See Uganda: the Economic Impact of AIDS, Populationand Hunan Reaourcesdivision, the World Bank, 1992

Public Disclosure Authorized

IQa

YE~~~~~~E usv Public Disclosure Authorized

QTh'~~~~~~~~~~KI

A~ ~~~~~~~~~~~~~~KN Public Disclosure Authorized

I~~~~~~~~~~~~IKWI.~~~~~~~~~~~~~~~i ~¾ra VTIt4

Public Disclosure Authorized VY%~.*.

A W OR L D B AN K CO UN TRY STUDY

Uganda

Growing Out of Poverty

World Bank Country Studies are among the many reports originally prepared for internal use as part of the continuing analysis by the Bank of the economic and related conditions of its developing member countries and of its dialogues with the governments. Some of the reports are published in this series with the least possible delay for the use of governments and the academic, business and financial, and development communities. The typescript of this paper therefore has not been prepared in accordance with the procedures appropriate to formal printed texts, and the World Bank accepts no responsibility for errors. The World Bank does not guarantee the accuracy of the data included in this publication and accepts no responsibility whatsoever for any consequence of their use. Any maps that accompany the text have been prepared solely for the convenience of readers; the designations and presentation of material in them do not imply the expression of any opinion whatsoever on the part of the World Bank, its affiliates, or its Board or member countries concerning the legal status of any country, territory, city, or area or of the authorities thereof or concerning the delimitation of its boundaries or its national affiliation. The material in this publication is copyrighted. Requests for permission to reproduce portions of it should be sent to the Office of the Publisher at the address shown in the copyright notice above. The World Bank encourages dissemination of its work and will normally give permission promptly and, when the reproduction is for noncommercial purposes, without asking a fee. Permission to copy portions for classroom use is granted through the Copyright Clearance Center, 27 Congress Street, Salem, Massachusetts 01970, U.S.A. The complete backlist of publications from the World Bank is shown in the annual Index of Publications, which contains an alphabetical title list (with full ordering information) and indexes of subjects, authors, and countries and regions. The latest edition is available free of charge from the Distribution Unit, Office of the Publisher, The World Bank, 1818 H Street, N.W., Washington, D.C. 20433, U.S.A., or from Publications, The World Bank, 66, avenue d'Iena, 75116 Paris, France.

ISSN: 0253-2123

Library of Congress Cataloging-in-Publication Data

Uganda: growing out of poverty. p. cm. p. cm. growing out of poverty. (A World Bank country study) "This report has been prepared by a team led by Kapil Kapoor"-CIP pref. Includes bibliographical references. ISBN 0-8213-2460-8 1. Uganda-Economic conditions-1 979- 2. Uganda-Economic policy. I. Kapoor, Kapil, Ph. D. II. International Bank for Reconstruction and Development. m. Series. HC870.U438 1993 338.96761-dc2O 93-17518 CIP - iii -

GOVERNMENT FISCAL YEAR July 1 - June 30

CURRENCY EQUIVALENTS Currency Unit: Ugandan shilling (U Sh)

Official Rate: US$1.00 = U Sh 1214 (December 1992) Foreign Exchange Bureau Rate: US$1.00 = U Sh 1248 (December 1992)

ACRONYMS AND ABBREVIATIONS

ACFODE Action for Development ACS Aid Coordination Secretariat DA district administrator DANIDA Danish International Development Agency DCAT District Community Action Trust DENIVA Development Network of Indigenous Voluntary Associations DHS Demographic and Health Survey FHH female-headed households HBS Household Budget Survey MAAIF Ministry of Agriculture, Animal Industry and Fisheries MHH male-headed households MLG Ministry of Local Government MLSW Ministry of Labor and Social Welfare NGO non-governrmental organization NIC newly industrialized country NRM National Resistance Movement NUDIPU National Union of Disabled People of Uganda OGL Open General License PAPSCA Program to Alleviate Poverty and the Social Cost of Adjustment RC Resistance Council SIP Special Import Program SSA Sub-Saharan Africa SWAA Society for Women and AIDS in Africa TOT terms of trade UCBHCA Uganda Community-Based Health Care Association UCOBAC Uganda Community-Based Association for Child Welfare UDC Uganda Dairy Corporation UNEPI Expanded Program on Immunization UNICEF United Nations International Children's Emergency Fund URA Uganda Revenue Authority WHO World Health Organization WID Women in Development - iv -

Contents

Preface ...... ix

ExecutiveSummary ...... xi

Part 1: A Profile of Povery ...... 1......

Chapter 1: A Profile of Poverty ...... 3 Background...... 3 Defining Poverty ...... 4 Poverty Indicators and the RegionalVariation in Poverty ...... 8 IntrahouseholdInequality ...... 10 NutritionalIndicators ...... 10 Social Indicators ...... 11 The Voice of the People ...... 14

Chapter 2: The Poor and the Vulnerable ...... 17 Who are the Poor and Vulnerable? ...... 17

Chapter 3: The Gender Dimension ...... 25 Profile of Women's Roles and Constraints ...... 26 Gender Issues for Poverty Reductionand Economic Growth ...... 28 The Value of Women in Ugandan Society ...... 29 Unequal Before the Law? ...... 31 The Diversity of Householdsin Uganda ...... 32 T'he Gender Divisionof Labor ...... 33 IntersectoralLinkages ...... 35 Conclusion ...... 39

Chapter 4: Impact of Adjustmenton the Poor ...... 41 General ...... 41 The Ugandan Experience ...... 42 Rural and Urban GDP Per Capita ...... 43 Changes in Domestic Terms of Trade (TOT) ...... 45 MacroeconomicPolicy and Its Implications ...... 46 Implicationsfor HouseholdWelfare ...... 49 What Has Been Happeningto the Incomesof the Poor? ...... 50 Distortionsin Product and Factor Markets ...... 51 Inflationand Its Impact on the Poor ...... 53 The Provisionof Social Services ...... 55 Conclusion ...... 55 - v -

Part II. A Strategyfor Reducing Poverty ...... 57

Chapter 5: AcceleratingEconomic Growth ...... 59 Introduction...... 59 Prerequisitesfor AcceleratedEconomic Growth ...... 60 How Did Uganda Fare? ...... 63 The Link Between Growth and Poverty Reductionin Uganda ...... 64 Structure of Production ...... 66 Sources of Future Growth ...... 67 Agriculture ...... 67 Industry ...... 71 Savings,Investment, and the Balanceof Payments ...... 73 Annex to Chapter 5: AgriculturalGrowth Prospects ...... 81 TraditionalExport Crops ...... 81 Matoke (Banana) ...... 83 Cereals ...... 83 Roots and Tubers ...... 84 Pulses ...... 85 Oilseeds...... 85 Industrial Crops ...... 86 Horticultural Crops ...... 86 Livestock and Dairy ...... 87

Chapter 6: Labor Market Policiesfor Poverty Reduction ...... 89 The Distributionof Labor Force ...... 89 Features of the Labor Market ...... 94 Which EmploymentCategories are Poor? ...... 98 The Objectiveof Employmentand Labor Market Policies ...... 99 Annex to Chapter 6: Urban Employment...... 103 A Further Disaggregation ...... 103 Growth in Public Sector Employment ...... 105

Chapter 7: ProvidingKey Services and Safety Nets ...... 107 Introduction...... 107 GovernmentExpenditure on the Social Sectors ...... 108 Incidence of GovernmentExpenditures ...... 109 Education...... 110 Health ...... 111 Family Planning and Environment...... 113 Developmentof the Rural Areas ...... 113 The Provisionof Safety Nets ...... 114 Keeping Afloat ...... 115 Is 'Mere Anybodyto Help? ...... 117 Safety Nets for the Future ...... 119 - vi -

Chapter 8: InstitutionalFramework Delivering Essential Services ...... 123 Background...... 123 GovernmentalInstitutions ...... 123 Self-HelpGroups ...... 125 Non-GovernmentalOrganizations ...... 126 Coordinationbetween Governmentand NGOs ...... 127 Creating a Frameworkfor Collaboration ...... 129 Decentralization...... 135 The District CommunityAction Trust (DCAT) ...... 135 Conclusion ...... 136

Chapter 9: A Strategy for Reducing Poverty ...... 137 A. Policies for AcceleratingGrowth ...... 137 Policies for AcceleratingAgricultural Growth ...... 138 Policies for AcceleratingIndustrial Growth ...... 139 Employmentand Labor Market Policies ...... 140 Policies for Gender-ResponsiveGrowth ...... 141 B. Policies for the Delivery of Social Services ...... 142 The Public ExpenditureAgenda ...... 142 InstitutionalIssues and the Provisionof Safety Nets ...... 143

SelectedBibliography ...... 145

Annex I StatisticalAppendix .151 Annex II Poverty: A Child's View .189 Annex Il Adjustmentsto the HBS Data .199

Text Tables

Table 1.1 SelectedCharacteristics of UgandanHouseholds. 5 Table 1.2 Poverty Indicators for Uganda. 6 Table 1.3 Expenditureon Different Food Items. 8 Table 1.4 Uganda Social Indicators.13 Table 4.1 Growth and Sectoral Distributionof GDP .43 Table 4.2 Foreign Trade Taxes, Trade Deficit and ExchangeRates .47 Table 4.3 Index of Real Farmgate Prices .49 Table 4.4 Real Wages of AgriculturalWorkers .51 Table 4.5 Fiscal and MonetaryPerformance .52 Table 5.1 Annual Rates of Inflation .61 Table 5.2 Domestic Investmentand Savings.61 Table 5.3 School Enrollment .62 Table 5.4 Average Annual GDP Growth Rates .63 Table 5.5 Average Annual SectoralGrowth Rates .64 Table 5.6 Years Required to Double GDP .65 Table 5.7 Distributionof GDP .66 Table 5.8 Present vs Potential Crop Yields .70 Table 5.9 Key MacroeconomicIndicators .74 Table 5.10 GDP By Sector in Constant Prices.76 - vii -

Table 5.11 GDP By Sector in Constant Prices (Annual Growth Rates) ...... 77 Table 5.12 GDP By Expenditure at Constant Prices ...... 78 Table 5.13 Balance of Payments ...... 79 Table 6.1 Distribution of Population by Primary Activity ...... 90 Table 6.2 Estimated Mid-Year Distribution of Labor Force, 1992 ...... 91 Table 6.3 Some Facts About Urban and Public Sector Employment ...... 92 Table 6.4 Percent of Male and Female in the Work Force ...... 93 Table 6.5 Relative Earnings of Labor in Different Activities ...... 97 Table 7.1 Sectoral Expenditure Performance of the Central Government ...... 108 Table 7.2 Poverty Focus of Government Current Expenditures ...... 109

Text Boxes

Box 1.1 Poverty Indicators ...... 7 Box 2.1 Orphans on Their Own ...... 18 Box 2.2 A Woman Who was an Orphan ...... 19 Box 2.3 I Used to Crawl ...... 20 Box 2.4 Meet the Core Poor ...... 21 Box 2.5 Life in Wikuk (Gulu District) ...... 23 Box 2.6 Everybody Has Horror Stories to Tell ...... 24 Box 3.1 No Time for Health ...... 28 Box 3.2 A View from the Top ...... 29 Box 3.3 Ties That Bind ...... 30 Box 3.4 Women and Legal Reform ...... 31 Box 3.5 Female Labor Time and Efficiency ...... 33 Box 3.6 Differential Incentives: Two Cases from Kenya ...... 34 Box 3.7 Poverty and AIDS ...... 36 Box 3.8 Domestic Relations in the Context of AIDS ...... 37 Box 3.9 Investing in Girls' Education ...... 38 Box 7.1 The "Rwot Kweri"...... 116

Figures

Figure 4.1 Real Per Capita GDP ...... 44 Figure 4.2 Domestic Terms of Trade ...... 45 Figure 7.1 Primary Education Expenditures ...... 110 Figure 7.2 Primary Health Expenditures ...... 112

Maps

Map 1.1 Districts with the Highest Incidence of Poverty ...... 9 Map 1.2 Districts with the Highest Incidence of AIDS ...... 12

Preface

The last Country Economic Memorandum for Uganda, Towards Stabilization and Economic Growth (Report Number7439-UG), was issued in September 1988. Since that time, despite a severe deteriorationin the country's internationalterms of trade and the exacerbationof the AIDS pandemic, the Governmentof Ugandahas continuedto implementa challengingeconomic reform program, with support from the World Bank, the IMF and the donor community.

This report documentsthe progress made to date and complimentsthe Governmentof Uganda for taking some bold economicdecisions. With the restorationof peace and security throughoutmost of the country, and the improvementin the investmentclimate, the time is now ripe for the citizens of Uganda to start taking advantage of the new economic environment. While poverty in Uganda continuesto be a serious problem, this report suggeststhat there is some evidence that, as a result of the policies implemented,Uganda is slowly on its way towards "growing out of poverty". The report cautions, however, that there is little ground for complacency. Reinforcing the recommendationsof the 1990 World DevelopmentReport, it advocatesa growth oriented strategy for poverty reduction in Uganda and advises the Government to "stay the course" and continue implementingthe set of policies which will augment the country's human capital and enable its citizens to participate equitablyin that growth.

This report has been prepared by a team led by Kapil Kapoor (Task Manager, AF2CO) and comprising Emmanuel Ablo (AF2CO), Maurizia Tovo (AFTSP), Lemma Merid (AFTSP), Mark Blackden(AFTSP), Hailu Mekonnen(PHRWD), Elisabeth Shields (EDICD), Mimi Klutstein-Meyer (AF2CO), Aziz Khan (consultant)and Mark Henstridge(Summer Intern). BonnieKeller (DANIDA) and Carol Carolus (USAID)participated in the main mission and contributedto the preparation of the report. The team worked in close collaborationwith a UNDP programmingmission on poverty alleviation and rural development,led by Prof. A. Mafeje and the Government's counterpartteam led by Keith Muhakanizi(MFEP) and consistingof Joseph Okune (PAPSCACoordinator), Damon Kitabire (MFEP), Margaret Kakande (MFEP), R.P. Tumusiime (MWIDCY), Moses Bekabye (MFEP), Harriette Mugerwa(MFEP) and Francis Wagaba(MFEP). The peer reviewerswere Helena Ribe (PHRPA) and Oey Astra Meesook (AF4CO). The document was produced by a team led by Kathryn Rivera (AF2CO).

The report was discussed with the Governmentof Uganda in March 1993.

- xl -

Executive Summary

1. With a per capita incomeof under US$170,Uganda today is one of the poorest countries in the world; indeed, it is a living testamentof the havoc caused by the political turmoil and economic decline brought about by more than a decade of despotic rule.

2. At independence(1962), Ugandahad one of the most vigorous and promising economiesin Sub-SaharanAfrica (SSA), and the years followingindependence amply demonstratedthis economic potential. Favored with a good climateand fertile soil, the country was self sufficient in food, with the agriculturalsector being a large earner of foreign exchange.The manufacturingsector supplied the economy with basic inputs and consumer goods and was also a source of foreign exchange earningsthrough the export of textiles and copper. Export earnings not only financed the country's importrequirements but also resulted in a current accountsurplus. Fiscal and monetarymanagement was sound and the domestic savings rate averaged about 15 percent of GDP, enough to finance a respectablelevel of investment.Uganda's systemof transportationwas widely regarded as one of the best in SSA and included an effective network of roads, railways, port and air transport.

3. Uganda's social indicatorswere comparableto, if not better than, most countries in Africa. The country's health service had developedinto one of Africa's best and pioneered many low cost health and nutritionprograms. There existed a highly organized networkof vaccinationcenters, and immunizationprograms reached as much as 70 percent of the population.Although school enrollment was still low, Uganda's educationsystem had developeda reputationfor very high quality.

4. The Amin regime radically reversed the economic and social progress attained since independence,and the ensuingcivil strife resulted in a tremendousloss of human life. It is estimated that as many as 500,000 Ugandanslost their lives during Amin's eight-yeardictatorship and as many as one million more were internally displacedfrom their homes and farms. A 1985 estimate by the U.S. Committeefor Refugees concludedthat in that year one out of every fourteen Ugandanswas either a refugee or was displaced; it was estimated that as many as 200,000 Ugandans had fled the country and were living in exile.

5. Economicmismanagement accompanied the civil war and professionalstandards deteriorated rapidly as skilled personnel fled the country. Between 1970 and 1980, Uganda's GDP declined by about 25 percent, exports by 60 percent, and import volumes by close to 50 percent. With large increases in defense expenditures,the governmentbudget became increasingly untenable and was largely financed by bank borrowing which resulted in average inflation rates well in excess of 70 percent. Economicmismanagement and abuse of human rights on a massive scale continuedduring the Obote regime in the early 1980s. By 1985 governmentexpenditure on education and health, in real terms, amountedto about 27 percent and 9 percent respectivelyof the 1970s levels. When the NationalResistance Movement (NRM) Governmentassumed power in January 1986, it inherited a shattered economy whose social indicatorstoday paint a dismal picture of the quality of life of its citizens and are indicativeof the extent of poverty within the country.

6. Giventhis legacy,this CountryEconomic Memorandum focuses on poverty in Uganda. Part I of the report presents a profile of poverty, using data from the recently concluded Household Budget Survey, the Demographicand Health Survey and the PopulationCensus. These data were supplementedby a Rapid Poverty Appraisalconducted by the mission in August, 1992, in an attempt to get a better understandingof the causes of poverty and what it means to be poor in Uganda. - xii -

Particular emphasis has been placed on the gender dimensionsof poverty, recognizingthe fact that poverty affects men and women in different ways because they play different roles, have different needs and face different constraints. Part I of the report also examinesthe policies that have been implementedby the NRM Governmentduring the past five years and, despite the data-poor situation, analyzesthe impact of adjustmentpolicies on the poor. Part II of the report is forward looking and, consistent with the recommendationsof the 1990 World DevelopmentReport, articulates a two- pronged strategy for poverty reduction. The first prong of the strategy outlinespolicies designed to increaselabor productivityand accelerateeconomic growth. The second prong recognizesthat while economicgrowth is a necessary condition for poverty reduction, it is, by no means, a sufficient condition. In order to ensure that the poor are able to participate equitably in such growth, it recommendsincreasing the share of public expenditureson critical social services in order to foster the developmentof human resourceswithin the country. Althougha special case could be made for targeted interventionsfor the vulnerablegroups in Uganda,the report recommendskeeping these to a minimum, given the tight budgetary constraintand poor administrativecapacity in the country.

A. Poverty Profile

7. This report has defined two relative poverty lines for Uganda.The first poverty line has been drawn at U Sh 6,000 per capita per month (approximatelyUS$110 per capita per year in 1989/90 prices), which is approximatelyequal to four fifths of the mean per capita monthly expenditurein 1989/90(U Sh 7,512). The secondpoverty line has been drawn at U Sh 3,000 per capita per month (approximatelyUS$55 per capita per year), which is approximatelyequal to two fifths of the mean per capita monthly expenditure. Ugandans falling below the U Sh 6,000 poverty line have been characterizedas the 'poor" and those falling below the U Sh 3,000 poverty line as the "poorest", or the core poor.

8. Preliminary estimates indicate that, in 1989/90, a minimum of U Sh 6,000 per capita per month would have provided a daily intake of 2,200 calories plus some reasonable nonfood expenditures(e.g., clothing, fuel, etc.). This does not mean that people with monthly expenditures higher than U Sh 6,000 in 1989/90lived comfortably;rather, it meansthat those with lower monthly expenditurescould not satisfy their basic requirements. The lower poverty line of U Sh 3,000 representsthe bare minimumfor adequatefood intake. At U Sh 3,000, if you are adequatelyfed, you do not have anything else at all, however essential. As the key social indicators discussed below indicate, access to basic social services is very low throughout the country, implying that some Ugandans living above the poverty line could be consideredpoor.

9. Using total expenditure as the measure of welfare and a poverty line of U Sh 6,000, 55 percent of Ugandans can be defined as being poor. The poor are disproportionatelyfound in rural areas, where about57 percent of the populationis poor, comparedto about 38 percent in urban areas. Ninety-two percent of the poor in Uganda live in the countryside, while only 89 percent of the populationis classified as rural. The discrepancybetween rural and urban levels of poverty is even greater usingthe lower or "core" poverty line; 96 percent of the core poor live in rural areas. When the data are analyzedto take into accountthe depth of poverty, it turns out that not only is poverty more widespreadin rural areas, but that it is deeper. In other words, rural people are more likely to be poor than urban dwellers, and their poverty is more severe.

10. Poorer householdsin Ugandatend to be larger, have older and less educatedhousehold heads, and are more likely to be headed by a woman. Not surprisingly, the dependencyratio, which is calculatedas the proportionof the householdpopulation younger than 18 and older than 55, is quite high and increases for poorer households.The difference in averageper capita expenditurebetween - xiii - urban and rural areas is significant, with rural people spending about 60 percent as much as urban dwellers. Accordingly, indicators generally associated with poverty have higher values for rural Uganda: larger householdsize, higher dependencyratio, higher illiteracy, etc.

11. The highest incidenceof poverty is in the northern part of the country. In the rural areas of the north, 81 percent of the populationhave a real per capita monthlyexpenditure of less than U Sh 6,000 and 42 percent have a real per capita monthlyexpenditure of less than U Sh 3,000. Such high incidencecan be ascribed for the most part to the ravages of the civil war. As persistent insecurity prevented the hardest hit northern districts from being included in the Household Budget Survey, it can be assumedthat the actual poverty incidencein the north is higher.

Social and NutritionalIndicators

12. Life expectancyat 47 for men and 50 for women is one of the lowest in the world, and it is difficult to see how the situation could improve in the near future with the AIDS pandemic threatening not only the adult populationbut also infants. AIDS has emerged as a significantcause of death and illness among young children. Thus, the already high infant and child mortality rates (respectively, 117 and 180 deaths per 1,000) can be expectedto rise. Areas where peace has not yet been completely restored are at a greater risk of HIV infection because of the presence of both regular and rebel armies.

13. Not surprisingly, insufficientaccess to health and sanitationservices is also reflected in the general mortality and morbiditypatterns. Uganda's crude death rate, at 20 per 1,000, is about twice the level of the average low-incomecountry (e.g., neighboringKenya) and considerablyabove the SSA average. While such high rates can be partly explainedby civil strife and AIDS, it is undeniable that poor health conditionsare part of the explanation. Malaria has been found to be the principal killer among adults admitted to hospitals, far ahead of AIDS. Diarrhea, pneumoniaand anemia are almost as common as AIDS as reported primary causes of death. Among children under five, who account for over half of hospital deaths, the main killers are malaria, pneumonia, diarrhea and malnutrition. Available informationon morbidity, as measured by outpatient statistics and surveys of mothers, confirms a similar pattern. Thus, the main causes of mortality and morbidity, with the exceptionof AIDS, appear to be related to a generallyunhealthy environment and a lack of routine medical intervention.

14. The Demographicand Health Survey (DHS) conductedin 1988/89found that 45 percent of children aged 0-60 months had stunted growth, as shown by their height-for-age. Severe growth retardation occurred in 20 percent of the sampled children. Stuntingreflects the cumulativeeffect of chronic malnutritionover a number of years, and it is typically associated with poor economic conditions(as illustrated above). Although it is conceivablethat some families will fail to provide adequatenutrition to their childrenout of ignoranceor neglect, it is very unlikely that this would be the case for the majority. This assumptionis supportedby the fact that stunting is almost twice as common in rural than in urban areas, that is, in areas that where the incidence of poverty is much higher. Thus, we can conclude that, when the effect of insufficient access to social services is combinedwith that of limited economic resources, the percentage of Ugandans living in poverty appears to be higher than what a poverty line at U Sh 6,000 would suggest.

15. There are, of course, regional variations. The DHS found the prevalence of stunting to be considerably higher in the southwest (54 percent), while Kampala had the lowest percentage of stunted children (22 percent). The northern districts could not be surveyedbecause of the insecurity in the area, but medical personnel in Gulu and Kitgum estimate malnutritionrates to be at least 60 - xiv -

percent. Acute malnutrition (i.e., wasting) is only 2 percent according to the DHS, but in Gulu, Kitgum and Karamoja it is estimatedto be as much as 30 percent.

Impact of Adjustmenton the Poor

16. Against this backdrop, and the seeminglyoverwhelming problems facing them, the NRM Government must be complimentedfor their tenacity and determinationto improve the lot of Uganda's citizens. During the past five years, the Government has implemented a far-reaching economicreform agenda whichhas today transformedtJganda into one of the most liberal economies in Sub-SaharanAfrica. With the liberalizationof the exchangeand trade regime, the abolitionof the Industrial LicensingAct, the promulgationof a new investmentcode, and the gradual liberalization of agriculturalpricing and marketing, the Governmenthas succeeded in establishingsome of the fundamentalpreconditions that are essentialfor sustainablegrowth.

17. Rural and UrbanGDP Per Capita. Since poverty in Uganda is largely a rural phenomenon and most of the poor are engaged in multicrop and mixed production, the trend in real agricultural GDP closely reflectsthe pattern of rural livingstandards. Moreover, since most rural land falls under the customarytenancy system where access to land is fairly open, and since over 85 percent of the farming populationoperates on less than two hectaresof land, the benefits of agriculturalgrowth are likely to be evenly distributedamong the rural population.Thus, averagetrends are expected to be a fair reflection of changes in rural living conditions. Aggregate real per capita GDP, which had declinedsteadily between 1983 and 1986, increasedsignificantly between 1987 and 1991. Real GDP per capita in rural and urban areas essentiallyfollowed the same trend. On average, the welfare of the rural and urban poor, in real terms on a per capitaoutput basis, has improvedby between 14 and 16 percent respectivelyduring the past five years, maintainingthe rural-urban incomegap at around 5.2:1 over the second half of the 1980s. Compared to the early 1980s, some income redistribution in favor of the rural areas seems to have occurred. It is importantto note, however, that averages mask the existence of wide intraregionaldifferences, particularlyin urban areas.

18. There is considerableuncertainty in determiningwhether these improvementswere entirely a result of adjustmentpolicies, other extraneousfactors, or some combinationof the two. Given that peace and security were regained at a time that the structural adjustmentpolicies were put in place, causes serious difficulties in disentanglingthe effects generated by the macroeconomicpolicies per se. The problem is also compoundedby the difficulties in isolating the effects of the price reforms (i.e., changes in the exchangerate, interest rates and productpricing) from the impact of the injection of large amounts foreign exchangesupport accompanyingthe reform program. To isolate, at least partially, the effects of the "peace dividend" from the effect generated by policy measures, the domesticbarter TOT for cash crops (traded), food crops (nontraded), and the agriculture sector as a whole have been calculated. These prices are influenced, in one way or another, by such policy measures as the changes in the exchangerate, price and trade liberalization,and the eliminationof marketing monopolies.

19. The terms of trade for cash crops fluctuated considerablyduring the 1981-86 period and experienced a sustained deterioration between 1986 and 1989. However, there was a dramatic improvementin the cash crops TOT in 1990 and 1991. Food crop TOT improved steadily between 1981 and 1987, largely as a result of insecurity, which prevented adequate supplies of food from reachingthe market. Since 1987, with the restorationof peace throughoutmost of the country, there has been a significantrevival in agriculturaloutput and in the marketedsurplus causing the TOT for food crops to decline. Given the dominanceof the food crop sector, this has resulted in the aggregate TOT shifting against agriculture.This somewhatsurprising result probablywarrants further research - xv - which present data limitationsdo not permit, i.e., a closer examinationof farmers' incomes and expendituresper capita. The key implicationof the decline in agriculture TOT is that increases in gross domestic incomes have tended to be smaller than the increases in gross domestic product in agriculture. It must also be pointed out that the improvementin the security situationhas resulted in a decrease in the farmers cost of production and so the decline in TOT probably overstates the loss in incomes.

20. MacroeconomicPolicy and Its Implications.A poor macroeconomicenvironment, in addition to providing the wrong market signals and thereby stifling economic growth, hurts the poor by indirectly taxing the resources that they possess. An enablingmacroeconomic environment, on the other hand, removesproduction disincentives,encourages optimal resource allocation,and improves income distribution.The impactof macroeconomicpolicies, particularlythose focusingon taxation and the exchange rate, is particularlypronounced in largely agriculturaleconomies such as Uganda and thereby warrants careful consideration.The exchange rate, in particular, produces the most pervasive impact that often outweighsthe other effects of crop-specificpolicies.

21. Througha series of changes in the foreign exchangeallocation system, the Ugandanshilling has depreciated significantly since 1987. In 1987, the US dollar was officially exchanged at an average rate of U Sh 44.7 and the spread between the official exchangerate and the free market exchangerate amountedto about 266 percent. Startingin 1988, the Governmentbegan implementing an aggressive program of adjustingthe official exchangerate with a view to compressingthe gap between the official and the foreign exchangebureau rate and in July 1990 it legalizedthe operation of the foreign exchange bureaus. By 1991, the Governmenthad adjusted the official exchangerate to an average of U Sh 750 per dollar, i.e., over a 17-folddepreciation. When the differential had declined to about 15 percent, the Governmentdecided to cease setting the official exchange rate administrativelyand introduceda foreign exchangeauction in January 1992.

22. It is importantto note that real devaluation,by itself, cannot provide positive incentivesand stimulate the supply response unless it is complementedby price and market liberalizationwhich allows market-determinedor border prices to be transmitted to the producers and reduces the monopoly rents enjoyed by marketing boards. Recognizingthis, the Government of Uganda has overhauledthe systemof marketing,pricing and taxing coffee. Startingwith the 1991/92coffee year (October-September),the Governmenthas moved to a system where only an indicative floor price for coffee is announced and the actual producer price is determined by market forces. The export monopolyof the Coffee MarketingBoard has been eliminatedand coffee exporting has been opened to competition.To counteractthe adverse impact of the decline in the internationalprice of coffee, the Government has virtually abolished all taxes on exports; the remaining export tax only applies to Arabica coffee which commandsa premium price on the internationalmarket. The Government has further eliminatedthe export monopolyof the UgandaTea Authorityand the Produce Marketing Board, although the Lint Marketing Board still enjoys a monopoly for exporting cotton. These changes have gone a long way in altering relative prices in favor of cash crops (tradables), and have had a beneficial impact on the poor, to the extent that they are involved in the production of such crops.

23. The over 50 percent decline in the internationalprice of coffee, as a result of the collapseof the International Coffee Agreement, has caused major foreign trade, fiscal and price instability, inimical to the rural as well as to the urban poor. As a result of the actions mentioned above, particularly since 1990, the Governmenthas been able to protect the price received by producers. Althoughthe real producer price of Robustacoffee declined in the early years of the program with the deteriorationin the internationalenvironment, in 1991the price was about 7 percent higher than - xvi -

the price in 1987, the year when the Governmentcommenced implementing the ERP. The farmers' share of the internationalprice of coffee in 1991 and 1992 was more than double their share in the mid 1980s. As a result of the adjustmentmeasures implementedby Government, relative prices, particularly since 1989, have changed in favor of coffee as compared to the prices received by farmers for growing food crops such as plantain (matoke), cassava and maize, which compete with coffee for farmer's resources,particularly labor. It is therefore apparentthat, throughmacroeconomic and agriculturalpricing policies,the Governmenthas been largelysuccessful in preventingthe coffee industryfrom virtuallycollapsing which would have occurred in the absenceof the structuralreforms implemented.

24. Implicationsfor Household Welfare. To keep real incomesrising in the face of the overall decline in the TOT of the agriculturalsector, various strategies are being adopted by farmers. As recent experiencein several areas shows, monetizationof the agriculturalsector is gradually on the increase as farmers shift resources, particularlylabor, from less paying crops to crops commanding higher prices at the farmgate level. In particular, incomesare being supported, in some areas, by the introduction of some high-paying export crops into the production mix. These encouraging developmentsare also being affectedby changesin the labor market. In recent years, labor shortages have started manifestingthemselves, particularly in the estate sector, as a result of which real rural wages have risen. Among the reasons for the labor shortage, particularly during peak season, is the decline in the supplyof migrantlabor from neighboringcountries and the impact of the AIDS crisis on the productivesegments of the population.Although this labor shortagehas adversely affectedthe up-keep of the coffee trees, a relatively labor-intensiveactivity, it has helped to supplementthe incomeof net-labor supplyinghouseholds further improvingrural welfare.

25. Inflationand Its Impact on the Poor. While adverse weather conditionsand the frequent adjustmentsto the exchangerate and to interest rates have all contributedto the escalationof prices, the imbalancebetween governmentrevenues and expenditures,and the link between the fiscal deficit and monetaryexpansion has been the primary cause for the high rate of inflationin Uganda. The tax effort in Uganda is still among the lowest in SSA, and total revenues are sufficientto finance only about 50 percent of government expenditures.While external loans and grants have traditionally financeda large part of the fiscal deficit, borrowing from the domesticbanking system has been an extremely importantsource of financingfor the Government, at least until 1988/89. Consequently, until 1988/89, the large-scale monetizationof the deficit resulted in triple digit inflation levels, thereby severely eroding real incomes in Uganda.

26. In order to stem this erosion, one of the key objectivesof the stabilizationcomponent of the adjustmentprograrn in Ugandahas thereforebeen to keep prices in check by mobilizingincremental fiscal revenues and restricting expendituresto budgeted amounts. Bringing inflation down to levels comparable with those of major trading partners is a declared objective and a continuingchallenge for the Government.Since 1990, the Governmenthas embarkedupon a concertedprogram aimed at mobilizingincremental revenues and controlling expenditures.The tax and tariff regime has been rationalized,the rate structure has been greatly simplified,and tax exemptionsminimized. With the establishmentof the UgandaRevenue Authority (URA) in September1991, the Governmenthas taken a bold step forward in overhaulingits system of tax administration.Furthermore, the Government has instituteda strict program aimed at keeping expendituresin line with the budget and has started paying back its arrears to the banking system. As a result of these measures, Uganda has been successful in reducing the average rate of inflation from over 200 percent in 1986/87 to about 42 percent in 1991/92. The average monthlyinflation rate between July 1992 and February 1993, has been -0.2 percent. - xvii -

27. TheProvision of Sodal Servies. Anotherchannel through which adjustmnentpolicies impact on the poor is through the restructuringof public expendituressuch that an increased emphasis is placed on the provisionof economicand social services which increasehuman capital. Expenditures in areas such as primary health, primary educationservices, agriculturalresearch and extension,etc. empowerthe poor, raise their productivityand augmenttheir earningpotential in the long run. While Ugandastill spends far less on economicand social servicesthan do most countries in the world and in SSA, the adjustmentprogram in Ugandahas nonethelessemphasized the need for and resulted in a visible change in public expenditurepriorities with expenditureallocations increasingin favor of social sectors in general and on education and health care in particular. Given the extremely low revenue effort mentioned above, the Government plans to continue increasing these allocations steadily, keeping in mind the macroeconomicimplications of the constraints imposed by the tight resource envelope.

B. A Strategy for Poverty Reduction

28. Despite the progress achieved to date, the unfinished economic agenda is large and there should be little cause for complacency.If the Government's overarching objective is to make a serious dent in poverty over the next decade, it will have to vigorouslyimplement the two-pronged set of policies alluded to earlier, i.e., (i) policies which will accelerate economic growth; and (ii) policies which will deliver key services to the poor and, by investing in human capital, ensure that the poor are able to participate equitably in that growth. While there is broad ownership of these policies within Government,the principalchallenge ahead lies in buildingthe capacityto ensurethat policy changes can be implementedefficiently. Furthermore, there is an urgent need to strengthen the databaseand develop a poverty monitoringsystem which can provide policymakerswith regular informationon the impact of economicand social policies on the lives of the poor.

Policies for Accelerating Growth

29. Experience from other countries has shown that macroeconomicstability, a high rate of investmentbacked by domestic savings, and high rates of literacy and numeracy are vital for rapid economicgrowth. Uganda continuesto do poorly in each of these three areas. Inflation continuesto be high and savings and investmentcontinue to be low as do rates of literacy and numeracy. The key to achievingmacroeconomic stability is to get firm control over the governmentbudget. Despite the rationalizationof the tax and tariff regime and the establishmentof the Uganda Revenue Authority, the Government's revenue effort remains one of the lowest in the world. In the coming years, it will be imperativeto further tighten tax administrationand to broaden the tax base. The Governmenthas made remarkable progress in keeping expendituresin line with the budgeted amounts and such discipline has to be maintained. However, given the low revenues, the Government will probably continue to be a dissaver for several years to come thereby necessitating that the private sector generate the large increases in savings that will be necessary to finance the investmentneeded for rapid economicgrowth. Experienceof countries like China, India, and Kenya has shown that poor countries are capable of saving 20 percent of GDP or more. This places a premium on the developmentof the financial sector, becausethe key to higher rates of saving is an efficient financial system capable of mobilizingsmall savings from a large proportion of the population.

30. Policiesfor AcceleratingAgricultural Growth. At the heart of its growth strategy will have to be the transformationof agriculture. The agriculturalsector has the potential to feed the country, to supply food to the regional market, to export horticulturalproducts in additionto the traditional export crops, to produce industrialraw materials (sugarcane, cotton, etc.) and generally to act as the engine of growth. The key to realizing this potential is increasingyields by raising the productivity - xviii - of the farmer. That means security of land tenure, investmentin research and extension, control of plant and animal diseasesand rural feeder roads. For some crops high-yieldingvarieties are available from local research stations or foreign stations. What remains is to adapt them as necessary, and propagate and disseminatethem. An effective extensionservice is needed to ensure rapid adoption of improved varieties by the farmers. Many countries in Asia have shown that rapid and sustained improvementsin yields are possible.The key questionis one of timing, i.e., how quicklycan Uganda put in place the infrastructureneeded to raise yields. The answer partly lies in targeting incremental nationaland donor resourcestowards agriculture,particularly research and extensionand rural feeder roads. Raising agriculturalproductivity, along with increased access to basic health and education, must take priority in the Government'sspending program. Markets, with the associatedinfrastructure of storage and refrigeration, also matter a great deal. This is a matter for private investorsrather than Governmentexcept with regard to governmentsupport for the collectionand disseminationof market intelligence.

31. In additionto raising yields, Ugandamust pay more attentionto the working of the domestic markets in agriculturalproducts. This is not an invitationto Governmentto intervene in the affairs of producers, processors and traders. Rather, it is to draw attentionto the need to use public policy and public expenditureto facilitate the proper functioningof markets in agriculturalproducts. Poor rural roads, for example, can place producers of certain crops, or producers in a particular region, at a disadvantagevis-a-vis producers in other regions or producers of the same or other crops. Or the way truck operators are licensed or regulated may have adverse effects on the markets for agriculturalcommodities. In Ugandathere is anecdotalevidence of very low farmgateprices for some crops relative to final consumer prices. It is the legitimateresponsibility of Governmentto find out whether the gap between farmgate and final prices is broadly reflective of costs or whether it is indicativeof a malfunctioningmarket.

32. Rural feeder roads constitute an essential element of Uganda's strategy for accelerated agriculturalgrowth. Roads are needed to bring in agriculturalinputs and implements,facilitate the work of extension staff, bring access to manufacturedgoods, create access to basic social services such as educationand health and, most important of all, provide access to markets for farm goods. During the early phase of the ERPthe Governmentplaced emphasis on the rehabilitationof the major trunk roads and much progress has been achieved in this area. The Governmenthas recently shifted attentionto rural roads and the donors have respondedenthusiastically with funds, equipment and technicalassistance. However, progress with the rehabilitationof rural roads has been slow. The main reason for this is the weak implementationcapacity of the Ministry of Local Government and the district authoritiesupon which falls the responsibilityfor the construction, repair and maintenance of rural roads. This is a task which will be gradually shifted to the districts, in parallel with the necessary financial and trained manpower resources. The decentralizationof responsibilitieswill probably start with feeder roads maintenance.

33. In the drive towards greater reliance on market forces and on the private sector, Ugandahas moved quickly to dismantle the monopoliesin the agriculturalsector. These monopolieswere the Coffee Marketing Board for coffee exports, the Uganda Tea Authority for tea exports, and the Produce Marketing Board for the export of food crops. In all three areas the emergence of competitionis acting as a spur to further market development. In food crops, for example, private traders have been the driving force behind the penetrationof the European and Middle East markets in simsim. Unfortunately,Uganda is not entirely rid of monopolies.The Lint Marketing Board still has the right to purchase all cotton for export. Ginneries are free to sell lint to the domesticcotton- processingmills. However traders in cotton lint and seed must be licensed by the Lint Marketing Board. The CottonAct, last revised in 1964, providesfor the zoning of cotton production, the setting - xix - of fixed seed and cotton lint prices, restrictionson the importationof, or trade in cotton and for the licensingof ginneries and restrictionson the siting of the same. As a result the cooperativeunions have a de facto monopolyof cotton ginning. There has been much talk of amendingthe Cotton Act and the Lint Marketing Board Act with a view to introducing more competition into the cotton industry but little action so far. Given the potential for fast growth of a once-thriving industry, Governmentmust set itself monitorabletargets for removingthe institutionalfactors constrainingthe revival of cotton.

34. Ineffectiveand inefficientfinancial intermediationhurts all sectorsof the economy, including agriculture. At present agriculture does not depend much on purchased inputs and implements, whether locally made or imported; nor is the bigger proportion of agriculturaloutput marketed or processed. The heaviest demand for credit in the agricultural sector has come from the agencies responsiblefor the procurement,processing and marketingof the traditionalexport crops, particularly coffee. Apart from the Rural Farmers Scheme that has been operated by the Uganda Commercial Bank, institutionalcredit has generally not been availableto the smallholder.As a result not enough is known about the capacity of the smallholderto absorb and repay loans. The usual presumptionis that credit is a constrainton production. This may well be true in Uganda but it does not follow that rushingcredit to the farmer wouldtranslate into increasedproduction. The first priority shouldrather be to get a better understandingof the different factors limiting the smallholder's ability to expand output.

35. Policiesfor AcceleratingIndustrial Growth. Althoughimport-substitution, certainly the old- fashioned, state-directedkind, is out of favor, it must be said that Uganda has substantialscope for replacingimports, providedthis is done efficiently.A wide range of basic products are still imported, includingcement, paints, biscuits, processedmilk, garments, blankets, tomato paste, and tinned fruit juices. Import-substitutionis more likely to succeed if it is part of an outward-orienteddevelopment strategy than one focusedentirely on the domesticmarket. The Republicof Korea, China, and other fast-growingcountries have demonstratedthe superiorityof outwardorientation. Thus, Uganda must exploit,to the full, the opportunitiesfor entering export markets. As far as manufacturedexports are concerned,there are no obvious winners on the horizon but these cannot be ruled out. It is private investors, not governments, who pick winners (or losers). The government's role is to ensure that the investment climate is conducive to attracting investors who can produce a wide range of manufacturedgoods for the local and export market. In this regard, the great success achieved by Mauritiusin identifyingand exploitingan export market niche (in this case, wearing apparel)provides an indicationof the opportunitiesthat Uganda can seize.

36. Given the dearth of long-termfinance, moderntechnology, knowledge of foreign markets and managementskills in Uganda, foreign investmenthas a crucial role to play as the catalyst for the transformationof the economy. In some countries minorities from a particular foreign country or region have performedthis role. In Ugandathe Asians fit the bill. A significantnumber of the Asians expelled by Amin in 1972 have become successful entrepreneurs in Britain, Canada and other countries. The courageousdecision of the NRM Governmentto return the expropriated properties to the owners opens the way for these entrepreneursto invest in Uganda. Havingbeen dispossessed once, they will most likely exercise maximum caution in committing resources to Uganda. Nevertheless, the early signs are encouragingand a number of Asians have embarkedupon a major rehabilitationof their properties once these were returned to them. Uganda needs to be forthright about the catalytic role that investors can play in the economy and it should mount investment promotionexercises directed at Asian and other foreign investors. -xx-

37. Employment and Labor Market Policies. The central objective of employmentand labor market policies for the reduction of poverty in Uganda should be to increase the earnings of labor in agriculture. From the objectivesof the land reform legislationproposed by the AgriculturePolicy Committee, it appears that the Governmentrightly wishesto pursue a strategy of rural development which promotes a system of smallholderagriculture by further consolidatingthe de facto universal access to land. For this strategy to work, the proposed land reform legislationmust find ways of guaranteeingtenancy rights to all the existingusers of land and by improvingthe land endowment of the very small farmers. To encouragethe smallfarmers to overcomethe absence of the advantage of economiesof scale, steps shouldbe taken to facilitatetheir access to an appropriatetechnological package. As has been stated throughout the report, Ugandan agriculture operates under primitive conditionsand the knowledgeof farming systemsand crop husbandry practices is rudimentary.

38. Labor Mobility. The operationof the rural labor market will continueto be constraineddue to the absence of a large and mobile rural labor force. This is not by itself a problem as long as the concentration of labor in smallholder agriculture is not artificially promoted by inappropriate incentives. Improved systems of information and infrastructural facilities, along with the steady improvementof productivityin smallholderagriculture, is likely to make as much labor availableto the rest of the economy as it can efficientlyemploy.

39. Government Role in the Formal Sector. Regulationof employmentand wages in the formal sector has traditionallybeen kept low in Uganda and the Governmenthas rightly resisted pressures to regulate formal sector wages, through such measuresas the legislationof minimumwages. As long as there is a healthy growth of earnings in the primary sector of the economy, the market might be trusted to ensure that the formal sector of the economy pays a price for labor that is adequate for living above the poverty threshold. Creatingtoo great a differentialbetween earnings in the primary sector and the earnings in the formal sector often leads to a lower than optimal rate of industrialization,an unwarranted influx to urban areas leading to an overcrowded informal sector, and a rising differentialin the earnings between the formal and informalsectors.

40. Education and Training.Education and training facilitiesmust adjust to make the allocation of labor more efficient. Improvedproductivity of agriculturallabor hinges criticallyon the expansion of primary education. Urban educationalservices would be better advised to reduce the focus on general higher education and instead increase the emphasison technicaltraining. Clearly the entire system of pricing of educationalservices is in need of a basic reappraisal. At present a primary school student in Uganda is charged a tuition which is often a significantproportion of the cash income of an average rural household whereasuniversity educationcontinues to be free of tuition, often with access to additionalsubsidies.

41. AIDS. The effects of the AIDS pandemic on employmentand the labor market in Uganda cannot be determined with any precision. It is however clear that AIDS, which disproportionately strikes people in their prime, will increase the dependency ratio and make the task of poverty reductionmore difficult. The pessimisticprognosis is heightenedby the fact that measuresto prevent the spread of HIV infection-critically important though they are-will have little effect on the incidenceof AIDS during the next decade. The Governmentof Uganda deserves credit for dealing with the AIDS issue in an open manner. Besidesconcentrating efforts on containingthe further spread of the HIV infection, e.g., by promoting the use of condoms, the Government, working with the NGO community, should explore the need to make targeted interventions in order to reduce the extreme effects on householdswhose labor endowmentshave been depleted by AIDS. - xxi -

42. Public Sector Employment. Expansion of public sector employment is not an effective method of poverty reduction. Indeed an expansion in public employment almost certainly hurts the cause of poverty reduction by appropriating resources that might be used to expand employment more productively elsewhere in the economy. The number of persons employed in the public sector has grown too rapidly over the years from the standpoint of the actual expansion of public services and the capacity of the Government to protect real wages of public employees from serious erosion. The result is widespread resort to corrupt and fraudulent practices (e.g., bribery, keeping "ghost" workers on the payroll) and moonlighting and second informal jobs. On the whole, the effective labor time spent in public employment may have declined at the same time that there has been a steady rise in the number of persons on payroll. Recently the Government has succeeded in reversing the trend and reducing the number of persons on payroll. The task is incomplete and should continue. To the extent that the reduction in government employment comes through the elimination of ghost workers and attrition (i.e., not replacing the low priority retirees) the process does not impose any burden on the current budget. A further opportunity may become available if the increase in capacity utilization in fonnal sector enterprises provides a leeway for productive expansion of employment. In that event positions in these enterprises might be filled by the redundant government workers. However, a reduction in public employment is by itself a rather minimal measure. A central objective of the presently underway civil service reform program must be to introduce a culture of improved job performance, matched by higher real earnings commensurate with skill levels. This in turn means that a way must be found to prevent employment in public sector to be determined by the supply of high school and university graduates in a system of irrational relative costs of different forms of education.

