, Table 3.2 indicates that inequality probably deteriorated between 1983 and 1 991.63 Estimates of consumption inequality based on expenditures indicate that the situation is no worse in 1995 than in 1991. However, the data for the two most recent years suggest a deterioration between 1993 and 1995. This indication of increased inequality in the rural areas in the more recent years is consistent with the household's own perception found by the 1995 PPA Survey (para. 3.14). 61 See World Development Report. World Bank, 1995. 62 The Gini Coefficient is a measure ranging from zero to one (with zero denoting complete equality) which captures all the information from the income distribution. 63 The estimates for the Gini coefficient in 1983 and 1991 may seem too high, but they are not unusual because they are based on income rather than expenditure. World Bank. 1993, "Tanzania: A Poverty Profile", Report No. 12298-TA, Washington D.C.  64 Chapter 3 Table 3.2: Gini Coefficient for the Distribution of Income in Tanzania, 1969-93 1983"7 1991a' 1991"' 1993D' 1995D' Change in (1) (2) (3) (4) (5) Inequality Rural .52 .72 .60 .35 .52 U All Urban -- -- .46 .42 -- Tanzania -- -- .57 .41 - a Using income. b Using expenditures. The results in column (I) are comparable with column (2). The results in columns (3), (4) and (5) are comparable with each other. 3.14 While household income and expenditures have steadily grown over the last decade, some groups in the bottom were unable to take advantage of the changing situation and may be falling further behind. During the 1995 PPA Survey, households were asked to rate whether they were better- or worse-off compared to 10 and four years ago (Table 3.3). The majority, including the poorest quintile, felt that they were definitely better-off now compared to 10 years ago (1985), when "shops were empty". However, a minority felt they were better off now compared to four years ago (1991). Important differences exist by poverty level. Overall, 49 percent felt that they were worse-off now than four years ago, compared to 35 percent who felt they were better-off. Among the poorest quintile, more than half (53 percent) felt that they had become poorer in the last four years. For the well off, 63 percent perceived an improvement in their own standards of living. Also, the majority of the rich and the very rich expected to do even better in the future, while the majority of the poor expected to do worse in the future. Table 3.3: Household's Perception of Changes in Inequality, 1985-1995 (percent) Compared to 10 years ago Compared to 4 years ago Poorest Richest All Poorest Richest All Quintile Quintile Quintile Quintile Better-off 51 37 47 35 63 35 Same 10 0 14 12 0 16 Worse-off 40 63 39 53 37 49 Total 100 100 100 100 100 100 Source: PPA Survey, 1995. 3.15 Various factors may account for the differences in the perception of well being between 10 years and four years ago. It appears that the deregulation and liberalization and the recovery of the economy, especially agriculture, from the highly depressed levels preceding reforms benefited the population broadly. The liberalization of agricultural markets was central to the reform process, and this led to the growth of production of food, cash crops and export crops. The availability of imported consumer goods from trade reforms reinforced this initial process. However, some households may have been able to gain more in recent years. Some factors are access to inputs, transport facilities, financial services and market information. Results from the 1995 PPA Survey indicate  Chapter 3 65 that although overall there was a slight increase in the use of inputs between 1990/91 and 1993/94, use declined among the poor and increased among the rich. In addition, the well-off who sell larger quantities of produce, and have greater access to transport facilities, sell primarily to private traders (51 percent). The poor, by contrast, use primarily local markets (25 percent) and cooperatives (27 percent), and only 15 percent use private traders. Village group discussions during the 1995 PPA Survey revealed that whereas rich farmers negotiate selling prices with private traders, the poor sell in small quantities and at lower prices. Another factor is vulnerability to drought. For example, in 1994, drought conditions affected agricultural production adversely, and the poor are the least able to cope. Also, the macroeconomic situation deteriorated over the last two years with an upsurge in inflation, which may have eroded the living conditions for the poor. C. POVERTY UPDATE Definition of the Poverty Line 3.16 Poverty has various dimensions. This section examines the poverty question based not only on consumption deprivation, but also on the 1995 PPA Survey based