Policies for Gender-Responsive Growth

43. The strategies adopted by the Ugandan Government to reduce poverty and to foster sustainable economic growth need to take explicit account of the gender dimension. In this respect, they must recognize and seek to address the asymmetries in the respective rights and obligations of men and women, and pay particular attention to the gender division of labor, and differential incentives and opportunities facing men and women, as the country embarks on monetization, diversification, and productivity enhancement in the critically important agricultural sector. It is, in particular, necessary that gender-responsive actions be undertaken as an interconnected package of measures which are mutually reinforcing. The priority areas requiring attention if Uganda is to enable both men and women to break out of poverty, and to contribute more fully to economic and social development, are:

(i) to promote, through literacy and education, and in conjunction with the vigorous pursuit of the gender-responsive legal (and customary) reform efforts underway, the legal rights and protections enabling women to benefit from their own labor and to have greater access to and control of economically productive resources, including capital and land, thereby raising the status of women to enable more equal participation in household-level, community, and national decisionmaking;

(ii) to raise the productivity of women's economic (paid) labor through investment in education aimed at overcoming social, financial, and cultural barriers to female participation, including at the post-primary level; through investment in basic, accessible, and affordable health care responsive to the wide range of women's health needs; and through targeted actions aimed at raising women's access to information, technology, inputs, credit, and extension services;

and markets), and in water supply and woodlots,that take explicit accountof female users' needs in design and implementation;

(iv) to provide maximumpolitical and financial support to the efforts, spearheadedby UNICEF, to reduce AIDS risk among young girls; and to protect the rights of children, including through institutionalmeasures in the RC system to ensure appropriate representation and articulationof children's needs.

C. Policies for the Delivery of Social Services

44. The Public Fxpenditure Agenda. The Government recognizesthat, since the single most important asset owned by the poor is their labor, the central element of its poverty reducing public expenditurestrategy should be to accord highest priority to developingtheir human capital. Human capital, more than any other factor, increases the income earning opportunities of the poor and contributesboth to individualand nationalproductivity. Accordingly, during the past two years, the Government has been attempting to foster such development in human capital by restructuring governmentexpenditures in favor of the social sectors and rural infrastructure. Notwithstandingthe severe resource constraints, the Governmentmust continue to ensure that these priority programs remain protected. Not only should more resources be channelled towards primary education and primary health care, but the efficiencyof these expendituresshould also be improvedby ensuringthat money is spent on high impact programs and that the combinationof expenditureswithin and across sectors are optimal. The Governmentneeds to critically review its portfolio of investmentprojects in order to ensure that it is responsiveto the country's changingneeds. In other words, Government needs to ensure that expendituresare made on a rational basis rather than the allocated on the basis of historicallevels. The combinationof capital and recurrent expenditurealso needs to be improved.

45. Both the quantity and quality of desired social services is adversely affected by a lack of sufficient funds. In the medium-to-longrun, mobilizationof domesticresources, through improved efficiency in tax collection and also through the judicious adoption of cost recovery schemes in appropriatesectors, will be essential.This will need to be augmentedby shiftingfunds from relatively unproductive areas such as defense, state farms, teacher training colleges, universities, curative health, etc. to areas where the economic and social returns are the highest, i.e., primary health, primary education, agriculturalresearch and extension, rural feeder roads and rural water supply.

46. Policiesto Controlthe Growthin Population.Uganda needs to slow down populationgrowth in order to reduce poverty within the shortest possibletime. At present, the Governmentof Uganda does not have an explicit population and family planning policy and lags far behind other SSA countries in encouragingits citizens to have small, manageablefamilies and informing them of methods of doing this. In addition to aggravating poverty, the growing population has adversely affected the environmentby increasingthe encroachmenton forests and by intensifyingfarming, resultingin soil erosionand stagnatingyields. It is therefore imperativethat the Governmentdevelop a national family planning program and closely monitorprogress towards decreasingpresently high fertility rates.

47. InstitutionalIssues and the Provisionof Safety Nets. Althougha strong case could be made for this in the Ugandan context, targeting the poorest and most vulnerable is far too expensive an option for the Governmentof Ugandato consider at present, given the extremely low revenue effort and the weak administrativecapacity. Instead, the focus should be on providing fundamentalservices in rural areas: primary education,primary and preventivehealth care, rural feeder roads, safe, easily accessible water, agricultural extension, and marketing assistance. Communitiescan help provide - xxiii - some of these servicesthemselves, with assistancefrom Governmentand NGOs. Government,donors and NGOs need to develop a tripartitesystem to support self-helpprojects, and to ensurethat priority investmentsare undertaken. If Government wishes NGOs to follow Govermmentleadership in a scenario in which Governmentformulates policy and sets standards,NGOs, includingUgandan ones, should be involvedin policy formulation.For the tripartite systemto work, it must be 'owned' by a substantial portion of the three sets of actors (Government, NGOs and donors). The policy developmentprocess in the areas of AIDS and child welfare could become a model for other sectors.

48. The programof decentralizationbeing adoptedby the Governmentoffers a unique opportunity to support communityand NGO initiatives,particularly in light of the fact that the Governmentlacks the capacity to deliver much needed services to most of its citizens. For this new model to work, it will be imperativefor the Governmentto educate the populationand to reorient and educate civil servantstowards its changingrole. The District CommunityAction Trust (DCAT), which shouldbe designedas a social investmentfund whichdisburses funds quickly for projects generatedby self-help groups, could emerge as an important instrumentto address poverty on a broad scale.

Part I

A Profile of Poverty

'There are three distinct components to thc measurement of poverty. First, we have to spec6fy what we mean by the standard of living. Second, we have to delineate a critical level of the standard of livng below which there is poverty, by defiition. Third, we need to compress information on the standards of living below the critical level into an index of poverty. ' (Ravi Kanbur, Poverty and Development, PPR Working Paper 618, World Bank, 1991, p. 3.)

Background

1.1 With a per capita income of under US$170, Uganda today is one of the poorest countries in the world. Having seen better days at the time of gaining independence, it is a living testament of the havoc caused by the political turmoil and economic decline brought about by several years of despotic rule.

1.2 At independence (1962), Uganda had one of the most vigorous and promising economies in Sub-Saharan Africa (SSA), and the years following independence amply demonstrated this economic potential. Favored with a good climate and fertile soil, the country was self sufficient in food, with the agricultural sector being a large earner of foreign exchange. The manufacturing sector supplied the economy with basic inputs and consumer goods and was also a source of foreign exchange earnings through the export of textiles and copper. Export earnings not only financed the country's import requirements but also resulted in a current account surplus. Fiscal and monetary management was sound and the domestic savings rate averaged about 15 percent of GDP, enough to finance a respectable level of investment. Uganda's system of transportation was widely regarded as one of the best in SSA and included an effective network of roads, railways, port and air transport.

1.3 Uganda's social indicators were compnrable to, if not better than, most countries in Africa. The country's health service had developed into one of Africa's best and pioneered many low cost health and nutrition programs. There existed a highly organized network of vaccination centers, and immunization programs reached as much as 70 percent of the population. Although school enrollment was still low, Uganda's education system had developed a reputation for very high quality.

1.4 The Amin regime radically reversed the economic and social progress attained since independence, and the ensuing civil strife resulted in a tremendous loss of human life. It is estimated that as many as 500,000 Ugandans lost their lives during Amin's eight-year dictatorship and as many as one million more were internally displaced from their homes and farms. A 1985 estimate by the U.S. Committee for Refugees concluded that in that year one out of every fourteen Ugandans was either a refugee or was displaced; it was estimated that as many as 200,000 Ugandans had fled the country and were living in exile.

1.5 Economic mismanagement accompanied the civil war and professional standards deteriorated rapidly as skilled personnel fled the country. Between 1970 and 1980, Uganda's GDP declined by about 25 percent, exports by 60 percent, and import volumes by close to 50 percent. With large increases in defense expenditures, the government budget became increasingly untenable and was largely financed by bank borrowing which resulted in average inflation rates well in excess of 70 percent. Economic mismanagement and abuse of human rights on a massive scale continued during the Obote regime in the mid 1980s. By 1985 government expenditure on education and health, in real 4 Chapter1

terms, amounted to about 27 percent and 9 percent respectively of the 1970s levels. When the NationalResistance Movement (NRM) Governmentassumed power in January 1986, it inherited a shattered economy whose social indicatorstoday paint a dismal picture of the quality of life of its citizens and are indicativeof the extent of poverty within the country. The followingsections present a brief profile of poverty in Uganda.

DefiningPoverty

1.6 Being multidimensional,poverty cannot be reduced to a single indicator. However, in order to estimatethe distributionand depth of poverty, it is generally consideredacceptable to use real per capita expenditureas a proxy for welfare. A yardstick is then needed to determine who is poor and who is not and, for this purpose, a poverty line is drawn which defines the cutoff living standard below which a person is classifiedas being poor. Two approachesare frequentlyused. An absolute poverty line can be calculated on the basis of the income needed to satisfy minimal nutritional requirementsand a non-foodcomponent. Alternatively, a relative poverty line may be used, whereby a percentage of householdsat the bottom of the income distribution is considered poor, e.g., the bottom40 percent. The advantageof using a relative povertyline is that it avoidscomplex, and often controversial, calculationsof a minimumfood or commoditybasket, while still providing essential informationfor policy action and targeting of specialprograms.

1.7 This report has defined two relative poverty lines for Uganda. The first poverty line has been drawn at U Sh 6,000 per capita per month (approximatelyUS$110 per capita per year), which is approximatelyequal to four fifths of the mean per capita monthly expenditure in 1989/90 (U Sh 7,512). The secondpoverty line has been drawn at U Sh 3,000 per capita per month (approximately US$55 per capita per year), which is approximatelyequal to two fifths of the mean per capita monthlyexpenditure. Ugandansfalling below the U Sh 6,000 poverty line have been characterized as the "poor" and those falling below the U Sh 3,000 poverty line as the "poorest"or the core poor.

1.8 Preliminary estimates indicate that, in 1989/90, a minimum of U Sh 6,000 per capita per monthwould have a daily intake of 2,200 caloriesplus some reasonablenon-food expenditures (e.g., clothing, fuel, etc.).' This does not mean that people with monthlyexpenditures higher than U Sh 6,000 in 1989/90 lived comfortably; rather, it means that those with lower monthly expenditures could not satisfy their basic requirements.The lower poverty line of U Sh 3,000 represents the bare minimum for adequate food intake. At U Sh 3,000, if you are adequately fed, you do not have anything else at all, however essential (see also para 1.11). As the key social indicators discussed

An averagc intake of about 2,200 calories per day would allow an adult to maintain reasonable health and performance tandards, according to the standard established by the World Health Organization. People of different ages, sizes, and occupation may require different levels of caloric intake to be adequately fed. An 'equivalence scale" is therefom often used in order to tranform the number of people in a household into an 'adult-equivalent' number of people. For example, women and younger members of the household are often counted as having consumption requirements that are less than that of an adult male. In the analysis presented in this chapter, such equivalence cales have not been used and the data have been presented on a per capita basis. While equivalence scales can be derived from household expenditure data, this can be a complex and time consuming press. The imposition of a scale from another country, or a made-up equivalence scale is often considered unnecessarily arbitrary. Work is presently underway to estimate a Uganda-specific equivalenec scale. It is important to point out that the per capita measure presented here may lead to larger households with more children appearing to be poorer than they actually are which is of particular significance in aewsing relative umal/urbanpoverty. It is expected that the analysis presently underway will shed further light in this area. Chapter 1 5

Table 1.1: SelectedCharacteristics of UgandanHouseholds

All Non- All All Female Male Average Uganda Poor Poor Poorest Rural Urban Headed Headed

Real Per Capita Household Expenditure 7,512 11,810 3,485 1,845 6,885 11,760 7,491 7,517 Household Size 5.4 4.8 6.1 6.4 5.6 4.5 4.5 5.7 Dependency Ratio (%) 44 38 51 52 46 37 45 44 Average Age of Household Head 42 40 43 43 43 35 44 41 Female Headed Households (X) 22 21 23 25 21 31 100 0 Household Heads Literate (%) 77 80 74 70 74 93 69 79

Percent Shares in Total Expenditure: Food 67 66 67 58 67 61 70 66 Drinkand Tobacco 5 6 5 3 5 6 2 6 Clothes 6 6 7 10 6 7 6 6 Rent 3 3 4 7 3 3 4 3 Fuel 2 2 2 3 2 3 2 2 Transport 0.3 0.4 0.2 0.2 0.4 0.2 0.1 0.3 Health I 1 2 2 2 1 2 1 Education I I 1 2 1 1 1 1

Food Expenditure, as Share of Total Expenditure: Market Purchases 26 30 23 19 23 53 29 26 Own Production 40 36 44 39 45 8 41 40

Source: Staff calculations based on HBS data. below demonstrate, accessto basic social services is very low throughoutthe country, implyingthat some Ugandans above the poverty line could also be consideredpoor. The role of the poverty lines defined in this report is not to present a precise definitionof poverty in Ugandabut rather to illustrate where the poor are and what their characteristicsare (see Annex III for a brief description of the adjustmentsmade to the HouseholdBudget Survey data to arrive at these poverty lines).

1.9 Table 1.1 shows mean per capita expenditure per month and a number of household characteristicsestimated on the basis of a HouseholdBudget Survey (HBS)conducted in 1989/90.2 The table reveals that the poorer householdstend to be larger, have older and less educatedhousehold heads, and are more likely to be headed by a woman. Not surprisingly,the dependencyratio, which is calculatedhere as the proportion of the householdpopulation younger than 18 and older than 55, is quite high and increases for poorer households.The difference in average per capita expenditure between urban and rural areas is significant,with rural people spendingabout half as much as urban dwellers. Accordingly, indicators generally associated with poverty have higher values for rural Uganda: larger householdsize, higher dependencyratio, higher illiteracy, etc. (additionaldiscussion on female-headedhouseholds is presented in the followingtwo chapters).

2 Data in the 1989/90 HBS conducted by the Statistics Depaitment of the Ministry of Finance and Economic Planning consist of a tratified sample of 4,500 houscholds across Uganda, except for eight districts in the Noith and East which were not sampled due to insecurity. Expenditures were calculated adding the value of purchased goods and the estimated value (at mariket prices) of the goods consumed out of own production. 6 Chapter 1

Table 1.2: Poverty Indicators for Uganda

Population P. Indicator Contribution to National Poverty Share P0 P, P2 Po P, P2

For a PovertyLine of U Sh 6,000

All Uganda 1.00 0.55 0.03 0.01 1.00 1.00 1.00

Urban 0.11 0.38 0.02 0.01 0.08 0.07 0.06 Rural 0.89 0.57 0.04 0.01 0.92 0.93 0.94

Towns 0.07 0.38 0.02 0.01 0.05 0.04 0.04 Central Urban 0.01 0.48 0.03 0.02 0.01 0.01 0.01 Central Rural 0.29 0.49 0.03 0.02 0.26 0.23 0.20 North Urban 0.01 0.55 0.03 0.02 0.01 0.01 0.01 North Rural 0.09 0.81 0.06 0.03 0.15 0.18 0.19 East Urban 0.01 0.30 0.02 0.01 0.01 0.01 0.01 East Rural 0.21 0.70 0.05 0.03 0.28 0.32 0.34 West Urban 0.01 0.30 0.01 0.00 0.01 0.01 0.01 West Rural 0.31 0.47 0.03 0.01 0.23 0.21 0.20

For a Poverty Line of U Sh 3,000

All Uganda 1.00 0.19 0.01 0.01 1.00 1.00 1.00

Urban 0.11 0.08 0.00 0.00 0.04 0.04 0.03 Rural 0.89 0.21 0.01 0.00 0.96 0.96 0.97

Towns 0.07 0.07 0.00 0.00 0.02 0.02 0.02 Central Urban 0.01 0.14 0.00 0.00 0.00 0.00 0.00 Central Rural 0.29 0.10 0.00 0.00 0.15 0.15 0.12 North Urban 0.01 0.15 0.00 0.00 0.00 0.00 0.00 North Rural 0.09 0.42 0.02 0.01 0.21 0.21 0.20 East Urban 0.01 0.04 0.00 0.00 0.00 0.00 0.00 East Rural 0.21 0.30 0.02 0.01 0.35 0.41 0.42 West Urban 0.01 0.04 0.00 0.00 0.00 0.00 0.00 West Rural 0.31 0.17 0.01 0.00 0.24 0.20 0.23

Source: Staffcalculations based on HBS data.

1.10 Scatteredevidence suggests that the high dependencyratio estimatedhere is likely to be only one side of the coin, althoughavailable national data do not allow further analysis. The other (hidden) side is that, as a result of civil strife and the AIDS pandemic, there are now entire householdsmade up of "dependents".These are grandparentstaking care of their grandchildren,or minors and elderly people living on their own (see followingchapter for further discussion). While the phenomenonis probably more pronounced in areas heavily affected by AIDS and civil war, under present circumstanceswe can expect the dependencyratio to grow, putting an increasinglyheavy burden on the shouldersof surviving responsibleadults. Chapter 1 7

I . : -- : ..-- . - --.. : ....- . .- :-.:.-. .-... .-. . 1'"'"'""''''';;"''""''"""'''''"''''":'... .'... . '. -' ';'.".--.....:''-'--'o' ,: ...-q .: -. ,.-. . . --- ,, - - -, .:: ...... '-'.'-'.'.-'.".'.:v...... :-x t P. O ...... c or...... '.'-' ...... -heiMPt. .. me eoft o _ ; i eich ...... -..-.,. fall belowth poveay lIne, o ...the ~hea--otis~: ,-.- ., --- -...... -.;ndez--...-.--..--.--,-...... -P~ .Thahe. . . . -ivida - number;ae of i a ling *linebelow 0 the divided by the ~~~~~~~~~~~~~~~~~~mlationof the mea under oonaideration~~~~~~~U..71"u lu .

- -; HoEswee,naeaueote=41im...... l _¢f D S r. W a poverythhe.o.. onrt oesano Ihwtedfho ''ply'a,'Z, ,'-,", 'w . d...... 4...... X. 'd,'rAM''

implya11 upethe dfec i e ethexpenditure increasres ad texpovete foro thoe wonpre betopo :~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.g*o:--d.. : ...... tueholidswhe. .hncveatheitots.. . . sallo ,.~~~~~~~~~~,,...oweyer,perihdaupsbec...... ~~~~~~~~~...... the share...... of efeot for, thel ndonpt WovartyhWs interb e the ..... asede ot uppteo .. :.w.the.-povcrt:ty lne .-:n i thn, wIiedb thv.c. y In i osdul40to,pmv.e tchemasre proportonlg-:l belowpoverth ar H-/aline ty s , theri. Thasferfroo t. someoe bf the lonestoaletheaon co-v .Wt:etha worse poof '-.:0Vi.."M'':"o th le Agratn ' ths-' p ~ea th tota -Prota shral,- als nown Xs the poeat gap - l -- t. Fot oopit iIy betwee dffDrt n- arieaof.U. tbea teuic .s iiddb- ah poplto in thtara e-

iniaigta h orethv od ihu oo ohv n o-odcosmto.Cohshv a fairlyshare contant o expenditur for:difrnWemnso h oulto,bttesaeseto fuelas grows incomes decline. disturbing finding is that expnditure on drinks and.toba.co i almostas high or higherOthan on,"esetaitessc 4sclothnkn et hl h ors pn proportionallyless than the others. on...... suc . luxuries thysil pn adsrssnl hg ereto iniatndrinkan thuadtoaccothe poorescompardR ihatve towneacto do wisthout ecntofoodes~i to hveaty. ikhhee noncon consengs hcr umption aeagalCltheshav h mostimpotantof proeinssoures lbr oung chilren. Wors still, te share sent on edcation.an healtcombine fisvabout haf teaonsptondikadtbco...f .. alm posttighonar hoghrtfhan for esse ndivdalitem hosuchl as declbdthoingand rent. Wh in ife thepostpenadd * ptesurdproportionaehofallythasn suchfoesas ftlhntheylz bowther line againiesdistidsinglhih3preto byllsen *rneand thepbuatcon compeared tn uestio, wercnh nmik,vheeannnidro po e ggsywhichreacs crnsareguardyth thetdistribtiontore of Asshow,prlaotenghsepains it1.24-.30 cn noforyon,g be assumd beildrn.W that expendiuresln.T oresil the ondiaoIshr healt dscpiedtd edcaio..... on ...tedluteaturn e low because... asd suchservicesfree areof availablecharge. ~ ~ ~ ~ ~ ~ ~ ...... healthombine prisabo ua trhglf of theamun ipndat on, drinlcusanod I tobatctho.m'deouoa Ti

eAnsthata the-.3 na,show foith whnole ofsumendchatepniuson beealthada podulation-wihean avrae lof eas

povertliearvieaso struggaing.lThee food shrcfrteharstiresghaeht.o hepo 8 Chapter

Table 1.3: Expenditure on Different Food Itaes (As a Percent of Total Expenditure)

Central Northern Eastemn Westen Rural Rural Rural Rural

Bread and Cereals 7.1 17.6 15.1 8.2 Meatand Poulry 4.8 6.5 8.6 4.5 Fish 4.2 5.4 4.5 2.1 Milk, Cheese& Eggs 2.5 1.1 3.0 3.0 Oils and Fats 0.8 1.0 0.9 1.1 Fruitand Veg. 15.2 18.4 19.0 14.4 Matoke 13.1 0.9 9.1 13.4 Potatoesand Tubers 12.3 20.4 13.5 11.2 Salt 1.9 1.2 2.4 1.0 Sugar 0.7 1.5 1.1 0.8

Source: Staff calcuklions based on HBS data.

Poverty Indicators and the Regional Variation in Poverty

1.12 At a poverty line of U Sh 6,000, 55 percent of Ugandans can be defined as being poor (see Table 1.2). As one wouldhave expectedfrom differencesin averageper capita expenditures,the poor are disproportionatelyfound in rural areas, where about 57 percent of the population is poor, comparedto about 38 percent in urban areas. The head count ratio (P0-see Box 1.1 for definition) indicatesthat 92 percent of the poor in Ugandalive in the countryside,while only 89 percent of the populationis classifiedas rural. The discrepancybetween rural and urban levels of poverty is even greater using the lower or 'core' poverty line; 96 percent of the core poor live in rural areas. Optimistic inferences about the welfare of urban dwellers, however, should be tempered by the realization that life in the towns is more expensive than in rural areas, so that at equivalent expenditurelevels, urban dwellers will face a harsher existence. The higher cost of securing decent housingand sanitary conditionsis a case in point. Also, homelesspeople are more likely to be found in urban centers, but they are not includedin the HouseholdBudget Survey.

1.13 When the data are analyzed to take into accountthe depth of poverty, it turns out that not only is poverty more widespreadin rural areas, but it is also deeper.In other words, rural people are more likely to be poor than urban dwellers, and their poverty is more severe. Thus, while the *contribution' to national poverty shown by the head count ratio (PO)for rural areas is 0.92, the poverty gap index (PI-see Box 1.1), which measures both the incidence and the depth of poverty, is 0.93. The fact that the P2 measure (see Box 1.1), which places even greater emphasis on how far people fall below the poverty line, is 0.94 shows that poverty is deeper in rural areas. In other words, the very poorest are in rural areas.

1.14 The poverty gap index also allows for the calculationof the amount of money that would be necessaryin order to eliminatepoverty if perfect targetingwere possible ( see Box 1.1). The amount of money needed to bring everybodyup to the U Sh 6,000 poverty line, if there were no leakages, was U Sh 3,840 million per month in 1989/90prices or about US$5.8 million per month at the 1989/90average exchangerate. However, this number needs to be interpreted with caution. While Chapter 1

SUDAN UGANDA

/ ~~~~~INCIDENCEOF POVERTY'

>/ / es ; .\ \ g9A - HIGHESTINCIDENCE \ EXCLUDEDFROM HBS

ZAIRE

_ ; v _ ^ @ ~~~~~~NATIONALCAPITALl DISTRICTBOUNDARIES /P_A - INTERNATIONAL BOUNDARIES

{-rlA - *- 0 ~~~KENYA

TANZANIA

Map 1.1: Distncts with tHighest Incidenceof Poverty a monthlytransfer of US$5.8 million in 1989/90would have enabled all Ugandan citizens to spend a minimum of U Sh 6,000 per month, this would still not have been sufficientto provide everyone with access to basic social services. For example, since about 80 percent of adult women had not completedelementary school in 1989,4this implies that many of the women above the U Sh 6,000 poverty line did not have adequate access to education and therefore providing such a transfer to Ugandawould still have been insufficientto financethe primary education of all its women.

1.15 As Table 1.2 reveals, the highest incidenceof poverty is in the northern part of the country (see Map 1.1 for districts with the highest incidence of poverty). In the rural areas of the north, 81 percent of the populationhave a real per capita monthlyexpenditure of less than U Sh 6,000 and 42 percent have a real per capita monthlyexpenditure of less than U Sh 3,000. Such high incidencecan be ascribed for the most part to the ravages of the civil war. As persistent insecurityprevented the hardest hit northern districtsfrom being includedin the HouseholdBudget Survey, it can be assumed that the actual poverty incidence in the north is higher (data from other sources support this assumption,as discussed in the followingchapter).

1.16 Table 1.3 shows the regionalvariation in the patterns of food expendituresfor the rural areas of each region. The variation reflectsboth differentdiet patterns and differentlevels of expenditures. Hence, expenditureson matoke are highest in the central region, where matoke is the staple food,

4 Emmnuael Kaijika, Edward Kaija, Anne Crou and Edilbeto Loaiza, Uganda Dcmographic and Health Swwvy 1988/89, Miry of Health, Entebbe, Uganda, October 1989. 10 Chapter1 while expenditureson cereals are highest in the north, where millet is the staple food. The share of expenditureon cheap and fillingfoods (e.g., bread, cereals, matokeand tubers) is generallyconsistent with the regional distributionof poverty as shown in the Pa indicators.

Intrahousehold Inequality

1.17 The choiceof householdexpenditures as an indicatorof individualwelfare, and therefore of poverty, assumes that resources within the householdare distributedequitably among all members. In the case of Uganda, there is some evidencethat this might not be true and that men, particularly in certain regions, consume a disproportionatelyhigh share of household resources, often at the expense of women and children.

1.18 It was noted earlier (see para 1.11) that the share of household expenditureson drinks and tobacco is higher than on any other basic necessity, with the exceptionof food, and that this holds true at any expenditure level. Children do not generally consume any tobacco or alcohol; which leaves women and men. The fact that female-headedhouseholds spend about half as much on drinks and tobacco than male-headedhouseholds is telling (see Table 1.1). From the HBS data, as well as informationprovided by NGOs and districtoffices, it appears that men's drinking is often a problem and that men spend much of the little cash they have on alcohol.

1.19 As pointed out by a number of NGOs, the most troublesome aspect of intrahousehold inequality, however, is the inequitable distribution of food among family members. In many traditionalfamilies, the man is presented with all the food, and only after he has taken as much as he wants will the other members of the householdbe served. Nutritional studies in Kamuli district show that the man usually takes (or is given by his wife) half of the food, irrespectiveof how much food is there to begin with and how many other people have to eat.5 Older boys eat second, some times sharingtheir father's dish. Women, girls and small childreneat last. It is clear that this practice could be detrimental to small children and women, particularly in times of weather-inducedfood scarcity. The resulting poor nutritional status of women has serious implications for maternal mortality, low birth weight babies, problematic lactation, and infant and child mortality. When mothers do not eat well, breast milk may not be enough and thereforemay need to be supplemented by cow's milk. But poor people can ill afford to buy milk and they will water it down, thus starting the malnutritionand diarrhea cycle.

Nutritional Indicators

1.20 Nutritionalindicators are both a measureof healthand of wealth. Unlike economicindicators, which are more open to subjectiveinterpretation, nutritional indicators are better indicatorsof poverty because they represent the result of insufficientfood intake and/or inadequateaccess to health care, and also reflect sanitary conditionsand (to a minor extent) mother's education. For example, food intake may become insufficient because of the onset of diarrhea which prevents the body from absorbing some of the food. Diarrhea, in turn, is often the result of unsanitary living conditions, especially lack of access to potable water. The inabilityto obtain appropriate medical care, either because doctors and medicines are not available within a reasonable distance or because they are

Louise Sserunjogi, 'Vitamin A Intake among Oaildren below Six - lamaui Distict", Child Health and Development Center, Mulago Hospital, Kampala, March 1992. Chapter1 11 unaffordable, will delay the return to a healthy state. Mothers' ignoranceof simple actions to limit damage (e.g., administrationof safe fluids) may make things worse.

1.21 The Demographicand Health Survey (DHS) conductedin 1988/89found that 45 percent of children aged 0-60 months had stunted growth, as shown by their height-for-age.6 Severe growth retardation occurred in 20 percentof the samnpledchildren. Stuntingreflects the cumulativeeffect of chronic malnutrition over a number of years, and it is typically associated with poor economic conditions(as illustrated above). Although it is conceivablethat some families will fail to provide adequatenutrition to their childrenout of ignoranceor neglect, it is very unlikely that this would be the case for the majority. This assumptionis supported by the fact that stunting is almost twice as common in rural than in urban areas, that is, in areas that where the incidenceof poverty is much higher. Thus, we can conclude that, when the effect of insufficientaccess to social services is combined with that of limited economic resources, the percentage of Ugandans living in poverty appears to be higher than what a poverty line at U Sh 6,000 would suggest.

1.22 There are, of course, regional variations. The DHS found the prevalence of stunting to be considerablyhigher in the southwest (54 percent), while Kampala had the lowest percentage of stunted children (22 percent). The northern districtscould not be surveyed because of the insecurity in the area, but medical personnel in Gulu and Kitgum estimate malnutritionrates to be at least 60 percent. Acute malnutrition(i.e., wasting) is only 2 percent according to the DHS, but in Gulu, Kitgum and Karamoja it is estimatedto be as much as 30 percent.

1.23 Seasonalvariations in agriculturalproduction, and therefore in income or expenditures,are also important. Hospital admissionrecords and field interviews suggest that malnutritionincreases dramatically in the preharvest season. This has important bearings for the interpretationof data on householdbudgets, as the followingexample illustrates. If rural householdshave expendituresof U Sh 2,000 for three (preharvest or dry) months a year and of U Sh 7,000 for the remaining nine months,the frequency distributionwill show 25 percent of householdsfalling below the poverty line of U Sh 6,000. However, all householdsfall below the poverty line for three months of the year.

SocialIndicators

1.24 As the above suggests, data on mean per capita expendituresare not enough to understand poverty. When expenditureson health are very low, for instance,does it mean that people are on the average quite healthy, that the state offers very cheap medical services, or that people can not afford to spend money on doctors and medicineshowever much they need them? Social indicatorsprovide at least a partial answer, because they measure effective use of social services (e.g., immunization, school attendance, etc.) and/or the result of the accessibility and quality of such services (e.g., maternal and child mortalityrates).

1.25 The extremely low level of Uganda's social indicators makes it abundantly clear that optimistic interpretations of low per capita expenditures in health and education would be inappropriate. Life expectancyat 47 for men and 50 for women is one of the lowest in the world (Table 1.4), and it is difficultto see how the situationcould improve in the near future with the AIDS

6 Stuntingis said to occur when height-for-ageis two or more standarddeviations below the mean of the reference population. 12 Chapter I

SUDAN UGANDA / Y 's q INCIDENCEOF AIDS

M HIGHESTINCIDENCE

ZAI R E

( NATIONAL CAPITAL DISTRICTBOUNDARIES - INTERNATIONAL BOUNDARIES

{ ; XENYA ..... lENY

LQ Jfk ~~~TANZANIA

Map 1.2: Districts with the HEghestIncidence of AIDS pandemic threatening not only the adult population but also infants.' AIDS has emerged as a significantcause of death and illness among young children. Thus, the already high infant and child mortalityrates (respectively, 117 and 180 deaths per 1,000) could possiblyrise. Areas where peace has not yet been completelyrestored are at a greater risk of HIV infectionbecause of the presence of both regular and rebel armies.

1.26 Not surprisingly, insufficientaccess to health and sanitationservices is also reflected in the general mortality and morbiditypatterns. Uganda's crude death rate, at 20 per 1,000, is about twice the level of the average low-incomecountry (e.g., neighboring Kenya) and considerablyabove the SSAaverage. While such high rates can be partly explainedby civil strife and AIDS, it is undeniable that poor health conditionsare part of the explanation. Malaria has been found to be the principal killer among adults admitted to hospitals, far ahead of AIDS. Diarrhea, pneumoniaand anemia are almost as commonas AIDS as reported primary causes of death. Amongchildren under five, who account for over half of hospital deaths, the main killers are malaria, pneumonia, diarrhea and malnutrition.8 Availableinformation on morbidity, as measured by outpatient statisticsand surveys of mothers, confirms a similar pattern. Thus, the main causes of mortality and morbidity, with the exceptionof AIDS, appear to be related to a generally unhealthyenvironment and a lack of routine

For a detaileddiscussion of AIDS in Ugandasee Uganda: The Economic impact of AIDS, Populationand Hunan ReaourcesDivision, The World Bank, 1992.

See Uganda: Social Sector Strategy, Population and Human Resources Division, Eastem Africa Department, World Bank, 1993. Chapter 1 13

Table 1.4: Uganda Social Indicators

PovertyLines 1990

Upper PovertyLine (U Sh) 6,000 HeadCount Index (%) 55 Lower PovertyLine (U Sh) 3,000 Head Count Index (X) 19 GNP Per Capita(USS) - 1991 170

196S 1970 1980 1990

Gross EnrollmentRatios Primary, Male 83 46 56 80 Primary, Female 50 30 43 63

Infant Mortality 119 109 116 117 Under 5 Mortality .. 180

Immunizedfor Measles (9G) 60 Immunizedfor DPT (%) .. 60

Child Malnutrition(96) 45 Stunting Prevalence(%) 2 Wasting

Life Expectancy: Male (yrs.) 47 50 48 47 Female/MaleRatio 1.05 1.03 1.04 1.02

Total Fertility Rate 7.0 7.1 7.3 7.3

Source:Data provided bygovernmnent aahorities. medical intervention.

1.27 In the absence of a data set with information about both health indicators and per capita expenditures,it is difficultto determineto what extentthe poor differ in their mortalityand morbidity patterns from the non-poor. Scattered evidence appears to confirm what common sense would suggest:poor people suffer from higher levels of morbidityand mortality(with the possibleexception of AIDS). To begin with, they often cannot afford to pay for treatment, as even in public health centers and clinics it is difficult to see a health professional or obtain needed medicines without paying a "fee". To the direct medical costs, one must add travel costs, which tend to be higher for poorer people, since they are more likely to live far from health centers and hospitals. Even when treatment is absolutelyfree, as in the case of a number of pediatric nutrition centers, mothers may be reluctantto take their malnourishedchildren because they will then not be able to feed themselves while their children are hospitalized.Medical personnel report that it is not uncommonfor mothers to end up eating part of the rations destined to their malnourishedchildren because otherwise they have nothing to eat. 14 Chapter 1

1.28 It is well known that maternal education is a major determinant of good child health, especially among the poor who cannot afford paying (literally)for the consequencesof ill-advised child care practices. For example,it is widely believedthat pulling off the roots of the lower canines of an infant at the first occurrence of fever will guard against future occurrences. The infections resulting from this operation can cause not only more fever, but also serious nutritionalproblems. Taboos related to nutrition, more commonamong the uneducatedpoor, may have deleteriouseffects on health. The widespreadbelief that chickenand eggs cause barrennessand thereforeare not suitable food for womenmeans that women and girls are deprived of a main source of protein. Uneducated people are also more likely to seek the assistanceof self-appointedmedical specialists and traditional doctors, whose practices are often questionableif not outright dangerous.9

1.29 But just how uneducatedare the poor? The DHS found that about 38 percent of the women had no educationat all, while 43 percent had received some primary educationand only 19 percent of the women surveyedhad completedprimary school. Since accordingto our core poverty line 19 percent of Ugandans can be classified as poor, it is very likely that almost all poor women are illiterate. The negative relation found between educationand expendituresconfirms this hypothesis. Low educationalattainment among the adultpopulation, however, is more an indicatorof past levels of access to educationthan of the present. Enrollmentratios, especially in primary school, therefore represent a better measure of current access to education.

1.30 Government as well as private sources agree that official data on school enrollment are inflatedin response to incentivesfor over-reporting.Knowing that the official enrollmentfor primary school is 80 percent for boys and 63 percent for girls, therefore, allows us only to say that the true enrollment is lower. How much lower is impossibleto say. Reported rates in the poorer northern districts, for example, are as high as 80 percent. But interviewsconducted by the missionwith local school authoritiesand with parents suggest that the true rate may be close to 40 or 50 percent, with between one fourth and one third of the children never attending school at all. The estimates of teachers and governmentofficials in Gulu and Kitgum put it at about 30-40 percent; in Karamoja, enrollmentis believed to be as low as 20 percent. Needlessto say, enrollment drops between grades P1 and P7, and it is significantlylower in secondary school. The gender gap widens followingthe same pattern.

The Voice of the People

1.31 Thus far, poverty has been measured and described above on the basis of criteria defined by various experts, that is, by people who study poverty but have no direct experience of it. But how do the poor themselvessee poverty in Uganda?What are the criteria they use to determine who is poor and who is not? What are the causes of poverty, in their opinion?This section tries to present poverty from the point of view of the poor. What follows, therefore, is highly subjectiveand it could not be otherwise. It is based primarily on evidence gathered in the field by the mission during a "rapid poverty appraisal", listening to what peasants living in remote villages had to say and observing the conditionsin which they live. Because evidence from participatory rural appraisals showsthat childrenare more likelyto offer unbiasedinformation than adults, rural schoolswere also visitedto obtainthe views of children. Older ones were asked to write and/or discusstheir thoughts, and little ones to draw pictures of the poor (see Annex II for how childrenin Uganda view poverty).

In some instances,traditional doctor. have appaentlytrted AIDS by bleedingall adultmembers of the village, using the sameblade for theincision. The consoquences can be easily guessed. Chapter 1 15

The understandingof people who have been working with the poor and listeningto them for years, such as missionariesand NGO workers, supplementsinformation collected first hand.

1.32 As a broad generalization,the poor were identified as those unable to pay for school fees, and/or to buy soap on a relativelyregular basis. Shoes were of course out of question, sugar was a rare luxury and so was meat. The average poor would have one set of clothes for every day wear, often in quite poor condition, and a better set for special occasions, e.g., going to church or to funerals. Some peasants thought that a poor family may be as rich as to own one or two heads of cattle.

1.33 In spite of the fact that Uganda is blessedby fertile land and a good climate, several people mentionedthat they did not get enoughto eat, particularlyin the northern districts. People in villages often identified the poor as having to skip meals more or less on a regular basis, and indeed during field work very seldom was anybody seen eating or cooking during the day."0 In the poorest districts, skipping lunch is normal for the majority of rural people.

1.34 During field work, it was not uncommonto detect an element of personal blame attached to poverty. More than once it was reported that poor people are lazy, that they drink instead of working hard, that their compoundis poorly kept, or that they make foolish choices (e.g., plantingthe wrong type of crop). Ubiquitouswas the blame on the civil war and a feeling that life today was harder than before."1 A number of farmers, many of them RC1 members, also felt prey of middlemenwho take advantageof the situation by offering very low producer prices, which have to be accepted for lack of any alternative.

1.35 Whilethere mighthave been somedisagreements in identifyingthe poor (e.g., can somebody who owns a bicycle be consideredpoor?), there were no doubtsin determiningwho were the poorest, neither in terms of their characteristicsnor in terms of their identity. In this sense, the commentof an NGO using a participatoryapproach is telling: "Why do you want to waste time coming up with specific numerical indicators?If you go to a village, everybodywill be able to tell you immediately who are the poorest and where they live. We have done spot checks, and we have found that the communitiesare much better than we can be in identifyingthe desperatelypoor" (see Chapter 8).

1.36 Interestinglyenough, the poor were also identifiedas people who worked on somebodyelse's land. Only under extreme duress would the average farmer consider working as a day laborer for wages; conversely, being able to hire someone is a sign of wealth. It appears, for example, that at least in some cases it would have been conceivablefor a family to scrape together enough moneyto send a child to school if the father had "lowered' himself to work for someone else in order to have

A nutritionist at Mulago Hospital explained why studies may portray a much rosier picture than the reality: "When you go for a nutrition survey, you ask people what they ate the day before at lunch, and they will describe for you a healthy meal. But then you look around and you notice that it is lunch time and there's no sign of eating-not a speck of fire, no dirty pots or dishes, not even smoke. They told you what they know they should be eating, not what they ate. Yet, the average enumerator will simply record the answer without realizing that it is inconsistent with what she sees." (Louise Sserunjogi, personal communication, August 1992).

However, it appears that in most cases people were comparing their life now with their life in the pre-Amin era. When pressed for a comparison with the more recent past, many admitted that at least now there were things available to buy, even though they were expensive. 16 Chapter 1 cash for the school fees (the mother generally would have no time for such extras). But the social stigma attached to working on somebody else's land would have been too much. Opinionswere dividedover whether it was more disgracefulto work on someoneelse's field or to be unable to send one's own childto school. One way or another, it was agreed that poverty takes a heavy toll on pride and self respect. 2

The Poor and the Vulnerable

'In this village,we are aU poor. 7hereare tines when evetybodygoes to bedhungry andfew of us can afford to sendour children to schooLHaving sugar in thehouse or a bar of soap is a sign of wealth.But it is truethat someare poorer than others. Youcan teU becausethey are dressed in rags and their houseshave no door and a kaking roof Thesepeope sue banana leavesfor blakets because they cannot even light a fire. (A poor farmer in Akworo village in Kitgwn district, as narrated to the World Bank mission,August 1992.)

Who are the Poor and Vulnerable?

2.1 Given Uganda's legacy of nearly two decades of civil strife and the severe abuse of human rights, the poorest segments of Ugandan society are the ones who were unable to escape the devastationof war and who, if they were lucky to come out alive, lost all sources of livelihoodand had to scratch a living from subsistenceagriculture. While they are now able to feed themselves, given Uganda's fertile soil and favorableclimate, the resurgenceof preventablediseases like malaria and diarrhea has meant that their bodies are usually not able to retain the calories they consume. Moreover, the AIDs pandemic has intervened to further exacerbate the situation. The following sections are devoted to the various groups of people who comprise the bulk of the poor in Uganda and the causes of their poverty.

2.2 Orphaned, Displacedand Abandoned Children. Resultsfrom the 1991 census identifyabout 784,000 orphans,"2 or 10 percent of all childrenbelow the age of 15. The majority of the orphans are believed to have resulted from AIDS and civil strife and are concentratedin the war-ravaged and in the areas with the highestincidence of AIDS, such as the districtsof Masaka and Rakai (see Annex 1, Table IX. 1). The situationis expectedto deterioratefurther, and accordingto projectionsprovided by the Director General of the National AIDS Commission,the number of orphans due to AIDS is expectedto increase five fold in the next five years.

2.3 The situationappears even more dramaticconsidering that childrenwho have alreadylost one parent can expect to lose the other within a few years. Thus, although cases of families headed by a minor are still relatively rare, they are likely to multiply, as dependencyratios climb to figures unmanageableby the survivingadults (Box 2.1). In addition, in some instanceschildren are advised to stay "in possession"of their parents' property, rather than to move in with relatives or neighbors, for fear that unscrupulouskin would misappropriatetheir inheritance.'I

2.4 Parental death is not the only reason for children living on their own. According to NGOs working with street children, economichardship is the main reason for the apparentincrease in their numbers, with certain societal attitudesmaking things more difficult. For instance, it is commonfor women who divorce to have to leave behind their children as a conditionfor remarrying (Box 2.2). If they can afford it, the childrenare put in a boarding school or in the care of relatives. However,

Deined as individuab below 15 yeas of age missing one or both parnts.

13 Misappropriation of inheritance is also common when the mother is dil alive because of the inferior legal iatus of women. Thus, children who lose their father are generally more at risk thn thoe who lose their mother. Legal ismes are further discussed below and in Chapters 3 and 7. 18 Chapter 2

,7.A,Xi,E .X-E-:i-E-SE-...... - . i , EE -- .- i .-. . .: .:-.... -E--,-... . -Si-. .-:..- .. t-'...---- .. --:-- : ...... ' .,. ,- ...0,,' , -'" " , " ...... ' .".'. '.. " ''f" ..'0" .. ... U" ...... ' .' ...... '' '..'..'.'".""..".'..

There are.. of:.v em . ..t... .gir... d o11...e y

comleed witho:utwndws, the kthe :ia colpez to: th gri n h ean of an M:doa stn tin0 the area behn h househef4 chlrnhv ml ltmesrn Gb ( ers whc thygo bananatsond the al.sogrowy ben n oaoss soo-naa thrpaens ie, hy drpe otit schoelof it:becuebrews0 monfeyl0The cely the kno :ii a iaeslucl .eatv holve nohr onty

. : : .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

funeral nIes f ttheircom onky.The tre eder cldentaket: ''rn in sleina 0 to kee com"pany

They arBehcke uponby neighors, recve git ore od- and a- agklua offe ha encouragedthe ;ito star i small h,o,ticultu enerpis to bringin se cah Thirlee of life onprafvrbywt -thatof thronrpa chidrn of the annieag and cicmtne fi the ara -hi-rt-not :tne

t--;Soure.oy ane Pier*00BlaIe, :-S hs l:rca19,0 (hW lrPrr0 London. these solutions are often not availableand the children are left to fend for themselves.

2.5 Other children may actually be sent to the streets by their parents or guardians to earn their living. For example, young girls (6-10 years old) are sent from city slumsto the marketto pick beans that might have fallen from lorries or under the stalls, which are then resold at a lower price; profit for a day's work averages U Sh 800-1,000 (about US$0.75), but may be as high as U Sh 1,500. A small survey of street children in Kampalarevealed that about two thirds of them were in regular contact with at least one parent (generallythe mother) or a guardian/relative,and often contributed money to the household;only 37 percent were completelyon their own and slept in the streets.14

2.6 Childrenon their own are a commonphenomenon in post-war (and war) situations.Finding themselvesviolently separated from their family and not knowing where to go, displacedchildren head for the cities in search of a way to survive. More than twenty orphanages were opened in the Kampalaarea during the civil war (1981-86) to shelter children escaping from the atrocities of the Luwero Triangle.1 5 Now that peace has returned to most of the country, displacedminors tend to be concentrated in the northern region, where there are still pockets of insecurity. It has been estimated'6 that there are about 2,000 street children in Mbale, most of whom have fled there to escape violence in Karamoja, Kumi and Soroti; it is also believed that large numbers of displaced childrenhave sought refuge in Lira. In Kampalait is estimatedthat there are over 800 childrenliving

David Muyayisa, tSfreet Children in Kampala", unpublished manuscript, August 1992.