on an alternative participatory methodology. The first requirement for measuring consumption poverty is to select and justify a poverty line, i.e., the level of expenditure or income below which people or households are considered poor. Currently, no official poverty line exists for Tanzania. To evaluate the trends in rural poverty between 1983 and 1991, the poverty line used was based on a per capita income of one US dollar per day, translated into Tanzanian Shillings in Purchasing Power Parity terms. This is discussed in the next paragraph. To evaluate trends in poverty for the other years (1991, 1993 and 1995) an absolute poverty line of T Sh 46,173 in 1991 prices updated for inflation was used. Another poverty line of T Sh 114,187 (1993 prices), estimated by Wagao, was also used.64 Rural Poverty Between 1983 and 1991 3.17 The household budget surveys from 1983 and 1991 provide an opportunity to evaluate the evolution of living standards over the period for rural areas. Although limited to the rural population, the analysis is justified by the fact that the rural population comprises about 70 percent of the population in Mainland Tanzania; and that the national household survey that was conducted in 1991 indicates that poverty is mainly a rural phenomenon. 3.18 Sixty-five percent of rural Tanzanians lived below the poverty line in 1983, compared to 50.5 percent in 1991. This corresponds to a 30 percent reduction in poverty, and is enough to reduce the population living in poverty from 10.8 million to 9.7 million 64 See Kigoda and Mwisomba, "Defining a Poverty Line and Alternative Measures of Standards of Living and Poverty". Background paper prepared for the Poverty Assessment. 1995.  66 Chapter 3 over the period. Over the same period, the number of people whose incomes were above the poverty line rose from 5.7 million to 9.5 million. Recent Evidence on the Incidence of Poverty 3.19 The overall (rural and urban) poverty situation is better now than it was in the early 1980s. Table 3.4 shows estimates of the incidence of poverty for different locations and different price deflators in 1993. Since it is conventional to use more than one poverty line, the analysis is based on estimations of three poverty lines. Line A is equivalent to T Sh 73,877 and was based on the 1991 poverty line of T Sh 46,173, updated for inflation, using the official inflation rate of 26.5 percent. Line B is equivalent to T Sh 83,111 and was also based on the 1991 poverty line, updated for inflation using the annual inflation rate of 34.2 percent, which parallels broad money growth.65 The two different annual inflation rates were used to check the sensitivity of the poverty line.66 A third poverty line of T Sh 114,187 (Wagao Line) based on a study by Wagao and adjusted for adult equivalent values was also used. Based on Wagao's poverty line of T Sh 114,187, which corresponds to 78 percent of rural average adult household expenditures, poverty was estimated in 1993 to be 41.6 percent in Mainland Tanzania, 49.7 percent in rural Tanzania, 24.4 percent in other urban towns, and 2.9 percent in Dar es Salaam.67 The depth of poverty-a measure of how far the average income of the poor is from the poverty line-is widest for rural Tanzania at 15.7 percent and lowest for Dar es Salaam at 0.38 (based on Wagao's poverty line). This means that those poor who live in Dar es Salaamn have, on average, incomes closer to the poverty line than those who live in the rural areas. The various levels of poverty in 1993 compare favorably with poverty in 1991 as shown in Table 3.5. 65 For example, between November 1990 and November 1991, open market prices for food rose 75 percent, while the official inflation rate was 22 percent (Collier and van der Brink, World Bank internal memo). Given the share of food in the National Consumer Price Index, this would require a deflation in the nonfood component of 75 percent. The official inflation rate is clearly underestimating price increases over the period. 66 The 26.5 percent inflation rate corresponds to the national consumer price index, while the 34.2 percent inflation rate corresponds to the growth rate of broad money. 67 Other studies also confirm the decline in the incidence of poverty. For example, Kigoda and Mwisomba estimate poverty in 1994 to be 42.7 percent in Mainland Tanzania (Kigoda and Mwisomba, op. cit.  Chapter 3 67 Table 3.4: Poverty Measures by Location and Poverty Line, 1993 Location Poverty Line Poverty Line Head Depth (in T Sh) a as % of Count (PO) (PI) Average Adult Expenditure Rural Tanzania Wagao 78 49.7 15.7 Line A 57 27.8 6.9 Line B 51 21.2 4.7 Urban Tanzania Wagao 50 24.4 6.1 Line A 36 9.9 2.2 Line B 32 6.9 1.5 Dar es Salaamn Wagao 27 2.9 0.38 Line A 20 0.21 0.06 Line B 18 0.21 0.04 Mainland Wagao 62 41.6 12.7 Tanzania Line A 45 22.3 5.5 Line B 40 16.9 3.8 a/ Wagao poverty line T Sh 114,187; Line A: poverty line of T Sh 46,173 in 1991, assuming inflation rate that parallels the growth rate of broad money; Line B: Poverty line of T Sh 46,173 in 1991 assuming inflation rate corresponds to the national consumer price index. Source: HRD Survey, 1993. Table 3.5: Evolution of Poverty, 1983-1995 1983-1991 1991-1993 1993-1995 1983-1995 Rural All Tanzania 3.20 The most recent PPA data for 1995, based on expenditures, yield a rural poverty index of 58.6 percent,68 which is slightly lower than in 1991, but higher than 1993. It has been noted earlier that while the evidence of average expenditure growth is strong, some groups in the bottom income level, who are isolated from markets and facilities and are characterized by lower investments in human capital, are unable to take advantage of the changing situation and may be falling further behind. Thus, despite higher expenditure levels on average in the more recent years, the deterioration in equality noted earlier was enough to off-set these gains, with adverse effects on poverty. Thus, overall (Table 3.5), the incidence of poverty declined between 1983 and 1991. It was better in 1993 than in 1991, but it appears to have deteriorated between 1993 and 1995. 68 This is based on updating the poverty line of T Sh 114,187 for inflation.  68 Chapter 3 Spatial Mapping of the Poor 3.21 Results from the 1991 and 1993 Figure 3.3: Distribution of Poverty by surveys show that poverty is mainly a rural Location, 1993 phenomenon. Rural inhabitants and people who rely on agriculture as their main source of income are far more likely than others to be poor and more vulnerable to exogenous Urban DSM shocks. Using any of the different poverty 8% 0 10% lines presented, the incidence of poverty is at least twice as high in rural areas as in urban areas other than Dar es Salaam, and many times higher than the incidence in Dar es Salaam itself. People whose main source of income is their farm are five times more likely to be poor than those who receive a Rural wage from the public or private sector. Ninety-two percent of the poor live in rural villages, 8 percent in urban areas outside Dar es Salaam, and just 0.1 percent in Dar es Salaam. These results are even more striking than those obtained in 1991, at which time 85 percent of the poor lived in rural areas, 13 percent in urban areas outside Dar es Salaam, and 2 percent in Dar es Salaam (Figure 3.3). 3.22 A review of the distribution of income by Figure 3.4 Distribution of Population by Welfare Level, location and income quintile 1993 in Figure 3.4 also clearly indicates that a 70 disproportionate share of the 602 rural population are in the 50 bottom quintile of the income 7 oAI distribution. For example, 48 , 40- percent of people living in 30- g Urban rural areas fall into the two DSM bottom quintiles, compared to 2 - about 23 percent of the 10. population living in other o towns, and 2.5 percent of the 1 2 3 4 5 population living in Dar es Welfare Quintile Salaam. However, for the __ richest quintile, 67.4 percent of the population is living in Dar es Salaam, compared to 30 percent of residents of other towns. A mere 13.4 percent of rural residents are in the top (most affluent) quintile. 3.23 However, urban poverty may be underestimated in household surveys because of the presence of recently-arrived, migrant households and the unplanned nature of settlements. To gain more insight into the nature and causes of urban poverty, a rapid  Chapter 3 69 assessment was conducted in three slum areas of Dar es Salaam: Buguruni, Vingunguti, and Manzese. Overall, 120 household heads were interviewed, all of whom had been identified as poor by local leaders contacted through the ward executive officers. The responses indicate the importance of the informal sector for the poor and the insecurity of their income. Forty-seven percent of household heads described themselves as self- employed (compared to about 15 percent of household heads nationwide and 22 percent in urban areas). Just 6.7 percent of household heads were in a formal job, and 7.4 percent sought income by casual labor. Twenty-seven percent of all household heads were women, higher than the average for urban areas of 16 percent. Regional Mapping of the Poor in 1993 3.24 The decomposition of poverty by region shows large variations in welfare across the country. Regions were ranked based on an aggregated welfare index comprising several non-monetary indicators: infant mortality, under five mortality, stunting and wasting indicators, incidence of poverty, access to water, household size, population per doctor, population per nurse, illiteracy, and educational