Hon. A.A..Bwani Bbaale,.- h Rolcof th UgandanGoernment in Nonal StrategyFormdadaion", paper preented to the Working Group on Children Orphaned by AIDS in Africa, undated. .By representatives.of the Ministryof Labor and SocialWelfare at the districtlevel in collaborationwith Fend of Children Asslociation. Chapter 2 19

:ox .M2; AwWoman WIo was anOrp .an '..':-' .'.-:.-' . ". ". ' ' . :- .:- :-.. -'' ' : ' : .'' ' - -: " ...... ante ma. He tol mywohrtath oil mai7 hq onl Yf-h :e id

:-Myonly other mlativ~ was aty granduxothot, an oldwotnan who died when. .wa... .. s.... y -own-my:-witap-o f .. U as: f - - - v .

.:"br'.,','s.....''''".X.'w'-'','''...... '.'I lie nearbyma d ey s waI Xbaain dO m otws. Faile to o. sa hewh'l f bin...... a crsed,.. I f ... .ey I':'

,-,',nothngaboutwohi gmy an,estors' ' ,' -'.o' 'y h. WA w af'-id 'le w b nosed mei'e "ent 'id o'r 'ir awy. ' who:m

'f.o,rme. P .. I " bad, no ,choiebut , ' toretun t thelandtha. - .' ' .y ' 'rnniteadef: ..... ' ... '...... Iwsveydapotd.. '" ''". ".--- and 'wa sr tha wo.l beiursadby. my aet. rwn -sm odo a ln n cln ::ia . .-off. e sas to m my.. -.-c...... --~~.. . a.. -.--.- --..--. . . Nby.-.- ---.-,..---....::.----. t a e n e ag -:y s-rvic as a hiousegms the me to, beas noIha chde to bo kt.Nwhe tisar -- u yeaaw:ld adlam uhn e itesaewya beore eaigicmfe o.y adn Sria ---tough for: and so far Ihaeno owof... :vr d.. g :. -. :: :

.": few. . menrA: have appwachcd mce,wanting -o marry but an ...... eh...lg.rn

Soxce: 'ch n ad Womenin A, A n4~F k¶. pa, 99.: p.-:.

on the streets.'7

2.7 The Handicapped. Partly a legacy of the civil war, but also the result of inadequateaccess to health care, the numberof handicappedpeople in Ugandais very large, althoughno firm estimates are available.'5 In 1977, WHO estimated that 10I~~~~~~~~~~~~~~~~~~~~~...... percent of the Ugandanpopulation was. disabled, but continuingviolence and the breakdownof immunizationprograms are likely to have caused an increase. The National Union of Disabled People of Uganda (NUDIPU), an umbrella organization for NGOs operating in this area, estimates the percentage to be close to 15-17, excluding the emotionallydisabled. Yet it is easy to imagine that the horrors of the civil war have seriously traumatized large numbers of people, who are now unable to lead normal productive lives. It is possible that the number of the disabled could be as high as 1 to 1.5 million.

7 CharlesTuhaise, "Need for Foster Care in Ugana in The VuhserableOald (UCOBACNewsletter), January 1992, p.4 .

The 1990 Census containeda questionon disabledpeople, but findingsare likely to be unreL;ablebecause the definitionused was vague ("unableto supporthimself'') and the enumesetorswere not properlytrained, e.g., they were instructedto rely on observationor on questionsto the head of household. 20 Chapter 2

l:...... -.E.E.-.-.t. -'.. ..-. 'h...... , ...... :h ...... Iwa...oy...... I o .poio..My...... faytouhti... .a.....h.lth...bu.w '. a se.:.ol

-a o di t-- Wa hfiseo.A wha ar yo dig tohl ii is~te thebss Tht' save my lie l wl takititso the hospta, heeit tunedout tht* alliil"0;I nede owakwsseprtc d tagsml ace,-That's- sidlSl needed t be likethe:others, util T..T.7.:..br;t ..:...... :..:.:.:...... :...... fbu. .. .w:1:s:...... fouticn spet :myidays segrgaeat...... hote,%-...... :T..

: -;0gktatLoo:00-- meXno. gIstuid,1gotaob Igtnrid a uprtnslsdyfml.l s.e venhelopout mv rentsfzmimeacXtotme. And lheysused te ea:shamed of:met IhkavebecomeTthcuf ltg -favorieso,adevsen wie's etivyle's -hfhavesatd to ccptme. !Hersstes se l:otinsh wascray lot0ilhav iarried-me,and teypMitdher. ut- now they Ssee Xthtw0have twyo heatycidrentatI doa't l lrn nd--gIi'm: kin to her,+0that I+ca- roidefo he an the0i;0i00-hitldren,00 lThey 0 envy-he!

2.8 What makes the handicappedparticularly vulnerable is not their disability, but the attitude of people. Accordingto traditionalbeliefs in several parts of Uganda, disability has a supernatural origin. Disabilitysymbolizes a curse which has strucka particular person but is most likely weighing on the whole family. As a result, those afflictedby disabilitiestend to be shunned by society, while their families try to keep them hidden to avoid the shame and possibly the ostracism of the crommunity(Box 2.3). Often, physical disability is assumed to go hand in hand with mental deficiency,so little effort is made to educate handicappedchildren or even to treat them as persons. Shouldthe attitudeof the family be positive, assistancemay still be lacking because of the multiple demands on the time of women, who traditionally are responsible for the care of the disabled.'9 Poverty compounds constraints to the provision of even the most basic assistance. For example, handicappedpeople, and children in particular, need to wash more often because they spend much time on the ground, but soap is a luxury item to be used sparingly for the poor. The situation of disabled women is even worse. It is probably not an exaggeration to say that they are likely to represent the most miserablegroup of society, doubly discriminatedagainst because of their gender and because of their disability.

2.9 Female-HeadedHouseholds.7' In both industrial and developingcountries, female-headed householdsare disproportionatelyrepresented among the poor. In the case of Uganda, while the HBS data reveal that approximatelyone fifth of all householdsare female-headed,it is difficult to test this hypothesis,as cultural norms discouragewomen from consideringthemselves without a partner even if that is the situationde facto. Polygynousrelationships further complicatethe issue. It is estimated that as many as a third of all marriagesare polygynous;it is not uncommonthat men would have an official wife in one village and other "wives" in other villages. Because of the general level of impoverishmentin the country,only the richest men are able to provide adequatefinancial, let alone

Herbert Muyinda and Tom Barton, "Soda-EconomiucInfluences on the Health and Rehbiltto of Physicully Handicapped RuralChildren", unpublishedmanuscript, Child Healthand Development Center, Makerere University, Pebruary 1992.

For a comprehensive discussion on women's issues, see Chapter3. Chapter 2 21

Box 2.4: Meet the Core Poor.

-t was diffi4ukto guess how ot 'she could have beenmand she did not knoWvAlR hei ten childrenm.i. were grown up, with their own children. Some migt even have grandchildr but she did not know> ::eausc she had not see any of then in years. Wat she did knw vwaspoverty.Absolute poverty. ..

Wh. n aikedto defu3ea poor Mron, shesimply said: 'Look at mc and look at whiaI havo", SheC" :.was very old, blind andA nlyIaty coe by rags. Her belongingi in the im down hut consistedof a .. papyrusmat to sleep onn.Nothing else. No clothes, no bance, no tools, To :eepherelf warm atnight she -usesbaa leaves. BecOAuseshe is blindand wk, she cannotfetch woodor keep a hre, sosfhecannoteoo In-anycase, she would:ot havepoa ia whih to cook..The neighbors bring her cooke food, bii ri the pOm harvest season, whenood is scarce -r evaybody. she y brdays uitht ybdy co see r with something to eat. She does try togrow sone milet, using a harppiece of woodas a ho (aiecanriot :afford a hoe), but in her oonditionthere iu little she can do. Foriunately: youthassociation-d by the RCcomestlo help het fromtitne.to time, bringig her water and working on het field. If it- iotter them,-she probablywould be dead.

Source:Rapid PovertyAppraisal, Aygioi 1992. emotional, support to more than one family. Most women in polygynousmarriages are, for all practical purposes, heads of their household,in the sense that they are principally-and often solely -responsible for the welfare of the family.

2.10 Widowed and divorced womencan confidentlybe assumedto be at risk because of legal and cultural traditions,which in many cases effectivelydeny them access to the resourcesonce controlled by the household. Whether these women are allowed to keep their children or not, their situation upon separation from their husbands tends to be critical. In the case of widows, relatives of the husband (generallyhis brothers)will claimrights on the householdproperty, often evictingthe widow and her children from their home. In some areas, widows are 'inherited' by brothers-in-lawalong with property, thus affording women the possibilityto maintain the usufruct of their household property and providing them the protection and status deriving from a husband. This practice, however, appears to be dying, certainly in part because of the AIDS pandemic. Unclear inheritance laws and the attitude of the courtsare a formidableobstacle for the few womenwho dare to challenge their in-laws. In the end, the big losers are the children, who find themselvesnot only without a father, but also destitute.

2.11 Older Peopleand ThleirDependents. There is generalagreement that the elderlyhave become a particularlyvulnerable group. Base-linestudies conductedin a number of districtsby various NGOs using different methodologies, all arrive at the conclusion that older people-alone or with dependents-are disproportionatelyat the very bottomof the socio-economicladder, barely eking a living (Box 2.4). How many of them are in a desperatesituation is impossibleto say, as counting all heads of household above a certain age would undoubtedlyproduce an overestimate. It is clear, however, that this is not a negligiblysmall group, that its size is growing, and that uncertainty about prevalence should not be an excuse for inaction.

2.12 Traditionally, older people have been cared for by their children, thus making specific assistance for the elderly unnecessary(of course, an argumentcan be made for the opposite, as the lack of reliable pensionschemes has made high fertility rates the only form of insurance for old age). But the situation is changing as a result of the AIDS pandemicthat is claiming most of its victims 22 Chapter 2 among the middle aged who are generally supposed to provide for the welfare of the elderly, therefore effectivelywiping out their 'pension scheme". To make things worse, many old people find themselveswith the added burden of having to take responsibilityfor their orphaned grandchildren. Already too weak to take care of their own needs, they are obviously unable to provide adequately for the needs of their grandchildren-both materiallyand psychologically.

2.13 Landless Peasants. Common wisdomhas it that land in Uganda is fertile and in abundant supply. While this may be true in general, there are areas of high populationdensity, especially in the east and near the Rwanda border, where shortage of land has become a serious concern. Unfortunately,it is difficult to estimate precisely how many people suffer from land shortages, as land tenure is not always clear and sometimessmall plots under cultivationmay be the result of insecurityor labor scarcity rather than lack of access to land.21

2.14 Nevertheless, there are sufficient indicationsof distress related to land scarcity to warrant special attentionto this issue. District authoritiesin Mbale report that crimes in connectionwith land disputes (mostlyhomicides) have shown a dramatic increasein recent years. A participatorypoverty appraisal conductedby Action Aid in Mubendedistrict found landless and semi-landlesspeasants to be a sizable group in the poorest of six categories identified by villagers.' Given current demographictrends, the situation could worsen in the future. It is worrisome, for example, that school children interviewedin a land-scarcearea near the Kenyan border showed full awarenessof the adverse consequencesof landlessness,but had firm plans to have at least six or seven children "because it is good to have large families".

2.15 Becauseland shortage is a limitedproblem, the solutionshould be at hand, as families could relocate (or be helped to relocate) to scarcely populated areas. Two factors, however, make the solution not as straightforwardas it would seem. To begin with, it is difficult for landlesspeople to even know about areas where land would be available, let alone to actually get there. If a trip to the county health post is only undertakenfor very serious illnessesbecause it is too expensive, moving a whole familyhundreds of kilometersaway (presumablyup north) is definitelyout of reach. To put it in economicterms, the rural labor market is plagued by informationand mobility imperfections, due partly to a weak road and transport infrastructure.23

2.16 Second, relocating is a very risky move for a poor farmer. Interviews with peasants in Bulambulicounty, Mbale district, where the average plot size is three fourths of an acre for a six- person household, made it clear that moving away from the village is consideredclose to suicide, because it meanslosing the safety net providedby the extendedfamily (see Chapter 7). Since poverty is perceived as having both an economic dimension (no resources) and a social dimension (no relatives), it makes little sense to try to escape one of its aspects by embracing the other.

2.17 The North. Identifying a whole region as poor and vulnerable is somewhat of a generalization, as not all of its inhabitants are destitute. But the difficult conditionsin which the

21 For a discussion of the land tenure system, see the Uganda: Agricukure Sector Memorandum, Agriculture and Environment Operations Division, Eastem Africa Departnent, World Bank, 1993.

22 Action Aid, Analysis of Resuki of ParticipatoryPoverty Assessmentin MityanaCounty, MubendeDistrict, mimeograph,July 1992.

23 Alison Evans, A Review of the Rural Labour Market in Uganda, University of Sussex, May 1992. Chapter 2 23

::: .2: .. :. W...... e u Distr.ct) :--:.-.. . . :...... --.:Y :-, - :. -. erebelsame weThe -ntok - bd l th i: eeyt y nt alre-d tak. When.they a the te -- :f*ore;, they too h thi NGOh gi en us, thir r-sdteent -ka a a u wo hoe. c mayb:1ay, ttss why hy . Th waing e NM s t k h w tns t lo. atI why tey m.-eTwi. tl --0 gasveus somethwig-and twace VW Wbels amea tokita . N eee frd f gkl hing fromXhe NGOS*ea t wi th-.:-the :-bels. , . ..-.----- . . ,.-...... -.:-...... ---::-:.-l I: Thsas tiimewhen fthrebels: cal It:ne :Wbwha Nit-S-.oVO Ueyltlebut he till#took it.:Theytake our:fod 1 so nowwejustke littlebimt' granar'-ecaUe ty o' : . t a we haveston y W..h i teded WOne me SIMah on a -(dad, hut there -sn't -nouXhbr he-whole lamlly. n. a g. .. --.. --..-... -. . . -. . . . -. -,. ..--,.- :Th,eyevntook myiwife' dres-tat why nbc Aifnake6djutwain ak.n6 hycoeand :ake. BVai Ourwom y. irs .of -14 la ar th,:atthe 'el wiXi m0 an d taeUthem awayo ttey haedonewhmany they:sle te b far rom lhehoes."

| UC"e: Pid Po-wo -A Appraira4 iut 1.99,

majorityof the populationin this region appear to live warrant this generalization.While the rest of the country is fast recovering from the ravages of wars and insecurity, parts of the north are still strugglingto satisfy even the most basic needs. Large scale numerical data are missing,2' but there is enough informationfrom NGO reports, church missionsand field interviews to piece together an overall picture. It is a picture of suffering of people who in many cases have been left with nothing (Box 2.5).

2.18 To be sure, the Governmentshould be commendedfor havingbeen able to curb violence and restore peace and security throughmost of the countryin a very short period of time. However, there continueto remain pockets of insecurity where reports of abuses committedby rebels and soldiers continue,taking a heavy toll on both physical and mental health. The resulting insecuritydisrupts the regular production cycle of peasant households.

2.19 For example, the raids of Karamojong cattle rustlers have resulted in much looting and violence, causing many people to move away from their fields to the relative safety of militarycamps and towns. District officials identifiedvictims of cattle rustlers as the most desperate cases in their jurisdiction, at times on the verge of starvation; a large number of displacedfamilies living in dire poverty is also believedto be in Lira. Althoughgovernment forces have managedto containthe raids, and efforts are presently underwayto promote peace talks between the two main conflicting groups of Bokora and Matheniko, there is still ample evidence of armed thugs taking advantage of the situation and perpetuatingacts of common criminality.The violence of the past few years has left a large number of mutilatedor otherwisedisabled people (Box 2.6).

It wiln be recalled that the north and northeast were not includod ui the two national surveys,which form the basis of the povertyprofile, becauseof securityproblems. Results from the IntegratedSurvey conducted under PAPSCA are not yet available. 24 Chapter 2

f....,..-.-,...'-.',..',,..'.'.,-.'"""-t'''"' . '''''...... '''...'. '" '- '. : ...... i.:'.", ;.,:,'f.s,u...... , ,T

,'.... ,. .' ' -'.'...'itD'"-:-"- ..-.. "' i'"E"-..-.

.--.-.S...... ,,,,,y...',.i,-z--.. .. W :-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~," 1.( . . .,a .. , , , . , o ,.0. .,.. :...... :...... !,,~~o 2.6 - :v.b4,: Urn Uoni StorE- to Te b...... ,- . . . . . - .... , -- , - , ...--".

2.20Becau ~~ e bef isiseuity,narmerath in the fh northtet ave~ anotben th ple tofego to thae sielsan d Kitum'fooshortages St...... Joseh.Hopitl,.teyiar~en chr odic7 in s mome a .ae.8treapercent.....Medical...personnel....estimates. Ato LacrHsptal, the lrgbest hodsppa..malnutrition....rate...of3 ital .... thfin oth, 8 60-65prcent,with aout 30 ercentbeing cute m lnutriton (wating).Needles....sa,.malntritio Kitgum'inrassdrmSet. JsephOO tialydrigth...a.stsasn.I.arm Hnosptl, they are 8lld0 werentMedicalpersohanel ja hentitinlstaindtroaeeoimae sa como.alnurton rateo durg...the .....dr . eao, syon aut migrate t mor fetl landw ith the cate leavigbhn the65lweak-women th0percentg ith with abus young ha0perent children, beineg theacteae sick, ..malnuatrto.the.elderly.(welltibg). Reuret Neetdless.rogtsan.heineurtYo say, malnutito whichprevented~~~~~~~~~~~~~~~~~~~. many h. dsenfrm.rngngbak.h.ctte ro ptdlo..poitcan.t.rqus (andobtain)whatease s dramatica the establishment~~~~~~~~~~~~~~~~~~..the suringadof...an...emergency..... prhrvs sesnAnKaanutrtio.ceter oa[hentiinlSiuto dtroae duringthe2.21 Many pe~~~~~~~~~~oplsacor pl of, fidtesleswtoteentem yas ayoungeaduelts migratedt m oreapeople.YO s a iccommdies dland Uwith the catte,and thereforeleayvringehinunable... tewealto beome c-wmewtelf-ufficent.Ther youngad'.Te chalden is uthegreat.need.forhoes..At.U..h3,000.or.more.their.cost.i mn aicksat esheedry.Recubrrlethntdoghs. ahndcthe insecrt prohiitiv for...... motfres.e ihu hm i ilb ifcl to. increase.production....even.. if whicpsecurity werearev tooepte manyimprove. Cerdnyhsmen Cattle rustling and looting.p..ple's..ability have basically...... destroyed to supportthemselves~~~~~~ The threat~~. of atrombingies....HIV inecioh aveea keis the lolkl cabels tromptsed o eutiloca spporiticirom iigVobdt,tuthe requeat (adobu]tain) furhethe eodngpepl'shsablishmeydntel-sffcinc. of Watn emer $ enyNKrMoj,itrtia nuetr itin ounter.utngpol. fighting,.. cultra ooyunetn raiiosan

2.21~ Man~out of Iextreme peolefindl poverty.": to brinuteve e sle the erctoLPrnehigmotstotw peasie domodities,ome anod Fon ftftheiforeual lrooibtinv fan rapist faremees yeeitoders, austthey atwlllobt smedindic alt theinrase eproductionavny i scurisere com animpoves Catthleputln and laoeeoaethings haete Thcalya desnh nthoedcrosple's abiliy t supr hvemevsnooy To hethrean towfr toV ineto saoolkl.orsuti iigmobdt,tu

general lack : of inratrctreairewn eauerbatin th1iutini htsagalytepoes rao

2.20unryIBease ofeathas ineuity,u fatrmers eolin the northhave not bee able togouto themfieldes,n pecet of children' adisonorvormluriin(permabur0pecntefryh.wr;a 3

The Gender Dimension

The essenceof women 's distinctiveness les in themultiplicity of their roles. Most men can confine themselves main(y to being producers. Most women, in addition to being heavily involved in economicproduction, take prime responsibility as home managers, child bearers, and carers ofthe children and elderly.... In consequence, women work longer hours than men, usually with smaller resources,fewer opportunities, and lower rewards. Inequities,infact, typify gender differences. X (Mary Chinery Hesse, et al, Engendering Adjustment for the 1990s, Report of a Commonwealth Expert Group on Women and Structural Adjustment, Commonwealth Secretariat, London, 1989, p. 3.)

3.1 The preceding chapters have addressed the economic and social dimensions of poverty in Uganda. The analysis of the incidenceand depth of poverty in Ugandasuggests that it is essentialto distinguishbetween various types of households,to assessdifferent expenditure patterns and the issue of intrahouseholdinequality, and to recognizethat people are affectedby poverty in many different ways, as the portrayal of the lives of poor people in the country has shown. These various factors illustrate, in short, the gender dimensionof poverty in Uganda. Furthermore, as this chapter will argue, gender is important not just in the assessmentof poverty, but in the formulationand design of poverty reduction and economicgrowth strategies for the country.

3.2 While it may seem self-evidentto state that, in Uganda as elsewhere, poverty, economic change and structural adjustment affect both men and women, what is often not recognized by policymakers-and certainly not explicitly-is that they affect men and women in different ways, because men and women play different roles, have different needs, and face different constraints in respondingto economicpolicy changes and to shifts in incentives.' This combinationof differences arises from fundamental imbalances, or 'asymmetry",' in the respective rights and obligations of men and women, and translates into men and womenhaving highly differentialeconomic capacities, as reflected in their access to, use of, and control over, economicallyproductive resources. These differences have implicationsnot only for economicequity, but also, and much more importantly, for economicefficiency and foregone economicoutput and income.

3.3 By analyzingthe gender dimensionof poverty and economicchange in Uganda, this chapter aims to illustrate the asymmetryin rights and obligationsof men and women and its effects on the country's economic and social development. Since development interventionstend to be at best "gender blind", the chapterdoes shiftthe focus of attentiontoward women and emphasizesthe roles, needs, and constraints of women. This is not to ignore the fact that our chief concern is with the sociallystructured (gender)relations of womenand men, for, as Hilda Tadria has argued, "household gender and economicrelations are interconnectedin the wider economicsphere.... There are many

This is the coneptual rationale for addressing "gender as a critical dimension of developmet. See, Caroline O.N. Moser, 'Gender Planning in the Third World: Meeting Practical and Strategic Gender Needs", World Development, Vol. 17, No. 11, 1989.

24 Ingrid Palmer, Genderand Population in theAdjustment of4fican Econonmes:Plauningfor Caange,Women, Work and Development Series No. 19, Intenational Labor Office, Geneva, 1991, p. 12. 26 Chapter 3 problems ... men and women experiencetogether. Their activitiescan only be understoodin relation to each other."'

3.4 From the outset, it is importantto stress three factors that limit the analysis in this chapter. The first is that Uganda is a country of some diversity, and any generalizationsmust necessarilybe treated with caution. In particular, notwithstandingmany similarities,men's and women's respective responsibilitiesand burden-sharing arrangementsdo vary to some degree across the country, in particular as a function of the broad distinctionsthat can be made betweenthe Bantu societiesin the south of the country and the Nilotic, Nilo-Hamiticand Sudanic peoples of the north. The second is that many of the phenomena we observe, especially the economic survival strategies such as increasingreliance on sales of food crops to meet householdneeds for cash, have changed in scale if not in nature in recent years; it is thereforetoo early to reach any firm conclusionson the trends that are emerging.The third is that data are scarce, and sex-disaggregateddata all the more so. It is therefore necessary to combine qualitative and quantitative data, and case study evidence, to substantiatethe analysis.

Profile of Women's Roles and Constraints

3.5 The situationof women in Ugandahas been described in various documents.' What follows is a brief summary of women's multipleroles and constraints, in relationto those of men, which are of particular relevance to understandingthe dimensionsof poverty, and to the formulationof poverty reduction and economicgrowth strategies for the country.

3.6 Agriculture. Womenare responsiblefor producing80 percent of Uganda's food, and provide about 70 percent of total agriculturallabor.' They are principally but not exclusively confined to the unpaid subsistencesector, and carry out their agriculturaltasks without benefit of technological innovation, inputs, or finance. Many of these problems, of course, also apply to poor male farmers in rural areas, though not in the same way, since men provide only around 30-40 percent of the country's agriculturallabor, and men are not constrainedby competing claims on their labor time. Evidence from the north suggests that basic agriculturalimplements (hoes, for example)are scarce and costly for the poorest farmers.'

H. M. K. Tadria, 'Changes and Continuities in the Position of Women in Uganda', in Beyond Crisis: Development Issues in Uganda, Paul Wiebe and Cole Dodge, Eds., Makerere Instituteof Social Research, Kampala, Uganda, 1987, pp. 79-80.

See, in particular, Eva Jarawan, Womenin Development:Current Issues and Agendafor FurtherResearch, Population and Human Resources Division, Working Paper, June 1991; UNICEF, Womenand Childrenin Uganda: A Situadon Analysis, Kampala, Uganda, 1989; UNICEF/ACFODE, Uganda: Women's Needs Assessment Survey 1988; Agricultural Sector Memorandum, Agriculture and Environment Operations Division, 1992; Social Sector Strategy,Population and Human Resources Division, 1992, World Bank.

Data on women's time allocation, and especiatly in relation to the agricultural division of labor, are scarce. The figures cited here are in Jarawan (op. cit. p. 20). The Siuation Analysis (UNICEF, 1989, p. 77) cites survey data indicating that women provide 68 percent of the labor for food crope and 53 percent of the labor for cash crops. Interestingly, the same survey indicates the disparity of women's controlover the income from their labor, as women sell only 30 percent of fooderops and 9 percent of cash crops.

'In many homes ther are not enough agricultural tools for ach adult to have a hoe.... In as many as 4 perent of aural households, there mnaybe no agricultural tools at all." Womenand Childrenin Uganda: A SituationAnalysis, UNICEF, Kampala, Uganda, 1989, p. 21. Chapter 3 27

3.7 Household Management. Women have primary responsibilityfor household management, child rearing, food preparation, care of the sick and elderly, and family health and welfare. These tasks are also carried out withoutbenefit of labor-savingtechnology (for example, in the north, it can take two hours to grind sufficientmillet for a normal family meal), or adequate social service and transport infrastructure. Women's primary responsibilityfor these tasks is difficult to document in the absence of time-usedata, or data on how long it takes to perform certain critical tasks (such as fetchingwater or wood). Surveydata reinforcethe asymmetryof male/femaleobligations with respect to householdtasks, where the greater share of householdtasks is performed by women/girls:cooking (86 percent); fetching water (70 percent); collectingfirewood (73 percent); child care (62 percent); washing clothes (88 percent); care for the sick and elderly (62 percent)."

3.8 Workload. This combination of responsibilities means that women have a very heavy workload: women work longer hours than men, between 12 and 18 hours per day, with a mean of 15 hours;' there are no comparable data on average male working hours. In Kenya, it has been shown that women work half as many hours again as men on both agricultural and nonagricultural tasks (12.9 hours compared with 8.2 hours).' Interviews with district officials and NGO representativesconfirm a prevalent view that the time womendevote to work is considerablygreater than that of men-indeed, some argue that women are essentiallyworking throughouttheir waking hours. Women are, therefore, if anything,overemployed, not an economicallyidle resource awaiting supply response incentives. This reinforces the argument made in Chapter 6 that a more accurate picture of labor force participation is made by including in the labor force those who are predominantlyengaged in "householdactivities." If "leisure time" is a valid indicatorof well-being, womenare very poor indeed.

3.9 Health. As a result of women's heavy workloadand time burden, women's own ill-health is especially significant. The Needs Assessment Survey established what it described as an "alarmingly high" morbidity rate for rural women-at 76 percent-in the two weeks before the interview.' This has particular significancewhen viewed in conjunctionwith the high total fertility rate at 7.3 live births (see Table 1.4 for selected social indicators).Contraceptive use is constrained to a low level (5 percentamong womenof reproductiveage), becauseof lack of access, and generally unfavorablemale attitudes. With the high fertility rate, most of a woman's reproductiveyears are dominatedby pregnancy, childbirth, and child care, leadingto health problems. Women's low health status greatly constrainstheir capacityto provide for the health of their families (which is principally the responsibilityof women), as well as their economicproductivity and potential. Women's health, not just maternalhealth, and nutrition,thus deserves particular attention. Women's lack of time and their lack of minimumeducation and informationare further important factors negativelyaffecting the health and well-beingof children (Box 3.1).

31 UNICEF/ACFODE,ibid, pp. 28-29.

32 UNICEF/ACFODE,ibid, p. 17.

33 CKatineSaito, Raisingdte Producdvityof WomenFannerr n Sub-Saharamnfca', Vol. 1, OverviewReport, PHRWD, September 15, 1992, p. 26. Her conclusion is almopettinet. 'An average of 13 hours' work a day is debilitating to body and spirit and cannot be expected to [lead tol satisfactory result for either farn output or household maagemnent."

34 UQNICEF/ACFODE,ibid, p. 3. 28 Chapter 3

...A 'a.stt.., :ua in...... fly in'os are.'f Lk 'r l-iv

we,'0-li%-V-:-V:RR:::9s-:-:---- E:-z ig ::.-..::.:..:..-..-..z as,he,'im,',b.-zE h:: :-- E-E ' -Eu::: ::'ohe to' ' - a he'oz- an bmtfoEE her ARrsxmnh,hwvr......

fiarErbc ir i-her lil s lEaves-e -Ehil. :li-: -him ... -,....i ;...... r..w;eaE-E:.-.-.-.,-. w. .i . . 'l:lotl... . s. was ill...... S...... tm de.. r.n.o... 'n...... t. .. la.t.h r

3.10 Education. Women's low levels of literacy as adults and the increasinggender inequity at different levels of the educationsystem compoundthe difficultiesthey face in meeting their multiple responsibilities;despite being the principal providers and promotersof good health at the household level, many are illiterate3 and have very little knowledgeof health, nutrition, and hygiene. More importantly,women are also poorly informed about other issues of importance to them and their families, notably their legal rights-see paras 3.17-3.19 below.

3.11 Isolation. Characteristicof women'slives in rural areasis theirrelative isolation. There are few channelsof communicationto and in rural areas, and women seem largely excludedfrom them. WVomenhave low rates of participationand involvementin formal groups or decisionfora (18 percent in women's groups; 2 percent in agriculturalcooperatives); women receive minimalattention in the form of visits from either extensionor health workcers(5 percent); they have low access to radio (34 percent), while most external contact arises through RC activities (38 percent). 3'

G;ender Issues for Poverty Reduction and Economic Growth

:3.12 The aboveprofile of women's multipleroles and constraintsamply illustratesthe asymmetries in rights and obligations of women and men, as reflected in significantlydifferent workloads, responsibilities,degrees of isolation and autonomy, and access to productive assets. This leads to identificationof a number of issuesthat need to be addressedin defining a gender-responsivepoverty reductionand economicgrowth strategy for Uganda. These are:

(i) the value (both economic and noneconomic)of women in Ugandan society, and how this affects women's capacity to fulfill their multiple roles, and by extension reinforces the impoverishmentof women and childrenin Uganda;

(ii) the extent to which laws and customs constitutea greater obstacleto women than to men in carrying out their economicand social roles, and inhibitpoverty-reduction efforts;

35According to the HBS, for Ugandaaic whole,26 percentof menarc illiterate,comprd with 40 percentof womecn.

36UNICEF/ACFODE, ibid. Chapter 3 29

:::: - - .- . : ~ : ~~.-:: .- . .:.:..::.-..:.: - - ...... ;...... ,...... -' ...... ~~ ...... ' ...... :Dit*-oxZ: . &.ES4 w frof ...X

tradiional~'alaswhch aBip ~vmy b~h ~aluoo ~~lld~a4ngandawa none at alto aythn lew

~~~~~~~~~~~. . .. -...... -..-.--,...... -.--:...... ,-..,,.,-.. haod f smare wrin oe orm etigineio ndiaeqae dcaiP andeeluall en ..mothet: .daushtr,fromgenerao to ron. Vlw~equwi:y, yo.a.u:find,- in s"Pome thltm,girilse ea .g -or awyi i W4at a in :a :

-SQuc:.-Janc uwz an ddr-s +ntnr n :tn,i Prvidr an Prmtr - et . O...b.. 79 ... m . .. * Ugand-.8 -)

(iii) the diversity in size and compositionof households in Uganda, and how this affects both analysis of poverty and design of poverty-reductioninterventions, especially as concerns intrahouseholdresource allocationand inequality;

(iv) the gender divisionof labor, and its implicationsfor poverty alleviationand economicgrowth strategies; and

(v) the intersectoral linkages that are particularly important from a gender perspective in addressingmultidimensional poverty.

The Value of Women in Ugandan Society

3.13 Uganda's women are principally valued for their role as mothers, which is reflected in the high total fertility rate at 7.3 live births, itself partly attributableto the widespreadpractice of early sexual activity, early pregnancy, and early marriage.37 This is not to say that the centrality of wvomen'srole in economicactivity (especiallyin food production) is not recognized, only that this role is taken for granted and undervalued(Bo0x 3.2).3'

3.14 In the economicsphere, men control the cash-basedeconomy, while women remain for the most part in the nonmonetizedsubsistence sector. Men are so closely identifiedwith the cash sector, to the virtual exclusion of women, that women who earn cash often hand it over to their husbands, and while men are usually paid in cash, womenare often paid in-kind, for example, with food, salt, or soap. This makes it difficult to identifypossible wage discriminationbetween men and women, though "distress labor sellingSof female labor at less than the village norm has been observed as particularlyprevalent during the 1991/92 season, in the face of drought and escalatingfood prices.3

37For a discussion of the implications of this in theceontext of AIDS,see: IVewPhase of UNICEFSupport for AIDS Control an Ugandla, sumniz Document,UNICEF, Kampala,Uganda, April 1992.

38Subsistence agriculture, based on HBS data, is now included in the calculation of GDP, and in this repect, womeon's economic activitq has begun to oven ome its invisibility. This cannot yet be said of women's work in the eommunity and around the household.

See AlisonEvans, A Rew of te Rura LabourMarket in Uganda,May 1992. 30 Chapter 3

0}:~~~~~~~~~~~~~~~~~~~..: ...... hu.....d a.req'e . h"',to ''"""'"' ' '' b e o'k ' 't s 'oa la -se,wall ---iiE:X.sev ..... ndw-t.. ... hpasc.epo... --t,;...... n.. r.i...... e a

say. .he... h husn. d..f t. .hu.ba.. di.d. sh. n.. n hth hrse lb ee......

lorc:E tsoLi:cg Agai' t- Woa -a V;anda" 4' Rwarc R'apo 'y Xs Va Wai and Ms 'Hope:;

3.15 Women's exclusion from the cash economy has far-reaching repercussions, both because men's and women's "vajue" to society tends to be measured as a functionof their contributionto the monetized economy, and because of ambivalentattitudes and shifting expectationstoward women takcingon greater responsibilityfor cash earning. "The differencesbetween men and women in access to cash produces and maintains significant gender inequities. A man who earns cash is highly regarded even if he cannotfeed his family, whereas a womanwho is "merely" a subsistenceproducer is undervaluedeven if she feeds her family. It follows ... that with the existingdivision of labor, a mnanis highly regardedbecause he is essentiallya cash earner (even if he is not earning cash) whereas a woman is underratedbecause she is essentiallya dependent(even if she is earning cash)."4

3.16 Women who do earn cash face a particular dilemma in a climate where men are "not keen on women who are economicallyindependent."' 1 Even though men are the principal beneficiaries of women's cash earning, they (men) are not supportiveof women in this respect. ACFODE, a local NGO, points out that in some cases men abdicate all responsibilityfor their famnilieswhen women earn money, so women "are not sure which is better, to remain being looked after by the man however bad this 'looking after' is, or to get money and do everythingin the home."42 This has far- reaching repercussionsin the contextof a developmentstrategy that emphasizesthe monetizationof the economy, and particularlyof agriculture. If women are to avoid being further marginalizedin low-technologyand low-productivitysubsistence agriculture, Uganda will need to take particular measuresto address this cultural dimensionof monetization,if the modernizationand diversification of agriculture-a key componentof the country's growth strategy-is to be undertakenin a manner that enables the majority of (women)farmers to participatein and benefit from it.

Tadria,4S op. cit., pp. 81-82.

41Action for Development (ACFODE), BackgrorindPaper on ACFODE's Objectivesand Activities, no date, p. 2.

42Ibid, Action for Developmnent(ACFODE). #t A~~~~~~~~~.

m-c~ ~ ~ ~ ~ ~ - 9...b J~ ......

'~~~~~~p ~~~~~~~2jil , :z: ...... u I *jj~~~~ .~. . ...~ ~~ r.a..... '.. AIIj: p...... ~~~~~.

0 0, . . j~~~~~~~~~~~~~~~J:,...... I...... ~...... 10. ~~~~~~~~~~~~A.~~~~ 32 Chapter 3 to perform their economicroles. This applies especially in the areas of access to and control of the physical and financial resourcesthey need if they are to play their multipleroles in both economic and social development.'3 Key issues concern property law and land use/tenure rights, inheritance laws and practices, and family law governing marriage, divorce, child custody, domesticviolence, and bride price (Box 3.3).

3.18 Asymmetryin the respectiveobligations of men and womenis characteristicof both the legal system and customary practice in Uganda. Indeed, while there are specific instances of sex discriminationin laws and regulations, 'most of the laws are gender-neutral, thus giving rise to de facto discrimination."" While this is an area requiring specific analysis, especiallywith respect to the impactof such discriminationon women's economiccapacity, gender bias exists in a number of areas, and derives from the fact that under customary law (and practice) women are essentially minors, without adult legal status or rights.' These include divorce, including definition of and compensationfor adultery; inheritance, since women are not their husbands' automaticheirs; and child custody laws, as women do not have custody of their children. There is also bias in property rights, even where no statute prevents women from owning property, since, by custom, property acquired during marriage belongs to the husband.'

3.19 To address these recognizedbiases in law and custom, Uganda is making a concertedeffort to reform its legal system, beginning with constitutionalreform, and women's groups are actively seeking to ensure that these reforms respond to the needs and constraints of women as well as of men. The most importantinitiatives concern the redraftingof the constitution,following a grassroots consultativeprocess in which women's groups and the Women in DevelopmentMinistry have been very active (Box 3.4), and the work of the Child Law Reform Commissionin seeking to articulate and protect the legal rights of children (especially important in view of the number of orphans in Uganda)as the country's laws and statutes are revised.

The Diversity of Householdsin Uganda

3.20 If the analysisof householdwelfare and behavior is to take accountof gender asymmetries. it must necessarilyreflect the diversityof householdsin Ugandaand the different roles played by men and women.The size, composition,and dependencyratios in Ugandanhouseholds vary considerably. Evidence from the HBS suggests that about one fifth of all households in Uganda are headed by women. Female-headed households (FHH) are generally smaller than male-headed households (MHH), comprisingon average 4.5 people, as comparedwith 5.7 in MHH. While it is importantto look at the educationand health statusof all householdmembers, the data indicatethat the proportion of female household heads among illiterate household heads is, at 30 percent, higher than among literate householdheads, where it is 20 percent. Illiterate female heads of householdsmay therefore

4'1 The linkage betweenlaw, gender, and economic developmentis exploredin Doris Martinand Fatuma Hashi, Law as an Iwtitudonalarnier to the EconomicProgress of Women;Sub-Saharan AMica: Gender, Evoluidon of Legal Insttuions, and EconomicDevelopment; and Womenin Development: he Legal Issuesin Sub-SaharanAfrca Today, AFTSP Division WorkingPapers, Nos. 2, 3, and 4, June 1992.

44 "Inidaland SecondReport by the UgandaGovernment on the Implmentationof the Convenionon the Elmination of AUForms of Discriminadonagainst Women", no date.

4' See Jarawan,op. cit., p. 8; and, more generally, the Fatuma and Hashi study cited above.

46 See UNICEF,Ouidren and Womenin Uganda:A SituationAnalysis, p. 75. Chapter 3 33

~~~~~~~~~~~~~~~~~~~~~~..,...... :...... _.'''iPanily or-.ffienc, or r..fl...... mae laborthmd fere1y......

: t...... w t.h.. ...f..... v. :.. a : ......

costtues anespcal foo vulnerablgseogroupof thecrd poor. u az ndrc,m~te iw.oaun

3.21oTe iewmportanceofexminigitaechden heimozatsoa.aed withte heloatiycardnnof boee wverkmhszd -lhelasoue alocatraitonproess wiotthin1houshoklds areflects bthestatus lborgainingpoer, aund optin oftheparties contatclne wo, vehich~trarein largely aulfuncltion aof Iconro overat setsbhandml incore A hasbee rshon i abprceing~ setonowvrwomen ~lovt tor genestallyt doi not and are naotexeteo controlcash V inome orenowichi asheots; othey re ewe ainid otheubsiasteance (onash)ectr, wheemeei constitut aotiuinese ia vulnerabluedgrupto thkencore porane.Hueodepniuepten,a 3.21le intheimprtnc confirexaminigene inhequity rithourcehocationd csannottbemoveremphrasned Themensiourcalocation procswswionfhouseholds rheflecitst the sotats baergawinhingpowser,oads optomns of te parties onerthnmed, adwhicentrre hlargenivralyaduntinsfcnrole oveirmosthes andbincomre.rA has bheensahown. ine faprctn sectdion,hnowever, wompien genrall doanot,icand sarenotfexpetedhto coxpnturolcshincm orHeconoiase; they remain inHthe ubsedishaten sis (1ncs, iseoner,ao wher thei ectonoal rontribufAtionisfot valudrbut tlakenmhssontenegrnedhouseholdsepediuespaterns,lfor aso rouevod-tealedcaintheHB,cnirm that gxendertinquitawithino thehosehmolditis andetremely mportant are visibl pooreriothandmen,t canwhre childre inaialfesrbtheir motersasreinnpore

The Gender Division of Labor

3.22 The gender division of labor is the central manifestationof the asymmetrical rights and obligations of men and women. While both men and women play productive, reproductive, and community managementroles in society, women, in contrast to men, must balance simultaneous competingclaims on limitedtime for each of these roles-women's time and flexibilityare therefore much more constrainedthan is the case for men. Accountmust be takcenof two aspects of the gender divisionof labor. The first derives from women's alreadyheavy workload,the significanceof which is that women's labor time is highly inelasticand that female labor availability(whether for economic

47 Uganda NationalPlan of Actionfor Children,Th Wey ForwardIVt Protits for Social Scncez Sector Devlopmen an thec19S90s, Ministy of Financeand EconomicPlanning, Kampala, Uganda, June S, 1992. 34 Chapter 3

pa itk uta ...... rly reEa. l...... -.--.... i is... f t...... in...d....b...... E ...... a.. the ,i:Adoption e....f...Te...... -x- ,abo Arcn....i -E,...... e.dow.e.. , ,he. td ... th. .. ftagointo a nKea afudtaaI fe .faotoha.ehae .. ouseh..lds ruralhouseholdsare female-headed this diminishedcapacity .Ken...... y..a...... n.can be substantial.The caseof...... te...... is

ID. .pnatt ...iEy..i...... ad...... t...... te..ra...... E.. ar eil lbrlast (tslelysglatIces...... ya,...... th..e...... f~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.,-.0.- .Y...... s.... .*;-...... , A;i .Ed.i ..- ...... -.S-S...... i.-i7.~~~~~~~~~~~~~~~~~~~~~~~~~~~......

l-EE-Elesslikeyi to:do so. Theia mplication is thfat ferneadedhbouseholda face Dome Nra roonstraintal aditi*naVl l aus tivityac-; for emalehededhoshls

Maz.;.:f-iWeein. 0;e llustaio.nof- a-m eti rightsad chliaion ihn d;ehouseol is give whore l Iii0womenwork onholdings thegoutputof!:!E:ESii-f-::f:-:::-:Ei::,l-. . .E--,-...... which : ii contrledby.-,: -- - - -men.:AWEA_-! :ecn:Kenyan : .: F....-...... sample survey-t-- :-----:--..-:t. i ...i: ....E

l hous:-olds.ols bo:thi ye of:househol :theewaretw wedn pr eon sadasiiweeing sgnilcaty ra;::ised:;yields.- 'However -wheesin fE: aehae houehd ths wedg ra'ised yi' d by"eret li t: male-beadedhloUsehlds thIncras i¢Cn:yicid wasonl5 iS ercent.Sine othe differces wereconnrlnLEd l r,the r otliely explnaio iasystmtcdfeec in efotdet difreta inetv. Thu, the0

thiit pesecie if0Sthe sampleis rereenatv of rua Kenyai the a tioa maiz los ro this

::::::::::::: :- :: ctfA d.t.o. :Womcn-RpriN . 891- WotdBank, ..14...... 09-pp. ::5::

activitiesor for community-basedmobilization for self-helpschemes) requires explicit assessmentof the opportunity costs involved. This is especially the case in reference to agriculture. Just as important, women's lack of access to labor-savingtechnology, whether for economic or domestic activities, coupled with the multipleconstraints and obligationsthey face, combineto diminishboth the productivity of women's labor time use and the flexibility or responsiveness to economic incentivesof their labor allocation(Box 3.5). This means that raising the productivityof Uganda's (women)farmers is a significantand multifacetedchallenge for Ugandanpolicymakers.

3.23 The second aspect is the still pervasive economicdivision of labor where men largely control the monetized (cash) sector, while women largely remain in the unpaid subsistence sector, as indicatedabove. The significanceof this is that womenhave extremely limitedaccess to and control of economicallyproductive resources, which is important for two principal reasons. First, women and men face differenteconomic incentives and opportunities-therefore, supply response cannot be assessed in a manner undifferentiatedby gender. Specifically,if womencontinue to have little or no control over the cash incomegenerated from export crops, a strategyencouraging farmers to diversify production away from food crops towards export crops will fail to attract women out of the subsistence sector (Box 3.6). Second, women are increasinglybecoming responsible for meeting household cash requirementsthat were once paid by men's cash income. For example, in 1988, 30 Chapter3 35 percent of women with children in schoolpaid all or part of their children's expenses.4' This is one of the contributingfactors to the adoption by womenof strategiesto generate income-such as sale of food crops, alcohol, labor, sex, and other services. These strategies, and the search for alternative and less damaging economic opportunitiesfor women, therefore require special attention both in terms of how womenfulfill their multipleroles and in terms of designand implementationof poverty reductionand economicgrowth strategies.

Intersectora1 Linkages

3.24 It is the interconnections,and inevitable trade-offs, among economic production, child bearing, caregiving, household management,and communityparticipation activities that take on particular significance,given the competingclaims on women's time. The linkagesbetween poverty, lack of access to and control of economic resources, and survival strategies have particular importancein the contextof two emergingdemographic and social trends. The first is the growing number of single female-headedhouseholds, which are substantiallymore cash-poorthan their male- headed counterparts.The second is the AIDS pandemic,and the attendant increasedburden and risk for women. Key issues which need to be addressedare:

(i) the interdependenceof economicallyproductive and social sector investmentsand programs, all of which affect labor time allocation;

(ii) the socio-economicimplications of AIDS for householdsurvival strategies, poverty reduction, and economicprospects; and

(iii) the problem of low female participationin education, and the nexus of low education and high fertility in Uganda.

3.25 The Interdependence of Economicaly Producive and Social Sector Investments and Programs. Hesse et al point out that the disproportionateburden on women in the area of household managementand family welfare means that womenalmost always face more severe constraints and harsher choices in their use of time than do men.4' Uganda provides many poignant illustrationsof these harsh choices. e.g., the trade-off between makling(principally female) child labor availablein the householdand sendingthat child to school. This is illustratedby Evans in her studyof rural labor markets, in discussionof strategiesadopted by householdsto address labor shortages.'

3.26 A second is in the area of health care. Many studies confirm that while first immunizations reach a relatively good proportion of children (60 percent), follow-up immunizations,and, more generally, routine preventive medical care, are not as successful.If a child is otherwise 'healthy' at the time a second or third immunizationis due, it is likely that a woman, faced with the competing claims on her time, and given that she probably has minimal mobility, will not be prepared to sacrifice a day or half-day to do somethingwhich does not appear on the face of it to be necessary,

48 Cited in Childrenand Womenin Uganda, A SiawJonal Analysis, UNICEF, Kampala, 1989, p. 78.

Mary Chinezy Hesm, t. al., Engendering Adiwhnen for the 1990J, Repoztof a Commonwealth Expeit Groupon Womenand Structural Adjutent, ComnmonwealthSecrariat, London, 1989,p. 3.