attainment. The results are shown in Figure 3.5. A higher score indicates greater welfare. On the basis of that index, Kigoma, Lindi, Mtwara, Rukwa, and Ruvuma could all be classified as especially poor. According to Lugalla (1993),69 the poorest regions in Tanzania are Dodoma, Kigoma, Rukwa, Lindi, and Mtwara.70 They do not produce export crops, and there is little official investment in roads, communications, or social services. As he says, "one can argue that, in Tanzania, poverty is rampant in non-export crop producing areas."71 Figure 3.5 gives some support for his selection of very poor regions. More research on the regional pattern of poverty will help isolate priority areas for investment in markets, infrastructure and communications, and to target subsidies for social services to the poor. J. Lugalla, "Poverty and Adjustments in Tanzania: Grappling with Poverty Issues during Adjustment Period', mimeo, University of Dar es Salaam, 1993. 70 According to the 1995 PPA Survey Morogoro, Rukwa, Lindi, Mtwara and Kigoma were also ranked among the poorest regions in Tanzania. 71 In the 1995 PPA Survey households identified the growing of cash crops as a defining characteristic of the top rural household expenditure quintile.  Figure 3.5: Distribution of Welfare Index TANZANIA WELFARE INDEX BY REGIO '. Map Features U Intemnational Boundaries - Region Boundaries angs ,.' \ nR;a~~~~~~~~~~~njaro Lakes Kigo ma __+ S>Xf Tabora = Tanga pPe~~mba T:) t abora Dodoma taZanzibar Dar Es Salaam The boundaries, colors, denominations Rukwa and any other information shown on this wan map do not imply, on the part of the World Bank Group, any judgment on the legal status of any territory, or any acceptance ooo or endorsement of such boundanes. Morogoro 4, . . .< 1 l 3il |l~~~~ringa/ iE Maria Island ' _ . f {g L~~~~~~~~~~~~indi: WelfarelIndex X : : 0 I Ito 3: :: : . : 0 4 to 9 - uvuma - _ KM _I " ~~~~~~~~~~~~~~20 to 30 0' 0 100 200 31 and Hig|er t,,  71 Chapter 3 D. THE HUMAN DEVELOPMENT DIMENSION OF POVERTY 3.25 This section will examine the development of social indicators. Despite recent income growth, indicators of basic human welfare-infant mortality, nutrition, housing conditions, and primary school enrollment-appear to be stagnant or worse, compared to the level of the 1970s or early 1980s. Improvements in the quality of life of the Tanzanian population have lagged far behind growth in income. After a review of some specific indicators, the possible reasons for the apparent lag in social indicators, despite rising incomes, are discussed. Access to Water 3.26 Access to safe water has Figure 3.6: Distribution of Population with Access to declined since 1976. At the time Safe Water of Independence, about 1 .4 million people, or 12 percent of the rural population of Tanzania, 100% oou had access to clean water. Today, 80% 76% in rural areas, only about 22 55_ 60% *% r percent of people have access to * urban safe water, marking a decline 40% 25 21 from the rate of 25 percent in 20% 1976. The urban population 0 n/0 faced even worse conditions: in 1969 1976 1993 1969. 90 percent of urban Year residents had access to safe water, b'ut by 1976 this had fallen to about 76 percent, and by 1993 the estimates of the proportion of people benefiting from safe water had dropped further, to 56 percent. 3.27 Though most households in Tanzania do not incur monetary costs to obtain their domestic water supply, they pay with their labor. The average one-way distance from the home to the preferred water source is 0.25 km for residents of Dar es Salaam, 1.1 km for residents of other towns. and 1.6 km for rural households. Rural households spend an average of 3.1 hours a day collecting water, with the brunt of the work borne by women.  72 Chapter 3 Box 3.1: Access to Water In most villages, access to water becomes an acute problem during the dry season. People cope with this water shortage by neither bathing nor washing. The entire village gets involved in water fetching, from distances up to six hours. Those who own carts, bicycles or donkeys, fetch and sell water to others. Water collection is a woman's job. They are assisted by children, particularly girls. Men only assist in water collection when the circumstances are particularly harsh-6-12 hours is required for a water collection round-trip-or when there are water vending activities, or if carts and donkeys are used. Consequently, women are most affected by water shortages, and the primacy of water in the lives of women emerges through the results of the PPA Survey (see table below). Pattern of Priorities (in percent) Men Women Transportation 28 (1) 4 (6) Farming Inputs 22 (2) 9 (5) Water 8 (5) 20 (2) Food Shortage 6 (7) 21 (1) Education I (9) 2 (10) Note: (j indicates ranking of prioriy Source: PPA Survey, 1995. Literacy 3.28 One of the key indicators of human capital is the ability to read and write. Literacy allows people to broaden their range of information, and increases their likelihood of employment. The HRD Survey (1993) shows a literacy rate of about 76 percent for people older than 14, which is impressively high in comparison to countries with similar income levels. Table 3.6 (based on rural household survey for four regions) shows the improvements that have been made in literacy over time. To the extent the sample is representative, in just one decade, from 1983 to 1993. rural literacy increased from 59 percent to 73 percent. Moreover, this improvement has reached both men and women, and the percentage rate of growth in literacy has been higher lor women. However, the decline in enrollment rates, which is discussed in the later section on education, implies that the extent of literacy may fall in the future.  Chapter 3 73 Table 3.6: Changes in Rural Literacy Rates" Rural Gender Tanzania Male Female 1983 1993 1983 1993 1983 1993 Read and Write 59.1 72.6 71.1 82.8 48.4 63.6 Read Only 6.2 1.0 4.5 0.8 7.6 1.1 Neither 34.8 26.4 24.4 16.4 43.9 35.3 Total l00 100 100 100 100 100 a Based on a comparison with the 1983 Rural Household Survey offour regions (see footnote 53). Infant Mortality 3.29 Nearly 12 per- Figure 3.7: Infant Mortality Indicator cent of children born (perl,000) in Tanzania today will not reach the age of Changes in Infant Mortality five. For every thou- sand children born in 300 Tanzania, 90 will die before their first birthday. Though 200 Infant Mortality Rate these numbers are 150 much better today than 100 U -._ Mortality Rate they were in the 1960s _ or 1970s, they are still 50 too high. Moreover, 0 the income growth that 0\ started in the mid- 1980s did not reduce infant mortality. Until the first birthday, a child born in Tanzania will have a life expectancy of 47 years. After the fifth birthday, a child's life expectancy approaches that of more developed countries. Distance to Major Centers and Access to Amenities 3.30 How far have people come in terms of their access to the facilities which signify higher living standards and better services? Table 3.7 shows the changes in proximity to key amenities: a water source, a government primary school, a dispensary or health center, and a market place for food purchases.  74 Chapter 3 Table 3.7: Evolution in Access to Services 1976-1993 Rural Urban Gap between Urban and Rural Water source less than 0.5 Km Wide Primary school less than 0.5 Km x Narrow Health facility less than 0.5 Km xx xx Wide Food markets less than 5 Km Narrow Note: Double arrow means strong increase. 3.31 Over 17 years, physical access to facilities has improved markedly, especially for education and health. Today, the situation for rural residents with regard to proximity to primary schools and health facilities is better than it was for urban residents in 1976/77. However, the disparity in access between rural and urban residents persists. Although the difference in service levels seems to have declined, it is still wide, particularly for water and to health facilities. Further, the quality of services provided has declined. For example, schools are run down; desks, chairs, and other inputs are missing; and health facilities lack drugs and qualified personnel. Tanzania and Neighbors 3.32 How is Tanzania doing when compared with neighboring countries? In several indicators Tanzania is outperformed by Kenya and Uganda. * The gross primary school enrollment rate of 69 percent is clearly lower than the rate of 95 percent in Kenya, and slightly worse than the 71 percent rate achieved in Uganda. * The Tanzanian secondary school enrollment rate-at 5 percent-is the lowest in the world. This compares to 29 percent in Kenya and 13 percent in Uganda. * Life expectancy in neighboring Kenya is 8 years higher, and the infant mortality rate (at 66 per one thousand live births) is one-third lower. Conclusion 3.33 In conclusion, the review of human development indicators shows that progress in basic welfare has been mixed. Rural housing conditions have improved, but infant mortality and life expectancy have shown little improvement, while primary school enrollment and access to safe water have both declined. How is this record to be reconciled with consumption data, which shows that rural income has grown strongly  Chapter 3 75 since 1982/83, and that per Table 3.8: Evolution of Social Indicators capita expenditures for all Tanzanians rose significantly between 1991 and 1993? Indicator 60's-70's 70's-80's 80's-90's Primary School 3 x X 3.34 The first reason for the Enrollment disparity between the two types Infant Mortality 4 of indicators is that social Calorie Intake 4 welfare lags behind current Life Expectancy 4 consumption. Welfare indicators depend in part on built-up stocks of physical and human capital. For example, access to safe water is linked to past investments and policy choices. Infant mortality is linked to maternal health, which is also related to past nutritional deprivation. 