JD Evans, op. cit. o~~~~~~~~~,-0 P

~~~~~~~ _ ~~~~~~~~~~~~~~~~~~~~~......

I ~~~~9OQOE CD~D

- 0 ~~~~~~~~~~~~~~: 0:

cn :3 ...... -

OQ~~~~~~~~~~~~~~~~~~~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~ ~ ~ w~~~~~~~~~~...... 0 ~~~ 0 c ', ...... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... ~ ...... - ...... cjj ,~ 0A~II

M .. . . . N~~~~~~~~~~~~~~~~~~~~~~C ...... CD.... a'~~~~~~~~~~~~~~...... ~~~~~~~~~~~~~~~~~~~~~.. . . .

*~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~. CA...... t2 ......

CDp0 ... .. 4.d 38 Chapter 3

...... j..,...... Z .i. . . .io ..

."'-."' ~~...... :i.E.,,,gg}s,g~...... ,,A.,:.''°~~~~~~~~~~~~~~~~~~~~~~S..g.''..... ~ i ~ ~ "-''g-:-X Yi 'lt° '''''' :z ;';

:. :i::. :: :S .: -.:.::E :. -: : i . .iE.i,. i-.- -....-. i.,. i. ...E..i-...... Economtaicfs~wt hoit emhasisvt no weAerovid Persa u wy nesadha in il outlay,directed at educatinggirls would yield aionnoiin economic and modal beneft.s. Fourth, the share...... amEdXuEprive of opw1ntts'aio'n "tpnya ':' l'int" t ua~ daon t tha 'her

it, thr is a hi-0n-it*ing prole o--fawess femal montl; in maydveog counriesIti h

cheaDlaboeratd~ Whe gil hi lbrhs iteeonmcvlu usd tehneTe r 3.2 Brced tonlrteymnads E aeunabl on tandth euptheirdof h ands.Teyhv mcore The percepthanthey .vomen wareless J Ucnomclyvlablen@theavynt~ menc hisdreflete is chorpeswtharespcd.iveteti

education.enyChildH a...... fairlisanYeeometCnr, .... faeucaed they diffvel aeeeUiversity chooic ofpnostubeing AVpalaable Utnd, tom seed Fbu avl chilrena92 tok inhol Wheie surviaA .e.lurt... aof thneyfra.,GrrPritneadTahrienveinPmaydutonnsucess. feeseymarylte ctd n ase ablepritakiparl resn horsehooldU...anda....U..A....,dropout, Tthey K..mpala. f chorse Ugadatnouhbtedflwe Marc p1992. hldya pant: cpanesta orelutneinee hchithe o ...... lo d's laof,...... eachhld. oheintret pre unaneyand smarrhave,gender hoiaons antelc d oftnhe esckageberoee povmaryeuainannoe

towardswom penadthoe value offmale educaftgion aUrtea sgiican 5W&. facitor thoe in o erahivemnt attainment, enolb'and entrteso giprls ~in achols. In thewords Sfoemialeparent, W[educatin of

3.29 i BarrersteofFmaney casiogth andy NaredufFetltand Educawao.Theh rihyerucption pil tate wome aorelssSooEconomically valuabenoJo thc chandmePsrflcen coice w*yithUresipett invstenaind eduationMany failie fRymacthe difiul hoice ofnAUgnot. beingable olt senduyndalchdrnd Tom school. Whilelac oflt mney forvcholomnfesi fenr citedasUiithe princial, reasnd for school droou,2ohe prgnnc Ande maruriagetal,gender crias,e b and the ack ofinkagies betee rimEprmarydo inUcatindand iS]Kncome

st.. y..a..n Uganda, ic Fl e Chapter 3 39

arrogant." Or, as one girl put it, "I believe that even if I had the fees [my parents] would not have permitted me to go [to school] simply because I am a girl."5'

3.30 The benefits of female education are receiving increasing attention among development practitioners, so much so that it has been argued that female education is the single most important investment that a country can make (Box 3.9). Girls' education, apart from any intrinsic advantages it may provide, has the beneficial effect of postponing marriage and first birth, and therefore contributes positively to the reduction of teenage pregnancy and high fertility. However, educating girls in itself is not sufficient, if cultural, legal, and other biases against women in the society at large continue to limit women's economic capacity. Consequently, the need to pursue gender equity beyond prinmary school, and in the society at large outside the school setting, becomes of paramount importance.

Conclusion

3.31 This chapter has sought to explore the particular cultural, economic, financial, and social constraints which women face in carrying out their multiple roles and responsibilities. While it presents a situation facing women that is very difficult, though not unlike many countries in Sub- Saharan Africa, it is important to recognize the progress that has been made, especially since the NRM Government came to power in 1986. The Ugandan Government clearly recognizes the importance of women's role in both economic and social development, and is aware of the critical need to raise women's status. It has taken some significant steps, some of which are unique in Africa, to promote women's participation in the political process, to establish institutional mechanisms aimed at addressing women's issues in public policy and resource allocation, and to begin to tackle an important area of systemic bias against women through legal reform. The extent of women's representation in political bodies and their participation in government and political fora, through the reservation of one position for women at every level of the RC system, and through affirmative appointment of female ministers in key portfolios, compares very favorably with other SSA countries, and constitutes a significant step forward, notwithstanding the difficulties women continue to face in effectively exploiting the opportunities afforded and in making their voices heard.57 The establishment of a Ministry responsible for Women in Development provides an important institutional focal point for addressing gender issues across a range of government activities, as well as an opportunity to tackle underlying systemic issues, such as legal reform. In fact, as indicated earlier, the Government is particularly aware of legal issues (including issues of custom and practice) affecting women both economically and socially, and is actively engaged in revising its constitution. Women's NGOs and legal organizations have been actively involved in this process.

3.32 While there has been progress, it is clear from the picture of women's status and the constraints they face presented in this chapter that much more remains to be done, if women are to be empowered to fulfill their economic and social roles. As the Government prepares its poverty reduction and economic growth strategy, it will need to address the gender dimension explicitly. While cultural practices and traditions are extremely difficult to change, a key priority for the

Ibid.

See, for example,the profilcof a femalecomunnity leader in UNICEF,Cdldrcn and Womenin Uganda:A Sita*ionAnalysis, p. 12; andthe discussionof culturalconstmints to womentaking part in RC meetings,and in publiclife outsidethe householdin DodothCounty, in AlisonLochhead, Gender and Developmentin Dodoth County,Karamoja, OXFAM, Kampala, Uganda, Januaiy 1991. 40 Chapter3

Governmentmust be to create a level playingfield wherebyboth men and women are affordedequal opportunity to participate in economic growth. This will necessitate that the Government follow throughwith the legal reform process that has been initiatedand ensurethat the legal rights of women are protected so that they may benefit from their own labor and have greater access and control of economicallyproductive resources. Through literacy and education,the Governmentshould endeavor to overcome the social and cultural barriers which limit female access to information,technology, credit, and extensionservices. Giventhe time constraintsfaced by women, the Governmentmust pay increasedattention to investmentsin femalelabor-saving technologies. This would includeinvestments in infrastructure,particularly feeder roads, water supplyand transportation.Finally, the Government should provide political and financial support to efforts aimed at reducing the risk of AIDS among young girls. 4

inpactof Adjustment on the Poor

'Any analysis of the effectsof structural ajustment on poverty mnst compare the outcome not with the preceding period, but with the outcomes that could be expectedfrom alternative polcies that would have been economicallyand policaly feasible under the dicwut conditios .... '(Lyn Squire, 'Poverty and Adjustment in the 1980s, The World Bank Economic Review, May 1991, p. 182.)

General

4.1 Adjustmentmeasures can broadlybe dividedinto two categories:stabilization measures which focus on financial disequilibriumwithin the economy, i.e., in the fiscal and external accounts, and are designedto reduce overall demand; and structural adjustmentmeasures which stimulatesupply and are aimed at 'restructuring the production capacitiesin order to increasetheir efficiency and to help restore growth'.' It is not always easy to disentangle the combined impact of adjustment measures from those of the preceding crisis and other long term factors and it is therefore critical to have a clear understandingof the economicand politicalconditions which existed in the countryprior to the introductionof the adjustmentprogram. To understand how adjustmentpolicies affect the welfare of the poor, it is necessaryto explore in detail the sources and patterns of their income and expenditures, and the changes in factor and product markets through which those policies are mediated.

4.2 The poor in Sub-SaharanAfrica (SSA), as in Uganda, are mainly rural and are engaged in a multitudeof economicactivities of which agriculturalproduction looms large. Since the bulk of the exportsof these countriescomes from the agriculturalsector, structuraladjustment policies, with their emphasison restoring an "equilibrium' level exchangerate and liberalizing agriculturalpricing and marketing,are posited to impactpositively on the agriculturalsector, and thereby, the rural poor. For example, to the extent that the poor participatein the markets for traded (tradable)goods as sellers, exchangerate adjustmentsare likely to move the terms of trade in their favor, resulting in an increase in incomes. Furthermore, stabilizationpolicies, which dampen the excess demand for goods and services, control fiscal deficits, reduce monetaryexpansion and thereby control inflation are similarly expectedto have a beneficialimpact on the poor. However,to the extent that the poor operateoutside the monetizedeconomy, and are absent (or left out) from the traded goods sector, their welfare would not be affected by adjustmentpolicies, which tend to change relative prices. Therefore, one needs to know the rate of market participationof the poor, and particularlythe extent to which the poor are involved in the production and marketingof tradable commodities.

4.3 The record of adjustmentpolicies in SSA countries indicates that in countries where such policies have been persistentlyfollowed, structural imbalanceshave been removed and reasonable rates of growth have been restored. In additionto the conducivemacroeconomic environment created by the adjustmentpolicies themselves, these results have, however, depended on such factors as the gravity of the pre-adjustmentsituation and the impactof other factors such as changesin international terms of trade; weather conditions; the 'adequacy' of external financing; and the emergence of a stable political environment,where this has been lacking. The impactof these policieson the welfare of the poor is howevernot adequatelyresearched, and the limitedfindings available are inconclusive.

so Ishmr Huain, Adjustnenand te Impac ou te Poor, h Caseof Mica, Word Bank, mimo, 1992. 42 Chapter 4

This is due to the weakness of the statistical agencies as a result of which a fairly long series of household data, which are crucial for such purposes, are lacking. Where the data are available, complicatedmodels are often needed to understand what would have prevailed had the adjustment policies not been implemented, given that most of these countries experienced unsustainable macroeconomic imbalances and weak or negative economic growth rates that precipitated the adjustmentprograms in the first place (i.e., the problem of the counterfactual).

The Ugandan Experience

4.4 As discussed in Chapters 1 and 2, with more than a decade of civil strife, political turmoil, and economicdecline up until the mid-1980s,Uganda's historical legacy has been much worse than most of its SSA neighborsthat have experiencedmacroeconomic imbalances and faltering economic growth. Upon assumingpower, the key objectives of the NRM Government were to build a broad- based government and restore peace and security in the country; to bring about internal financial stabilityby loweringthe rate of inflation and reducing the imbalancesin the economy;and to lay the basis for the alleviationof poverty by promotingrehabilitation and economicgrowth. Accordingly, starting in 1987, the Government began implementing its far-reaching Economic Recovery Programme (ERP) supportedby the World Bank, the InternationalMonetary Fund (IMF), and the donor community.

4.5 To achieve its stated objectives, the adjustment measures implementedby Uganda have contained both sets of policies alluded to above, i.e., stabilization policies, designed to restrict demand to the confinesof the overall resource envelopeand thereby restore financialequilibrium and structural policies, designed to increase efficiency, stimulate the supply side of the economy and encouragegrowth. On the stabilizationfront, the Governmenthas adjustedthe value of the Ugandan shillingand movedto a marketdetermined exchange rate; containedthe fiscal deficitby implementing numerous revenue enhancing and expenditurecontrolling measures; and implementedmeasures to curb inflationary crop financing. On the structural front, the Governmenthas liberalized the trade regime by abolishingexport and import licensing; dismantledall price controls, which were few to begin with; repealed the Industrial Licensing Act, promulgated a new investment code, returned properties expropriatedby the Amin regimeand commencedprivatizing public industrialenterprises; made important strides in abolishingexport and distribution monopolies;embarked upon a major overhaul of the civil service; restructuredthe tax system and improved tax administration;and has made an impressive start in restructuringpublic expenditurestowards critical economic and social services. While the economycontinues to be plagued by severe capacity constraintswhich make day- to-day implementationof the reform agenda very difficult and often result in lapses and slippages, the outcomeof the adjustmentprogram has, on the whole, been very encouraging.There has been a substantialdecline in the rate of inflationsince 1986 and the economy has consistendyregistered a healthy rate of economicgrowth.

4.6 This chapter analyzes the impact of the adjustment policies implemented by the NRM Governmenton the citizens of Uganda and, in particular, on the lives of the poor. Data limitations make this an extremely difficult task. From the outset, it should be noted that, like other SSA countries, Ugandalacks a series of household-leveldata. While Uganda has completedseveral useful surveys in a short period of time, and others are underwayto bridge the data gap, these data have very limited use in capturingthe evolutionof poverty over time. In the absence of fairly long time series data on the volume and sources of income and expenditures,one can only resort to indirect ways of assessing the changes in living standards. The analysis presented below is therefore partial in nature and is only intended to present the general direction of changes and broad orders of Chapter4 43

Table 4.1: Growth and SectoralDistribution of GDP

SeckwalDbtribudoa GrowthRate aIbet perent 1961-4S 1966-91 198145 196-91

Agriculture 63.2 61.2 2.3 4.2 Industry 7.3 8.4 0.8 11.8 Servioe. 29.5 30.4 2.7 5.8

CashCrops 6.5 6.7 6.4 7.1 Food Crops 93.5 93.3 2.0 5.1

AggregateGDP 100.0 100.0 2.2 5.3

So&wceData prowded by Goverem ALhoritses.

magnitude.In this pursuit, the chapter evaluatesUganda's experiencein the recentpast by examining the structure and trends in aggregate GDP, and the trends in rural and urban real per capita GDP. These are, in turn, supplementedby an examinationof some of the underlying factors that influence changes in real incomes.

Rural and Urban GDP Per Capita

4.7 In order to determine the implicationsof policy-inducedchanges in relative prices on the structure of output, it is necessary to assess the degree to which the economy produces traded (tradable)goods, which relative price changes are intendedto support.' As shown in Table 4.1, the Ugandaneconomy is dominatedby the productionof food crops which, although tradable in theory, are mostly grown for subsistencepurposes. Consequently,for the analysis in this chapter, the traded goods sector refers to the cash crops sector and the manufacturingsector.f° Despite the relatively faster pace of growth that the traded goods sector has experienced,the share of traded goods in GDP has remained less than 10 percent.

4.8 Since poverty in Uganda is largely a rural phenomenonand most of the poor are engaged in multicrop and mixed production, the trend in real agriculturalGDP closely reflects the pattern of

The degree of tradability,constiwuting both impotabl andexpotable goods, refers to a hypothdicalstructure of outputthat would emerge under condition wheo all ditoin ameremoved, andgoods and factorsare tbed freely.

The so defined teded goods sector a probably undemeimdW her mce only the traditionalexport crops are countedas cah crops in the nationalaccounts. In reen yea, wveal taditonl food crps, such asminmim and bean., haveemerged as imporant uoures of cashfor thefaner. Nevertheles, the hameof the traded good sector still renmin. under 10 perct of GDP. 44 Chapter 4 rural living standards.6" Moreover, since most rural land falls under the customary tenancy system where accessto land is fairly open, and since over 85 percent of the farming populationoperates on less than two hectares of land, the benefits of agriculturalgrowth are likely to be evenly distributed among the rural population. Thus, average trends are expectedto be a fair reflection of changes in rural living conditions.

4.9 Figure 4.1 presents the trends in REAL PER CAPITAGDP real per capita GDP, broken down into its In 1987 Prices urban and rural components. Between 1983 and 1986, aggregate real per capita GDP declined steadily, reaching a bottom A of U Sh 13,184 in 1986. Startingin 1986, / \ it grew steadily to reach U Sh 15,802 in 3 '°° 1991. Real GDP per capita in rural and / / urban areas essentiallyfollowed the same \ ,- trend. Real GDP per capita in rural areas i increased from U Sh 9,269 in 1986 to U Cal end9a7Yew Sh 10,543 in 1991, while real nonagricultural real GDP per capita rose U- from U Sh 47,236 to U Sh 54,682, f respectively. These numbersdemonstrate that, on average, the welfare of the rural - and urban poor, in real terms on a per j 40\ capita basis, has improved by between 14 647 and 16 percent respectively during the s * * * 1967 1 199 past five years, maintaining the rural- cmndw YW urban income gap at around 5.2:1 over OMERA" the second half of the 1980s.' e- Compared to the early 1980s, some incomeredistribution in favor of the rural / areas seems to have occurred. It is 14 importantto note, however, that averages , mask the existence of wide intraregional 1 differences,particularly in urban areas. '* Om* *w 19a7 1W0 190 Canca Yew 4.10 There is considerableuncertainty in determining whether these improvements were entirely a result of Vigure4.1: Real Per Capita GDP adjustment policies, other extraneous

61 Real per capita agricultural GDP, used as a proxy for rural red per capita GDP, is arrived at by dividing real agricultural GDP by the rural population. This overates the rural per capita GDP to the extent that urban farming is important. On the other hand, it will be an underestimate provided that nonagricultural activities are impoitant in rural areas. In this computation, the two sources of bias are anumed to offset each other. A simihr reasoning is mnadefor the red per capita GDP in urban arean.

Two asnumptions implicit in this analysis are that agricultue's terns of trade vis-a-vis the nonagricultural eotor remnainedunchanged as did the distribution of income in both rural and urban areas. If agriculture's terms of trade improved (deteriorated) there would be a relative imnprovement(deterioration) in agricultural income. See the discussion on the trend in the terms of trade below. Chapter4 45

DOMESTICTERMS OF TRADE

W^SH&SOD CROPS TOTALAGPCULlURE

115 99

Si95 f 85

~75 FOc ~75

______6 51...... 1981 193 1951987, 199 199 1991 1983 1965 1987 IN9 199 C4l&yY CaBlnd'Y

igure 4.2: DomesdicTerm of Trade factors, or some combination of the two. During the second half of the 1980s, Uganda was just emerging from the political chaos following the Obote II regime, which had brought the urban economy to a virtual standstill. The large-scale destruction and heightened insecurity impacted significantly on rural areas as well; market links were disrupted as marketing institutions and infrastructurecollapsed forcing the rural economyto become self-reliant.Thus, given that peace and security were regained at a time that the structural adjustmentpolicies were put in place, causes serious difficultiesin disentanglingthe effects generatedby the macroeconomicpolicies per se. The problem is also compoundedby the difficulties in isolating the effects of the price reforms (i.e., changes in the exchangerate, interest rates and product pricing) from the impact of the injection of large amounts foreign exchangesupport accompanyingthe reform program.

4.11 To isolate, at least partially, the effects of the "peacedividend" from the effect generatedby policy measures,the followingsection examines some policy-induced determinants of rural and urban income such as the trends in the domesticterms of trade and real wages.

Changes in Domestic Terms of Trade (TOll)

4.12 Figure 4.2 presents the domesticbarter TOT for cash crops (traded), food crops (nontraded), and for the agriculture sector as a whole. The TOT have been computedby dividingthe cash crop, food crop and agriculturedeflators by the nonagriculturalGDP deflator in an attemptto comparethe 46 Chapter4 prices of the goods that farmers sell to the prices of the goods that farmers buy.' These prices are influenced,in one way or another, by such policy measuresas the changes in the exchangerate, price and trade liberalization,and the eliminationof marketing monopolies.

4.13 As the figure reveals, the terms of trade for cash crops fluctuated considerablyduring the 1981-86period and experienceda sustained deteriorationbetween 1986 and 1989. However, there was a dramatic improvementin the cash crops TOT in 1990 and 1991, the reasons for which are discussed below. At the same time, the TOT for food crops, which dominateagriculture, have been on a steady declining trend since 1989. Food crop TOT improved steadily between 1981 and 1987, largely as a result of insecurity,which prevented adequate supplies of food from reachingthe market. Since 1987, with the restorationof peacethroughout most of the country, there has been a significant revival (the present drought in some parts notwithstanding)in agriculturaloutput and in the marketed surplus causing the TOT to decline. Given the dominanceof the food crop sector, this has resulted in the aggregateTOT shifting against agriculture.This somewhatsurprising result probably warrants further research which present data limitationsdo not permit, i.e., a closer examinationof farmer's incomes and expendituresper capita. As is discussed below (paras 4.24-4.25) the increase in agriculturaloutput was accompaniedby an increasein the real wages in the agriculture sector further corroboratingthe findingthat the benefits of growth were substantiallyshared by the smaller farmers. The key implicationof the decline in agriculture TOT is that increases in gross domestic incomes have tended to be smallerthan the increasesin gross domesticproduct in agriculture. It must also be pointed out that the improvementin the security situation has also resulted in a decrease in the farmers cost of production and so the decline in TOT probably overstates the loss in incomes.

Macroeconomic Policy and Its Implications

4.14 Macroeconomicpolicies, includingagricultural marketing and pricing policies, togetherwith external factors, affect both tradable and nontradablecrops by influencingthe TOT for these crops and thereby have an impact on the structure of incentivesand income distribution.The price that a farmer receives for growing a cash (tradable) crop, when it is not arbitrarily fixed by the Government,is generallydetermined by the internationalprice of the product, costs of transportation from farmgate to processingcenters and then to ports of exit, processingand marketingmargins, the exchangerate regime, and explicit or implicit export taxes. The farmgate price for food crops, on the other hand, largely reflects domestic demand and supply conditions, including the flow of substituteimported food crops, and marketing margins.

4.15 A poor macroeconomicenvironment, in additionto providingthe wrong market signals and thereby stiflingeconomic growth, hurts the poor by indirectlytaxing the resourcesthat they possess. An enabling macroeconomicenvironment, on the other hand, removes production disincentives, encourages optimal resource allocation, and improves income distribution. The impact of macroeconomicpolicies, particularly those focusingon taxationand the exchangerate, is particularly pronounced in largely agricultural economies such as Uganda and thereby warrants careful consideration.The exchangerate, in particular, producesthe most pervasive impact that often out-

Four variants of the domertic batr TOT wemecomputed. Thea includedTOT computedon the besi of the deflatorsfor industry,aervicer and the nonfood coAun_ r priceindex in the denominator,rpectively. The rewults fromtle alternativemeauru werenot sipificantly diffoent fromtho analysispresented above. Chapter 4 47

Table 4.2: Foreign Trade Taxes, Trade Deficit and Exchange Rates

Nomina REER Ebqt Spread Bdetween Cakendar Exbuhane 1987=100 Tax/ Offrial & Free Market Year Rate U Sh/US$ lndex GDP (%) Rate (%)

1981 0.5 3.06 1.2 1982 0.9 1.07 2.9 1983 1.5 0.84 4.8 1984 3.8 0.57 7.2 52.1 1985 7.0 0.74 5.8 145.1 1986 14.2 0.78 3.3 329.0 1987 44.7 1.00 2.0 266.3 1988 106.0 0.90 1.0 289.8 1989 222.0 0.77 0.9 156.2 1990 433.0 0.47 0.9 61.3 1991 750.0 0.36 0.4 24.7 1992' 1000.0 0.332 0.0 15.02

IEstimated. 2 Annualaverage as of June 30, 1992.

Sources. Minustryof Fuance and Economic Planning,IMF and staff estimates. weighs the other effects of crop-specificpolicies.'

4.16 Through a series of changes in the foreign exchangeallocation system, the Ugandanshilling has depreciatedsignificantly since 1987 (Iable 4.2). In 1987, the US dollar was officially exchanged at an averagerate of U Sh 44.7 and the spread betweenthe official exchangerate and the free market exchangerate amountedto about266 percent. Startingin 1988, the Governmentbegan implementing an aggressive program of adjustingthe official exchange rate with a view to compressingthe gap between the official and the foreign exchangebureau rate and in July 1990 it legalizedthe operation of the foreign exchangebureaus. By 1991, the Governmenthad adjusted the official exchangerate to an average of U Sh 750 per dollar, i.e., over a 17-folddepreciation. When the differentialhad declined to about 15 percent, the Governmentdecided to cease setting the official exchangerate administrativelyand introduceda foreign exchangeauction in January 1992.

4.17 It is importantto note that real devaluation,by itself, cannot provide positive incentivesand stimulatethe supply response unless it is complementedby price and market liberalizationwhich allows market-determinedor border prices to be transmitted to the producers and reduces the monopoly rents enjoyed by marketing boards. To the extent that the international prices are not passed on to producersbut are retained by the state as tax and by marketing boards, the benefits to

A numiberof oauntly cametudies conuinaioned by the World Bankhave demonrated that the impat of exchange rate Lmaignmton rculkual incantives is mor wevce as compared to direct crospecific policies. A wunmary of themecane dtdies can be found in A. 0. Krueger, M. Schiff and A. Valdes, "Agricultura Incatives in DevelopingCountries: Measuring the Effect of Sectondand EconomywidePolicies", 7he World Bank Economic Pn*ew, Vol. 2, No. 3, 1988. 48 Chapter 4

the farmer are limited.'5 Recognizingthis, the Governmentof Uganda has overhauled the system of marketing, pricing and taxing coffee. Startingwith the 1991/92coffee year (October-September), the Governmenthas movedto a systemwhere only an indicativefloor price for coffee is announced and the actual producer price is determinedby market forces. The export monopoly of the Coffee Marketing Board has been eliminated and coffee exporting has been opened to competition.In the early years of the program, all exporters were required to surrender their foreign exchangeto the Governmentat the appreciatedofficial exchangerate. The Govermnent now permits all noncoffee exportersto either retaintheir foreign exchangeearnings or sell these to the foreign exchangebureaus at the market determined exchange rate. Coffee earnings, while still sold to the Govermnent, are transactedat the foreign exchangebureau rate. To counteractthe adverse impact of the decline in the internationalprice of coffee, the Governmenthas virtually abolished all taxes on exports such that export taxes declined from about 34 percent of total revenues in 1986/87 to less than 2 percent in 1991/92.The remainingexport tax only appliesto Arabica coffee which commandsa premium price on the internationalmarket. The Governmenthas further eliminated the export monopoly of the Uganda Tea Authority and the Produce Marketing Board, although the Lint Marketing Board still enjoys a monopoly for exporting cotton. These changes have gone a long way in altering relative prices in favor of cash crops (tradables),and have had a beneficialimpact on the poor, to the extent that they are involved in the productionof such crops.

4.18 Impacton the Fanngate Price of 7hdable Crops. Uganda's exportsare dominatedby coffee which accounts for the bulk of the country's foreign exchangeearnings and, until recently, its fiscal revenues. Due to its heavy dependenceon a single export commodity, Uganda's macroeconomic performance, and the welfare of the poor, has suffered considerably as a result of the adverse movementsin internationalprice of coffee. The over 50 percent decline in the internationalprice of coffee, from US$2.60/kg in 1985to less than US$1.00/kgin 1992, as a result of the collapse of the InternationalCoffee Agreement,has causedmajor foreign trade, fiscal and price instability,inimical to the rural as well as to the urban poor.

4.19 As a result of concerted action, particularlysince 1990, the Government has been able to protect the price received by producers. Despite the deteriorationin the internationalenvironment, the real producer price of Robustacoffee in 1991 was about 7 percent higher than the price in 1987, the year when the Government commenced implementingthe ERP.'5 The farmers' share of the internationalprice of coffee in 1991 and 1992 was more than double their share in the mid 1980s.

4.20 Table 4.3 reveals that, as a result of the adjustmentmeasures implementedby Government, relative prices, particularlysince 1989, appear to have changedin favor of coffee as comparedto the prices received by farmers for growing food crops such as plantain (matoke), cassava and maize, which compete with coffee for farmer's resources, particularly labor. The farmer's relative price between coffee and maize, coffee and cassava, and coffee and matoke, for instance,has increasedby 227 percent, 272 percent, and 119 percent, respectivelybetween 1989 and 1991. It is therefore apparent that, through macroeconomicand agriculturalpricing policies, the Government has been largely successful in preventing the coffee industry from virtually collapsing which would have occurred in the absence of the structural reforms implemented.

altrt Huain, ibid.

Unlikefte nonagncultumdGDP deflatorused earlier, the real price eries in this section ix computedusing the aggregateCPI as th deflator.Recomputing the real pricesusing the nonagriculatulGDP does not changethe trend, althoughthe ngnitude are slightlydifferent. Chapter 4 49

Table 4.3: Index of Real Fanugate Ptices 1987 = 100

Ralo of Relative Prices: Robusa/Food Crops Seed Greenle Robusta Cotton Tea Millet Maize Sorghum Casava Plantains RobWt/ Robust/ Robust/ (Kiboko) (AR) Maie Csava Plant

1982 69.0 69.7 33.1 50.8 77.3 47.1 24.9 30.6 89.3 276.8 225.7 1983 116.0 131.9 111.4 50.6 139.2 45.6 35.9 81.4 83.3 322.9 142.5 1984 176.9 191.6 161.8 163.0 235.9 152.5 121.8 130.7 75.0 145.3 135.4 1985 153.5 132.1 109.8 77.0 196.0 70.3 64.1 102.5 78.3 239.6 149.8 1986 106.1 63.1 149.8 125.5 109.9 89.3 72.8 179.8 96.6 145.7 59.0 1987 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 100.0 1988 84.4 142.2 67.5 83.7 84.4 83.7 81.0 103.9 100.0 104.2 81.3 1989 52.3 143.1 73.2 67.0 244.0 67.0 64.5 80.4 21.4 81.0 65.0 1990 78.5 181.9 55.0 66.9 130.9 66.8 64.1 84.6 60.0 122.5 92.9 1991 107.3 219.5 73.6 52.2 153.3 67.0 35.6 75.5 70.0 301.4 142.2

Note: Until 1990 the GovernmentestabUshed producer pricwfor coffee, cotton and tea which have since been liberalized.

Source: AgriculturalSecretariat, Bank of Ugandaand African DevelopmentIndicators ffor nDiet, sorghmn and cassava).

4.21 The real farmgate price of cotton, the other important export crop which accounts for about 5 percent of total exports, registereda much strongerincrease, with the price increasingby about 119 percent between 1987 and 1991. While the real farmgate price of tea declined by about 25 percent during the 1987-91 period, tea presently accounts for a very small fraction of Ugandan exports. The upward trend in the TOT for cash crops duringthe past two years, as illustratedin Figure 4.2 above, is therefore largely the result of the combinedimpact of coffee, cotton, and a few newly emerging nontraditionalexports such as beans and sesame.

4.22 Imnpacton the Farmgate Prce of Nontradabk Crops. Tbe major nontradable crops in Uganda constitute plantain (matoke), cassava, maize, millet and sorghum. The trend in the real farmgateprice of these crops, as shownin Table 4.3, supports the deteriorationin food crops TOT. On average, the real producers' price of these five crops droppedby about 25 percent between 1987 and 1992. Since nontradablecrops accountfor about 70 percentof agriculturalproduction, the trend in the real farmgate price of such crops has more than offset the increase in the farmgate price of tradablesand resulted in the decline in the overall TOT of the agriculturalsector.

Implications for Household Welfare

4.23 While combiningthe trends in real farmgateprices with existingdata on prevailingfarming systems can provide some idea of the regional evolution of household welfare on account of agricultural activities, lack of details of agricultural income by district makes this an extremely difficult exercise.For example, in the high rainfall areas aroundLake Victoria, where Robustacoffee is predominantly grown and where poverty is comparatively less pronounced, it is likely that householdincome from cash crop production would have increasedin recent years on account of the 50 Chapter 4 improvementin the cash crops terms of trade. This would also be true for households of the Teso system in eastern Uganda where cotton is grown and in the mountainareas of the west where Arabica coffee is grown. However, since these householdsare also engaged in the production of food crops, whose terms of trade have declined,the net results are ambiguous.It, however, bears repeatingthat with the significantincrease in agriculturaloutput in recent years, welfare of the rural poor appears to have improved although, given the weight of nontradablefood crops in total agriculturaloutput, incomesprobably did not increaseas much as did output. What is unambiguous,however, is the fact that policy changes have helped the cause of poverty reduction.

4.24 To keep real incomesrising in the face of the overall decline in the TOT of the agricultural sector, various strategiesare being adoptedby farmers. As recent experiencein several areas shows, monetizationof the agricultural sector is gradually on the increase as farmers shift resources, particularly labor, from less paying crops to crops commandinghigher prices at the farmgate level. In particular, incomes are being supported, in some areas, by the introductionof some high-paying export crops into the production mix. These encouragingdevelopments are also being affected by changes in the labor market. In recent years, labor shortages have started manifestingthemselves, particularly in the estate sector,' as a result of which real rural wages have risen (see below). Among the reasons for the labor shortage, particularly during peak season, is the decline in the supply of migrant labor from neighboring countries and the impact of the AIDS crisis on the productive segmentsof the population.Although this labor shortage has adversely affected the up- keep of the coffee trees, a relativelylabor-intensive activity, it has helped to supplementthe income of net-labor supplyinghouseholds further improvingrural welfare.

What Has Been Happening to the Incomes of the Poor?

4.25 While there is no direct evidence available of the change in the incomes of the poor households, a look at the trend in real wages of agriculturallabor may shed some light on the trend in the living condition of the rural poor. Table 4.4 shows the estimates of real wages of casual, permanent and contract labor in rural Uganda between December 1984 and February 1992. There are obviousquestion marks about these estimates.' And yet the overall positivetrend in real wages is strong enough to defy all the doubts and imperfections of measurement. If the obviously questionableobservation for December 1987 is excluded,there was a steady and strong increase in real wages of casual labor until 1990. Real wage for contract labor has been more volatile, but an increasing trend continued through 1991. For permanent labor the increase in wages continued through 1992 with the exceptionof the questionableestimate for 1990.

4.26 Of the three, the trend in the real wage of casual labor is probably the most reliable indicator of the trend in earnings of the poorer households. This is because the permanent workers are not principally drawn from the rural households (see Chapter 6) and the contract wage rate, being

67 Alison Evans, A Revew of the Rural Labour Market In Uganda, University of Susx, May 1992. It should be noted, however,that the estatesector accountsfor a very smallproportion of Ugandanagriculture.

In a highlyinflationary economy, adjutments for price changesare problematicin anycase. In Ugandathe difficulty is conpounded by the fact that the only availableconsuner price index(CPI) is for the capital city, Kampalaand a few other urban centers.Moreover, therewas a changein the weightsof the indexdarting in 1988. We have splicedthe old andthe newindex together. Thero arm also some eriousquestion maks aboutthe nominalwage data. For exampl, why wasthere a suddenupward urge in casual andcontract wage ratesin 1987?Why was there an absolutedecline in permanentwage rate in 1990?Why w thee a shap discontinuityin contractwage rateafter 1986?Satisfactory answers to thesequestions arm not available. Chapter 4 51

Table 4.4: Real Wagesof AgriculturalWorker (Real Wages in Ugandan Shillg at December 199l Prices MoneyWages in Current UgandanShillings)

Date Caxual Labor Daily Wage Permanent Labor Monthly Wage Nominal Real Nominal Real

Doc 1984 1 48.08 - Doc 1985 3 52.54 90.00 1576.18 Doc 1986 15.00 107.53 - Dce 1987 100.00 259.20 750.00 1944.01 Dec 1988 165.00 165.00 2000.00 2000.00 Dec 1989 300.00 191.31 4000.00 2550.86 Doc 1990 400.00 208.37 3500.00 1823.20 Dcc 1991 500.00 197.01 7000.00 2758.08 Feb 1992 600.00 208.99 10000.00 3483.23

Contract Labor (Per Ha) Kampala CPI (Base Dce 1987) Nominal Real Doc 1984 40.00 1923.08 2.08 Doc 1985 100.00 1751.31 5.71 Dce 1986 260.00 1863.80 13.95 Dcc 1987 3500.00 9072.06 38.58 Dce 1988 5200.00 5200.00 100.00 Dec 1989 12500.00 7971.43 156.81 Doc 1990 17500.00 9116.01 191.97 Doc 1991 30000.00 11820.33 253.80 Feb 1992 30000.00 10449.69 287.09 May 1992 - - 359.40

Source: Nomial wagerates arefrom Agricukdral Secretariat, Bank of Uganda.Ihe CPI hasbeen prepared by splicingtogether the OldKampala Cost of Livng Indexforlow incomegroups (1984 to 1988)and the New KampalaCost of Lving Index (1988onwards). The indexhas been"rebased* upon December1988 by dividingeach of the tw indicesby theirDecember 1988 valus. expressed in terms of payment per hectare, is difficult to interpret. Looking at the real wages of casual workers, it appears that the relatively rapid growth of Ugandan agriculture, particularly between 1986 and 1992, was substantiallyshared by the smaller peasants and that this led to a rise in the supply price of their labor and, consequently,a reduction in rural poverty.

Distortions in Product and Factor Markets

4.27 Distortions in product and factor markets, such as quantitativerestrictions on imports and importand export licensing,confer significantincome on a favored few, mainly throughthe creation of rent income. Insofar as this income is derived from unproductiveactivities, the distortionstransfer incometo higher incomegroups. The removalof these distortions, therefore, reallocatesresources

For examle, contract wage rato ixs pecific to the type of tak per ha and nothing is known about the composition of tank over time and/or themethod of weighing the wage rtes for individual tak in arriving at the overall wage rat for contret labor. 52 Chapter4

Table 4.5: Fiscal and Monetary Performance

1986/871987/88 1998/89 1989/9" 1990/91 1991/92

Flcal Ratio Total Revenue/Total Expenditure 52.6 50.9 52.2 54.0 50.8 32.0 Tax Revenue/TotalRevenue 83.2 87.4 84.5 91.6 93.7 93.4 Coffee RevenuedTax Revenue 41.3 32.0 13.3 17.3 9.9 1.2 Total Revenue (inci GrantsyGDPMP 4.6 6.9 7.0 8.3 10.8 13.5 Total Revenue (excl Grants)/GDPMP 4.2 5.4 5.4 6.8 7.1 6.6 Total Expenditure/GDPMP 7.9 10.5 10.3 12.7 14.1 20.6 Capital ExpenditurewTotal Expenditure 37.5 44.2 33.5 43.8 51.5 44.5 Current Expenditure/Total Expenditure 62.5 55.8 66.5 56.2 48.5 55.5 Current A/C Balance/GDPMP (-= deficit) -0.8 -0.5 -1.5 -0.3 0.3 -4.9 Overll DeficitWGDPMP(+ = deficit) 3.8 5.5 5.3 6.5 7.6 14.0

Finaning: Grants/Total Financing 10.7 28.3 30.2 22.6 48.5 49.5 Foreign Borrowing (nct)/Total Financing 7.6 37.8 41.3 102.2 45.0 36.2 Domeatic/Total Funncing 81.7 33.9 28.5 -24.8 6.5 14.3

Moneta Ratios Claini on Government/Domestic Credit 50.0 49.2 38.6 11.5 10.7 31.9 Crop Financing/Domeatic Credit 18.3 18.7 28.4 30.6 33.6 20.6 MI/GDPMP 5.4 5.8 6.1 5.9 6.1 6.2 M2/GDPMP 6.2 6.4 6.8 6.8 7.2 7.9 M1/M2 88.4 90.3 90.2 86.2 83.8 79.2

Infltion: CPI Kampala 216.5 167.9 130.5 45.4 24.6 42.1

Note Capitalexpenditures and total expenditaresusclide not knding.

Source:Muinutsy of Fuance and Econonuc Planning, IWF and stqf estimates. away from relatively well-off groups, causes increases in productiveactivity, and therefore benefits a broader cross-section of society, including the poor. Since the introduction of the ERP, the Governmenthas made considerableprogress in reducingsuch distortionswithin the economy. With respect to the trade regime, the process which began with the introduction of the Open General License (OGL) and the Special Import Program (SIP) has now culminatedin a system whereby the Governmenthas abolishedall importand export licenses and replacedthese with an automaticexport and import certification system, subject to a narrowly defined negative list. Furthermore, on the domesticfront, the restrictiveIndustrial License Act has been abolishedand the InvestmentAuthority established as a one-stop shop for encouraging private investment and promoting the newly promulgatedinvestment code. All of these measuresare expectedto have a significantpositive impact on the business environmentwhich should lead to productivejob creation and an increase in wages and thereby in the overall standard of living. Chapter 4 53

Inflation and Its Impact on the Poor

4.28 As over 90 percent of GDP in SSA countries goes to consumption,and more so in poorer countriessuch as Ugandaand by the poorer members of the population,bringing inflationinto check and reversing the trend is prudent macroeconomicpolicy and is good for poverty reduction. Inflation essentiallytaxes all thosewho participatein the market as buyers of goods and services, and generally distributesthe benefits in favor of government.The merits of this inflation tax are questionableon efficiency and equity grounds; particularlysince the benefits of the expendituresgenerated through the inflationtax tend to be narrowlydistributed while the adverse effects are more pervasive.

4.29 In Uganda, own-productionaccounts for about 40 percent of the total consumptionof the rural poor. However, to the extent that the poor have to depend on the market for over 60 percent of their consumptionneeds, their welfare is affected adversely by the high rates of inflation. The negative impactof inflationon the rural poor is further aggravatedby the fact that food crops, whose terms of trade have declined, dominatethe marketablesurplus from the agriculturalsector, while a few critical 'traded' consumables,such as sugar, soap, vegetableoil, etc., figure large in household purchases.Therefore, the positiveimpact that inflationhas had on agriculturalincomes has been far less than its negative impact through expenditures.Furthermore, as a result of infrastructuraland transportation bottlenecks, the marketing margins of intermediaries tend to be high, further dampening the incomes of agriculturalproducers while inflating their expenditures. For instance, between 1987and 1990, the margin betweenproducers' and consumers'prices for matokeincreased from 17 percent to 45 percent, while that of maize increasedfrom 14 percent to 81 percent, over the same period.

4.30 While adverse weather conditionsand the frequent adjustmentsto the exchangerate and to interest rates have all contributed to the escalation of prices, the imbalance between government revenues and expenditures,and the link between the fiscal deficit and monetary expansionhas been the primary cause for the high rate of inflation in Uganda. The tax effort in Uganda is still among the lowest in SSA,1 and total revenuesare sufficientto financeonly about 50 percent of government expenditures.While external loans and grants have traditionallyfinanced a large part of the fiscal deficit, borrowing from the domestic banking system has been an extremely important source of fmancingfor the Government, at least until 1988/89. Consequently,until 1988/89, the large-scale monetizationof the deficit resulted in triple digit inflationlevels (Table4.5), thereby severely eroding real incomes in Uganda.

4.31 In order to stem this erosion, one of the key objectivesof the stabilizationcomponent of the adjustmentprogram in Ugandahas therefore been to keep prices in check by mobilizingincremental fiscal revenues and restricting expendituresto budgeted amounts. Bringing inflation down to levels comparable with those of major trading partners is a declared objective and a continuingchallenge for the Government.Since 1990, the Governmenthas embarkedupon a concerted program aimedat mobilizingincremental revenues and controllingexpenditures. The tax and tariff regime has been rationalized,the rate structure has been gready simplified,and tax exemptionsminimized. With the establishmentof the UgandaRevenue Authority (URA) in September1991, the Governmenthas taken a bold step forward in overhaulingits system of tax administration.Furthermore, the Government

In 1990,total revenue to GNP ratior for Malawi,Zaire, Kenyaand Sierra Leone have bee 23 percent, 12 percent, 23 percent,ad 9 perct respectively.World Dewlopnvnt Repori, 1992, Tho World Bank, Washington,DC. 54 Chapter4 has instituteda strict program aimed at keeping expendituresin line with the budget and has started paying back its arrears to the banking system.

4.32 While day-to-dayimplementation problems abound, the overall results of the Government's efforts have been very encouraging.As illustrated in Table 4.5, government revenue efforts have increasedfrom about 4 percent of GDP in 1986/87to about 7 percent of GDP in 1991/92. During the first three monthsof 1992/93(i.e., July to September1992), the Uganda Revenue Authorityhas mobilizedover U Sh 60 billion in revenues as comparedwith U Sh 31 billion for the corresponding period in 1991/92, implyingan annualizedtax-to-GDP ratio in excess of 8 percent. The efficiency of tax collectionhas also increasedgreatly with URA's recurrent costs, as a percentage of revenues mobilized, declining to about 5 percent and projected to decline further to 3 percent by 1994/95. While expendituresas a percent of GDP appear to have increasedsignificantly, these largely reflect an effort on the part of the Governmentto consolidatethe budget by includingall expenditureswhich were previously off the budget. As a result of these measures, claims of the banking sector on the Governmenthave continuedto decline. Governmentclaims as a percentageof total domestic credit decreasedfrom about 50 percent in 1986/87to under 11 percent in 1990/91, increasingthereafter to about 32 percent in 1991/92. The increase in 1991/92was the result of a temporary lapse in fiscal managementwhen the Governmentfailed to cut expenditures in line with the shortfall in import support assistancefrom the donor community.The Government,however, correctedthe situationin the fourth quarter of the fiscal year and its financialprogram is back on track. Consequently,Uganda has been successful in reducing the average rate of inflation in recent years; inflationdeclined from 216.5 percent in 1986/87to 24.6 percent in 1990/91,before it reversed to 42.1 percent in 1991/92, on account of the drought and the temporary slippage in fiscal management.Largely as a result of the renewed financial discipline,the average monthlyinflation rate between July 1992 and February 1993 has been -0.2 percent.

4.33 Seasonai*y of lation and Intra-Year Vulnerability.The harvest period in Uganda is different dependingon the crops harvested and geographicdistribution of the farming activity. While crops like matokeand cassava can be harvested throughoutthe year and there are multiplecropping seasons in several parts of the country, there are areas, particularly in the north and the east, which are characterizedby a single yearly harvest. For farming households, harvest is usually rich, and post-harvestcan be lean even for those who are otherwise well-to-do.The well-beingof net-buying households in rural areas, rural and urban wage earners, and otherwise well-to-do farmers (in the leaner months) suffers when prices swing upwards seasonally on their long-term trend. In addition to the agriculturalcycle, seasonalprice variabilitycould be conditionedby short-term macroeconomic managementsuch as the seasonality in bank credit including those destined for coffee purchases. Understandingthe underlyingfactors for the seasonalvariations in prices can help in facilitatingthe design of short-term poverty sensitive macroeconomictargets within the perimeters of the annual targets.

4.34 Using monthlydata for the decadebetween 1982-92(using the old, the new and the combined series), inflation in Uganda appears to follow a tri-modal pattern. In the first cycle, inflation starts to pick momentumin November and reaches an apex in January. After a brief period of lull, the second cycle comes in April and May. The third is a short spell in September. These peaks match the coffee harvestingseason well as coffee deliveriesrise between Novemberand January and again between July and August. Seasonalfluctuations in aggregatedemand caused by coffee crop finance could be partly responsible for the seasonal variations in the aggregate price level. The seasonal fluctuationin prices assumesspecial relevancegiven the weak post-harveststorage conditionsand the importanceof perishableproducts in the product mix. Householdvulnerability to inflation in Uganda Chapter 4 55

would therefore oscillate seasonally, particularly given that the poor have limited assets, liquid or otherwise, serving as a hedge against bad days.

The Provision of Social Services

4.35 Another channel through which adjustment policies impact on the poor is through the restructuring of public expendituressuch that an increased emphasis is placed on the provision of economic and social services which increase human capital. Expendituresin areas such as primary health, primary educationservices, agriculturalresearch and extension, etc. empowerthe poor, raise their productivityand augmenttheir earning potential in the long run.

4.36 While Uganda still spends far less on economic and social services than do most countries in the world and in SSA, the adjustmentprogram in Ugandahas nonethelessemphasized the need for and resulted in a visiblechange in public expenditurepriorities with expenditureallocations increasing in favor of social sectors in general and on education and health care in particular (Chapter 7). Between 1989/90 and 1991/92, there has been a 49 percent real increase in locally funded expenditures on primary education whose share in GDP has increased from 0.4 percent to 0.5 percent. Similarly, locally funded expenditureson primary health increasedby about 238 percent in real terms over the same period. Given the extremely low revenue effort, discussed above, the Government plans to continue increasing these allocations steadily, keeping in mind the macroeconomicimplications of the constraintsimposed by the tight resource envelope.

Conclusion

4.37 Despite severe data constraints,which limit the analysis presentedin this chapter, there is no denying the fact that the last five years have seen an increase in production, a decrease in inflation, and a marked improvementin the overall economicand political environment. To that extent, the population of Uganda has benefitted. With the return of peace and security to most parts of the country, Ugandan citizens are once again able to feed and clothe themselves and plan for a better tomorrow. Notwithstandingthe 'peace dividend", had the Government not embarked upon the adjustmentprogram when it did, it is our conclusionthat the economicand social climate would have continuedto deteriorate, with the gains from peace being eroded by the continuedravages of triple- digit inflation,resulting from irresponsiblefiscal policy. The coffee industrywould not have survived the developments in the international market; foreign exchange shortages would have continued thereby stiflingagricultural and industrialdevelopment; and incomedistribution would have worsened with significant rents accumulating to the favored few. The Government should therefore be applaudedfor implementinga courageousand demandingeconomic reform program.

4.38 It is often argued that there can be severe short-term 'social costs" associatedwith structural adjustmentprograms which traditionallyemphasize the eliminationof price controls, severe cuts in fiscal expenditureswhich are often accompaniedby cuts in wages and salaries and the elimination of subsidies.While this mightbe true in some cases, in the Ugandancontext, the adjustmentprogram has not called for wage freezes (insteadthe emphasishas been on increasingsalary levels); reductions in subsidies(which were few to start with); or the eliminationof price controls (which were limited and were almost always nonbinding).As has been demonstratedabove, there appears to have been a general improvementin the quality if life in Uganda and most Ugandansappear not to have been adverselyaffected by the social costs of adjustment.While it is too soon to evaluatethe results of the recently started civil service retrenchmentprogram, the discussionof labor markets and the salary structure in Uganda (Chapter 6) indicates that the costs associated with such retrenchmentwill be 56 Chapter 4 mitigated by the fact that the recent shortage of labor will enable retrenched civil servants to be absorbedin other sectors of the growing economy and that the householdfarming sector and/or the urban informalsector is able to provide returns to labor which are comparableto what the retrenched civil servants were receiving from the Government.

4.39 However, having said that, there appears to be very little cause for complacency.Living standards, as was discussed in Chapter 1, are still far below the levels achievedin the 1960sand the early 1970s. While the Governmentis on the right track, it is imperativethat the Government "stay the course' and continuevigorous implementationof the adjustmentprogram which is designed to transform the subsistence economy into a strong, monetized economy. The poor in Uganda are located in areas with little access to roads, transport, communication,agricultural services and marketing facilities. Overhaulingthe incentive system and allowing the markets to operate may be necessary conditions for restoring incomes, but these are not sufficient. To the extent that infrastructuralbottlenecks prevent Ugandansfrom participatingin an increasinglymonetized economy and the majority of the people continueto eke out a living in the subsistenceeconomy, the benefits of adjustmentwill, for the most part, only have an indirect impact on their existence. Adjustment policies can only have a beneficial impact if the people are empowered to participate in the new economicenvironment which results from the policies implemented.As the Government continues implementingits adjustment program, it is imperative that it also devote adequate resources to develop and maintaina poverty monitoringsystem which will enable it to quantify and measure the impact that the adjustmentpolicies are having on the lives of the poor on a continuingbasis. The Integrated Survey presently underway is an excellent example of the kind of data that needs to be collected on a fairly regular basis.

4.40 The fbllowingchapters of this report are forward-lookingand present a vision for the future economic developmentof Uganda together with a set of policies designed to make that vision a reality. At the heart of the vision is the overarchingobjective of reducing poverty through sustained economicgrowth. Part H

A Strategy for Reducing Poverty N1 5

Accelerating Economic Growth

T7hecountries that have been most successful in attacking poverty have encouraged a pattern of growth that makes efficientuse of laborand have invested in the human capital of the poor. Both elementsare essentiaLThe first provides the poor with opportunities to usetheir most abundant asset-labor. Thesecond improves their immediate well being and increases their capacity to take advantage of newly created possibilities. ' (World Development Report 1990, World Bank, p. 51.)

Introduction'

5.1 In almost every country reducing poverty is a key objective, if not the raison d'etre, of developmentpolicy. For a countrylike Uganda,whose GNP per capita is estimatedat under US$170, making it one of the poorest countries in the world, the imperativeof reducing poverty cannot be overemphasized.Experience from a cross-sectionof countries has demonstratedthat economiesthat have achieved high rates of growth (e.g., Botswana, Mauritius,Thailand and Indonesia)have been most successfulin reducingpoverty. Overall economicgrowth results in a rise in incomes, which is a necessary but not sufficient condition for reducing poverty. Poverty in Uganda is largely a reflection of the fact that the economy has failed to achieve sustained growth for a quarter of a century (1965-1990).During this period Uganda's GDP per capita declinedat an average annual rate of 2.4 percent, compared to a weighted average annual growth of 1.7 percent for low-income countries.' It must be added that Uganda was not alone; only a handful of SSA countries bucked the trend towards economic retrogression.'

5.2 The pattern of economicgrowth also matters and this means that incomes which are rising as a result of economic growth must also be accompaniedby a reduction in income inequality. Internationalexperience has shown that the growth that benefits the largest number of people is that based on labor-intensivemethods of production. Labor is the most important asset, sometimesthe only asset, owned by the poor. To enable the poor to participateand share in the growth, labor and product markets have to function efficiently.Public policy affects the way these markets work and, at a broader level, determinesthe rate and pattern of growth. Policy has an even more critical role to play in improvingthe social conditionsof the people, principally through meeting basic needs, particularly in health and education.

5.3 This chapter advocatesa growth-orientedstrategy for reducingpoverty in Uganda. It reviews Uganda's recent performance on the growth score, suggests alternative growth scenarios and examinesthe sectoral sources of growth. In doing this one must bear two points in mind. The first point is that there has been a general slowdownof growth throughout the world. In the 1965-1980

This chapter dmws on Uganda: AgnculturalSector Memorandum, Agriculture and Environment Operations Division, World Bank, 1992 and NationalAgrickural ResearchStrategy and Plan, Vol. H: Prnorties and Programs, Government of Uganda.

Unles otherwise specified, the comparative data used in this chapter come from the World DevelopmentReport, World Bank, 1992. Also, wherever the expresion low income countries appears in this chapter it excludes China and India.

7n The notable exceptions are Botswana, Mauritius, Burundi, Lesotho, Cameroon and Congo. 60 Chapter 5

period 42 of the 78 low and middle incomecountries for which data are available achieved average annual GDP growth rates of 5 percent or higher. During the next decade (1980-1990)only 15 out of 86 low and middle incomecountries managedgrowth rates of that magnitude.Clearly, the global recessionand the internationaldebt crisis of the 1980shave taken their toll. The second point is that growth actually accelerated in East Asia and the Pacific and in South Asia during the 1980s. This suggests that, in spite of the odds, rapid growth is attainable, provided certain basic conditionsare met (paras 5.6-5.10).

5.4 While the imperativeof reducing poverty in Uganda calls for rapid growth, the past failed efforts directed at promotinggrowth and alleviatingpoverty in SSA seem to argue for a long-term perspectiveon poverty reductionin Ugandaand other SSAcountries. Skepticismregarding the ability of Uganda and other SSA countriesto break out of the low or negative growth conundrumdoes not seem to be entirely misplaced. Still, as will become clear later, Uganda stands a good chance of followingthe fast track to povertyreduction. Moreover, there is encouragingevidence from a number of countries to the effect that the developmentprocess can be telescoped into a much shorter time frame. The outstandingexamples are the Republicof Korea and China which, in recent years, more or less doubled per capita output in just a decade. Two SSA countries have done just as well. Botswanaand Mauritius also virtually doubled per capita output within the short space of ten years. This helps to counter the view that SSAcountries are condemnedto slow growth. Still one must raise the question:why have a few countriesbeen so successfulin increasingoutput while so many others have failed miserably?There is obviouslyno blueprintfor rapid growth but a number of ingredients for success can be identified. These are discussed below.

5.5 Before turning to the prerequisitesfor rapid growth, it is useful to ask: is the performance of the fast-growingcountries explainedby special factors, i.e., small or large population, more or less homogeneouspopulation, small land area, a vast mineral endowment,etc.? Obviously,there are circumstancesand conditionspeculiar to each country and it is to be expectedthat these would come into play as a country tries to grow and develop. However, special factors by themselvescannot explain rapid growth. After all, special factors can be as easily squandered as taken advantageof. Take one example.Rapid growth in Botswanahas been built upon diamond exports while many SSA countries similarly endowedwith mineral wealth have failed to prosper. Moreover, special factors can be adverse rather than favorable. Thus, Botswanahas achieved rapid growth in spite of being a land-lockedcountry and being buffeted, until recently, by apartheid related turmoil in the region.

Prerequisites for Accelerated Economic Growth

5.6 The experienceof the economicallysuccessful developing countries shows that three factors namelymacroeconomic stability, high rate of investmentbacked by domesticsavings, and a high rate of literacy and numeracy, are absolutelynecessary for achieving rapid growth over the long term. Macroeconomicstability is needed to give people the confidenceto undertakelong-term investments, particularlythose involvingtechnological change in industryand agriculture. The main indicatorsof macroeconomicstability are low fiscal deficits (or budget surpluses), low and stable inflationand a stable exchangeand trade regime. Countrieslike Botswanaand Mauritiushave been constantin their pursuit of policies that ensure the maintenanceof macroeconomicstability. In Table 5.1 inflation is used as a proxy for macroeconomicbalance. In the earlier period (1965-1980)the world economy experiencedgenerally low inflationand Uganda's averageinflation rate was not far above the average of 17.3 percent for the low-incomecountries. In the ensuing decade (1980-1990)Uganda clearly failed to achieve macroeconomicstability and inflation surged, while the successful comparator countries either reduced or kept inflationlow. Chapter5 61

5.7 Another fundamental prerequisite to Table 5.1: AnnualRates of Inflation rapid growth is a high rate of investnent. That Ugandaand ComparatorCountries means investmentin farms, factoriesand other (Percent) productive assets as well as investment in human capital and in economic and social 1965-1980 1980-1990 infrastructure.Such investmentmust in turn be backed up by high rates of domesticsavings by and 21.4 107.0 either the public sector or the private sector or Ghana 22.9 42.5 both. Of course, foreign savings and investment BovAw 8.4 12.0 have an importantrole to play in development, Mauritius 11.8 8.8 both in the rich countries and the poor ones. Indonesia 35.5 8.4 But in general, foreign investment often Thailand 6.2 3.4 follows, rather than leads, domesticinvestment. Table 5.2 shows that Uganda lags well behind the successfulcomparator countries on the saving and investmentscore.

5.8 High rates of literacy and nwneracyare Table 5.2: DomesticInvestment and Savings also fundamental to the achievement of rapid Ugandaand ComparatorCountries and sustained growth. Literacy and numeracy (Percentof GDP) have important externalities which justify investment in these programs. But on the Domestic Domestic narrower grounds of promotinggrowth there is Savings Investment a strong case for heavy investmentin education, 1965 1990 1965 1990 particularly primary education. The ability to read and write is crucial to the acquisitionand Uganda 11 5 12 11 applicationof modernproduction and marketing Ghan 18 15 8 11 skills. In agriculture, for example, the farmer's Botswana 6 40 -13 30 capacity to adopt yield-raising innovations Mauritius 17 30 13 21 depends on his or her ability to at least read the Indonesia 8 37 8 37 messages directed at him or her. This is why Thailand -2 37 19 34 countries that have made rapid growth a key objective of national development policy have invested heavily in educationat an early stage of their development.Table 5.3 shows that Uganda again does poorly on this score.

5.9 However, lessons from internationalexperience show that the three fundamentalsdiscussed above are also not enough. Five additional ingredients for accelerating economic growth are highlightedhere:

(i) Transfoimation of Agricuture. This involves the modernizationof agriculture through the introductionof new technology,accompanied by the provisionof adequateinfrastructure and in some cases, land reform. The transformationof agriculture offers many advantages.It results in growth that is egalitarian, except where land ownership is highly concentrated;it initiatesthe process of diffusionof modern productionskills to large numbersof people; and it creates a ready linkage between agriculture and industry. The conventionalwisdom has been that agriculturalsector growth tends to be limitedby factors such as the slow growth of domesticdemand for food, the protectionistpolicies of the rich countries and the slow pace of diffusion of agricultural technology. While this is true to some extent, this type of 62 Chapter5

Table 5.3: School Enrollment pessimismcan be overdone. It has been Ugandaand ComparatorCountries partially disprovedby the experienceof (Percentageof Age Group Enrolled) a number of countries. China, for example, registered an average annual § "y Seeondary Tertiary growth rate of 6.3 percent in 1965 1989 1965 1989 1965 1989 agriculture during the 1980-1989 period. Other countrieswhich are much Uganda 67 77 4 13 0 1 smaller than China (Morocco, Togo, Ghan 69 75 13 39 1 2 Bhutan and Burkina Faso) achieved Botswana 65 111 3 37 .. 3 Mauritius 101 103 26 53 3 2 comparablerates of agriculturaloutput Indoneia 72 Ill 12 47 1 growth during the same period. For a Thailand 78 86 14 28 2 16 country that is heavily dependentupon agriculture for employment, exports and growth it makes sense to invest in agriculture as a prelude to industrialization.

(ii) TechnicalEducation. General education is essential but it is not enough; it does not equip the work force with the abilities needed to master modern production techniqueswhich are critical to raising productivity. Rapid growth demands more and more technicallytrained people, with training in mathematics,engineering and the sciences. The experience of the Republicof Korea, Malaysia, Thailandand other fast-growingcountries confirms this.

(iii) Exports. The fast-growing countries have also emphasized exports, and have adopted a vigorous program of export promotion,supplemented by specific governmentinterventions (export processingzones, fiscal incentives,the provisionof business-friendlypublic service, etc.). While governmentinterventions have no doubt played a role in the export expansion of the Newly IndustrializedCountries, one must take care not to exaggerate the range and efficacyof these interventions.International experience shows that rapid export expansioncan be based upon minerals, agriculture, manufacturesor, more often, some combination of these. In some cases (for example in the Republic of Korea and Mauritius) rapid export expansion has meant dramatic rates of growth of manufactures. However, in many cases rapid export expansionhas been resource-based(Malaysia, Thailand and Indonesia,to name a few). With regard to agricultural exports, the trick to success involves three elements: modernizetraditional export crop production; diversifyexport crops; and increase domestic value added through the production of high value crops and increased processing of agriculturalcommodities.

(iv) T7heCatalytic Role of Foreign Investment. This role can be played or supplementedby particular communitiesof settlers (for example, the Asians in East Africa and the Chinesein Southeast Asia). The foreign investor often brings with him or her capital, technology, managementtechniques, skills, quality standards and markets. As partners, competitorsor suppliers, local investorstherefore benefit from the presence of foreign investors.

(v) Problem-solving Partnership Between the Public and Private Secdors. Fast growing economieshave been characterizedby an effective, problem-solvingpartnership between the public and private sectors as opposed to an adversarial relationshipbetween the two. Such a partnership is built on the premise that the Government is willing and able to help individualsand businessesto grow and becomerich. Chapter 5 63

5.10 Uganda fares poorly on the prerequisitesand the other essentialingredients for accelerating growth, discussedabove. First, althoughwithin sight, macroeconomicstability has yet to be achieved. The key to achieving macroeconomicstability is firm control of the government budget. Second, saving and investmentrates are very low. For the domesticsavings rate to improve significantlythere must be in place a financial system capable of mobilizing savings from the vast majority of the people, which does not yet exist in Uganda.Third, illiteracyrates are too high, especiallyfor women; in 1990 adult illiteracy was 52 percent for the total populationand 65 percent for women. Only an accelerated program of universal primary education can remove this handicap. Fourth, there is insufficienttechnical education. Deliberate efforts need to be made to reorient educationtowards the technicalsubjects. Five, as will become clear later, agricultureis still very backward. High-yielding crop varieties are yet to be introduced;few modem inputs are used; only the simplestof implements are used; and the rural road network is in an appallingstate of disrepair. Six, the export base is too narrow and fragile, dependent as it still is on coffee for about 75 percent of export earnings. A determined push to expand exports is called for. Seven, although investor interest appears to be on the rise, only a trickle of foreign investmentis coming into the country. The Government needs to address systematically the problems of poor infrastructure and bureaucratic impediments to investment.Finally, there is no effectivepartnership between the publicsector and the private sector. There is a need for a partnership which provides a forum for problemsto be aired, followedup and solved.

How Did UgandaFare?

5.11 Uganda's overall economic growth Table 5.4: AverageAnnual GDP GrowthRates record is compared to that of a number of Ugandaand ComparatorCountries countries in Table 5.4. The table shows that (Pecent) while countries like Uganda and Ghana missed the opportunityto grow for the greater part of 1965-1980 1980-1990 the past quarter century, others raced ahead. This comparison is not entirely fair to Uganda, given the fact that for much of this period the Uganda 0.6 2.8 country was engulfed in chaos. Indeed, it can Botswana 13.9 11.3 be argued that Uganda's capacity to grow has Mauritiuw 5.2 6.0 not really been tested up to now. Performance Indonesia 7.0 5.5 under the ERP over the long term will provide Thailand 7.3 7.6 such a test. So far the results have been encouraging;the past five years of adjustment have witnessed an average annual GDP growth rate of 5.3 percent. Admittedly, there were some special factors behind this achievement.The growth resumed from a depressedbase and the return of peace and security to large parts of the country after more than a decade of civil war made it possible to bring idle capacity,be it an industrialplant or farm land, back into production. However, the high rates of growth recorded over this period cannotbe wholly ascribed to these special factors. The policies and measures implementedunder the ERP, with their emphasis on greater reliance on market forces, have played an importantpart in stimulatingthe growth of production. Other measures includethe injectionof more foreign exchangein the form of import support into the economyand the rehabilitationof the infrastructure,particularly roads. Growth during the 1987-1992period, in fact, came in a spurt of an average of 7.2 percent over the first three years, followedby a slowing down to 4 percent during the next two years. The recent decelerationof growth is indicativeof the fact that the impact of the special factors on production has started to wear off. But the lower rates of growth recorded in 1990 and 1991can also be explainedto some extent by poor rainfall in those 64 Chapter 5

Table 5.5: AverageAnnual Sectoral Growth Rates Uganda and Comparator Counties

Agricukure Industry Manufacturing Srvices 196540 1960-90 196580 1980-90 196580 1980-90 196540 1960-1990

Uganda 1.2 2.5 -4.3 5.5 -3.7 5.2 1.1 3.3 Ghana 1.6 1.0 1.4 3.3 2.5 4.0 1.1 5.7 Botuwana 9.7 -4.0 24.0 13.0 13.5 5.3 1 1.5 11.9 Mauritius .. 2.6 .. 9.2 .. 10.8 .. 5.1 Indonesia 4.3 3.2 11.9 5.6 12.0 12.5 7.3 6.7 Thailand 4.6 4.1 9.5 9.0 11.2 8.9 7.4 7.8

years, since Ugandan agricultureis entirely rainfed. Except in the drier, semi-aridareas of the north and Karamoja, rainfall is normally adequate for growing crops but once in a while the rains have been known to be inadequate, late or poorly distributed.

5.12 Uganda's record on sectoral growth is comparedto that of five other countries in Table 5.5. Uganda's performancein agriculture and industry during the 1965-1980period was dismal. In this period agriculturalproduction failed to keep pace with populationgrowth while industrydeclined by over 4 percent each year. The following decadesaw an improvementon the past 15 years but even then agricultural output barely kept pace with populationgrowth. Ghana was in the same league as Uganda and their performancecontrasts sharply with that of Indonesiaand Thailand which together provide a good example of rapid and broad-basedgrowth.

5.13 The main reason for the poor performanceof the various sectors in Uganda (and hence the economy as a whole) during the period under review is the general instabilitythat the country went through. Insecuritydisrupted agriculturalproduction and shut down markets while industrialplants were looted or abandoned. The country was isolated from the rest of the world for many years. Fortunately, starting in 1986, Uganda has largely returned to normality. Improved security and law and order, together with other factors, have had a positive impact on the performance of all the sectors. In reflection of these developmentsagriculture grew during the 1986-1991 period by an average of 4.2 percent, industryby 11.8 percent and services by 5.8 percent (Table 5.11).

The link Between Growth and Poverty Reduction In Uganda

5.14 Each successivePolicy Framework Paper (PFP)that Ugandahas negotiatedwith the IMF and the World Bank has targeted GDP to grow by 5 percent per annum. There appears to be really no basis for this growth rate other than that, if attained, it would permit a modest increase of about 2 percent per annum in per capita output.74 Looked at in relation to Uganda's growth record over the long term (minus 2.4 percent during 1965-1990)a growth rate of 5 percent seems robust. But when viewed in the contextof poverty reduction,5 percent is low (see below). Ugandacan chooseto stick to the PFP growth scenario of 5 percent per annum for the next 10-15 years. Alternatively,it could

This growth asaniption find common uage in Sub-SaharanAfrica PFPs. Chapter 5 65 lower its sights in line with the recent slowdownin growth to around 4 percent. Althoughnot ruled out, the scenario involvinga GDP growth target below that incorporatedin successive PFPs is not discussed in this report. Rather, what the report tries to do is to challengepolicymakers in Uganda to aim higher; the higher growth scenario presentedhere is a target and not a prediction.

5.15 How high can Ugandaaim? This report suggests that Uganda shouldaim to push the average annual growth rate towards about 8 percent by the end of the projection period. The choice of 8 percent is bound to be largelyarbitrary. It has, however,been dictated by the imperativeof reducing poverty in Ugandain the shortestpossible time. Countriessuch as Thailand, Botswana,Mauritius and the NICs generally have consistentlyachieved growth rates of 8 percent or higher for long periods. Can Uganda do the same? The initial conditions(literacy rate, economic and social infrastructure, population growth rate, macroeconomicpolicies, etc.) were not more favorable in the comparator countries than in Ugandatoday. It must, however, be admittedthat the case for high rates of growth in Uganda depends largely on the speed with which agriculture can be technologicallytransformed. The natural resource endowment is so favorable as to give cause for hope that rapid growth is possible in Uganda.

5.16 The base GDP growth rate (5 percent) Table 5.6: Years Required to Double GDP and the high growth rate (8 percent) have very (Per Capita) dramatic implications for poverty reduction. Attainingthe former would mean that per capita PopulationGrowth Rate GDP Growth Rate output rises by 2.5 percent per annum while (%) 5% 8% achieving the latter would cause per capita output to increase by 5.5 percent per year, 3.0 35 15 assumingthe 1980-1991intercensal population 2.7 29 14 growth rate of 2.5 percent holds for the next 2.3 27 13 decadeor longer. There is significantdifference 2.1 25 12 between the two growth scenarios when it comes to the number of years it would take to double GDP per capita (Table 5.6). With the economygrowing at 5 percent it would take Uganda 29 years to double GDP per capita which would still leave it below today's average GDP per capita for low-income countries. Were Uganda to achieve a growth rate of 8 percent per annum it would cut in half, to 14, the number of years it would take to double GDP per capita. DoublingGDP per capita by the year 2005 (or approximately US$340 GNP per capita) would still leave Ugandaa very poor country but it would begin to make a dent on the difficult issue of reducing poverty.

5.17 The Population Question. The number of years needed to double GDP per capita is sensitive to the populationgrowth rate used. Table 5.6 shows that, assumingpopulation grows at 3 percent instead of 2.5 percent, with the economy growing at 5 percent per annum, it would take 35 years, as opposed to 29 years, to double output per capita. Conversely, reducing populationgrowth to 2.1 percent while the economygrows at 5 percent would shave four years off the time it would take to double GDP per capita. The calculationis less sensitiveto the populationgrowth rate at the 8 percent (or higher) GDP growth rate than at the 5 percent growth rate. With GDP growing at 8 percent it would take just three years more to double per capita GDP if the population were to grow at 3 percent instead of 2.1 percent. The lessonto draw from this is not that at the higher GDP growth rate Uganda can afford a rapidly-growingpopulation. On the contrary, Uganda needs to slow down populationgrowth in order to reduce poverty withinthe shortestpossible time. In this regard, it bears repeatingthat, even if Uganda is able to double GDP per capita (resultingin about US$340GNP per 66 Chapter 5

Table 5.7: Distributionof GDP (Percent, in CurrentPrices)

Agriculture Industry Manufacturing Services 1965 1990 1965 1990 1965 1990 1965 1990

Uganda 52 54 13 10 8 4 35 35 Ghana 44 48 19 16 10 9 38 37 Botswana 34 3 19 57 12 6 47 40 Mauritius 16 12 23 33 14 24 61 55 Indonesia 51 22 13 40 8 20 36 38 Thailand 32 12 23 39 14 26 45 48

Note: Botswanais a drought-pronecountry where rainfaUcan fail abnosttotaly, thereby wipingout crops. capita) within the next 12 to 15 years it would remain one of the poorest countries in the world. There can therefore be no trade-off between economic growth and population growth.

Structure of Production

5.18 Before examining future sources of growth in Uganda it is necessary to review briefly the present structure of production. A country's structure of production is expected to change during the process of economic development. The universal pattern is for the share of agriculture in GDP to decline over the long term, with corresponding increases in the contribution of industry and services. This kind of structural change has been delayed in Uganda and other countries that have stagnated or actually declined over the past quarter century. In contrast, in the fast-growing economies, the transition to less dependence on agriculture proceeded smoothly (see Table 5.7).

5.19 As Table 5.7 shows, the Ugandan economy is still predominantly agrarian. While the contribution of agriculture to GDP increased from 52 percent in 1965 to over 60 percent in the mid- 1980s, it has now started to come down; in 1991 it was 51 percent. It is hard to say whether or not this is the beginning of the expected structural change. The structural adjustment that Uganda has been undergoing since 1987 has probably started to have an effect but it is too early to measure the effect of this on something as fundamental as the structure of production. As can be seen from the table, the fast-growing economies followed the normal pattern; i.e., the contribution of agriculture to GDP declined during the course of economic development. In Thailand the share of agriculture fell from 32 to 12 percent between 1965 and 1990; in Indonesia from 51 to 22 percent during the same period. It also implies that the share of food crops, which have accounted for 3545 percent of GDP in recent years, will have to decline. Future GDP growth will involve such intersectoral as well as intrasectoral output shifts.

5.20 The share of industry in GDP is exceptionally low for Uganda and the contribution of industry to GDP actually declined from 13 percent in 1965 to 10 percent in 1990. For low-income countries the average share of industry was 34 percent in 1990, having risen from 20 percent in 1965. The manufacturing subsector in Uganda underwent the same sort of contraction as industry as a whole, with its share of GDP falling from 8 percent to 4 percent. This means that Uganda would embark upon a strategy of rapid growth with an extremely small manufacturing base; this presents challenges as well as opportunities. Chapter 5 67

5.21 The share of services in Uganda's GDP was unchanged (at 35 percent) between 1965 and 1990. From international experience the general tendency is for the contribution of services to increase as economic growth proceeds. This tendency is confirmedby the data for Indonesia and Thailand, although Botswanaand Mauritiusbuck this trend. In the case of Uganda, some services, especially commerce, actually increased but growth in this and other areas was cancelled by the effects of the collapseof commodityproduction in the industrialsector and the growing retreat by the farmers into subsistencefood production.

Sources of Future Growth

5.22 The followingsection presentsa projection of the future sources of growth for the Ugandan economy. The basic aim of the exercise is to provide a sense of the rates at which individual subsectorswould have to grow in order for Ugandato aspire to an overall growth rate of 8 percent. There is no pretense that the actual outcome will correspondto this projection; rather, this section presents a vision of pushing Uganda's production possibilityfrontier to the limit and the policies needed to transform this vision into a reality. In agriculture, estimates have been made regarding future area expansion and gains in yields, based on recent performanceand expectationsabout the developmentof a particular crop. There is very little hard knowledge about the industrial sector; consequently,in this case, the projectionsassume that, as private investmentpicks up, the industrial sector would return to the high growth rates seen at the beginningof the ERP. Even less is known about the services sector. Nevertheless,the assumptionmade here is that as growth picks up in the commodity-producingsectors, the demand for more and better services would increase. Services are therefore projected to grow at much higher rates than in the recent past.

5.23 The review of the present structure of production serves to highlight the centrality of agriculture in any developmentstrategy for Uganda. To achieve an overall average rate of growth of 8 percent a year over the next 10-15 years Ugandawould have to raise agriculturalgrowth rates well beyond its long-termhistorical average of about 2 percent per annum. Since this is a relatively short period (less than a generation) in the life of a nation, expectationsregarding economic and social change within this time frame have to be tempered with realism. However, the economically successful countries have shown that per capita GDP can be doubled within the space of ten years and significant progress made in reducing poverty. The sections below explore the question of whether Uganda is capable of achieving this kind of progress. The discussionof sources of future growth centers largely on the agriculturalsector for three reasons. The first, obvious, reason is that agriculture is the dominant sector; given Uganda's resource endowment that is where, in absolute terms, the potentialfor rapid increases in output lie, at least over the time horizon considered in this report. The second reason is that even where industryis concernedthe best prospectsare to be found in agro-processing. The third reason is a practical one: a great deal more is known about the agricultural sector than any other sector. The emphasis on agriculture, however, should not be interpretedto mean that Ugandashould not push for industrialization.The only caveat to be sounded is that this should not be done at the expense of agriculture.

Agriculture

5.24 Importance and Salient Features. Agriculture is, and will remain, the mainstay of the Ugandan economy for a long time to come. The sector now accounts for 51 percent of GDP, generatesabout 90 percent of export earnings and employs 80 or more percent of the labor force. It also providesthe base for a numberof manufacturingand processingindustries. Agriculture therefore represents a potent force for poverty reduction in Uganda. There are a number of reasons for this 68 Chapter 5

claim. Firstly, 85 percent or more of the populationlive in the rural areas where they are engaged wholly or predominantly in agriculture. Secondly, smallholdersaccount for practically the entire agriculturaloutput; it is estimatedthat there are 2.5 million farm households, of which 80 percent cultivate less than 2.5 hectares of land each.' Thirdly, landlessness is not a major problem in Uganda.Fourthly, smallholderagriculture relies primarily on labor-intensivemethods of production. In view of the preponderanceof smallholders,the absence of landlessnesson a large scale and the labor-intensityof production,the benefits of rapid agriculturalgrowth shouldin principlebe equitably distributed.Unfortunately, such a benignoutcome cannot be taken for granted becausethere are other factors at work. The distributionof the benefitsof agriculturalgrowth are affected by such things as the functioningof product and labor markets, the allocationof research and extension resources to different types of crops and the regionaldistribution of physical infrastructure,particularly roads.

5.25 The Compositionof Agnicultural Output. Uganda is endowedwith a wide range of soils, rainfall and altitudes which enable it to produce a diversity of crops and livestock. It also has large fish resources in Lake Victoria and other lakes. The country has a total land area of 19.4 million hectares which are distributedas follows:4 millionha for annual crops, 1.5 million ha for permanent crops, 5 million ha for pasturesand grazing land, 6.5 million ha for forests and 2.4 million for very marginal use. Crop yields per hectare are very low right across the board (Table 5.8). This is so principally because producerslack access to improvedtechnologies and inputs necessary to increase yields. In spite of the havoc wrought by civil wars and the low yields, Uganda has more or less remainedself-sufficient in food. However, occasionalfood shortagesin some parts of the countryon accountof poor rainfall are not unknown.

5.26 The compositionof agriculturaloutput is given in Table 5.10. There are three striking things about the agriculturaldata. The first is the oversized contributionof food crops to agriculturaloutput; during the 1987-1991period food crops accounted for between two thirds and three quarters of agriculturaloutput. The slight decline in the food crop share seen over the past two years probably reflects the effect of adverse weather conditionsmore than anything else. Second, nearly 55 percent of food crop production is considerednonmonetary, i.e., for subsistence.Third, the contributionof cash or export crops to agriculturaloutput is exceptionallylow; it ranged between 3 percent and 7 percent during 1987-1991.? In Ghana, cocoa alone still accounts for 13 percent of agricultural output despite the sharp decline in the role that this crop plays in the economy.

5.27 The next importantsubsector of agriculture is livestock.During the past five years livestock contributed, on average, 17 percent of agriculturaloutput. Livestock is an integral part of farming systems in Uganda and about one third of farm householdsdepend on livestock for a major part of their incomes. Data on the numbers of livestock are poor. However, the Ministry of Agriculture, Animal Industry and Fisheries (MAAIF) estimatedthe 1990/91livestock population as follows: 4.3 millioncattle, 3 milliongoats, 0.8 millionsheep, 0.6 millionpigs and 10-12million poultry. The rest of agriculturaloutput comes from forestry and fishing, with the combinedcontribution of these two subsectors close to 8 percent.

7S Large tea and sugar entate and plantations exst but the total number of hectams under these crops is only about 40,000. In addition to these estates, then are a few large cattle rmnches, wbhoeherds were docimatedby war and disas.

The national accounts equate cash crops with export crops. As nontraditional exports have startedto grow strongly, therm has been a tendencyto under edimnatecash crops. Sinsim illustr the point: it is still classified as a food crop althoughan increasingproportion is being exported. Chapter 5 69

5.28 Past Developments in Agricultural Growth. A realistic assessmentof the growth potential in the agriculturalsector must start with a long-termview of past achievements.The average annual growth rate of the sector during the period 1965-1980was 1.2 percent, which was well below the 2.6 percent achieved by low-incomecountries (rable 5.5). In the 1980-1990period the average annual growth rate of agriculturein the low-incomecountries went up to 3.9 percent, while Uganda only managedto raise the average annual growth rate of agriculturalproduction to 2.5 percent. At 2.5 percent agriculturaloutput growth in Uganda barely kept pace with the growth of population. Uganda's agriculture was no doubt badly disrupted by civil conflict and the attendant economic difficulties.

5.29 In recent years (1987-1991)the agriculturalsector has shown signs of a strong recovery. The average annual growth rate during this period was nearly 5 percent (Table 5.11). The rapid growth registered in the sector over the past five years, made possible by the recovery of the food crops subsector,owed a great deal to the improvementin security. It also reflectsthe fact that output started recoveringfrom a low base. Butthat is not the entireexplanation. The improvementin the underlying conditionsneeded to spur growth also had somethingto do with it. As far as agricultureis concerned the most relevant factors have been the rehabilitationof the roads (especiallythe trunk roads), the rapid increase in the fleet of vehicles on the roads, the dismantlingof price and distributioncontrols and the greater availabilityof inputs. There is also no doubt that Uganda is in the enviable position of being favored with good rainfall. However, food production is sensitive to the vagaries of the weather. During the past two years, agriculturalsector growth decelerated, to less than 3 percent, as production, particularlyof food crops, was adversely affectedby poor rainfall.

5.30 The increased food crop production in recent years has come almost entirely from area expansion. Fortunately, except for a few heavily populatedareas, land availability is not a problem now or in the foreseeable future. It is estimatedthat, averagingover different soil classes, only 30 percent of the cultivableland was used in 1991.7 However, yields are very low and have remained so since the early 1970s. If anything, several crops have registereddeclines in yieldsduring the 1970s and 1980s,mainly on accountof the collapseof the agriculturalresearch and extension infrastructure and the lack of technologicalinnovations.

5.31 Although not entirely inimune, the food crop subsector proved remarkably resilient during the years of political and military turmoil, in sharp contrast with the virtual collapse of cash crop production during those years. In the case of the tree crops (principally coffee and tea) the area planted remained largely unchanged while for the annual cash crops the cultivated area shrank drastically. Cotton best illustratesthe decline in the area under cash crop cultivation:from a peak of 677,500 hectares in 1978the planted area fell to 67,000 hectaresin 1990.Cotton production actually dipped below the 1987 low of 5,800 tons before recoveringto that level in 1990. Similarly, coffee production declined from 160,300 tons in 1987 to 147,400 tons in 1991. Althougha recovery of cash crop production is underway, the factors that have adversely affected cash crop production includelow farmgate prices and delayedpayments to the farmers, insecurity, inadequatesupply of inputs, poor roads and the general breakdown of the agricultural research and extension services. Cotton has experiencedonly a weak recovery owing to the severe institutionalconstraints affecting the production of this crop. The constraintsinclude the virtual monopolyof the cooperativeunions on ginning and the export monopolyof the Lint MarketingBoard.

77 Sec Uganda:Agricultural Sector Memonrandum,ibid. 70 Chapter5

5.32 The numbers of livestockof all types declined during the 1980s, with cattle suffering the heaviestlosses. The cattle populationis estimatedto have declined by about one third on accountof the breakdownof disease control programsand increasedcattle rustling and looting. A slow recovery of cattle numbers has just started. In poultry, however, production has rebounded strongly since 1987, thanksto the improvedavailability of livestockfeed, importedfeed additivesand new breeding stock.

5.33 Future Sources of Agicultural Growth. The source of growth in agriculture has to come from the nontraditionalexport crops, technologychange in food production, and increaseddemand due to increased incomesfrom growth in the export and industrialsectors. Over the past five years the growth in urban demand for food has been an important driving force behind increased agricultural production. Based on population and income growth over the next decade or so the domestic market for food is expected to expand by 3-4 percent per annum. Production has concentrated on the traditional low-value, often bulky, crops such as matoke and cassava which suggeststhat there is room, even as far as domesticconsumption is concerned,for productionto shift over time towards higher-valueor more nutritiouscrops (beans and vegetables,for example)as well as livestock. The potential for import substitutionis alreadybeing tapped in sugar, dairy and other areas; it is almost exhaustedin tobacco. Further scope for importsubstitution probably exists but has yet to be identifiedand exploited.Furthermore, there is now overwhelmingreliance on the marketing of unprocessedfood, even in the urban areas. This can be expectedto change so that increasedfood processingbecomes the cause and effect of further monetizationof the agricultural economy.

5.34 As mentioned earlier, the bulk of the Table 5.8: Present vs Potential Crop Yields increase in agricultural production since the (Metric Tons per Hectare) launching of the ERP has come from the expansion of the area under cultivation. Even Present Potential then the total cropped area is still below the level of the late 1970s. While not a limiting factor on agricultural growth in the country as Coffee (Robusta) 0.6 1.5 a whole, land scarcity is a constraint in some Tea 1.4 3.0 parts of the southwest and northeast. However, Matoke 5.5 8.0 land availabilityis not enough. The fertility of Finger milet 0.3 1.5 the soil is very important.In many parts of the Maize 1.5 2.0 countrythe natural fertility of the soil cannot be Sorghum 0.8 2.0 relied upon to produce high yields. This is Cssava 9.0 12.0 especiallytrue since high-yieldingvarieties are Sweet potatoes .0 20.0 presently virtually unknown in Ugandan Irish potatoes 8.5 20.0 agriculture. As shown in the annex to this Beans 0.8 1.2 chapter, by any standards, crop yields are very Groundnut 0.8 2.0 low which in turn suggests that present yields Simsim 0.4 0.8 can be raised substantially. To say that is not to S Gvernen of Uganda, National Agiculural underestimate the resources and the commitment Research Strategy and Plan, Volume n1:Prioritie needed to raise yields. There are, however, and Programs, 1991and staff estimates. encouraging examples from international experience in this area.

5.35 The domestic market has already been alluded to above. Markets are indispensableto any growth strategy in Uganda, whether agriculture-ledor not. In agriculture it is not too difficult to envisage a situationof very rapid increases in output based on the widespreadadoption of improved Chapter 5 71 varieties and the associated technological packages. Under this scenario, lack of markets could become a major disincentiveto production. To augmentdomestic demand, Uganda will have to look to two outsidemarkets namely,the regionalmarket and the internationalmarket. The regionalmarket in agriculturalproducts is presently small but has the potential to grow, based on the growth of incomes, differences in food preferences, the vagaries of the weather, etc. As the country with probably the best agriculturalpotential in the region, Ugandaowes it to itself to be in the forefront of regional integrationaimed at opening up markets. The internationalmarket offers, in principle, vast opportunitiesfor Ugandanproducts. However, protectionand fierce competitionis the name of the game in these markets . Outsideof the traditionalexport crops, Ugandamust achievemuch higher agriculturalproductivity in order to be able to compete effectivelyin the internationalarena. The focus will be on horticulture, spices and other nontraditionalexport crops.

5.36 As far as the traditionalexport crops are concerned,a strong case can be made for a big push into cotton production. Current production is just about one seventhof peak production. Knowledge of the crop and what it takes to grow and process it have survived to some extent in the growing areas. The way forward is to address systematicallythe problems plaguing this crop (poor quality seed, a weak or nonexistentextension service, ineffectiveprocurement by the cooperativeginneries, etc.). A relatively large domestic processingindustry could be built upon the fortunes of this crop. In addition, Uganda could count on exports of cotton and cotton products.

5.37 Over the mediumterm, with increases in productivitytogether and with an increase in the area under cultivation and in markets, the agriculturalsector can be pushed to grow at an average annual rate of between 4 and 5 percent (Table 5.11). Despite this sustained growth, however, the share of agriculture in GDP is projected to decline from its present level of 50 percent to about 45 percent in the year 2005, in constant 1987 prices.

Industry

5.38 Industry and Economic Development. Industry plays an important role in economic development. In fact, hardly any country has "developed" without becoming industrialized. Industrializationis needed to absorb increasingnumbers of people releasedfrom the land as a result of rising agriculturalproductivity and it provides the principal vehicle for the diffusion of modern productionskills. Furthermore, the experienceof the Asian and other NICs has shownthat, with the right kinds of policies and adequatemarket access, very high rates of growth (10 percent or more) can be achieved in industry for long periods. The Republic of Korea, for example, achieved an average annual industrialgrowth rate of 16.4 percent during 1965-1980,followed by a rate of 12.2 percent during 1980-1990.For Botswana(which started from a very low base) the comparableannual industrialgrowth rates were 24 percent and 13 percent for the two periods. In Mauritius it was 9.2 percent during the 1980-1990period. Agriculturehas not been known to grow at such high rates on a sustainedbasis. Developingcountries therefore have a justifiable reason for pursuing policiesaimed at promotingrapid industrialgrowth, especiallythe growth of manufacturing.They realize that rapid growth of overall output (say, in the 6-10 percent range) cannot be achieved without a much faster growth of the industrial sector. The mistakethat many developingcountries have made is to pursue industrialgrowth without due regard to the health of the agriculturalsector. In Uganda, as in many countries, it is not a matter of either agricultureor industry but how both can be developed.

5.39 Past Industrial Performance. Uganda has witnessed a very rapid growth of industrial production since the launchingof the ERP, mainlyon accountof the growth of constructionand, to a smaller extent, manufacturing.The index of industrialproduction has risen strongly from 100 in 72 Chapter 5

1987 to 178.4 in 1991, implyingan average annual rate of growth of 13.7 percent, and the share of industryin GDP has risen from about 8 percentto just under 12 percent over the same period. There were special factors behind this high rate of growth. In the first instance it reflected the fact that growth resumed from a depressedbase. Secondly, it was relativelyeasy to reactivate idle productive capacity to meet pent-up demand for basic consumer goods. There has been a mini boom in construction, driven by rehabilitation investments in the public sector and new investments in residential buildings. However, the share of manufacturingin GDP has gone up by less than one percentagepoint over the past five years which serves to drive home the point that the industrialbase in Uganda is too small to be the driving force behind the rapid growth strategy proposed in this report. The developingcountries that have achievedrapid growth of manufacturingoutput in recent years started off with a larger initial base than Uganda's at the present time. In 1965 Indonesia, Mauritius,Thailand and the Republicof Korea, for example, already had manufacturing-GDPratios of 8, 14, 14 and 18 percent, respectively.

5.40 The current efforts to revive productionat the Kilembecopper/cobalt mine notwithstanding, Uganda has no miningsubsector to speak of. The output of public utilitiesalso remains tiny as they are confinedto the main urban centers which they are not even able to serve adequately. Industry is thus made up largely of manufacturingand construction. As is often true of countries at the early stages of industrialization,manufacturing in Uganda is dominated by the followingproduct groups (with the weightsin brackets): drinks and tobacco(26 percent), food processing(21 percent), textiles and clothing (16 percent) and chemicals,paint and soap (12 percent). Together these product groups account for three quarters of manufacturingoutput.

5.41 Most of the plant and equipment in the industrialsector was installed by the Asians in the 1950sand 1960s.With the expulsionof the Asian owners in 1972the factories, whether in the hands of the public sector or the private sector, fell into disuse and misuse or were lootedduring the civil wars. Thus, followingthe recent rapid growth of industrialoutput some firms are already reaching the limits of their capacity. This is one source of investmentdemand. The other is demand for new investmentin order to produce goods not presently made locally or to expand the production of those already being produced.

5.42 Future Sources of Industrial Growth. Given the structure of output, industrial development in Uganda must initially be agro-based. Uganda needs to step up the processing of agricultural commoditiesthat are produced now and the much larger quantitiesthat it is capable of producing in the future. The demand for processed agricultural products will come from three sources: one, subsistencefarmers drawn more and more into the market economy; two, urban consumerswhose ranks are growing rapidly; and three, foreign buyers.

5.43 Industry must not only be agro-based; it must also be relatively labor-intensive.This is in accord with Uganda's resource endowment.While there are some indicationsthat labor might be in short supply on account of the civil strife, the AIDS epidemic and the easy access to land which encourages Ugandans to eke out a living from subsistence agriculture (Chapter 6), the bulk of Uganda's labor force is underemployed and inefficiently utilized with substantial scope for redeployment.Only the most basic industrialskills are availablein the economy and even these are in short supply. The skills needed to handle advanced technologysimply do not exist, even if the importationof such technologymade economicsense. Furthermore, the initial thrust of industrial growth is bound to center on the usual list of basic consumer goods requiring simple production skills: food, drinks and tobacco, textiles and clothingand so on. This is the establishedpattern in all Chapter S 73

countries that have undergone the industrializationprocess. In Uganda the situation is made more favorableby the fact that the domesticresource base exists to support such production.

5.44 For a variety of reasons (greater reliance on technology, rapidly changing tastes, etc.) the manufacturingsector is normallythe most dynamicsector of the economy. Manufacturingin Uganda is now too small to impart much dynamismto the rest of the economy. To begin to do that it must grow very rapidly. In order to attain the high GDP growth rate (8 percent) postulatedin this report, the manufacturingsector would have to grow by at least 10 percent per annum over the next 10-15 years. If this happens, the share of manufacturingin GDP would double every eight years or so. Even so, manufacturingwould accountfor less than 20 percent of GDP two decades from now. The main factors that are likely to constrain the growth of the manufacturing sector are inadequate infrastructure,especially utilities, and the shortage of industrialskills. Governmenthas a major role to play in removing these constraints. In the case of commercialinfrastructure (telecommunications, water, electricity)there is an emergingtrend in several countriesof government-ownedprivate sector managed operations.' Uganda should seriously consider this new approach. The Government will also have to embark upon a major program of production-orientedskills training, with the active participationof industry.

Savings, Investment, and the Balance of Payments

5.45 Savngs and Investment. Projectionsof GDP by expenditure,in constantprices, are presented in Table 5.12. Rapid economic growth requires a rapid increase in savings and investment. Consequently,in order to achieve a GDP growth rate of 8 percent, gross domestic investmentas a share of GDP is projected to more than double, from 11 percent to 23 percent, between now and the year 2005. The share of public investment,which is presently largely donor-financed,is expectedto increasefrom its current level of 5 percent of GDP to about 8 percent by the end of the projection period, as domesticrevenues pick up. However, the private sector is projected to account for most of the increase in the share of investmentin GDP. That means gross domestic savings must rise correspondingly,if Uganda is to reduce its reliance on foreign savings. Domestic savings are now very low; under 6 percent of GDP. Given the exceptionally low domestic revenue effort, the Governmentremains a dissaver. The prospectof the Governmentbecoming a major saver duringthe next 10-15 years is not bright. The private sector would therefore have to carry the burden of generatinglarge increases in domesticsavings. While Ugandawill remain a poor country for a very long time to come, it is capable of saving a much larger proportionof GDP. This is confirmedby the experience of poor countries like China, India, and Kenya with saving rates of 20 percent or more. This places a premium on the developmentof the financial sector because the key to higher rates of saving is an efficient financial system capable of mobilizingsmall savings from a large number of the population. In Uganda a rising saving rate means a steady decline in the share of private consumptionin GDP, from around 90 percent at present to about 74 percent by the year 2005.

5.46 ThseBalance of Payments. A viable balance of paymentsis another essentialelement in any strategy to achieve rapid growth. Table 5.13 shows the balance of payments scenario underlying the GDP growth targets presented in Table 5.11. At present Uganda's external position is very fragile, mainlyon accountof its weak export base. Despitesevere importcompression, it had a large negative current account balance in 1991/92 (over US$340 million) which was financed largely by official

76 TheMBotswanw Telcommnuiation Corpoation ha opated undaerach an arrwuant for manyyac. 74 Chapter S

Table S.9: Key M o Indicators

1987 1991 1993 1995 2000 2003

Growth rates: GDP at factor cost 6.7 4.1 6.0 6.2 7.3 8.0 Gross domesticincome 3.1 4.2 5.8 6.4 7.1 7.5 GDY per capita 0.6 1.7 3.3 3.9 4.6 5.0 Private consumptionper capita 2.9 -1.6 1.8 0.1 2.9 3.2

Debt Service: Debt service, USS million 195 215 18S 197 189 195 o/w Interest, USS million 52 74 83 93 75 91 Debt servicelXGS' 41.0 71.4 49.5 42.0 22.0 11.7 Debt service/GDP 3.9 7.2 5.2 4.5 2.6 1.5

Ratios to GDP, constantprices: Domestic savings 3.7 5.8 7.1 10.8 15.5 21.3 National savings 4.9 5.8 7.1 10.8 15.7 21.5 Public savings -0.7 -2.9 0.2 3.0 5.5 7.0 Private savings 5.6 8.6 6.9 7.8 10.2 14.5 Gross Domestic Investment2 9.2 10.7 13.0 16.2 18.9 23.4

Ratios to GDP, current prices: Government revenues 6.4 7.2 8.4 12.5 15.0 16.5 Government expenditures 12.4 19.0 17.8 19.5 18.6 18.6 Overall deficit 3 -6.1 -11.8 -9.4 -7.0 -3.6 -2.0

Miscellaneous: Inflation 199.8 28.1 23.5 9.0 5.0 5.0 Export growth rate 21.3 -8.3 10.9 11.4 13.2 15.3 Exports/GDP 4 7.3 5.0 5.2 5.7 7.0 9.8 Import growth rte 11.6 -13.2 15.6 5.0 5.0 7.3 Imports/GDP 4 12.9 8.2 8.6 8.3 7.1 6.8 Terms of Trade Adjustment 0.0 -2.2 -2.5 -2.7 -3.2 -5.2 Current account, USS million ' -218.0 -393.5 -507.5 -568.0 -573.5 -601.0 Current account/GDP 3 4 -4.4 -3.2 -3.4 -2.6 0.0 3.2

'Exports of goods and services, includingprivate transfers. 2 Sinscethe data are presented in constantprices, the trmnsof trade a4justmentmust betaken into accoutfor the natonal income identties to hold. Excludes grants. 'Expressedin constantprices to removethe inpact of exchangerate adjustments.While the balance of payments current account remainsnegative over the projectionperiod in currentprices, it is projected to turn positive, in constant 1987 prices after the year 2000.

Source: Staff estinates based on data provided by Governmentauthorities. assistancefrom the multilaterallending institutionsand bilateral donors. Uganda cannot rely entirely upon these sources indefinitely.At any rate, competingdemands for such resources have emerged from the former SovietUnion and Eastern Europe. Improvingthe balanceof paymentsmust therefore focus on two aspects namely, export expansionand foreign investmentpromotion. Chapter5 75

5.47 Exports are projected to increase their share in GDP from their present levels of about 5 percent of GDP to slightly under 10 percent by the year 2005, implyingan annual average rate of growth in excessof 10 percentfrom a severely depressedbase. Whilethe internationalprice of coffee is projected to recover from the present depressedlevels, the growth in export earnings is expected to materializefrom the re-emergenceof traditionalexport crops like cottonand tea, and nontraditional exports such as simsim, beans, vanilla, and cashewnuts. The level of imports has declined significantlyover the past few years, partly as a result of the tight foreign exchangesituation until 1991. As economicgrowth picks up, the importcompression is expectedto reverse itself. However, import growth will be moderatedby the dominant share of agriculture in GDP, a sector which is neithercapital-intensive nor foreign exchange-intensive.Consequently, the projectedshare of imports in GDP, while increasingin the initial years, declinesmarginally to about 7 percent in 2005.

5.48 Foreign investmentwill also play a major role in the transformationof Uganda. The Uganda InvestmentAuthority is reporting growing foreign investor interest in the country. However, the actual amount of foreign investmentcoming into the country remains a trickle. That has begun to change in the aftermath of the return of large numbers of expropriated properties to the owners. Althoughthe investmentclimate has improved a great deal, it will take some time to rebuild investor confidence.Significant amountsof foreign investmentare therefore not expected during the rest of this decade. It is projected that by the turn of the century, Uganda could see annual investment inflows of about US$30-40million, growing thereafter to over US$100 million by the year 2005. However, for this to happen, Ugandawill need to ensure that the foreign investor is not looked upon in suspicion,as has been the case in the past, is made to feel welcomeand is treated as a partner in development. UGANDA TABLE 5.10

GDP BY SECTOR IN CONSTANT PRICES Percentage of GDP at FC Prelim <------Projected ------> 1987 1988 1989 1990 1991 1992 1993 1994 1995 1998 2000 2003 2005

Agriculture 62.2 61.4 60.9 60.2 59.3 . Cash crops 2.6 2.3 2.3 2.2 2.6 25 25 24 23 21 2~ 17 15 Food crops 45.6 45.2 45.3 44.6 43.3 01w Horticulture/spices 8.3 7.8 7.6 7.8 7.8 G . . . 07 1. 26 1. Livestock 9.8 9.8 9.5 9.5 9.4 ~ ~ . i . . . Forestry 1.7 1.6 1.5 1 5 1.5 Fishing 2.5 2.4 2.2 2.4 2.4 ~4:::::::::::2423 2 . . 5 1

Industry 7.8 8.4 8.8 9.1 9.6 U 1$ US 17 1$2 11 Mining&quarrying 0.1 0.1 0.1 0.2 0.3 03 0 . 3 03 0 . . Manufacturing 3.6 4.1 4.6 4.7 5.2 Coffee,cotton, sugar 0.2 0.3 0.4 0.5 0.7e7timate Food products 0.3 0.4 0.4 0.4 0.5 05 0 . 5 0.A Miscellaneous 3.0 3.4 3.8 3.8 4.1 42 4 . 5 5 . , . Publicutilities 0.4 0.4 0.4 0.4 0.4 . , ...... 6 0. 05 Construction 3.7 3.7 3.7 3.7 3.7 7 3 3- 3 .. $. Services 30.0 30.2 30.3 30.7 31.1 ~I 38 3 1 ~2 45 3. Trade 13.5 14.1 14.5 14.7 14.9 49 4. 5 1. 16 74 184 13 Transport & communication 3.9 3.8 3.7 3.7 3.7 3, 37 3. 37 S 37 37 30 Communityservices 10.2 10.1 9.9 10.2 10.4 ~A4 14 r. 0 ( 26 1 owner-occupied dwellings 2.4 2.2 2.2 2.1 2.1 ~ 20 19 1 . . . GDP at factor cost 100.0 100.0 100.0 100.0 100.0.181W #718110 1 1 1 Netindirect taxes 5.2 4.5 5.1 6.4 6.3 6 . 4 7 0 1 8) 1

GDP atmarket prices 105.2 104.5 105.1 106.4 106.3 1

Source: StatisticsDepartment and staff estimtes. UGANDA TABLE 5.11

GDP BY SECTOR IN CONSTANT PRICES Annual growth rates Prelim <------Projected ------> 1987 1988 1989 1990 1991 1992 1993 1994 1995 1998 2000 2003 2005

Agriculture 5.2 6.2 6.4 2.9 2.5 .7 4 Cash crops 4.0 -4.4 7.3 -4.2 25.0 24 3 24 4 22 2 O 2 Food crops 5.5 6.7 7.6 2.5 1.0 50 7 54 I 4* 4 3 1 o/w Horticulture/spices 0.0 0.0 5.0 7.5 3.0 0 0 1 1 1 ( 1 2 14$ Livestock 4.3 7.7 3.5 4.0 4.0 4. 4. $0 5 50 4. 30 Forestry 6.7 6.2 1.1 3.9 3.8 5 5 3 3 2. 25 2. 25 Fishing 4.1 2.5 -0.4 14.8 4.0 4

Industry 23.0 15.0 12.7 7.5 10.3 6 7 ~ ~ 1~ 2 44 1 Mining & quarrying -17.3 -5.3 15.4 88.5 21.4 11 9 0 55 4 0 4 Manufacturing 16.8 22.7 19.1 7.5 14.1 1 2 9 ~ 5.5 Coffee, cotton, sugar 7.7 36.0 44.7 21.2 447 Food products 34.5 27.8 1.9 10.7 14.4 7 2 3 Miscellaneous 15.9 21.0 18.8 5.6 10.2 7 0 0 4 2 5 6 Publicutilities 6.2 2.9 6.7 7.9 81. . . Construction 35.1 9.7 6.0 4.4 5.0 44 4 62 7 9 9 13 13

Services 6.2 8.5 7.6 5.3 5.4 5 1 7 2 9 34 9 Trade 9.2 12.1 10.2 5.4 5.5 1 . 0 8 1fi 1 I ilkO Transport & comnunication 6.1 6.6 5.0 3.7 4.2 4 4 7 7 7 7 Conmmunityservices 3.2 5.8 6.1 6.3 6.3 $ 6 7 0 10 10 10 1. Owner-occupied dwellings 2.7 2.7 2.7 2.7 2.7 3 0 0 2 0 0

GDP at factor cost 6.7 7.6 7.3 4.0 4.1 0.6.0.62.62......

Source: StatisticsDepartment and staff estinates. UGANDA TABLE 5.12 4

GDP BY EXPENDrrURE AT CONSTANT PRICES Percentage share of GDP at MP Prelim <------Projected ------> 1987 1988 1989 1990 1991 1992 1993 1994 1995 1998 2000 2003 2005

Consumption 96.3 96.7 95.3 93.0 92.0 9(4 #4 6A # L3 7 $. Private 89.8 91.2 89.4 87.1 84.5 83 2 141 16 14 18 67 6 Public 6.5 5.5 5.9 5.9 7.5 I j2 2 $ $ $ Gross domesticinvestment 9.2 8.9 9.7 10.4 10.7 ~ ~ ~ 6 ~ 4 R 1 3 Fixed investment 9.2 8.9 9.7 10.4 10.7 4 1.G 4. 1$ f L9 2.7 24 Private 3.3 3.6 3.8 3.9 5.2 S 5 9~14 18 1. 36 1~ Public 5.9 5.3 5.9 6.5 5.5 $1$ V 4. 1 .* 14 Exports, G+NFS 7.3 6.6 6.2 5.6 5.0 5~ $ $ 5 $ ~ Imports,G-4NFS 12.9 12.2 11.3 9.8 8.2 1 6 8 3 1 1 * Terms of trade adjustment 0.0 -1. 1 -1.7 -2.4 -2.2 23 ~ 4 .~*. Import capacity 7.3 5.5 4.5 3.3 2.7 2. 2. 29 4 ... .3 4 Resourcebalance m-7-5.5 -5.6 -5.2 -4.2 -3.2 24

Net factor service mncome -1.0 -1.1 -1.2 -1.1 -1.4. .:7:.:.j:::,$: . 9 4 Net current transfers 2.2 2.0 1.6 1.2 1.4 1 ~ 1 % 1 1 4 . Current accountbalance -4.4 -4.7 -4.8 -4.1 -3.2 30 4 39 6 11 019 2 GDP at market prices 100.0 100.0 100.0 100.0 100.0 19 0 G0£9 00 19 00 19 GNP 99.0 98.9 98.8 98.9 98.6 92 983 94 15 99 9A 94 5

Gross domestic savings 3.7 2.2 3.0 4.7 5.8 56 7 90 Gross national savings 4.9 3.2 3.4 4.8 5.8 5 . 3~1. 91~,. Gross domestic income 100.0 98.9 98.3 97.6 97.8977 95 94 93 Od 9t 97 9. Gross national income 99.0 97.8 97.2 96.6 96.3 959 59 9.1 5. 960 5. 9.1 9. Source:Bank of Uganda,IMF and staff estimates. UGANDA TABLE 5.13

BALANCE OF PAYMENTS In millionsof US dollars Prelim < Projected> 1986/87 1987/88 1988/89 1989190 1990/91 1991/92 1992/93 1993/94 1994/95 1997/98 1999/00 2002103 2004/05 Exports(g+nfs) ~~~~~406324 304 246 200 :5 2 3 4 13 96 35 ExoMerchgandise(fob) 38.4 298 282 210 177 40 35 . olw Coffee 365 286 276 159 126 1~1923 24 34 Non-factorservices 22 26 22 36 2325 2 7 45 t14

Imports(g+~ S 600 682 712 676 658- Nonfactoriseices 864 137 1502 592 108 olW Petrol 63 69 76 78 87 ~ Resourcebalance -194 -358 -408 -430 -458 32 9 5# 57 60 60 .4R 45 Net factorincome -47 -57 -67 -77 -66 1~ 3 19 .5 14 14 .S 4 o/w Netinterest -47 -57 -66 -77 -62 7 7 -1 4s Currentprivate transfers 100 120 114 78 80 17 13 1 S 7 9 0 1 CIA Balance(exci graLnts) -141 -295 -361 -429 -4 4 -~ 2 5 $ 49 -5 ~0 CIA Balance (minegrants) -101 -203 -230 -276 -239 2 3 $ rni$ #4 41 Officialtransfers 40 92 131 153 205 27 ~ * * 4 ~ ~ ~ o/w IMprt support 0 33 49 29 87 17 U ZS k9 UlO92 5 NetM&I.Tf o0ans 45 101 125 215 131$320 4 24 Disbursements 135 186 211 292 217..4 9 0 Projectloans 135 141 143 12 118 7 .4 14 15 fl 23 ZI 34 Import support loans 0 45 68 167 99 9 31 141 5 Repayments 90 85 86 77 86 712 13 $ Foreign investment 28 10 13 6 15 5 3$1 16 Short-term, net -31 37 -17 12 2 Errors/valueadjustment -20 -10 7 0 0 aSSSS Overall balance -79 -65 -103 -44 -105 I - 1 9? 43 7351 Finnin:79 65 103 44 105 I1.1 4*35 5 1 Monetaryauthorities ~~22 -14 18 10 41 ... 7 . $ . 9 . rosarservehanges 33 -4 12 11 -11 -3 5 1$5 179 500 IMF,net -3 -17 7 -1 52 2 02 5 *4 SAF/ESAF and purchases 57 34 94 42 89 5 5 2 Other,nct -8 7 0 0 0 000 0000 Shortterm/commercial 10 -10 6 12 -2 -41 00 Externalarrears -45 47 18 -19 65 14 000000 Exceptional financing 92 41 61 41 1 .) 13 9 5 5 3 0 . Reschedul'mg ~92 40 57 40 0 0 5) 1 Debt caneffition0 0 3 2 1 0 0 00000 Residualfinance gap 0 0 0 0 0 . . 317 3 . 3) 051 Memo item: Grossreserves (EOP) 37 41 30 19 30 37 12 9 14 13 21 31 in monthsof imports 0.9 0.9 0.6 0.4 0.6 5 14 .5 6 20 22. 2. Source: Bank of Uganda, IMF, and staff estimates.

Chapter5 81

Annex to Chapter 5: Agricoltural Growth Prospcts

A5. I This Annex examinesbriefly the status and prospects of each of the major crops as well as livestock. The most striking thing to emerge from this analysis is that in crop after crop Uganda has a vast scope for raising yields. This requires painstakingand sustained effort over the long term. In the long term, area expansionis not a substitutefor such an effort.

Traditional Export Crops

A5.2 The traditional export crops are coffee, cotton, tea, tobacco and cocoa. Of these only coffee has survived as a major export crop. The production of the rest declined to the point where Uganda was no longer a serious supplier in foreign markets. For cotton and tea things have begun to change in that production has started to pick up and Uganda is slowlyregaining some of its lost markets.

A5.3 Coffeeis the number one export crop. It occupies about 240,000 hectares of land. The entire crop is produced by smallboldersand it is estimated that 1.2 million farm householdsare engaged in coffee production. In two thirds of the country's 38 districts, coffee is a major source of cash income. At the beginningof the ERP, coffee accountedfor nearly 95 percent of total export earnings. That share has now fallen to about 75 percentpartly as a result of the difficultiesfacing the industry (principallythe collapseof internationalcoffee prices since July 1989 and domesticsupply problems) and partly in reflectionof the rising trend of nontraditionalexports. Coffee was also, until recently, the mainstay of governmentrevenue, contributing, in the form of an export tax, to over a third of total domesticrevenue. Again becauseof the problems plaguingthe industrythe export tax on coffee was eliminated in 1992/93.

A5.4 Uganda's coffeeis predominantlyof the Robustavariety, accountingfor 90-95 percentof total production. Robusta is in oversupplyon world markets. However, that does not mean that Uganda should not increase production so as to regain its position in the world market. The balance of production is Arabica coffee whose internationalmarket prospects are more favorable. It is grown in the high altitude (1,500-2,000 meter) areas, mainly in Mbale. While there is some scope for increasingthe area under Arabica cultivation(presently 30,000 ha), suggestionsof doublingand even tripling the area have so far not yieldedmuch results.

A5.5 The present coffee yields are very low; about 0.6 tons/ha and 0.5 tons/ha of clean coffee for Robustaand Arabica, respectively.Yet six clones of Robustacoffee capableof giving yields in the range of 2-3.5 tons/ha of clean coffee, with some applicationof fertilizer, have been availablefrom the research program since the mid-1970s. Only small quantities of these clones have been forthcomingfrom the 28 existingnurseries. Pardy as a result, the yields obtainedfrom the research trials have yet to be fully validatedat the farm level. A major replantingprogram is needed since the coffee trees are not only very old (30-60 years) but have been neglected.Aside from introducingnew high-yielding varieties, Uganda must tackle the problems posed by disease and pest, low soil productivityand poor managementpractices. 82 Chapter5

A5.6 Cotton was once a major export earner for Uganda, second to coffee. The industry was at its best around 1970 when the area planted reached 900,000 ha and production peaked at 469,730 bales (86,900 tons) of lint.' From this strong position the industry went into a long period of decline. By 1988 the farmers had practicallyceased growing cotton: only 1.8 tons were producedin that year. A steady recovery has been evident since then and, in 1992, some 60,000 tons of cotton were produced.

A5.7 Although cotton is cultivatedin many parts of Uganda, the main growing areas are Teso, Busoga, Lango and Bukedi districts as well as parts of the western and northern regions. It is a smallholdercrop grown at altitudesof less than 1,500 meters, requiring annual rainfall of over 800 millimeters.The medium-longstaple cottonproduced in Uganda is of high quality but the yields are very low. Average yields are estimated at 0.25 tons/ha of lint cotton. An important reason for the low yields is the lack of improvedvarieties. The present standard varieties (SATU and BPA) are no longer pure. They have become susceptibleto a host of pests and diseases. Institutionalweaknesses in the industry, as exemplifiedby the lack of competition among ginneries and exporters, also constitute a major constraint on production. In the absence of these constraints and with good management,it is possible to achieve yields of 0.75-1 tons/ha; yields of 1 ton/ha and above have already been achieved in some field trials.

A5.8 The other major traditionalexport crop is tea. Productionpeaked in 1972 at 23,000 tons, of which 21,000 tons was exported. Production then dropped off dramatically during the years of political and military upheaval; in 1979 only 1,500 tons were exported. Being a tree crop, the tea bushes more or less survived and thus are only in need of rehabilitation.Approximately 22,000 ha of land are presently under tea, of which about half has already been rehabilitated. The gestation period for rehabilitation investments in tea is very short-about three months. This means that productionfrom existingtea gardens can be increasedvery quickly;output of tea is already nearing 10,000 tons.

A5.9 Tea requires good soils, altitudes of over 1,200 meters and rainfall of more than 1,200 millimeters per annum. The best growing conditions can be found in the districts of Bushenyi, Rukungiri,Kabarole, Hoima and Masindi as well as the northern shores of Lake Victoria, especially near Lugazi. Tea is produced on smallholder farms (accountingfor 9,400 ha) and small-to-large estates which accountfor the rest of the area planted. In the past Ugandaachieved high yields in tea; average yields in 1972 were 12 tons/ha and 5 tons/ha for the estates and smallholders,respectively. Yields dropped precipitously over the years and are now in the range of 1.4-1.6 tons/ha. With improvedmanagement of the plantationsand the introductionof improved clones over the long term, Uganda should be able to regain the previous yield levels and even surpass them. Full rehabilitation of the remainingtea plantationswould enable Ugandato attain the previous peak production.

A5. 10 The other traditional export crops are cocoa and tobacco. Cocoa has always been a minor crop in Uganda. Presently only 300 tons are exported in a year. Given the oversupply from the traditionalproducers, cocoa is not destined to become a major export crop in Uganda. Tobacco,on the other hand, is a major import-substitutioncrop which is grown by smallholders.The crop has been revived, thanks to BAT, the cigarette manufacturer. Aside from procurement, BAT provides a full range of research and extensionservices to the farmer. Production is now around 5000 tons per annum. About half of this production goes to meet domestic requirements and the rest is

79 A beleweighs 185 kg. Chapter 5 83

exported. Exports are, however, not expectedto increase much beyond this level primarily because of the worldwidehealth concerns about smoking.

Matoke (Banana)

AS.11 Matoke is the major staple food crop in the high rainfall areas of the country. It occupies the single largest cultivatedarea, estimated at 1.5 million hectares in 1992. The yield, presently about 5.5 tons/ha is low, because of decliningsoil fertility and problems causedby disease and pests. Given the favorableclimatic and soil conditions,the yield could be increasedto 10-20 tons/ha with the use of disease-freeimproved clones, better managementpractices and fertilizer application.There is no prospect of matoke exports on any significantscale, even to markets within the subregion. Domestic (human)consumption therefore sets the upper limit to production.' Accordingly,growth of matoke output can be expectedto equal the populationgrowth rate. The first priority is to stabilize yields at current levels. This in turn would suggest that research and extensionactivities should be focusedon making present varieties pest resistant and capable of giving sustained yields under variable conditions. The expansion of the cultivated area would thus continueto account for much of the growth of matoke output. The next stage is to initiate programs aimed at raising yields. This is necessary in order to free some of the land occupiedby matokefor other crops.

Cereals

A5. 12 The main cereals produced in Uganda are millet, maize, sorghum and rice." Finger millet is extensively cultivated in the drier and less fertile areas in the east and north. Bulrush millet is grown mainly in the semi-aridareas of Karamoja. Millet is an importantpart of the diet in the areas where they are grown, have a high nutritivevalue, have high tolerancefor drought, and they can be easily stored. They also form the basis of an important local beer brewing and food processing industry.The productionof finger milletdeclined sharply between 1970and 1982, from 783,000 tons to 401,000 tons, althoughproduction has since picked up to just under 600,000 tons in 1992. Yields have always been very low. On average farmers obtain 0.8 tons/ha while research stations have achieved yields of 3 tons/ha with improved varieties and better managementpractices. The main constraintsto increasedoutput are low-yieldingvarieties, poor husbandry practices, degraded soils, low adoption of improved varieties, competitionwith other crops and disease and pests, including birds. Given its place as a staple food in mainlythe low-rainfallareas of the country, milletdeserves high priority in Uganda's agriculturaldevelopment strategy.

A5.13 In terms of overall output of cereals, maize ranks second to millet. Although historically maize has not been part of the Uganda diet, the cultivationof this crop has expanded rapidly over the past ten years. The area under maize increasedfrom 260,000 ha in 1981 to about 440,000 ha in 1992. The cultivatedarea and output actually peaked at 527,000 ha and 674,000 tons, respectively in 1976. Except for a sudden burst of output expansionin 1989 to 624,000 tons, maize production declined or stagnatedduring much of the 1980s. Maize yields have remained very low, fluctuating between 1.2 and 1.5 tons/ha. Although climatic conditionsin Uganda are very favorable for high maize yields, productionis constrainedby a numberof factors. One of the most importantconstraints is the lack of suitable improvedvarieties. The present cultivatedvariety, a Kawandacomposite, was

so This doea not, of course, ruleout the emergenceof industrWialusea for matoke.

91 Wheatis also grownin Ugandabut on a very small scal (around 10 MT perannum). 84 Chapter 5 released in 1971. It is a tall late-maturingvariety which is susceptibleto maize streak virus. Other constraintsinclude inefficient farming practices, nonapplicationof fertilizer, poor harvesting, drying and storage. The productionpotential for maize is very good as shownby the fact that new selections from the ongoing maize research program at Kawandahave given yields of 3-4 tons/ha, based on trials in farmers' fields. Moreover, maize varieties with a higher yield potential(5-7 tons/ha) can be obtained from some of the InternationalAgricultural Research Centers (IARC).

A5.14 Another important cereal is sorghum. It is a staple food as well a raw material for beer brewing. It is grown mainly in the drier areas of Karamoja,Teso and the highlandsof Kabale. The area plantedunder sorghum has remainedmore or less stagnantfor a long time while productionhas been on the decline. In 1975, 467,000 tons were produced; by 1992 productionwas down to 375,000 tons. Present yields are very low (0.8 tons/ha) for the familiarreasons of low-yieldingvarieties, poor production practices, poor soils and pests and disease. Improvedvarieties which mature faster and are resistant to drought are availablefrom the IARCs. Under improved managementpractices these varieties can give yields as high as 3 tons/ha.

A5.15 The other cereal worthy of mention is rice, which is gaining in importancein the urban diet. Domestic demand used to be met largely from imports but the situation is now quite different. Productionhas expandedto about 60,000 tons per annumto meet domesticrequirements. Aside from increasingdomestic demand, there is export potential in regional markets. The climate and soils in many parts of Uganda (particularlythe swamps)are very favorablefor rice cultivation.The present low yields of 1-1.5 tons/ha can be increasedto 3-4 tons/ha with the aid of high-yieldingvarieties and improved productionpractices.

Roots and Tubers