3.35 Second, a sustained improvement in many welfare indicators will require a reallocation of government expenditures to support social sector objectives. Compared with neighboring countries in the region (Kenya and Uganda), Tanzania is spending less in absolute and relative terms on the social sectors. For example, Tanzania devotes 15 percent of the recurrent budget to education, while Kenya devotes 35 percent; and Uganda, 22 percent. In monetary terms, Tanzania spends about US$3.28 per capita on 72 education, compared to US$14.00 in Kenya and US$6.00 in Uganda. In recent years, social sector spending has increased, but has not kept up with growth in the economy, so that social spending has fallen as a percentage of GDP. During fiscal years 1981-85, expenditure on education was 3.48 percent of GDP. This fell by more than one-half, to just 1.66 percent of GDP in fiscal years 1986 to 1990. Education spending has since recovered but has not reached its earlier level, averaging 2.73 percent for fiscal years 1991 to 1994. The low revenue mobilization and persistent fiscal deficits have made it difficult to provide adequate support for critical social programs.73 E. POVERTY, INEQUALITY AND GROWTH: OUTLOOK FOR THE FUTURE 3.36 This section evaluates separately the impact of changes in average income and income inequality on poverty. If inequality remains constant, poverty will decrease faster than the rate of growth, since poverty is highly sensitive to a growing economy. For every 1 percent increase in income, the headcount index of poverty will decrease by 2.3 percent. Furthermore, for higher poverty lines, poverty is even more sensitive to economic growth. Therefore, growth oriented policies that will-at least-keep constant the share of income accruing to the poor will have a large impact on poverty reduction. 72 Additional information on the allocation and level of government expenditures is given in Chapter 2. 73 See Chapter 2 for an in-depth discussion of allocation of government expenditures to the social sectors. 4 Assuming a poverty line of T Sh 128,109.  76 Chapter 3 3.37 Holding income constant, increasing inequality raises poverty significantly. If the Gini Index of the income distribution increases by I percent, ceteris paribus, the poverty index will increase by 2.8 percent for the higher poverty line, and 7.2 percent for the lower poverty line.75 The "ultra-poor" are considerably more affected by changes in income inequality than by changes in mean income. If inequality increases with economic growth, there is the potential for increased poverty. This underscores the importance of achieving broad-based growth. Poverty and Growth 3.38 Given the present income distribution, Tanzania would need to Figure 3.8: Poverty and Growth sustain consistently high growth to reduce poverty effectively by the year 2015. To illustrate, with a Population population growth rate of about 3 in Poverty percent a year, a growth in GDP of 50 4.5 percent per annum would be 40 GDP required to raise per capita income 30 Growth by 1.5 percent a year. If this rate of Rate growth can be sustained, then by 20 2015, using 1993 exchange rates, 10 average per capita income in 0 03% Tanzania will be approximately 1993 2000 2005 2010 2015 US$345. If, instead, the growth rate Year in per capita income could reach 3 percent, then an average Tanzanian could expect to have US$477 dollars a year to spend by 2015. Neither of these income targets is 'high' by international standards. The reductions in poverty that will follow from the growth in per capita incomes are shown in Figure 3.8. For a growth rate in per capita income of 1.5 percent per year, the percentage of the population in poverty can be reduced to 15 percent by 2015 if equity does not get worse. If a higher growth rate in per capita income of 3 percent can be achieved, then a lower percent of the population will remain in poverty. These figures are illustrations, since the final outcome will depend very much on changes in income distribution. Poverty and Inequality 3.39 Figure 3.9 shows what would happen to poverty by the year 2015 given different outlooks for equality. Changes in the income distribution in both directions (i.e., less and 75 For poverty lines lower than the average income in the distribution, higher income inequality will always lead to greater poverty.  Chapter 3 77 Figure 3.9: Poverty and Inequality 25.00 -- 20.00 0fi 15.00 10.00 *2000 .