AS.16 The main crops in this categoryare cassavaand sweetpotatoes. In terms of output the former ranks second and the latter third among the food crops grown in Uganda. Cassava is a staple food in the eastern, northern and northeasternparts of the country while sweet potatoes are grown mainly in the south and west. Betweenthem cassava and sweet potatoestake up nearly one million hectares of cultivatedland; this is another measure of the importanceof these crops in Ugandanagriculture. There has been a steady improvementin cassava yields, from under 5 tons/ha in 1970 to around 9 tons/ha in recent years but Ugandan yields still fall far below those prevailing in many other countries. Ugandashould at least stabilizeyields at current levels and possiblyseek to achievehigher yields by tacklingthe main productionconstraints which are: nonavailabilityof high-yielding,disease resistant varieties, the prevalence of pests and poor agronomicpractices. The diseases that pose the greatest threat to cassava are the African cassava mosaic virus and cassava bacterial blight.

A5.17 There is also big potential for the cultivationof sweet potatoes. Compared to those of other developing countries, the present sweet potato yields are exceptionallylow. Yields are generally around 5 tons/ha when it has been demonstratedin the east and north that yields of 35 tons/ha can be achievedby plantingcurrently availablehigh-yielding, disease resistant varieties and introducing improved cultivationmethods.

A5.18 A recent additionto the list of roots and tubersgrown in Ugandaare Irish potatoes. The crop now occupies 40,000 ha of land which produce about 300,000 tons a year. Kigezi and Ankole account for the bulk of the production. While soil and climatic conditionsare favorable for potato growing, present yields are very low: only 8.5 tons/ha. Other countries achieve yields which are several times higher than this. Moreover, using new selections,the Kalegyere Research Stationhas Chapter 5 85

demonstratedthat yields of 25-30 tons/ha (four times current levels) are possible. The production constraintsthat need to be addressedare lack of improved varieties and plant diseases.

Pulses

A5.19 Of the pulses, the only one grown on a large scale in Uganda are beans.' The area planted under beans now stands at over 500,000 ha and the annual harvest is around 400,000 tons. Productionis concentratedin the westernparts of the country but significantquantities are also grown in Buganda,the east and the north. Beansprovide a cheap source of vegetableprotein in the Ugandan diet. Beans also have export potential. In spite of its attractions, bean production has either been stagnant or has declined during the past two decades. The main constrainton production is the lack of improved varieties together with the associatedimproved farming practices. The present variety grown in Uganda (K20) was released as far back as 1968 and now gives very low yields of 0.6-0.8 tons/ha. Uganda is consideredcapable of achievinga yield of 1.5 tons/ha. Still higher yields may be possible since research stations have been able to achieve yields of over 2 tons/ha. With improved varieties two and possiblythree crops can be grown each year. To realize this potentialUganda must tackle the other constraintson production;that is, eradicateviral, bacterialand fungal diseasesas well as the bean fly pest.

Oilseeds

A5.20 The most importantoilseeds grown in Ugandaare groundnut(peanut) and simsim. Groundnut is both a food crop and a cash crop and is a source of protein and vegetableoil. It is grown all over the countrybut more so in the north than elsewhere. In the 1960ssignificant quantities of groundnut were exported. However, there was a sharp decline in area planted between 1979 and 1985, with a correspondingdrop in annualoutput to as low as 70,000 tons. Since then steady increasesin area and output have been recorded. Productionhas almost returned to the 200,000 tons level that had been achieved in the mid-1970s. However, output remains constrained by the lack of high-yielding varieties, the ravages of disease and pests, poor husbandrypractices, nonapplicationof fertilizer and unreliable rainfall in some areas in the north. The present yield is estimated at 0.8 tons/ha of dried pods. This could be raised to 2.5 tons/ha, a result that has already been achieved on experimental plots in some parts of the country.

A5.21 Simsim (sesame), the other important oilseed, is grown mainly in northern and eastern Uganda. Production was adversely affected by insecurity in these areas but it has been boosted in recent years by the return of peace and security and growing export demand. The area under cultivationis approaching150,000 ha and annual output is in the 60,000-70,000tons range. At about 0.4 tons/ha the yield is very low. Aside from the absence of high-yieldingvarieties, production is plagued by the shattering of mature pods, leaf disease, insect pests and poor agronomicpractices. Given improved varieties and farming practices, yields can be raised significantly.Research trials have already shownthat yields can be doubled.

A5.22 Among the oilseeds, soya beans and sunflower are also grown in Uganda. During the past three years the area plantedunder soya bean has expandedrapidly to about 60,000 ha at present, with

The otherpulse. are field peas, cow pea and pigeon peas which are produced in comparatively small quantities. Thecombined area occupied by the these threc crops is less than 150,000ha. 86 Chapter5

a corresponding rise in output to a little over 50,000 tons. With regard to sunflower, only insignificantquantities are being produced at present.

Industrial Crops

A5.23 Aside from cotton, the only industrialcrop of significanceis sugarcane.Favorable climatic conditionsfor cultivationexist on the northern shores of Lake Victoria and there are three sugar factories (Kakira, Lugazi and Kinyala)which process the raw cane. Productionof processed sugar peaked at a little over 154,000tons in 1970. However, with the expulsionof the Asians in 1972 and the general chaos prevailing in the country, the sugar industry suffered such a sharp decline that in 1985 only 1,000 tons of sugar were produced. The cane is produced on large estates owned by the sugar companiesas well as on small outgrower plots.'

AS.24 The cane yields and sugar yields are low. In 1991Kakira and Lugazi achieved cane yields of 70 and 83 tons/ha, respectively. The respective sugar yields were 8.6 percent and 6.9 percent. These cane yields could be raised substantiallywith the help of improved varieties, better crop and soil managementpractices and adequatepest and diseasecontrol measures. A major rehabilitationof the sugar plantations and the factories at Kakira and Lugazi has been underway since 1986. As a result the annual output of processed sugar has climbed up to a little over 30,000 tons, enough to meet about one fifth of the domestic demand which is estimated to be in the range of 150,000- 200,000 tons. Domestic consumptionis expected to increase to 250,000 tons by the mid-1990s, implyingthereby that even when production capacity is fully restored (to around 170,000tons per annum) it would still not be adequateto meet domesticrequirements.

Horticultural Crops

A5.25 A large variety of horticulturalcrops are grown in Uganda.These includefruits, vegetables, spices, flowers and ornamentals.There exists a wide range of soil and climatic conditionswhich are ideal for the cultivationof these crops. The main vegetable-growingareas are Kabale, Kasese, Mbale, Masaka,Jinja, Mpigiand Kampala.Fruits are grown in smalllots in most parts of the country. There are no reliable figures on horticultural crops, on either area planted, total annual production or marketedproduction. However, it can be said that horticulturalcrops make only a minor contribution to agriculturalGDP. Encouragingly,small but increasingquantities of pineapple,green beans, chilies, passion fruit, ginger and citrus are now being exported to Europe and the Middle East.

A5.26 The potential for increasedproduction of horticulturalcrops, especiallyfor export, does exist but there are formidable obstacles. To start with, being a landlocked country, Uganda faces high transport costs. Then there are problems of poor produce quality, poor post-harvesthandling, weak distribution and marketing infrastructure and inadequate or nonexistentrefrigeration facilities. In additionto fresh produce, Ugandacould at a later stage enter the marketfor processedproducts such as juices, concentratesand canned products.

It i. etinume that there an 130,000outgrowr enachwith 2-3 hectda of muga:ane. Chapter5 87

Iivestock and Dairy

A5.27 Poultry, sheep and goats are sources of cash income for large numbersof farmers. They are also a source of protein in the Ugandandiet. However, in numbersand value the livestocksubsector is dominatedby cattle. The zebu, an indigenousanimal, accounts for 70 percent of the cattle, the sanga (Ankole), another indigenousanimal, 15 percent, the cross-bred animals, 13 percent and the exotic breeds, 2 percent. The indigenousanimals are adept at survivingin the local environmentbut they are also susceptibleto epidemicsof rinderpest, contagiousbovine pleuropneumonia(CBPP) and trypanosomiasis. The exotic and cross-breeds are prone to catch tick-borne diseases and must therefore be kept in fenced farms or some other enclosure. Livestock depend almost entirely on natural grass-dominatedpastures. The productivityof indigenousanimals is low. Cows mature in 3-4 years, calve every second year and produce about 350 liters per lactation. The mortality rates are high; for example, for calves it is 25-30 percent annually.

A5.28 Annual domesticmeat consumptionis estimatedat 55,000 tons which is equivalentto about 3 kg per capita. Domestic demand is expectedto increase in line with the growth of populationand per capita income. Basedon the 1989-90Household Budget Survey, the incomeelasticity of demand for livestockproducts is put at 1.5. This would implya strong growth of domesticdemand for meat products in the future. Regional markets also offer opportunitiesfor increasedlivestock production and offtake. Moreover, Ugandashould in principlehave access to overseasmarkets too, particularly the lucrative EEC market governed by quotas. Access to foreign markets outside of the region will require the attainment of rigorous quality and health standards in the livestock industry, from production and processingto marketing.

A5.29 Milk is the only dairy item of any importancethat Ugandaproduces. About 350-400 million liters of milk are produced each year. The bulk of this production is consumedon-farm and in the rural areas. About 10 million liters are collected, processed and marketed by the Uganda Dairy Corporation(UDC). UDC's output falls far short of the demand for milk in the urban centers. This means that additional capacity for collectionand processingis needed. This is turn should stimulate milk production at the farm level.

Labor Market Policies for Povery Reduction

7Thecountries at have succeeded in reducing poverty over the long term have encouraged broadly basedrural developmentand urbanemployment, thereby increasing the returnsto smail farm producton and wage labor. ' (World DevelopmentReport, 1990, World Bank,p. 56.)

6.1 The purpose of this chapter is to identifypolicies and programs for the improvedfunctioning of factor markets from the standpointof facilitatingeconomic growth and reducing the incidenceof poverty in Uganda, particularlywith respect to the functioningof the labor market, although access to land and capital are also discussed. The first section puts together a picture of the actual pattern of the distributionof employmentand labor use in Ugandatoday. The second section analyzessome of the distinctive features of Uganda's labor market. The final section outlines policies that would help improve the productiveuse of labor resources in Uganda within the framework of a pattern of distributionof benefits of increasedproductivity that wouldhelp reduce the incidenceof poverty in Uganda.

The Distribution of Labor Force

6.2 Uganda does not have a recent labor force survey. However, the HBS provides some basic informationabout the distributionof labor force by gender, location and sector of occupationwhich is summarizedin Table 6.1. Combiningthis informationwith the estimatesof populationby gender and location, Table 6.2 provides the bare outlinesof the allocationof Uganda's labor force, along with its gender distribution, into three broad sectors-agriculture, industries and services-in each of the urban and rural economies.'4 Table 6.3 attemptsto further disaggregatethe urban labor force into employmentin the public sector, the formal private sector and the informal sector by pooling the information from the Census of Civil Service (1987), the National Manpower Survey (implementedin 1988 and published in 1989) and government payroll statistics.' Together these three tables provide a picture of employmentin Uganda, whose broad features are discussed below.

6.3 Sectoral Compostion of Employment. Most Ugandans work in the rural economy and are predominantlyemployed in agriculture. Eighty nine percent of the members of the Ugandan labor force are employedin the rural economy, a slightlyhigher share than that of the rural sector in total population, reflecting a higher activity rate in the rural sector than in the urban sector.' Nearly 80 percent of all workers and more than 87 percent of the rural workers are employed in agriculture. Even in the urban economy, agriculture employs 18 percent of all workers, a third more than are employed in all industrial sectors, broadly defined. Most of the agricultural employment is in household-farmingbased on a very high degree of labor intensity. On average, a member of the agricultural labor force works on 0.9 ha of land. It is estimated that the amount of land per agriculturalworker could be raised to 3 ha if labor and/or complementaryresources to bring land

As is the case with *imilarinfonnation from labor force ruveys eblewhere, these ar distzibutions of individuals by 'pnncipal occupation", not distzibutions of acual bor time.

See Annex to thischapter for details.

Thew termnactivity raw' is usd to mean the atio of laborforce to population. 90 Chapter 6

Table 6.1: Distribution of Population by Plrimay Activity (Percentof Population in Each Category)

Activity Urban Rural Male Female Male Female

Members not in labor force 56.6 69.4 58.6 64.9 o/w 'Homemakers" 1.3 19.8 1.3 13.2

Membersin labor force 43.4 30.6 41.4 35.1 o/w employedin: Agriculture 5.6 7.5 33.7 33.0 Industries 8.7 1.3 1.3 0.2 Services 27.6 20.4 6.3 1.8 Unemployed 1.4 1.5 0.1 0.1

Note: Members not in the kaborforce includethose who are too young or old to work, students, those attendingto householdwork (rhomemakers), pensioners,renters and a malU0. 2 percent of total) undefinedresidual category of "others".Agriculture includesfarming,fishery, animal husbandry and forstry. Industrics include both modern and traditional (cottage) manasactuing as well as construction,electricity, gas and water. Servicesconsist of all the rest. As noted in the text, these are contributons of idividuals by "pruwipaloccpan, snotdistribut of labor time.

Source: The HouseholdBudget Surveyfor 1989-90.

under cultivation were available.'7 Relative to land, labor is in this sense a scarce factor of production. Only a tiny proportion of the agricultural labor force is employed in large-scale commercialfarming for which no reliable recent estimate is available.

6.4 Servicesrank next to agriculturein terms of the share of the total labor force, employing16 percent of all workers. Their incidenceis however vastly different between the urban and the rural economies:65 percent of all urban workers and only 10 percent of the rural workers are employed in services. Services include a wide variety of activities, ranging from the relatively highly productive, but numerically tiny, tertiary services (e.g., modern finance, transport and trade) to rudimentary trading in the urban informal and rural subsistence sectors. Employment in urban services is dominatedby public sector workers (see below).

6.5 Industries-broadly defined to include both modern and traditionalmanufacturing as well as construction, electricity, gas and water-employ only 3 percent of all workers; 13 percent of the urban workers and 2 percent of the rural workers. Formal manufacturing enterprises-those employingfive or more workers-together account for a mere 53,500 workers, 0.8 percent of the labor force." Thus a very large proportion of those employedin industriesare actually engaged in informaland rudimentaryactivities providing low output and income.

87 An estimated4.6 millionha wer wudercultivation in 1990which i. believedto be only 30 percentof total cultivable ara. See Uganda:Agacultul SectorMemorandwm, ibid.

This infomnationis based on the NationalManpower Survey and refe to the year 1988. Chapter 6 91

Table 6.2: Estimated Mid-Year Distribution of Labor Force, 1992 (thowand Persons)

Urban Uganda Rural Uganda Category Male Fanle Total Male Female Total

Employedin: Agriculture 53.6 75.8 129.4 2530.0 2573.1 5103.1 Industries 83.3 13.4 96.7 98.4 17.9 116.3 Services 264.2 206.7 470.9 473.3 139.7 613.0 Unemployed 14.1 14.7 28.8 9.8 6.2 16.0 Total 415.2 310.6 725.8 3111.5 2736.9 5848.4

Total Uganda Category Male Female Total

Employedin: Agricultue 2583.6 2648.9 5232.5 Industries 181.7 31.3 213.0 Services 737.5 346.4 1083.9 Unemployed 23.9 20.9 44.8 Total 3526.7 3047.5 6574.2

Note 1: TheJanuary 1991 censuspopulaton of 16.6mis asswinedtohave grown to 7I3mby July 1992 . e., an annualgrovwh rate of about 2.8 percent). A variation of growth rate within the aboverange will not have a significant effecton th numbers.The growth in urbanpopulation betweenJanuary 1991 and July 1992 is assumedto have been higher (5 percent over 18 months) than the growth in Uganda's population. T overall gender diuribution is asswunedto be the sane as in the 1991 census.The gender distributon of urbanpopulaion is assumedto be the sanc (48.6 percent mak) as in municoialitiesand towncouncils shown in Table35 of the Bwkcground to the Budges,1992-93. The genderdistribution in rural Uganda is residually obtained. Thepopulation for July 1992 by genderand location are thus: urban. 1.970m(0.956m male and 1.Ol4nfemale) and ruraL 15.314mn (7.512mmale and Z802mfemale). Ratios in Tabl 6.1 have been applied to thesepopulation estimates.

Note 2: OveraU activity rates (excluding the "homemakers'from the labor force) are: 38 percent (41. 7 percentfor the male and 34.6percentfor thefemale). theef ho7 makesre included asmembers of the laborforce the acnvity rates turn out to be 42.9percentfor the male and 48.6 percentfor the femae.

6.6 Gender Distributionof Employment. If labor force is conventionallydefined to excludethose who are predominantly occupied with 'household activities' then the male activity rate in Uganda would be higher than the female activity rate. On this basis 42 percent of the males and 35 percent of the females are members of the labor force (Table 6.4). The gender gap in activity rates thus defined is greater in urban areas than in rural areas.

6.7 There is however littlejustification for adheringto the above conventionaldefinition of labor force. Those who are predominantlyengaged in householdactivities shouldbe included in the labor force because: (i) "householdactivities" by and large represent worthwhile labor that produces welfare, thoughconventional national accounting does not captureit; and (ii) women "predominantly" 92 Chapter 6

Table 6.3: Some Facts About Urban and Public Sector Employment

A Further Classification of Urban Employment, Mid-1992

Thousand Pesons Percent of Total

Public Sector 314.7 43.4 of which: Traditional civil service 97.9 13.5 Other govermnent/parastatals 216.8 29.9

Formal Private/Cooperative Sector 80.4 11.1 Informal Sector 330.7 45.6 Total Urban Employment 725.8 100.0

Growth of Government/Public Sector Employment (Persons)

July 1987 Civil Service Employment 239,528 January 1988 Government Employment 244,195 Parastatal Employment 53,593 Total Public Sector Employment 297,788

July 1990 Public Sector Payroll 320,669

Growth of Traditional Civil Sevice (Persons) July 1987 89,750 July 1992 98,873 August 1992 97,854

Note: See Annex to this chapterfor soarces and cxpuanaon.

engaged in household activities also engage in more conventional kinds of productive work. In an economy like Uganda-without an overt abundance in the supply of labor-both these phenomena are stronger than elsewhere in a typical labor abundant economy. For a Ugandan woman being "predominantly engaged in household activities" almost always means having two sets of occupations: domestic work and work in the normal economic enterprise of the household. The fact that the former is the dominant activity does not mean a lower overall intensity of work, but its exact opposite (see discussion in Chapter 3). Although no systematic documentation can be presented to justify this claim there is little doubt about its general validity.

6.8 If those 'predominantly' engaged in household activity are included, the female activity rate substantially exceeds the male activity rate: 49 percent of the women and 43 percent of the men are engaged in productive labor (Table 6.4). There is little doubt that women contribute much more to social labor than men. A higher proportion of women than men are members of the labor force when the latter is appropriately defined. Secondly, it is virtually certain that, on average, a female member Chapter 6 93 of the labor force works more hours than does a male member, especially if domestic work is included.

6.9 Discrimination against women as Table 6.4: Percent of Male and Female members of the labor force does not simply in the Work Force consist of extracting more labor out of them. They are also relegated to less productive Excuctingthose in "PrincipallyEmployed sectors and activitiesand more often kept out of in Household Work" directly remunerative activities than men are. For each man principally engagedin household Male Female work-which is not directly remunerative, nor Urban 43.4 30.6 conducive to control over household Rural 41.4 35.1 income-there are more than elevenwomen. In Overall 41.7 34.6 industriesand services-sectors withhigher than average productivity-for each female worker Inlnudingthose wPrincipallyEmployed there are respectively 5.8 and 2.3 male in HouseholdWork" workers.' Women account for only 20 percent Male Female of employment in the formal sector and 26 Urban 44.7 50.4 percent of employment in the skilled Rund 42.7 48.3 categories.90 Overall 42.9 48.6

6.10 Dominance of Public EmWloyment in Source:Tables 6.1 and 6.2. the Fonnal Sector. The public sector accounts for less than 5 percentof aggregateemployment in Uganda, but a staggering 80 percent of the urban formal sector employment.'9The public sector also accounts for 43 percent of all categories of urban labor force.

6.11 It is difficult to determine,a priorn, if a certain level of public sector employmentis too high or too low. The appropriatelevel of publicemployment would dependon numerousfactors, e.g., the structure of the economy, the need for public services and the composition of public services (especially between directly productive and 'unproductive' services). Considering the extremely undiversifiedstructure of the economy and the low level of public services (especially in directly productiveand production-promotingactivities) in Uganda, even the five percent share of aggregate employmentwould appear to be too high a share for the public sector.

Noneof these broad sectorsis homogeneousin tems of productivityand/or earnings.Thexe is no infonmationabout the genderdistribution of employmentin homogeneousactivitics within ecwh broad sector.

Manpower Planning Depailmeat, Ministry of Planning and Economic Development, Manpower and Enmploymentin Uganda,Report of the NaionalManpowr Swwy, 1989,Kampala, xi.

Is ther a non-urban fomul sctor? In principle, tho very lage commercial farms nay be considered to form that category. The National Manpower Survey included some government, partal and private agriculturmlenterprise (together employing 30,578 persona) as a part of the formal sector. To the extent that these arc lcated in the rural area, the urban fomWl (informal) ector employmeat in Table 6.3 is overstated (understated). The extent of modem large-scale enterprises in the rural conomy that have not been captured by the National Manpower Survey is not known, but it is Ikely to be small. 94 Chapter 6

6.12 Of far greater concernhowever is the overwhelmingdominance of public employmentin the formal sector of the economy. Modern economicactivities under non-governmententerprise together employ only 1.2 percent of the nation's labor forcel

6.13 The Annex to this chapter discusses how difficult it is to estimate the growth in public employmentin Uganda in recent years. In the five years since 1987public employmentin Uganda has increased by somethingbetween a tenth and a third, not a very helpful range of estimate. The point, however, is that even the lower limit of this range would appear unwarranted in the context of the rudimentarylevels of public services and their failure to achieve a meaningfulexpansion.

6.14 The Low Incidence of Unemployment. The HBS data indicate that unemployed workers represent 1.5 percent of the urban labor force and an incredibletenth of one percent of the rural labor force.' By and large, these estimates are consistentwith the phenomenonof relative shortage of labor in the economy. The shortageof labor is relative to the comparativeabundance of land and the virtual absence of complementaryinputs to cultivate land. The result is a great deal of unassisted labor required to eke out a modestliving under almostprimeval conditions.There is little opportunity for labor to remain idle althoughthe return to it is modest.

Features of the Labor Market

6.15 Nearly UniversalAccess to Land. In agriculture, which employs most Ugandans, there is nearly universal accessto land and a negligibleincidence of landlessness.The HBS reports that there are 2.9 million households(2.5 million rural households)and that 2.4 million householdshave farm land. Approximately2 percent of agriculturallabor force is located in urban areas. If one assumes that landownershipis uniformly distributed among the rural households and those of the urban householdsthat supply agriculturalworkers, then approximately95 percent of the rural households have land. If one assumesthat only rural householdshavefarm land (i.e., urban agriculturallaborers work on others' land or on nonfarm land) then 97 percent of the rural householdshave land.

6.16 The HBS also shows that most households have small farms, i.e., 62.2 percent farm householdshave 1 hectare or less land; 85.1 percent farm householdshave 2 hectares or less land; and 95.5 percent farm households have 4 hectares or less land. These refer to the distributionof operational landholdingunits. Ownership distributionof land is much more skewed, although no quantitative measurement of it is available. Large-scale commercial farms-including plantations-command very small proportionsof both agriculturalland and labor.

6.17 The inequalityof the distributionof land ownershipby itself is not a major source of income inequality.This is because land rent is negligible.Kibanja tenants of mailo land do not pay any rent. Even informal tenants of mailo land and the sub-tenantsof the freeholders pay very little rent.9

The NationalManpower Suvey claimsmuch higher rate. of unemployment,5 peraent in urbanaae and7 pe:ent in Niumlam0. Since thir survey wan limitedto the fornul sector and to a very rudimentaryenumeution of the urban infonTalsector, it is had to asign k a muchcredibility u to the HBS. An notedin the Annexto this chapter, a number of eutimatesmade by th NationalManpower Survey appear implasible and/or inconisent with ual expectation.Its claim of a higher rate of unanploymentin zuralasm than in urban Areasis yet another exanple of such implausibility. n ThereiB no informationavailable on theu rentalrates. A field trip to some villagesin Luweo in August 1992, yieldodevidence of annual rentper ha beingas little U Sh 200. Chapter 6 95

6.18 The Low Incidence of Wage Labor in Agrcidture. Universalaccess to land means that most membersof the rural labor force are predominantlyengaged in agriculturalwork on their own farms. There is no significantrural proletariat, i.e., the landlessworkers, who are perenniallyavailable for wage employment. Tbis means that wage labor is limited and that a very small proportion of agriculturalwage labor consistsof permanentworkers.

6.19 There are three types of wage employmentin agriculture:(i) casual laborers hired and paid on a daily basis; (ii) permanentworkers paid on a monthlybasis and hired for even longer durations; and (iii) contract workers who are paid at piece rate specified for given tasks per ha. Most of the wage labor takes the form of casual labor or contract labor. In the past, permanent workers, employed by large commercial farms, were mainly drawn from immigrant workers from adjacent countries. At present, however, this source of labor supply is insignificant.

6.20 There is little quantitative informationavailable on the incidence of wage labor in rural Uganda. A recent study states that 32 percent of farmers (28 percent in the western region, 37 percent in the central region and 28 percent in the eastern and northern regions) use hired labor at some time to meet their labor requirement.9' The critical question is what, on the average, is the proportion of hired labor to total labor used by the labor hiring farms. Given the extraordinaryfact that 68.3 percent of farms use no hired labor at all and the implausibilityof a clear dichotomy between farms using hired labor and farms using family labor, it appears that the average share of hired labor in the labor hiring farms can not be very high. It may be reasonableto considerone third as the upper limit. Indeed it may be much lower. Thus hired labor would at most account for only about ten percent of all labor used in agriculture.

6.21 Anothersource of informationis the recent UNDP survey of 334 rural householdsin western, southwesternand eastern regions.' These are the relativelyland scarce and labor abundantregions of Uganda. Ninety two of these households, about 27.5 percent, reported wage labor of all kinds (i.e., includingwage labor in nonfarm activities)as their primary economicactivity. Once again, the questionsare: what proportionof the labor of these householdsis used as agriculturalwage labor and what proportion of labor of the remaining households (for which wage labor is not the primary economicactivity) is used as the same. Assuming, quite arbitrarily, that the former is two fifths and the latter is one fifth, the incidenceof wage labor in these regions turns out to be about a quarter of total labor used in agriculture [0.4(27.5) + 0.2(72.5) = 25.5 percent]. Since these regions are relatively labor abundant, the average for rural Uganda will be lower, let us say less than (perhaps far less than) 20 percent.

6.22 Where within the above range the figure actually lies is impossibleto ascertain. On the basis of the availablerecords of visual reports of observershowever one would tend to tilt in favor of the lower limit of the range.

AlhonEvans, A Revew of theRural LabourMaret us Uganda,School of Africanand Asian Studies, University of Suasex,May 1992,Table 5. The authordoes not cite the moumeof the data, nor the time period to which they refer.

95 UNDP, Preparatory Study on Poveny Aaeion and Rumi Developmnnt in Uganda, Draft Report, Kampala, July 12, 1992 (preparedby Managemat Systens& EcononmcConsultants Ltd.). 96 Chapter 6

6.23 The two conclusionsthat are suggestedby the above may be summarizedas follows:

(i) In agriculture, where most Ugandanswork, the incidence of wage labor is low, perhaps not much more than a tenth of total labor used.

(ii) Of the total wage labor employed in agriculture, casual and short-term contract workers account for a very high proportion. The incidence of permanentworkers is very low. This is because most members of the rural labor force are predominantlyoccupied with work on their own farms. Very few are availablefor employmenton a long-termbasis.

6.24 The labor market is constrainedby the fact that the supplyof wage labor comes from farmers who are tied to their own farms. Thus differencesin the regional balance between the demand for and the supply of labor does lead to substantial differences in wages, often between adjacent regions.' This has probably been exacerbated by civil unrest, the lack of information and the absenceof infrastructurenecessary to promotegreater labor mobility. The criticalconstraint however is the institutionof small scale farming that permits nearly universal access to land.

6.25 Labor Allocation and Returns to Labor. Small-scalefarming still appears to be about the most profitableeconomic activity for Ugandaso that it has not been possiblefor other parts of the economy to offer high enoughwages to pull labor out of it. The vast rural householdfarming sector is the primary pool of labor supplyin the economy. Any movementof labor between this sector and the rest of the economy would depend on the relative remunerationof labor in this sector and in competingactivities. Labor could be pushed out by low return to work in householdfarms or pulled out by high earningsoffered by alternativeactivities that are profitable enoughto be able to do so.

6.26 Alternative activitiesto householdfarming exist in both rural and urban economies.Within the rural economy these consist of: (i) working for other household farms on a casual basis as a source of income supplement; (ii) working for large-scale commercialfarms as an alternative to householdfarming; and (iii) finding alternativework in other, often 'informal", activitieswithin the rural economy either as the principaloccupation or as a source of supplementaryincome.

6.27 The alternativeto remainingwithin the rural economy is to migrate to urban areas where three kindsof employmentopportunities might be distinguished:(iv) entry-levelunskilled employment in the formal sector of the economy; (v) government employment;and (vi) employment in the informal sector either as a transitional stage before finding more permanent formal/government employmentor as a sufficientlyremunerative permanent activity.

6.28 The return to labor in householdfarming is perhaps the standard in relation to which the remunerationin all the alternativeactivities are determined.This is becausethe householdfarms can potentiallycompete for the labor of the small pool of permanently availableworkers and the casual workers are principally employedon householdfarms. The relationshipbetween the return to labor in householdfarms and wages in alternative activitiescan however be very complex. Permanently availableworkers may acceptpermanent employment in large-scalecommercial farms at a lowerdaily wage equivalentthan the wage rate for casual workers in order to have the assurance of continued employment.The wage rate for casual labor can fluctuate quite a bit around the return to average householdlabor dependingon the return to labor of the particular groups of peasant householdsthat

96 See Alion Evanu, Op. Ct. Chapter 6 97

Table 6.5: Relative Earnings of Labor in Different Activities (Ugandan Shillings)

1. Return to family labor in householdfarming per day 1991 season 606 2. Wage of casual agriculturallabor per day Febnrary 1992 700 3. Wage of permanentagricultural labor per month February 1992 10,000 4. Wage of agriculturalcontract labor per ha Febuary 1992 30,000 5. Urban formal sector wage for unskilledlabor per month July 1992 6,000 - 15,000 6. Averagewage of group workers in govermmentper month August 1992 10,263

Sources: 1. Agrcultural Secretariat(Bank of Uganda),Repoal ox Expofl Crops and ProAederPr1es 1992. The conversionof the figure at February 1992 prices has been made by using the price indexfor Kampala betweenJuly 1992 and February1992 (27 percent increase).

2,3,4. AgriculturalSecretariat, Bank of Uganda.

5. Data gatheredby mission in thefield.

6. Office of the Commissionerof Data Processing,Ministry of Fhaance. supply casual labor, the pattern of seasonalityand the extent of distress supply of labor in order to meet cash needs. To induce labor to move to urban areas, urban wages must be significantlyhigher than the return to labor for the relevant groups of household farms with potential migrants to compensatefor the increasedcost of urban living.

6.29 Table 6.5 summarizessome information on the relative earningsof labor in differentactivities and sectors. Even after allowanceis made for significanterror in the measurementof these indicators, it is quite clear that work in householdfarms is, on the average, the most remunerativeoccupation for an unskilledperson. If an average worker ventures out of the householdfarm for cash earnings as a casual worker in agriculture,he/she has to settle for a lower daily earning at the margin than the average return to a day's work on the householdfarm. This means that casual labor is supplied by peasant households with less than average return to labor and/or that casual labor is supplied by householdsthat would encounter sharply diminishingreturns by trying to use available additional labor on householdfarms.

6.30 More intriguing is the fact that, on a daily basis, the wage rate for permanent agricultural employmentis lower than the wagerate for casual labor. At this wage rate it is still possibleto attract the small pool of permanently availableworkers (the landless, includingthe surplus" members of householdsendowed with large supplyof labor relative to the supplyof land). But this wage rate is lower than what would be necessary to induce workers on any but the very marginal of household farms to move out. It is not surprisingthat the estate sector has been finding it hard to attract as much labor as it needs. 98 Chapter6

6.31 The entry-levelwage in the urban formal sector has a rather wide range. Even at the top end, the earnings are only modestly above that in household farming.' Unskilled wages in the public sector are significantlybelow the averagereturn to labor in householdfarming on the assumptionof a reasonable intensityof work.' Thus, it would appear that, none of the activities in the formal sector has the dynamismand productivityhigh enoughto enablethem to offer labor an earning higher than what labor on the average receivesin householdfarms. The silver liningto this gloomy scenario is that, despite the overwhelmingdominance of the public sector, the formal sector of the economy has avoided foisting an artificiallyinflated structure of wages. As a result the incentiveto migrate to urban areas has remained low and the urban informalsector has remained modest in size.' Another implicationof this is that the suffering of the civil servants retrenched by the civil service reform program, which the Government has recently embarked upon, will be mitigated by the fact that householdfarming and/or the urban informalsector can offer them returns on their labor which are comparableto what they were receiving from the Government.

Which Employment Categories are Poor?

6.32 A direct ranking of the employmentgroups in terms of the incidence of poverty is not possible. It is, however, quite obviousthat the greatest concentrationof poverty is in the group of smaller householdfarmers, among those 62 percent of farms that have a hectare or less land. This statement may appear to be in apparent conflict with the preceding argumentthat return to labor is about the highest in householdfarming. But that comparisonwas with the average reotrn to labor in the entire householdfarming sector, a finding that need not have any conflict with the proposition that the per capita incomeof the very small householdfarms would be much lower than average.

6.33 Virtually nothing is known about the incomes of the nonfarm households in the rural economy.They accountfor about 13 percentof rural households.In view of the relativelywidespread access to land in rural Uganda it does not appear that nonfarm occupationsgenerally consist of residual activities providing very low incomes to which the landless are driven. It is nevertheless possible that these groups have a high incidenceof poverty. Even so, the poor among them must be a small proportion of all the rural poor simplybecause these groups represent a small proportionof the rural population.

97 Assumrethat an average laborer on a household farrn has 200 daysof work per year. At the average estirated daily return of U Sh 770 this works out at U Sh 154,000 per year or U Sh 12,833 per nonth. The upper end of the range of wages for unskilled categories, according to data mnadeavailable to the mission (aee annex to this chapter), was U Sh 15,000 in July 1992, perhaps barely enough to copensato for the difference in cost of living.

How has it been possible for the Government and the formal sector to get workers for emnploymnt? One can even obwerve mnoderatequeues for such enployment which need to be expained. Tentatively, the following explanations might be advanced. First, to a degree this is due to 'market segentation". School leavers in urban areas rule themselves out of the rural labor market and offer thenselves for employnent only in the urban formul sector, especially for public ector jobs. Secondly, as uggested by the Ministy of Labor, it is possible to rcruit at the bottom end of the range of unsklled wage in the fornal sector if the enterprises making recruitment are located near depressed rural areas from which labor is drawn. Thirdly, a further flexibility is available in the public sector in the form of the opportunity for moonlighting and effectively reducing the contribution of labor (i.e., increasuingthe earning per hour of actsal work done).

In the recent past urban crowding was further relieved by civil strife that encouraged migrtion out of urban areas. Chapter6 99

6.34 Those employed in the urban formal sector are probably protected from extreme poverty. There are of course several instancesof very low wages for the unskilledcategories in some of these enterprises, but these probably represent cases of low supply price of labor (due to the proximityof the location of employmentto the workers' homes in rural areas) or the failure of the inspectorsof the Ministry of Labor to account for all the benefits. It is nevertheless possible that a certain proportion of these workers would belong to poverty groups although it is hard to imaginethat their number would be large.

6.35 Earnings in the urban informalsector were estimated by the National ManpowerSurvey to be only about 13 percent below the average earnings in the formal sector.' It is therefore reasonableto concludethat the incidenceof poverty in the urban informalsector is only moderately higher than the incidenceof poverty in the formal sector.

6.36 To conclude, there probably are pockets of poverty among the rural nonfarm households, householdsof the low paid workers in the urban formal sector and the families employed in the informal sector. But the greatest concentrationof poverty in Uganda is among the smallest half or so of the peasant households in terms of the size of land holding. Focussing attention on this group is also critical in view of the earlier conclusion that a rise in the earnings in the peasant sector is likely to push up earnings throughoutthe economy.

The Objective of Emnploymentand Labor Market Policies

6.37 The above analysis suggests that the central objective of employmentand labor market policies for the reduction of poverty in Uganda should be to increase the earnings of labor in agriculture. This will directly reduce poverty because most of the poor in Uganda are employedin agriculture. This will also help alleviate whateverpoverty exists elsewhere in the economy because a rise in the earningsof the agriculturalworkers will filter down to the rest of the economyand raise the earningsof the workers in the other productivesectors.

6.38 The overall context of the strategy of development within the framework of which the objective of increasingthe earnings of agricultural labor is to be pursued needs to be spelled out clearly to make it possibleto determinethe elementsof employmentand labor market policies in a consistent and coherent manner. In principle, there is no unique path along which to pursue the objective. One might illustrate alternativepaths by citing the following polar cases.

(i) One possible strategy is to promote a system of peasant agriculture by further consolidating the de facto universal access to land. In this case the proposed land reform legislationmust find ways of guaranteeingtenancy rights to aU the existingusers of land and by improving the land endowmentof the very small farms. This must be supplementedby the institution of a system of supplyof technology,extension services and credit that is appropriatefor the small peasants and adequate in enabling them to overcome the absence of economiesof scale.

(ii) The polar opposite strategy is to create a framework for the exploitationof economiesof scale by promotingas much of large-scalecommercial farming as possible. In this case the

ManpowerPanning Depamnt, Miniuy of Pannig and Econonm Dcevpmt, Mmpowr and EMy1Aw ui Uganda, Reporl of th Nadond Mawpowr Surwy, 1989, KTmp, p. I81. 100 Chapter 6

proposed land reform must perform a very different function. It should limit tenancy right only to those cases in which the right has already been granted de facto (e.g., the kibanja tenants on mailo land). Elsewhere the landowners might be encouraged to resume their ownership by evicting the informal tenants and converting them into rural proletariat. Incentivesfor large-scaleagricultural entrepreneurship might be strengthened.Employment and labor market policies might be backed by technologicalpolicies for an appropriate expansionof demand for labor. Conceivablyreal wages in agriculture would rise as fast as agriculturalproductivity to help reduce poverty.

6.39 While in principle both of the above strategies can lead to poverty reducing growth of the rural economy, their practical feasibilityneeds to be carefullyanalyzed. Also the choice among these strategies is not simply a matter of economic calculations.There are important social and political issues to be considered. The point however is that the overall strategy needs to be clearly outlined and agreed upon to provide a consistentframework for the determinationof employmentand labor market policies.

6.40 It does appear from the objectivesof the proposed land reform legislationthat Uganda wants to take the first route to rural development.This paper cannot possibly outline the componentsof such a strategy in any detail,101 but it is useful to highlight the following three main sets of its elements.

6.41 First, the proposed land reform shouldclearly determinethe institutionalframework for rural development.Freehold rights must be guaranteedto all cultivators,not merely those who have formal tenancycontracts.'02 Thus, for example, in the mailo areas, it appears that in additionto the kibanja holders there exists a large category of tenants who rent and borrow land without any written acknowledgementof arrangementsfrom landlords. On such land often the landlords determine the crops to be planted. This category of virtual squatters must also be given freehold rights if the strategy of peasant farming is to be successfullyimplemented. It is not clear from the documents related to the proposed legislationif adequateprovision is being made for this. Land endowmentof the smaller size groups should also be improved both by investment in land augmentationand by changes in the system of incentivesto make it unprofitablefor the large land owners to leave land unused, a practice that seems to be widespread.'0

6.42 Secondly,to enablethe small peasantsto overcomethe absenceof the advantageof economies of scale steps should be taken to facilitate their access to an appropriate technologicalpackage. At present Ugandanagriculture operates under almostprimeval conditions: the limitedspread of tractors and ox ploughsthat took place in the past has been wiped out, fertilizer use has dropped to virtually nothing, there is an acute absence of basic implements,and the knowledgeof farming system and crop husbandrypractice is rudimentary.Basic equipmentssupply (weeder, planter, thresher for grain

10° For further dtails of the and tenurc system and the proposed hnd reform, see Uganda: Agricuaure Sector Memorn1dum,ibid.

'U The purpose of this paper is not to provide a blueprint for land rtform, which involves complex ime, but merely to emphasize thatunivermal acces to landshould be insitutionalized.In the mailo amethis can clearly be done beat by Srantig freehold rightsto the tenants.The question of guarunteiSu access to land in tho areas of traditional land tenure-where communal contrl over lnd prevails-is obviously more complex and needs careful analysis.

'U A system of progressive and tax, if feasible, would help ensure this. Chapter6 101 crop areas, harvestingtools, and a reintroductionof ox ploughsand selective access to tractor service where essential)needs improvement.Selection and multiplicationof appropriateseed varieties for the use of the peasants is of high priority. Soil fertility measures and the introductionof fertilizerdeserve attention. All these need to be backed by the disseminationof knowledgeabout farming systems and crop husbandry practices. The emphasis should be on the provision of these services through the market, backed by an access to credit. The role of the public sector should be to help augmentthe supply of the technological ingredients by providing appropriate incentives to private producers/suppliersand build a network of extensionservices.

6.43 The third major element of the strategy is to ensure an appropriate compositionof peasant output that would help avoid the constrainton demand. If agriculturalgrowth takes place withoutany change in the compositionof output, soon the production of most subsistence crops and many traditional cash crops would face some kind of demand constraintthat would lead to reduced prices and profit. For the earningsof the peasants to continueto grow, it would be essentialto bring about a change in the compositionof output in favor of products for which demand is not a constraint. In the case of Uganda it would mean looking for agriculturalproducts that would end up as exports, with or without processing. An agriculture consistingof small peasants would need outside help in anticipatingthe appropriate pattern of demand and in obtainingthe infrastructure necessary to get access to new markets.

6.44 While the emphasis of the developmentstrategy in Uganda should be on agriculture with a focus on the small peasants, the process of gradual loweringof the share of the labor force employed in agriculturemust begin to take place. The necessarycondition on the supply side would be created by the improvedproductivity of the agriculturallaborers. On the demand side, nonfarm activitiesin which Ugandahas comparativeadvantage must be identifiedand developed.Several categoriesstand out as potential candidates.First, agriculturaldevelopment should create demand for a variety of both producer and consumer goods. This should create a basis for the growth of rural nonfarm employment.Secondly, a part of this demand should spill over in the form of demand for the output of industries and services in the formal sector. Existing industries, reported to be operating at a fraction of capacity, shouldbe able to productivelyabsorb more labor. There should also be a basis for the creation of production capacities in new activities, i.e., to supply producer and consumer goods to the rural economy and to process the agricultural products both for export and for absorptionin the domestic market.

Chapter6 103

Annex to Chapter 6: Urban Employment

A Further Disaggregation

A6.1 The purpose of this Annex is to disaggregateurban employmentshown in Table 6.2 into employmentin public sector, the formal private sector and the informal sector. There are several estimates of public sector employmentfor recent years althoughit is not always easy to reconcile them.

A6.2 First, there is the Census of Civil Service for July 1987 which shows a total government employmentof 239,528. Second, there is the National Manpower Survey of January 1988 which shows a total formal sector (enterprises employingfive persons or more) employmentof 378,227 divided into the following categories:

Governmentemployment 244,195 Parastatal employment 53,593 Formal private sector employment 80,439 (includingcooperative and 'other")

A6.3 The same survey also implementeda rather limitedenquiry of informal sector employment in four cities. It arrived at an estimate of total urban informal sector employmentof 7.136m which must be rejected as too high."l Clearly, there was an illegitimateextrapolation of the incidenceof informal activities in four major cities in which there is a concentrationof such activities. Finally, there is an estimate of government/publicsector employmentof 320,669 for July 1990based on the payroll statistics.'0

A6.4 The problem is that it is not altogetherclear if governmentemployment according to the three sources are quite comparable. For example, does the payroll estimate for July 1990 include employmentin parastatals? If it does, employmentin the public sector in two and a half years since the ManpowerSurvey in January 1988 went up by only 8 percent. If it is assumedthat the July 1990 estimateexcludes the parastatalsthen governmentemployment over the same two and half years went up by nearly a third! It is impossibleto know which is the right assumptionand what happened to employmentin parastatals if the July 1990 estimatesexclude them. This paper makes the assumption

A4 ManpowerPlanning Depaitment, Ministy of Planningand EconomieDepautmt, Mawpowerand Employmentin Uganda,Repor of the AdonalManpower Swrvy, 1989, Kwampha,October 1989, says that 'the informalsector ... is estimatedto be providingwork oppomnitiesto about13.7 pereent of the labour force' as comparedto a fornal ector employmentof 5.3 percent of the labor force (pp.ix-x).The impliedetimate of the labor force is 7.136m whichitself is too high for 1988.There wasno nationwidesurvey of labor force at the time and hence one must concludethat this estimatewas basedon a presumedsat of activityrates which aretoo high. Moreover,the estimateof informalsetor employmentis far greaterthen aggregateurban employment basod on the ratiosof HBS, the only sourceof laborforce participationra for recentyea. Thus this estinmteof informalsector employment must be rejectedas beingtoo high.

IN5 Quotedin an unpublishedpaper by the ILO consultantQ.U. Khan,Enployment Situation h GowenmmentSector, in the ManpowerPlanning Department (undated typencript). 104 Chapter6 that the 1990 payroll statistics actually include the parastatals which are excluded from the 1987 Census of Civil Service.106

A6.5 To arrive at 1992 midyear estimatesthe followingadditional assumptions are made:

(i) Formal private-cooperativesector employment remained unchanged between 1988 and 1992.107

(ii) For mid 1992 the only part of public sector employment for which a direct estimate is available refers to traditional civil service, employing 97,854.101No estimate is available of total employmentin governmentand parastatals.We make the plausible, but unconfirmed, assumption that this is 314,669, i.e., 6,000 fewer than the level in July 1990. It is well known that there has been a reduction in public employmentin recent period and the figure that is often mentioned is 6,000 although it is not at all clear to what time period and/or categories this number refers to.

(iii) All formal sector employmentis located in urban Uganda, an assumptionthat seems to be implied by the National ManpowerSurvey.

A6.6 These assumptions, together with the estimate of total urban employment in Table 6.2, provide the following distributionof urban employmentin mid 1992 (in thousand persons):

Public sector 314.7 o/w Traditional civil service 97.9 Other government/parastatals 216.8 Formal private (& miscellaneous)sector 80.4 Informal sector 330.7

Total urban employment 725.8

Besiderreiterating that great uncertaintysurrounds all these data, we would like to argue that this interpretationis additionallyjutified by the widespreadbelief that the 1990public employmentestimates uffer from a serious upward bis due to the high incidenceof ghost' worken. We wantto avoida possiblefurther overestinationof public sctor employmenteven at the risk of undersatingit somewhat.

A uvey of 26 enterprises-employinga total of 15,511persons in 1991-based on the inspectionreports of the Ministryof Laborleads us to concludethat there is no way to judge whathappened to employmentin private formal ectorduring the period under roview.No changeseema to be as good an asumption as any other.

This estimateis providedby the offce the Commisionerof Data Procesing, UgandaComputer Services, Ministry of Finance ad Planning. Excluded categoriesae teaching services, so-calledgroup employment(temporary workers),and panutatals. Chapter 6 105

Growth in Public Sector Employment

A6.7 The above sources of data shed a limited amount of light on the trend of public sector employment. Indeed the only part of public sector employment about which the change in recent years can be establishedis traditionalcivil service in which employmentchanged as follows:'0

July 1987 89,750 July 1992 98,873 August 1992 97,854

A6.8 There was a 9 percent increase over the five years since 1987. As a result of the recent governmenteffort to reduce public employmentthe number of traditionalcivil servants declined by just over one percent."'1

A6.9 Precise changes in total public employmentare much harder to determine. It is quite clear that there was a substantialincrease between 1987 and 1990 althoughthe precise measurementof the change is obviated by the problem of comparabilityof data over time. If the 1987 census of civil service accountedfor the parastatalsthen public employmentduring the next five years increasedby more than a third. If the 1987 census excluded the parastatals and our assumptionthat the 1992 payroll data includethe parastatals is correct then the increasewas of the order of 12 percent.

A6.10 The comparisonover time is made even more difficultby the problem of ghost workers. It appearsthat there is a widespreadpractice of inflatingthe payroll by includingnonexistent workers. No one knowshow their incidencehas changed over time. There is an ongoinggovernment program to weed these ghosts out. It seems highly likely that the recent decline in employmentin traditional civil service was largely the result of the eliminationof some of these ghosts. Currently an attempt is being made to eliminatethem from teaching and other servicesin whichtheir incidenceis believed to be greater.

TheA ouroe are as folow.: July 1987:Census of Civil Servce; the,other dates:The Office of the Commissioner of DataProcesing.

10 The offce of the Commisioner of DataProcesing gave the impresion that the decline in the number in the trditional civil servicepayroll between July and August1992 acully took place over a longer period. Thos rurAchod in recet times were not taken out of the payroll urtil the month of Augu"t.

Providing Key Services and Safety Nets

Growth is vital to reducingaU aspectsof absolt poverty.... Butgrowth unaccompaniedby other measuresmay neitherboost the incomes of thepoor much,nor leadto muchprogress on the non- incomeaspects of poverty.On both counts, humandevelopment programs have a part to play " (Poverty and Hwnan Developmnt,World Bank, 1980, p. 63.) Introduction

7.1 As the World DevelopmentReport 1990 stressed, economicgrowth is a necessary but not a sufficientcondition for the reductionof poverty. It is imperativeto ensure that the poorest segments of society are able to participate productivelyin such growth. Consequently, public expenditures, specially on social sectors, have an important role to play in tackling poverty and reinforcing a growth-orientedpoverty reductionstrategy.

7.2 The most importantgoal of economic development is the improvementin the standard of living of the population. In a poor and resource-constrainedcountry such as Uganda, where socio- economic indicators reveal extremely low levels of welfare (Table 1.4), such improvement in the standard of living requires substantialinvestment in human capital (education and health care) and in rural infrastructure (rural feeder roads and clean water supply). It is also importantto insure that the growth in populationis not so rapid as to underminethe growth in incomeand the generationof investable surpluses. The development of human resources is both a means and an end of the developmentprocess. It is a means to an end in that investmentin human resources and economic infrastructureis a prerequisitefor increasingproductivity and expandingincome-earning opportunities both at the individual and national level. It is an end because those with a well-developedhuman capital base can insure a better qualityof life for themselvesand their children. Investmentsin human capital and rural infrastructure,however, have long gestationperiods and therefore require activeand sustainedgovernment involvement in these sectors for long periods of time.

7.3 While public expenditures have an important role to play, the Government's role must necessarilybe guidedby cost-effectivecriteria. It is essentialthat the Governmentallocate its limited resources to those activitieswhich provide the maximumsocial and economic rates of return. For example, it has been shown that the average social rates of return to primary education for Sub- Saharan Africa, Asia, and Latin America and the Caribbean, are 26 percent, 27 percent and 26 percent, respectively; these are roughly twice the social rates of returns from higher education. Furthermore, experiencefrom several countrieshas demonstratedthat it is the poor in general, and girls in particular, who benefit most from expenditureson primary education whereas it is the nonpoor who benefit disproportionatelyfrom higher education. The nonpoor also generally have sufficientfinancial resources to bear the cost of such higher education. Consequently,where the beneficiariescan afford to pay for services rendered, it is advisablethat cost recovery measures be instituted. The Government of Uganda, therefore, faces a number of challenges in allocatingits limited domestic and foreign resources efficiently and in implementing cost-recovery schemes successfullyso that economicgrowth is facilitated, poverty is alleviated, social equity is enhanced and the welfare of its people is maximized.

7.4 The following sectionsbriefly review the level and compositionof govermnentexpenditure in key economic and social sectors, particularly primary education, primary health, and rural 108 Chapter 7

Table 7.1: Sectoral Expenditure Performance of the Central Government (Locaiy Funded Expenditue)

As a % of GDPMP aS%of Total Expenditures

Prelim. Budget Predm. Budget 1989190 1990/91 1991/92 1992/93 1989/90 1990/91 1991/92 1992/93

Social Services 1.6 1.8 2.5 2.4 20.0 23.2 29.7 33.4 Education 0.9 0.9 1.3 1.1 10.5 12.0 16.0 15.6 Health 0.4 0.4 0.5 0.6 4.7 4.9 6.3 8.5 Local Gov't 0.2 0.4 0.5 0.6 2.5 5.0 6.3 8.5 Other 0.2 0.1 0.1 0.1 2.2 1.3 1.0 0.9

Economic Services 1.3 1.4 1.1 0.9 15.7 17.3 13.5 12.5 Agriculture 0.2 0.3 0.3 0.2 2.9 3.3 3.7 3.5 Infrastructure 0.7 0.8 0.6 0.5 8.5 10.7 7.2 6.4 Other 0.3 0.3 0.2 0.2 4.3 3.3 2.6 2.5

Defense 3.1 2.8 2.3 1.5 38.7 36.3 27.8 21.7

Public Admiinistration 2.1 1.8 2.4 2.3 25.6 23.2 29.0 32.4

TOTAL 8.1 7.9 8.4 7.0 100.0 100.0 100.0 100.0

Sources: MimisLtyof Funaneeand EconomicPlanning, Background to the Budget (July 1992) and staff estimaes. infrastructure."'

Government Expenditure on the Social Sectors

7.5 A review of the structure of public expendituresin Uganda indicates that the expenditure pattern is characterizedby a low level of aggregatepublic expenditureeven by Sub-SaharanAfrica's standards. Constrainedby an extremely low domesticrevenue effort, public expenditureas percent of GDP was only about 10.5 percent in 1987/88rising to 20.6 percent in 1991/92,'12as compared to an average for Sub-SaharanAfrican countries of 30 percent (the average for developingcountries in general is about 21 percent). Between 1987/88and 1991/92,revenues collected as percent of GDP increased slightly from about 5 percent of GDP to 7 percent, as compared to an average revenue effort of about 20 percent for Sub-SaharanAfrican countries. Consequently,the weak revenue effort and the serious internal security problems until the late eighties, which required high expenditures on defense, have preventedthe Governmentfrom allocatingsufficient funds to priority sectors such as education,health and rural infrastructure.It has also forced the Governmentto rely on foreign aid to finance a significant proportion of its expenditures;in 1991/92, foreign aid financed nearly 60 percent of government expenditures. The Government's locally funded sectoral expenditure

For a detailed asament of the policies, priorities and expenditures in the areas of health, education, nd population, the intreWed reader in referred to Uganda: Sodal Seor Shrtegy, 1993, Report No. 10765-UG, World Bank.

112 There is general agment that data prior to 1987/88 are not very reliable. Therefore, the base year used here for the purpose of evaluating changes in public expenditures is 1987/88. Chapter7 109

Table 7.2: PovertyFocus of GovernmentCurrent Expenditures (As a Percent of Total Current Expenditures)

1989/90 1990/91 1991/92 1992/93 Actuab Actub preim. Budget

Human Resource Development 20.4 24.6 32.5 35.1 Education 13.9 16.2 20.8 19.5 Health 4.6 5.4 6.3 8.3 Local Govemment 1.9 3.0 5.4 7.3

Economic Services 5.9 5.7 6.8 7.7 Agriculture 2.5 2.7 2.9 2.8 Infastructure 3.4 3.0 3.9 4.8

TOTAL 26.3 30.3 39.3 42.8

Note: Actualsfor 1991192are prelminamy.In addiion to the Ministricsof Healthand Education,the Mmnistry of Local Governmentalso provideshealth and educationservices.

Source: Ministry of Fhanceand Economic Pklnning. performance is summarizedin Table 7.1. As the table shows,the share of expendituresallocated for defense purposes has traditionally been several times higher than those for the social sectors, reflecting the security concerns of the Government. However, expendituresfor 1991/92 indicate a noticeablereduction in defenseexpenditures as percent of GDP (from 3.1 percent in 1989/90to 2.3 percent in 1991/92), a real decline of 14.5 percent, reflecting the improved security situation throughout the country. Defense expendituresare projected to further decline to about 1.5 percent of GDP in the 1992/93 budget.

Incidence of Government Expenditures

7.6 It is possible to evaluate the effectivenessof poverty-reducinggovernment expenditure programs by determining what fraction of government expenditure in any given sector actually reaches the poor. Such analysis, however, requires reasonablyprecise informationof who the poor are, what proportion of the total populationthey represent, where they live, their sources of income (agriculture, informal sector activity, wage labor, etc.), the level of their education (primary, secondary, secondary,tertiary) and the types of diseases(preventable, curable) that they are afflicted with. Detailed householdbudget survey data, which contain the relevant questionsto enable a full- fledged incidenceanalysis, can usually be used to determine such information. Unfortunately,such data do not exist for Uganda and the analysiswill have to await the results of the Integrated Survey which is presently underway. Although data limitationsmake it extremely difficult to analyze who the actual beneficiariesof these expenditureswere, the following sections try to infer this from the existingbits of informationand from the evidence in other countries.

7.7 Table 7.2 examinesthe relative share of broad categories of poverty-orientedprograms in current expenditures.The table shows that while the share of expenditureswith the potential of reducing poverty in Uganda remains low, recent years have seen a significant improvementin the right direction, reflectinga recognitionon the part of the Governmentof the importanceof economic 110 Chapter 7

PRIMARYEDUCATION EXPENDITURES 11~

3.3- NyuX qwvt .1)2 -# Fisca YeAK

Figure 7.1: Primary Education Expenditures and social sectors. The table shows that in 1991/92,approximately 33 percent of total locally funded expenditureswere allocatedfor the purpose of enhancinghuman resource development,while about 7 percent went towards critical economicservices.

Education

7.8 The role of education in increasing economic productivity and improving the quality of peoples's lives through better health and nutrition has been well documented."' As mentioned earlier, it has been demonstratedthat the economic and social returns to primary education are significantlyhigher than the rates of return for secondaryand tertiary education.Policies that expand access to good qualityprimary educationcontribute significantly to poverty reduction.Furthermore, even if they never enter the labor force, the returns for expenditures on educating women are substantiallyhigher than those for men. Studies in Bangladesh,Brazil, Kenya, and Columbiahave shown that families tend to be better fed and the children less likely to die, the more educated the mother. Countriesare justified in giving priority to primary educationboth on equity and efficiency grounds."' The Government of Uganda appreciates the importance of education and its expenditureson educationas percent of GDP increasedfrom 0.9 percent in 1989/90to 1.3 percent in 1991/92, implyingan average real growth ratwof over 20 percent.

13 For examples,sec PnimaryEduoo A WorldBank Poluy Paper, World Bank, 1990; for the socio-economic benefitsof educatin wonmn, weahivesng in AUdw People:FWcaing Women in DevelopingCowaries, LAwrence Summer, 1992 (sumnarizedin Box 3.9).

114 The economifcand socialreums of pimary educationare wellknown. According to the WorldBank's policypaper on prnimryeducation, four year of educationincreased Umal-fam productivity by 7 pernat acros 13 developing countris and by 10percent in countri wheo new aSriculturl tecluiqueswen beingitroduced. Prinmryeducation is also observedto have positivesocial effects. For itance, womenwith normthan four year. of educationhave 30 perent fewer childrenthan womenwith no educationand their childrenhave motlity rates only half a high. Chapter7 111

7.9 The Governmentalso clearly recognizesprimary educationas a priority area in its poverty reduction strategy and its stated objective is to achieve universal primary education by the year 2001/02. The strategy, as set out in the White Paper on EducationPolicy, is to shift the distribution of expenditure away from secondary and tertiary institutions towards the primary levels."1 5 At present, secondaryand tertiary educationcontinue to receive financial resources which are in excess of what can be justified, given the resource constraintand the Government's priorities. While the share for secondary educationhas remained stable at an average of 0.2 percent of GDP during the period under review, the share of expenditureon tertiary education, as percent of GDP, has grown from 0.3 percent in 1988/89to 0.8 percent of GDP in 1991/92.This represents an increase of 170 percent as comparedwith the 131 percent increase in primary education.In addition, on a per pupil basis, governmentexpenditure on secondaryand tertiary educationis quite high relative to primary education.Expenditures on universityeducation and secondaryeducation were 157 times and 3 times that of primary education respectively in 1991/92. Generally, post-primary students come from relatively better-off families, who can afford to pay for the advanced education of their children. Given that the rich benefit more than the poor from expenditureson higher education and that the returns to higher educationare less than those for primary education,prompt action to increasecost recovery at the tertiary level, as recommendedby the White Paper and the Report of the Makerere University VisitationCommittee, is justified.

Health

7.10 It is commonknowledge that poor health reduces the capacity to work, constrainsthe ability to increaseincome and adverselyaffects the qualityof life. The link betweenpoverty and poor health, therefore, is direct. Investmentsin health and nutrition have a variety of socio-economicbenefits, includinggains in productivity,learning and life expectancy.The gains from adequatehealth care are particularlysignificant for women of child bearing age.'1' It is, therefore, encouragingthat, in the Three Year Health Plan, the Government is aiming at increased expenditure on preventive and primary health care, while selectivelyintroducing cost recovery for curativehealth services. It is also equally important to insure that the poor, irrespective of gender and place of residence (rural or urban), receive their share of the benefits from investmentsin health.

7.11 The overall resourcesallocated to the health sector in Uganda are low; in 1991/92,about 0.5 percentof locallyfunded expenditureswere allocatedto the health sector, and this amountrepresented a 28 percent real annual average increase over 1987/88. When donor funding is taken into consideration, expenditures on health increased to about 2.5 percent of GDP in 1991/92. It is generally accepted that the absolute minimumrecurrent expenditurenecessary for effective primary and secondary health care is equivalentto about US$34 per capita per annum.117Uganda recurrent expenditureon health is less than US$2 per capita per annum as comparedwith Kenya which spends US$6, Zimbabwewhich spends US$14 and Botswanawhich spends US$29. An importantfeature of expenditureson health is that the share of expenditureson preventive health care is very low as

IIS See almoRantgrowd to dse Budget, 1992-93,Minjity of Financeand Economic Plnning, Juno 1992.

lid Accodilg to the Wold Bank policy pyer on pnmmayeducation, mo than one fourth of all dea±h of women of child-beaing age am cloacy linked to ciildbiith. It i pomible to avoid anmostall of dwe dilalthawh baic family plannig, pratl cam ae asad ta. with childbuth.

117 Se Public Oce for PnMmt I ws: Po,itizinw Public E.qwra&ws for Swutainabk and Equitabk Gn%id in Uganda, World Bank Repoit No. 92=3-UG, FebNuy 12, 1991, p. U3. 112 Chapter7

PRIMARYHEALTH EXPENDITURES 1.

su X

10.8

w90LYS Slj2 -> Fisl Yew Lody_ed

Figure 7.2: Primary Health Expenditures comparedto expenditureson curativeservices. In general, only about 10 percent of the expenditures on health go to preventive health care while curative care receives the remaining 90 percent."'1 During 1989/90-1991/92,the share of total expenditureon preventivehealth care, as percentof GDP, averaged0.1 percentper year, while the correspondingshare for curativehealth care during the same period averaged 1.3 percent, i.e., 10 times that for preventivehealth care. Since curativehealth care often benefits wealthier members of society in general, and those who live in urban areas in particular, the poor benefit disproportionately less from such allocation. A poverty-conscious program, therefore, shouldmake more resourcesavailable to primary and preventivehealth care than curativehealth care. Chargingwealthy members of the Ugandansociety for curative medical services and using funds mobilized thus to improve access of the poor to preventive health care would be an appropriatestrategy.

7.12 The Governmentrecognizes this and its commitmentis reflected in the high growth rates for the budgetary allocationsto this subsectorin the last few years. In 1989/90, the total actual central governmentexpenditure on preventivehealth care (both recurrent and capital expenditure)amounted to U Sh 1,396 million. The corresponding amount for 1991/92 represents a real increase of 54 percent, with the budgeted amounts for 1992/93representing a further real increase. Despite these improvements in the allocation of funds to the health sector, Uganda's per capita recurrent expenditures on primary and secondary health is very low compared to a number of African countries. The recent resurgence in preventable diseases is a clear indicationof these low levels of expenditure. In 1990, preventable diseases like malaria, AIDS, diarrhea, pneumonia, anemia, and meningitis alone contributed to 53 percent of the deaths reported in 20 hospitals."' Nutritional levels, particularly of children, are also unacceptably low. Low nutritional levels expose a

u2 Accordingto the World Developmat Report 1990, on the aveAge,an eaimat 70-85percent of the developing world's expenditur on health(both public ad privat) goes to curative health cam. It appearsth Ugawda's expenditureon this ectoris even higher than thatfor the developingworld.

For details ee UganwdSocial SectorStvegy, World Bank,1993. Chapter7 113 considerableproportion of the populationto killer diseases. It is alarming to note that 54 percent of the deaths in hospitals are related to children under five years of age and that 55 percent of these deaths are due to malnutrition. Improvementsin the quality of health conditions,therefore, have important implicationsnot only on the welfare and productivity of individualsbut also on national welfare and productivity.

Family Planning and Environment

7.13 Chapter 5 highlightedthe quantitativeimplications of the high rate of populationgrowth on poverty reduction in Uganda. The population growth rate, 2.5 percent according to the recent populationcensus and higher accordingto the Social Sector Strategy report, is extremelyhigh. The fertility rate of 7.3 per woman places Uganda amongcountries with the highest fertility rates in SSA (the other countries are Ethiopia 7.5; Cote d'Ivoire 7.3; Yemen 7.7; and Malawi 7.7). With such a high fertility rate, the averageUgandan women would spend about 22 years of her life with a child under six, as compared to only 10 years for the average Thai woman." Projections made by the PopulationSecretariat of Ministryof Finance and EconomicPlanning indicatethat if the fertility rate of 7.3 per women continues during the period 1988-2018, recurrent expenditures on primary educationwould climb from U Sh 9 billion in 1988 to U Sh 31.5 billion in 2018, in constant 1988 prices; reducing the fertility rate to 4 children per woman during the same period would cut these expendituresby more than 50 percent. Similar projectionsfor expendituresin primary health care indicatethat the Governmentwill save about U Sh 30 billion over the same period if the fertility rate is reduced from 7.3 to 4 childrenper woman.

7.14 The growing populationhas increasedthe demand for charcoal and buildingpoles and this has resulted in an increasein encroachmenton the forests. In certain areas, populationpressures have resulted in an intensificationof farming resulting in soil erosion and stagnating yields. Population pressures, supplementedby several years of war which resulted in a breakdown of government control and abandonmentof natural resource managementand policies have taken a further toll on the environment.

7.15 The Governmentof Ugandadoes not have an explicitpopulation and family planning policy and lags far behind other SSAcountries in encouragingits citizensto have small, manageablefamilies and informingthem of methodsof doing this. It is therefore imperativethat the Governmentaccord high priority to developinga nationalfamily planningprogram and closely monitorprogress towards decreasingthe presently high fertility rates.

Development of the Rural Areas

7.16 Uganda is a predominantlyrural society, with a high dependenceon agriculturefor income and employment,and the poor are largely concentratedin rural areas. Consequently,emphasis on agriculturalresearch and extension, rural roads and rural water supplyshould be areas of high public expenditurepriority in order to impact positively on poverty reduction. As has been discussed in Chapter 5, Ugandanagriculture is characterizedby extremelylow productivitywhich can more than double if adequateprovisions are made for effectiveand sustainableresearch and extensionsystems. Upgrading existing technology through adaptive research and supplementing this with efficient

Effective Family Planning Programw,The World Bank, 1992,p.2. 114 Chapter7 agriculturalcredit, marketing, input distribution, and other support services will go a long way in facilitatingagricultural growth and reducing rural poverty.

7.17 Despite significantprogress made over the last few years, a large proportion of the 20,000 kms of feeder roads in Ugandacontinue to remain in a state of acute disrepair. The rehabilitationof feeder roads should be a high priority as this will facilitate the inflowof essentialfarm and nonfarm inputs and the outflow of farm output and will acceleratethe process of the monetizationof the rural economy. Moreover, it will also encourage an expansion in the area under cultivation.While full rehabilitation of the entire feeder road network will not be possible as a result of economic and institutionalconstraints, even over the medium term, the rural feeder road rehabilitation program shouldbe accordedhigh priority and shouldreceive real increases in annual fiscal allocations.Where rehabilitationcannot be justified, the roads shouldbe kept passable by making spot improvementsso that smaller traffic can access the roads at all times. This is particularlyimportant from the poverty reductionobjective in that a significantamount of produce is transportedvia bicycle and motorcycle, particularlymatoke, milk, poultry and charcoal.

7.18 At present, only about 20 percent of the entire populationand about 15 percent of the rural populationhave access to clean, safe water. While the Govermnent's long-term goal for the water and sanitation sector are well defined and seek to provide safe water and effective sanitation and health education for all who can be practically serviced, based on at least full cost recovery of recurrent costs, no formal strategy has been adopted to bring this about in any determinate time frame. Moreover, allocationof resourcesto the ministrydealing with water and sanitationhave been extremelylow historically. For instance,in 1990/91and 1991/92,the share of ministerialexpenditure (capital and recurrent) for water and sanitationas percent of total expenditure,was only 6.2 and 8.2 percent, respectively;these shares are extremelylow when comparedwith other neighboringcountries in SSA. Furthermore, the problems have been compoundedby an extremely weak institutional structure characterizedby poorly remunerated and motivatedstaff.

The Provision of Safety Nets

7.19 As discussedin Chapter 2, among the poor there is usually a segment of the populationwho need special assistance because they may not be able to take advantage of 'regular' poverty alleviationmeasures or because these measures may not be enough. Social safety nets are intended to fill this gap. They includeassistance (e.g., incometransfers) to those whose poverty derives from a chronic inability to work (too old, too young, or handicapped). They also include temporary transfersto those affectedby naturaldisasters or economicrecessions. By their very nature, therefore, safety nets must be targeted. But targetingcan be difficultand costdy,as it requires careful monitoring to ensure that the assistanceprovided reaches all of those for whom it is intended and nobody else. Indeed, the administrativecost of targeting may be so high as to be unjustifiable. The following sections attempt to assess the effectivenessof existing safety nets and make suggestionsfor future interventions.

7.20 Generallyspeaking, all people falling in the categoryof the core poor are likely to be in need of special assistance, as they are unable to secure adequate nutrition and access to basic social services. In some cases, appropriate interventionsmay enable them to break the cycle of poverty within a relatively short period. For example, landless peasants may be given land and farm implements, and become self-sufficientwithin a couple of (good) cropping seasons; similarly, appropriateassistance to retrenchedcivil servantsand demobilizedsoldiers shouldenable them to find Chapter7 115 new ways of earning a living. But in other instances,the assistanceneeded is likely to be long term, as in the case of orphans, or permanent, as in the case of the severely disabled.

Keeping Afloat

7.21 Poor and vulnerablepeople are constantlyobliged as a matter of routine survival to devise their own coping strategies. This is particularlytrue of situationsof increased stress and risk, such as those faced by the groups describedearlier. They have no choice but to find their own solutions, becausethey can not afford to wait for assistancewhich may or may not come from outside. It should not be surprising, therefore, that in Uganda, as in many African countries, extended families have traditionallyrepresented the main safety net. This safety net is most efficient in the rural areas, but reaches also into peri-urban and urban areas thanks to strong urban-rural linkages.

7.22 The role of the extended family in securing survival is best exemplified in the way communitieshave reacted to the orphan crisis. Throughoutthe country, orphaned childrenhave been taken in by relatives, to the point that finding a familyonly caring for its own offspring is becoming a rarity. This is true in urban as well as in rural areas, and among all socio-economicstrata. 'In Rakai and Masaka children can be found with their paternal grandparents, uncles, aunts, and with their single parents, and maternal relatives. These established practices are now being used extensivelyand there are even examplesof unusual and newly-emergingadaptations of this kind of child care such as divorced stepmothers and neighborsvolunteering to look after children upon the death of parents.""2' The fact that in Uganda the number of street children is not particularly high for African standards (although it is growing) in spite of the large number of orphans indicatesthat communitiesare doing a commendablejob in coping with the situation.

7.23 A number of factors, however, appearto threaten the continuedeffectiveness of such systems. First and foremost is the general impoverishmentfelt by people who find themselvesstruggling for survival and unable to offer assistance to their less fortunate relatives. In rural areas, such impoverishmentappears to be caused by a combinationof low agriculturalproductivity (because of lack of technologyand labor shortage), a rise in the price of basic items, and localizedland shortages. In the urban areas, householdsare physicallyseparated from their larger families, and difficulties in communicationand travel may be daunting. Should distance not be a problem, dependenceon a (generally meager) wage has made it harder for householdsto afford responsibilityfor members outside the nuclear family. The AIDS pandemic has exacerbated all of the above. As the most productivemembers of societydie leavingbehind their childrenand parents, the survivingadults find themselvesunable to care for an ever increasingnumber of dependents.

7.24 When the burden becomestoo heavy, two reactionsare possible. One is to literally give up, as in the case of the man in Rakai who abandonedhis whole family when he realized that he had nothing to feed them. These cases are very rare. The most common reaction is to lighten the burden artificiallyby concentratingassistance selectively. As food and other resourcesbecome scarce, those outside the nuclear family are increasinglyneglected. As a farmer put it, 'if the choice is between feeding my children or my husband's mother, I will feed my children". In spite of the traditional respect accordedto older people, it appears that the elderly are the first to be left behind, but there are also reports of orphans who were refusedby their indigentrelatives. In the poorest areas, widows

121 Alden,John, GeraldSalole and John Williamaon,Managing Uganda's Orphwu Qis, repoitprepared for USAID, July-August1991, p. 26. 116 Chapter 7

,,.'.'..,, .., -.'.',.'..,''......

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... *......

wi ge{=~~~~~~~...... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ .,,.,,,.,...... :. m. b .. sty . - i- . --:- :: .... . a:v.:-i:: . ... d:b ...en, who... ha... the.:

wor t~ierhiea dAysuvEa4wexek,,w bit thelh a*dollgreUjIi that W lp *oedit nrejbca r ess OtILI depei: enbo the orpir paup bi--; sobgta hy aet-hmevs pi nt w,bc-om ...... ;......

o dstuon.t oastyearweveny eldpedrs..olw wh a teat. r w " ca.s. .. .elt heroud hosioldrear e abenoloffer cona,soiterefohrs wsisltancdb te reood ctivs,obecausoe t odeirca inot

seriousJmenoug tolawaran dio kvd"~vertinscarce resureswy fromwhtheepi nucelear, faha.t hils noclar whtiwhenthigareshodifficuLno teape wnethol wthiselwctibecaussisanwemhelpcheanism,hericaho tnanykasei theialaeneraise relinutishb the tirvatinyoblgatipots. only in kextremcaases.e lTkespoint s thuattherei heaso eivesto tat extreme caseswoadre bodcomhl ingmorte mmon,as wtnahessor to isisent. rehord

7.25 Peorpleinhery urbn arwasn arelpincrlei~asndlhexperi eakingdoa worknsith toopobe dpriey o whatoisthesuportebodmassistancein thoreshclid cse uth etensdwd of death.andfunerals.marking famiy e thei VepiotalnvrbdyhshdtaMon or.tt theinssltivessistane achiev common... obecivs mecanism,whcheor oavreyfifra,osuh. swelprtetono inan cse,'s

ChristineObbo, Women, Children~~ ... a...i...... n...e....Hasn ndMihalTwdde,ad. avndhi ..children.....ngn artvtesN da Inomu.ain...s.lnearcnsieed... Ka.al...... 00) bxcigfe asstnebyrelaties,for becaustethelircaise istenotbl sertioletenoughre tor warat dficeringlscadricersucs pawayy froplmenthe nulare ofaiymot cisvno lerarts

lrikely, whtbeal qIt sbective.. Boruptitnisceaihtsti a roleactiondictated Uadesperatin by aostUgandanshv wrouldenerll resinqus eirthge .tb nil obligaionsg nl ihnomexteecae.Thnono iseng that il thereisor reton"ua ddibelev to thatetemecsesoo arebeomievng more ommon aees)witnessedng byr insistentreport ofx thedetiutowngo manyiecold perSaons. 51 erodgrsi sdcmetr ntesot

supp ortuaeyfromssacepoiddb the extended family.Vrulyerboyhsadtresor ofavaitye o inforemal,eor evemniegal,atvtes.mechncomes. eanedstpatby exacting feesfrservices thadtiothe pomublictiwostensibl eugntildatofor fexape,oruarffcall coxednpries partly suppalementgetherwagteson ofmotcii servanto.v frsietnd,wha shallof deat?'.. Corrfuptls,onis aclctieve disease injeUgandatuha mostwlprtcitizens hav prblm passingrjubo,gement Childrc and emergving phge'nHomgeno isetht ofateenagegirhals Twholc resort

Buf8 freaganda, ple uraln Pcmuite would becalepogthr9yth-sud1f00du)t.gv Chapter 7 117

road maintenance. Communal systems of labor sharing are frequent especially in the Northern districts, and are known by different names depending on the type of labor performed and on the region-'rwot kweri", 'rwot awor", 'kalulu', "awak", "amuti", etc. (Box 7.1). Usually,they consist of a group of farmers (often of the same gender) who will get togetheron a regular basis to work on the field of each one of the members in turn.

7.27 In the absence of in-depth studies, it is difficult to quantify the contribution of these arrangements to indigenous social security systems."2 It appears, however, that communal labor systemsplay a very importantrole, because they enable poorer families to share labor, agricultural tools and some times food (generally'superior' food rich in proteins) with more fortunatefarmers. As labor shortage and lack of access to even the simplesttechnology are two of the main causes of poverty, it is clear that a system whichgives the poor regular access to a pool of labor and tools acts as a powerful safety net. What is more, labor sharing clubs are not only an effective redistribution mechanism of key resources, but also a form of insurance against illness or other forms of incapacitation,because they guaranteecontinued agricultural production-and therefore survival-for their members.

7.28 Other community-basedsafety nets have started to operate thanks to the initiative of RCs. While the record of RCs as effective communitymobilizers is mixed at best, there is no doubt that in many instancesthey have managedto organizeyouth clubs, womensavings groups and other types of associationswhich function either as a safety net for the members themselves, as in the case of women's groups, or as a charity institution for the destitute, as in the case of youth clubs lending their labor to widows and old people (Box 2.4). There are also reports of savings associationsbased on occupation,e.g., among school teachers and among city prostitutes.

7.29 The positiverole that RCs have the potential to play should not be underestimated.The deep emotionalscars left by the horrors of the civil war and the disastrous spread of AIDS are manifesting themselvesin withdrawalfrom communalactivities, and generallyspeaking in lack of interestin life. Indeed, the majorityof NGOs consultedfelt that the biggest obstacleto poverty alleviationis people's lack of hope and the breakdownof traditional solidarity networks. In this context, the RC system, particularlyat the lowestthree levels, can provide a much needed catalyzingeffect by forcing people to come out of their shell and search for common solutionsto shared problems.

Is There Anybody to Help?

7.30 Responsibilityfor assistance to the most vulnerable groups falls mainly on the Ministry of Labor and Social Welfare (MLSW)'2 ' and the Ministryof Local Government(MLG). The former is mandatedto assist vulnerablechildren (i.e., orphan, displaced, delinquentand abused minors), and also to handle distress cases, such as victimsof naturaldisasters and abandonedmothers and children. The latter is in charge of communitydevelopment in general and of disabled people (but disabled childrenare under the jurisdictionof the MLSW). Displacedfamilies and Ugandanrefugees returning from abroad are handled by the MLSW, while foreign refugees in Uganda are handled by the MLG.

Fora quantificationof 'labor tmnafen' and a discussionof theirimplication in an Africancontext mm Jocchin von Bmun, 'Social Securityin Sub-SahaanAfrica: Reflectionson Policy ChallenSes",in E. Ahmnd at ml.,Sodal Secuiy in DevelopingCownhres, Oxford UniversityPress, 1991.

14 The Miniaryof Labor and Social Welfareis the reult of the union of the Minizty of Labor andthe Miniatryof Rehabilitationand Socid Welfare. The later wa establishedin 1979to cope with the impactof the civil war. 118 Chapter 7

7.31 Bothministries have staff in each district, with a district officer in the district capital and field personnel (social workers, communitydevelopment assistants) to cover the rural areas. In reality, many posts are vacant (either de jure or de facto), and personnelare concentratedin towns. Lack of transport makes it impossibleto do field work, so only those able to report to the town office receive assistance. Even then, little assistancecan be provided in the absence of a proper operatingbudget. District welfare officers, for example, have an imprest account of U Sh 70,000 per month for operatingexpenses. This sum shouldinclude an emergencyfund for public assistance,but most often it has to be spent to keep the office running.

7.32 In the end, it tends to be up to the resourcefulnessof individualcivil servantsto find ways to provide effective assistance. That is, it is a question of contacts with NGOs who can provide the needed support. Indeed, workingwith or through NGOs is so commonfor staff of MLSW and MLG that, without exception,those interviewed(both in headquartersand in the field) admittedthat without NGO's help they would not be able to do their work. It must be recorded, however, that personnel in these two ministries appear to be very competentand surprisinglydedicated, given the difficult conditionsunder whichthey operate. One can not but sympathizewith the frustrationthey expressed at their inabilityto carry out their mandateeffectively.

7.33 A couple of examplesmay illustratethe degree to which the Ugandan Governmentdepends on NGOs to provide safety nets to its people. When a displacedchild comes to the attention of the social worker, there is seldom money to pay for public transport to take the child back to his/her relatives, let alone to pay for meals to be eaten while traveling. In any case, often children are not sure of where to go or public transport is not available, so they need to be accompaniedwith a vehicle. But usually the ministryhas either no vehicles or no money for fuel. At this point, NGOs and donors are asked to step in and provide needed transport. It is estimated that about half of the operatingbudget of the Rehabilitationand SocialWelfare Departmentcomes from UNICEF and Save the Children.

7.34 The situation is similar for the handicapped.As it was mentioned above, disabled people (excludingchildren) are the responsibilityof the CommunityDevelopment Department. Each district, therefore, has a communitydevelopment assistant in charge of such cases. But with no transport and pitifullylittle money, all the communitydevelopment assistant can do in responseto requestsfor help is to refer cases to NGOs, hoping to find one able to provide fundingfor the necessarytreatment and aids. This happens in about 10 percent of the cases-90 percent of Uganda's disabled will never receive any assistance.

7.35 As a result of the inabilityof governmentfinances and institutionsto provide safety nets, a large number of internationaland local NGOs (sometimesbacked by donors) have steppedin to give much needed help. While it is difficult to quantify the proportion of assistance provided by NGOs and by the Government,there seemsto be wide agreementthat NGOs are deliveringabout 90 percent of the assistance. The quality and type of safety nets offered by NGOs vary considerably, and it is beyondthe scopeof this documentto offer an inventory. Someare very localizedand specific, others cover entire counties with integrated development programs.'" But the sheer number of NGOs (over 500 registered) and the fact that they appear to have taken responsibilityfor the delivery of almost all assistanceshould not be interpreted as an indication that help will be there for those who

us FuutherdiawuWon of the moleof NOOsis prswAtedin Chpter 8. Chapter 7 119 need it-regardless of the quality and scope of the services offered by different NGOs, only a very small proportion of those in need are reached.

7.36 In addition, NGOs have a tendencyto concentrateon some vulnerablegroups (generally,the most visible or the most "attractive"for sponsors) and in some areas. For example, last year there were 32 NGOs working in Rakai district," but only three NGOs active in Kitgum and two in Karamoja.Similarly, there are over 70 NGOs and donors focusingon orphans, but we could not find a single organization concentratingon the elderly. Thus, it appears that whatever help is available from non-governmentalsources is spread unevenly.

7.37 Asking whether it is preferableto have the bulk of public assistancedelivered by the private sector rather than by the Governmentmay be a rhetorical question. NGOs and governmentofficials alike agree that it would be unrealisticto expectthe MLSW and MLG (or any other ministry, for that matter) to be able to take primary responsibilityfor the deliveryof assistanceto the most vulnerable groups. The Governmentnot only lacks the resourcesto implementdirect interventionprograms, but lacks the flexibilityto respond rapidly to emerginglocal needs. NGOs are filling a need that cannot be met in any other way.

Safety Nets for the Future

7.38 What then shouldthe role of the Governmentbe? There seems to be wide agreementamong civil servants and developmentworkers that the MLSW and MLG (and other ministries)should focus on three responsibilities,i.e., providingpolicy guidance,coordinating and monitoringNGO activities, and ensuring that governmentbureaucracy does not get in the way.

7.39 The recently completed NationalPlan of Action for Children is an exampleof the key role the Government could play in ensuring that national policy is sensitive to the needs of the most vulnerable groups. The Plan is the result of a collaborative process which involved different governmentbodies and NGOs and will carry a very importantmessage, as it will be the first case in Africa of a complete rewrite of laws for the rights of children-and inevitably,of mothers. The point here is that, while it is true that the Ugandan Government Oike many other African governments)can do little to provide concrete safety nets, it can at least prevent some people from ever needing safety nets by safeguardingtheir rights. In this vein, a revision of inheritance and marriage property laws to strengthen the position of women and children would represent a substantial contribution, at a relatively low cost, to the welfare of orphans and widows, and of divorced women and their children.

7.40 The need for Government to coordinate and monitor NGO operations can hardly be overemphasizedand will be discussed further in Chapter 8. While most NGOs are doing excellent work, the activitiesof some local NGOs (the so-called "briefcaseNGOs") appear to be determined by politicalor financial ambitionmore than by the desire to help the needy. In other cases, intentions may be good, but a limited understandingof the situation or mere incompetenceproduce projects which are actually harmful. A case in point are orphanageswhere childrenhave been found to live in unsanitary conditionsor without proper care. Not only were these orphanagesunable to provide appropriate assistance, but their very existence undermined traditional coping mechanisms, as

John Alden d al., op. cit., p. 36. 120 Chapter 7 communitieswill tend to slacken their efforts for self-relianceif an institutionalalternative is readily available.'2

7.41 Given the proliferation of NGOs, and the doubtful nature of some of them, the recent establishmentof some umbrella NGO organizations(UCOBAC, NUDIPU, etc.) to coordinate the activities of all NGOs operating in a specific field should be welcomed. But these umbrella organizations cannot be a substitute for government leadership. The Government must take responsibilityin ensuringthat the efforts of donors and NGOs produce the best possible results given the means available.This requires coordination,close monitoring,the settingof standardsand, where needed, guidance.

7.42 The resources presently availableto the MLSW and the MLG are not enoughto allow them to carry out effectivelythe responsibilitiesdescribed above. Notwithstandingthe fiscal constraints discussed above, their budget needs to be increasedto pay at least for the travel and communication necessary for effective monitoring. Coordination would be facilitated by the creation of a computerizeddatabase on projects intendedto provide safety nets, includingtheir scope, location and target group. This would avoid wasteful duplication,help spread scarce resources more evenly and identify neglected areas for future interventions.An in-depth study of how effective NGOs operate would make it possibleto draw general guidelinesfor implementationand allow the Governmentto establish (and possiblyenforce) standards for the operationof NGOs. For example, the MLSW and MLG are keenly aware of the importanceto base outside assistanceon existing coping mechanisms at the communitylevel, so as to foster self-reliance.Identifying the key features of NGO projects which have successfullyused such an approach would enable the MLSW and MLG to capitalizeon the experienceof these projects and to guide other NGOs in a similar direction.

7.43 Better coordinationamong governmentbodies is also essential. While the MLSW and the MLG have the major responsibilityfor the provisionof safety nets, their work should be facilitated and complementedby the concerted efforts of a number of other ministries. The computerized database on NGOs mentionedabove, for instance, could only be created with the assistanceof the NGO Board in the Ministry of Internal Affairs. Similarly, a revision of existing laws to strengthen the position of women and children necessitatesclose collaborationwith the Ministry of Justice, especiallyfor its enforcement.The provisionof basic social services which are presently beyondthe reach of the most vulnerable groups is primarily the responsibilityof the Ministries of Health and Education,while the Ministryof Informationand Broadcastingcould play a critical role in changing societal attitudesand behaviors which contributeto, or even cause, the misery of the neediest (e.g., shunningof the handicapped).

7.44 Regardlessof the efforts of Governmentand NGOs, the bulk of the work will still have to be done by the communitiesthemselves. Communitymobilization is therefore imperative."2The question is how to bring it about, especially because the capacity to mobilize people tends to be determined more by personal qualities than training or position. Some NGOs have employedvillage development workers with good results, while others have relied on community development

127 For examplesand furtherdiscussion, see John Alden ct al., op. cit., and Abby Nalwanga-Sebinaand James Sengendo,Ophaned and Disabkd Jaildrcninlawero andKabak Districts,and in UgandaCMild Care Iwtistuions, mimeograph,December 1987. us Too often mobilizingthe communityturns out to mean mobilizingwomen. Giventhe altdy very hcavy demands on womentime, any effortto mobilizethe communityshould be mindfulof women'stight schedule. Chapter 7 121 assistants(whose work is facilitatedby providing transport or other perks). Indeed, experiencefrom other poor African countriessuggests that one of the best ways of mobilizingcommunities to improve their own welfare, and to do so in a sustainablefashion,is to work through local people reputed for their honesty and their leadershipskills."' Teachers, traditional chiefs and RC1 members are likely candidatesto play such role, providedthere are mechanismsto prevent them from taking advantage of the increasedpower they will inevitablyacquire. Biannualpopular electionsof RCs are one such mechanism,as demonstratedby the considerableturnover observed in the second round of elections.

7.45 Self-help initiativesat the communitylevel can go a long way, but where absolute poverty is widespread, little can be achievedwithout extra resources. In extreme cases, it is a question of handouts such as resettlement kits. But in most cases, the challenge will be to find a way of reinforcingthe capacity of communitiesand families to care for their less fortunate members while at the same time providingoutside help for those who are likely to fall throughthe local safety nets. Integrated projects run by a number of NGOs have chosen an approach whereby communitiesare mobilized and their efforts complementedby the provision of in-kind inputs, either material (agriculturaltools, blankets, iron sheets)or technical (training,preventive care). But NGOs can only reach a small minority,so an alternativemechanism for giving communitiesaccess to extra resources must be found.

7.46 Becausehardship cases are generallywell knownin the village, the communityitself is in the best position to identifythose who most need help. Thought shouldbe given to the establishmentof a village developmentfund to be used in projects that would increase the effectivenessor coverage of safety nets at the community level, either by assisting the disadvantagedpeople directly (e.g., giving hoes and fertilizers to widows)or by enablingthe communitymembers to provide assistance (e.g., giving hoes and fertilizersto those who take responsibilityfor abandonedold people). The fund shouldbe made availabledrecly to the communitiesand accountabilityat the local and nationallevel shouldbe firmly established.The proposed CommunityAction Trust appears to be conceivedalong these lines and deserves closer consideration(Chapter 8).

in A word of cautionis neededhere. Althoughcommniity effos armrecommnuded, they shouldnot be purued at theexpene of people'swodk for theirown sutenee. We heardcomplains of peants who hadto contributetume to commuiityprojects in spite of the fact that their ficldsneoded immediate atetion. DAtthe RCs had premised to the NGOr,and refusingto participatein the communalwork would have meanthaving to pay a fine.

Institutional Framework for Delivering Essential Services

'Instkutonal developmentconcerns the creationof organizadonalcompetenaes and values thaxare fictional to development.And jf developmentu to be accompaniedby poverty reduction, those organizatinal competencesand valuesnuut also be attwuedto the needsof thepoor. (Lawrenwe Sabnen, 'Institutional Dimensionuof PovertyReduction', PRE WorkingPaper 411, May 1990,p. 16.)

Background

8.1 The current situation in Uganda offers unique opportunities as well as challenges. The aftermathof decades of war, continuingsecurity problems in some areas, AIDS, and drought magnify the general problems of poverty. Despite the advances made by the Government to improve the security situation, the rural poor (indeed, the rural populationin general)receive few, and in many cases, no services from the government administration.While the RC system is bringing rural Uganda back into a coherent polity, there is still little government infrastructureon the ground. If the poor of Uganda are to make advances in the medium term, capacity must be strengthenedor created to provide them with key social services. Buildingsuch capacity is dependenton the needs and characteristicsof the ultimateclients-the citizensof Uganda,the communitiesin whichthey live, and the institutionswhich already serve them, or mightbe expectedto serve them. These institutions includeorganizations such as governmentagencies, NGOs, and commercialservice providers, as well as institutionsnot embodiedin organizations,such as customarylaw, and shared labor and revolving credit associations. These inter-related institutions form networks which can help or impede communitiesin reducing poverty.

8.2 Given the urgency and magnitude of the problems faced by the poor of Uganda, the Government has recently shown interest in a model of developmentoriented toward government support for communityinitiative. Supported by the Government'sstated objectiveof decentralization, the emerging model is one where communitieswill be given increasingresponsibility to formulate their own developmentactivities, which will then be supportedby the district administrationswith the help of NGOs and other communitybased organizations.The objective of this chapter is to lay out some of the issues which need to be addressedto make this model work, and to suggest some initial steps.

Governmental Istitutions

8.3 The Ugandan administrationcan currently do little more than it has been doing: pay minuscule salaries (often late), provide equipment and supplies occasionally, and in far smaller quantitiesthan needed, and pay allowancesfor meeting attendanceto Kampalaand Entebbe based civil servants. There are staff on the ground, but the Governmentis without the means to pay them a living wage or to provide fuel and vehiclesfor them to visit their 'clients."'3 When they do reach the communitiesthey are meant to serve, it is almostalways through donor projects-with donors in

130 -Clientsc will be uced here inud of "beneficiafies," which hex a connotation of paive receiver of goods and ervices. 124 Chapter 8

Uganda consisting of both official bilateral and multilateral aid agencies and of non-governmental organizations.

8.4 Many civil servants at all levels have been able to survive only by supplementingincome from governmentjobs. Often they do this through outside business activities, or through having salaries supplementedby donor and NGO projects. The 'tipping' system also seems extremely widespread.In effect, these tips represent user fees. Now, in addition to the civil service, there are reports that some RC members also must be "tipped' for various functions.

8.5 The net result is that many staff are unable to perform their duties on a regular basis. This system will not end among civil servants at least until they are paid a living wage. Despite the resulting low morale, NGOs and donors who have worked with Ugandan staff in the field indicate that most are eager to do their jobs, if only they can be assured of a reasonable salary in return. When field staff are drawn into adequatelyfunded projects, they have proven, in most cases, to be competentand dedicated.

8.6 While no reliable informationis availableon how many communitiesare being served in this way, observers agree that most of rural Uganda is not systematicallyreceiving basic services that contribute to poverty reduction. In recent policy documents, the Government has emphasized refocusingthe institutionsof Governmenton basic poverty-reducingor -alleviatingservices such as community-basedprimary health care, clean water and basic sanitation, primary education, feeder roads, and integrated agricultural extension services, but Government presently and for the foreseeable future lacks resources for staff and equipment to implement this policy change. Considerableemphasis must be put, for the short and medium term, on projects and programs of official donors and NGOs.

8.7 Outside agencies,both official and NGOs, have been implementingservice delivery projects directly through parallel systems, such as project implementationunits, often using their own personneland equipment.These are managedoutside the mainstreamof governmentministries so that donors have more control over the outcome. Often these agencies are motivated by their need to account, in turn, to criticalstakeholders: their own financiers.However, with such a system in place, government agencies and staff are shut out from the learning experience they could be getting in planning,managing or producingthe services. IndividualUgandans may acquire experienceand skills while working for these projects, but institutionalproblems cannot be solved simply by improving the skills of individuals;instead, the patterns of relationshipsand behavior (includingaccountability) at the organizational level must change. This is what is missing when quasi-independent implementationunits are set up.

8.8 Another problem has been the fragmentationof government ministries: donors (again, includingsome NGOs) fund vertical interventionswhich ultimatelywork against focusingministries on integrated efforts to relieve poverty. The result has been a series of unconnectedservices offered to communitiesby different individualsin an uncoordinatedway; the total may be less than the sum of the parts, rather than more, as some of the interventionsmay work against each other. Rather narrow objectivesmay be achieved, but at the expenseof even more basic concerns.'31

131 For example,the French fund a FrenchTeachers Program and FrenchTeacher Training. While this might be an imporant initiative,it does not noceauily reflectthe prioritiesof a countrywhere there are no texts in ruralschools, and whereteachers' salaries have to be paid by communities. Chapter 8 125

8.9 An unintendedeffect of projects like the Program for the Alleviationof Poverty and the Social Costs of Adjustment (PAPSCA) may have been to increase the perception of poverty alleviationas somethingwhich is not done through the core functions of line ministries. Originally meantto be a focal point within Governmentfor the coordinationand monitoringof poverty reduction work, PAPSCAhas come to be widelyregarded not as part of the Ministryof Finance and Economic Planning but as an autonomousWorld Bank project (even though other funders are also involved). Its physical location in a suburb of Kampalaaccompanied by a general lack of informationabout it (among other govermnentbodies as well as NGOs) reinforces its isolation.

8.10 While PAPSCA was set up with the intention of facilitating anti-povertyefforts, the long, slow procedures and detailed reporting required have made it difficult for NGOs to work with PAPSCA. As implemented, it has underscoredthe fact that, at present, providing essential social services e.g., in the areas of primary health and education, is not the core function of Ugandan ministries, but rather the province of a few scatteredprojects. The design team, includingthe NGOs on it, each put in a piece in their area of action; the few donors who contribute to the project have completed the 'Christmas tree" effect by funding projects in their area of interest. The result is certainly not a national program; rather, it is a marginal collectionof projects of varying quality spread far too thinly to have a broad impact on poverty.

8.11 The overall result of the "projectization'of ministrieshas been that staff are gettingpaid and doing work, but the senior ministry officials are little involved in prioritizing, deciding the mix of projects, or in realistic strategicplanning, and they have no responsibilityfor seeing that the complex system moves overall in the desired direction. They also get very little experience in managingthe budget, whereas the project directors have much more control of resources. In this context, there is little incentive for governmentplanning to be anything more than devising a "wish list." While the arguments for donors keeping control of the money can be compelling for the donors, and for the sake of the desired programs, the problem of governmentincapacity will never be solved if this state of affairs continues.

8.12 There are signs that this trend is being reversed both within Governmentand among donors. The Government has recently embarkedupon a comprehensivecivil service reform program aimed at "professionalizing"the civil service and the beginning of movement toward strengthening governmentministries to carry out core poverty alleviationfunctions has been perceptible. Among the donors, UNICEF, for example, has consolidateddiverse projects into one project called "Capacity Buildingfor SustainablePrimary Health Care," to support the Government'sdesire to put primary health care at the core of the ministry's functions. Ultimately,Government and donors will have to agree togetheron the priority actionsfor investmentin the productivesectors and in human resources, as well as on how to fund safety nets in the interim. Better coordinationof aid to Government (includingaid from the large NGOs) is a prerequisiteto long-termpoverty reduction.

Self-Help Groups

8.13 At the same time, Governmentrecognizes that self-helpgroups have often organizedto meet their own most pressing needs; for example,communities have kept schoolsrunning through parent - teacher associations (PTAs) which have supplementedand paid teachers' salaries and overhauled facilities, and families have expanded to take in AIDS orphans. Policy documents emphasize the Government'sdesire to design a systemwhich will support these communityinitiatives. A major step in this direction is the decision to decentralizemany governmentdecisions to the district level, and to redeploy much of the administrationto district and local field offices. Certain policy documents 126 Chapter 8

have also referred to Government's intention to establish policy and mobilize resources for the provision of services, but to arrange with other bodies to actually deliver the services. The Govermmentcan contract with either voluntary or commercialsector organizations`2to actually run the services. Indeed, the Government increasingly recognizes the important role played by the voluntary sector,13 whether organized into community-based groups or non-governmental organizations.The next sectionswill discussfeatures of communityself-help groups and NGOs which affect the degree to which their capacityto serve the poor can be increasedin the short and medium term.134

8.14 Community organization has been responsible for keeping some services-most notably primary education-running in Uganda. The developmentalresources represented by communities have always been underutilizedbecause developmentwas conceivedof as coming from the outside, 'modern sector. The Governmentnow recognizesthat it should support the efforts of the ordinary people of Uganda, rather than expectingthem to wait for government-initiateddevelopment activities, which may be long in coming and unsuitablefor prevailing conditions.

8.15 Those communitiesthat have received support services from Government and donors have often received them through vertical projects, with one agent prepared to offer only a narrow, if essential, service-vaccinations or drugs or nutrition, advice on livestock or on crops. Both Governmentand donors have stated their intent to move toward integrated service delivery, where the same person could be a resource for all basic health servicesor for all of a farming system. This would ensure that separate interventionsdo not work against each other, and that clients can have "one-stopshopping."

8.16 The Government's new position of supporting community action would ensure a more demand-drivensystem, with communitiesempowered to say what they need. The resultingprojects would be better adapted to their micro-enviromnents.At the same time, community-managed developmentprojects, whether providing communityservices or upgrading returns from economic activities, are labor-intensive and require time. Community health management committees, for example, have a benefit (over the long term, better health means less time spent caring for the sick and more people engaged in economicallyproductive activities which improve family quality of life) and a cost (membershave to attend meetings,work out arrangementswith health care providers, and manage the productionof health services).

Non-Governmental Organizations

8.17 Governmentis increasinglystating its commitmentto support communityself-help, but NGOs have long experience in doing so. The state of knowledgeabout NGOs in Uganda is typical for

Together, theae are often refened to as the "privaw mector,but sine. that term momntimeincludem and ometime excludes the voluntary metor, they will be referred to by the morc specific 'voluntary' and 'conumercial' wectori here. The voluntary ector imalso omtinem referredto am'non-governmental".

See for exanple the Health Three Year Plan, cited in the Social Sector Memorandum,or the National Plan of Action for Children.

134 There ir no legal defirition of the difference between themetwo kinds of goupr. However, infornmaly, there meem to be agreenert tht, in the Ugandancontext, comnunuitymeIf-heip groups arm temporary and infonrial; if they expect to become permanent,and if they have a fornrl trmture, they hould registeras NGOs. Chapter8 127 developing countries: we do not even klnowhow many development NGOs exist, much less how many communitiesthey serve, what kinds of activitiesthey engage in, how effective they are, and how much non-governmentalfunding they mobilize.There is not even an acceptedvocabulary to help distinguish between different kinds of NGOs (some organize charitable relief; others work with communities to implement community development projects; others supply credit or other developmentservices; somecarry out advocacyor policy-advisingfunctions; some faith-basedNGOs have been operating health and education services for decades, while newer secular NGOs are developingnew programs)." However, some general points can be made which have a bearing on what kind of partners NGOs can be in Ugandandevelopment.

8.18 In some circles, NGOs are respectedfor those qualities that represent NGOs at their best. Religious NGOs run the best medical facilities in the countryside, and supply more health care of better quality than the Government. Many NGO staff work long hours for little money because of their commitmentto development.They work in communitieswith groups that present the toughest development challenges: the disabled, AIDS orphans, war widows. Their work is based on a commitmentto communityparticipation and pluralism,rather than dependencyon Government.They are closer to the clients and know more about their needs than remote officials. They can tailor services to niches, whereas Government is set up to offer a standard package which may not meet all needs; they pilot new technologies.Their costs are lower than those of Government.

8.19 In other circles, it is suspicion,based on experienceof some 'bad NGOs, that dominates. Bad NGOs are fronts for private gain rather than communal benefit; they are private businesses seeking the NGO label to get tax exemptions;they are agents of external actors linked to political factionalismin Uganda; they don't have anythingto contributeto policy, since they don't understand the macro aspects of development;they mobilize money from outside but are not required to account for it; they work to benefit special interest groups, whereas the Government has the responsibility to ensure services and benefitsto all Ugandans.They have a reputationfor low-costservice delivery, but in fact there is little informationavailable on costs.

8.20 Both of these attitudes are based on experience;there are NGOs doing excellentwork, while others callingthemselves NGOs are incompetentand/or fraudulent.These are the extremes;probably most Ugandandevelopment NGOs are somewherein betweenon competenceand accountability,with hundreds who are at least trying to do good developmentwork. While NGOs can choose their own beneficiaries,those they choose are sometimesthe most vulnerable.

8.21 It is incontestable,however, that NGOs will be an importantelement of the poverty-reducing institutionalframework. The challengeis to develop a frameworkfor partnership amongGovernment (both administrationand RCs), NGOs and communitygroups to enhance competence,transparency and accountabilityin NGOs and communitygroups as well as Government.The recently established District (or National)Community Action Trust could be one such instrument(see below).

Coordinatlon between Government and NGOs

8.22 The Government of Uganda has set up registration and coordinationsystems to weed out fraudulent NGOs and strengthenlegitimate NGOs. The intentionhas been admirable, but there are

The Minity of Fin nd Plnning, with WorldBan* wppoit, is ng a dy to awew ame of doe quesion. 128 Chapter 8

now many governmentbodies involvedin NGO coordination,with no one having sufficientresources and staff to offer real assistanceto NGOs. The result is that rather than assisting NGOs, the system confuses them, with different actors each having their own requirements.

8.23 Ugandan NGOs are required to register with the National Board for NGOs, housed in the Ministry of Internal Affairs. InternationalNGOs first have to work out an agreement with the most relevant line ministry. To register, Ugandan NGOs explain their objectivesand plans to their local RCs. If they can get endorsementsfrom the RCs 1-3 in the area in which they intend to work, the applicationsthen go to the National Board through the district administrator. NGOs report that recently an extra step, not specifiedin the legislation,has been required. The district administrator, instead of forwarding applicationswith endorsementsfrom the RCs to Kampala, gives them to the Internal Security Organization,which goes back to the RCs 1-3. If no problems are found, the applicationis sent to the NationalBoard.

8.24 The Board consistsof a two members of the public and one representativefrom each of 12 ministries/governmentagencies." 3 ' Each NGO is required to present its organization chart, work plan, and constitutionto the Ministry of Finance and Economic Planning (MFEP), which has to certify that the NGO's program is in line with national priorities. It is recommendedthat NGOs take this material in person to the MFEP; those which are sent from the NationalBoard may languish in someone's in-box. In practice, then, the system works best if someonefrom the NGO can come to Kampalato hold the necessary discussionswith officials in MFEP, and often with the relevant line ministry or ministries as well.

8.25 After registeringwith the NationalBoard, NGOs may be able to obtain tax exemptionsfor imported goods and work permits for foreign staff. There are long delays in processing these, and they are not alwaysgranted. As a result of the systemof 'tipping" officis who provide service (such as signatures), NGOs may also need to make payments at any point in this process. When such payments are large-reportedly 10-15percent of foreign currencyexpected-NGOs inflate itemssuch as staff salary in their accountsfor donors, since donors will not supplymoney for "tips," and NGOs cannot afford them from their own resources.

8.26 Records on NGOs are kept in different locationsin Government. The NationalBoard keeps the applications,but work plans, constitutionsand organizationcharts, as mentionedabove, are sent to the MFEP. The Aid CoordinationSecretariat (ACS) in the Prime Minister's Office has been designatedas the lead governmentagency for NGO coordination,and it has periodicprogress reports which some NGOs send. Many NGOs are never heard from again at the nationallevel until it is time to re-register. They should,however, make annual reports to their District DevelopmentCommittee.

8.27 Both the ACS and the National Board try to monitor NGOs activities, the ACS through requestingregular reports and the Board throughoccasional field visits. Fuel, vehicles and allowance are necessaryfor the latter, and they are rarely available.The District Administration(DA) may also have progress reports or otherwisebe well informedabout what NGOs are doing in the district. Some DAs have held coordinationmeetings including NGOs; others have soughtto control NGO activities too closely, and have lost the cooperationof NGOs for these meetings.

16 See Non-GovernmenalOrgaizatiom Regidmton Sbtute, 1989; StatutozyInmt,menta 1990 No. 9, The Non- Governmenal Organizatin Regulations, 1990; Report of the Vice Pmident's Select Commritteeon Non- GovenmWentalOrganization, April 1992. Chapter8 129

8.28 With formal authorityfor NGO coordination,the ACS holds monthlydiscussion fora. The ACShas also tried to mediate in NGO-financierdisputes. The ACS seems to be generally considered by the NGOs as trying to offer assistance, even if this is limitedby resource constraints. Officialsof other ministries also have an interest in coordination: Ministry of Local Government in the coordinationof NGO activitieson the ground; MFEP in advising on how foreign currency should be used; the Ministry of Foreign Affairs in the programs of internationalNGOs and local chapters of internationalNGOs (such as the YMCA). The Ministry of Labor and Social Welfare may try to revive the National Councilof Voluntary Social Services, a body which used to have some of the functionscarried out by the ACS. In additionto all these agencies,NGOs may need multiplecontacts in the line ministries as well.

8.29 As of July 1992, 506 NGOs were registered, of which 104 were listed as foreign NGOs. Estimatesof the total number of NGOs in Ugandarange from 800 to 1,000; some may be registered at the district level without being registeredat the national level.

Creatinga Frameworkfor Collaboration

8.30 The institutionalchallenge facing Uganda is to create a framework in which Government, NGOs, self-help groups, and donors all make a contributionto poverty reduction. Uganda,like many countries, is a conglomerationof micro-environmentswhich makes it difficult to apply blanket solutionsto the whole country. Activitiesin one sector may be organized quite differentlyfrom those in another, e.g., a systemfor providingrural credit can be quite different from one providingprimary health care or education. The system that is ultimately needed is a "learning system", in which numerous small activities develop effectiveness and then efficiency before expanding."' These grassroots-levelactivities should be closelyadapted to local needs and preferences,and supportedby governmentpolicies creating favorableconditions for increasingproductive activities and investment in human resources at the local level.

8.31 Assisting CommunityProjects. In the short term, there is littlequestion that with Government and NGO capacityfor deliveringservices able to satisfy only a portion of need, communityself-help will be the vehicle for many communityimprovements. In line with the intentionto support self-help efforts of Ugandan communities,the 1992/93 budget speech introduced the idea of a District Community Action Trust (DCAT), a fund which would finance community projects throughout Uganda. Developmentefforts would not have to go through the long, complexdonor project cycle, and money could be released quickly for small projects. Communitygroups would be the primary users of such funds, but they could call on NGOs for technicalassistance if needed. The funds would be administeredat District level, and not necessarilyby Government.

8.32 There is experience in Uganda with this sort of assistance to small, local NGOs and community-initiatedprojects. The experience of the Small Funders' Network and some NGOs is particularly relevant, since many already provide this kind of support. The Community Action Program in the Northern Uganda ReconstructionProject will generate further lessons about setting

i37 See, anong others, David Korten, 'Community Organization and Rural Development: A Learning Process Approach,' PublicAdministration Review, Sept.-Oct. 1980: 480-511, and Larry Salmen, "ReducingPoverty: An bishtitonal Perspectve", Povertyand Social Policy Scries, Paper No. 1, ProgramDesign and Implementation, Washington, World Bank, 1992; Shaon Holt, "7heRole of Insttuions in Poverty Reduction: A Focus on the ProductiveSectors," PRE Working Paper No. 627, Washington,World Bank, 1991. 130 Chapter8

up a nationalprogram. The rest of the developmentcommunity needs to listen carefullyto and learn from these funds as these are the groups who have the experienceof funding communityprojects.

8.33 Typically, the fund is meant to respond to communitydemand. Communitiesarticulate their needs; the funder investigatesthe feasibilityof the project, and then provides materials and/or small amountsof money. However, as one of the officialsrunning a fund for smallprojects said, "We have concludedthat pure responsivenessis not responsible."Communities often present the kernel of a good idea, but need much assistancein making it into a workable activity. Even before getting viable project proposals, there is also a need to help communitiesanalyze their situation and prioritize actions. Getting involvedin this kind of communitypreparation is labor-intensive,both for members of the community and for the outside facilitators who are needed to help the process (supplying informationas well as process facilitation).The small funders' limits are set less by the amount of money they can mobilize than by the small numberof projects to which staff can give the necessary attention. A recent workshopon community-basedwomen's income-generating activities showed that much of the women's profits were investedin communityservices, but the women's own longer-term needs were neglected.

8.34 While the DCAT would make funding available, the problem of technical assistance to communitiesin decidingwhat to do remains. Many NGOs and smallfunders in Ugandaare convinced that it is a mistake to begin developmentactivities without prior preparation of communities,in particular to identifystrategic actions. In their experience,this preparatoryphase-which might once have been called consciousnessraising-can itself be slow. This stage requires outside facilitation, meaning that there must be considerablestaff on the ground. These efforts should be supportedby informationsharing so that communitiescan learn from successfulactivities in other places.

8.35 The Role of Infonnation Systems. Communityexperiments can fail if the policy environment does not support them. Policymakers need to know what is happening at the grassroots. But developmentcannot be only bottom-up,just as it cannot be only top-down;a mutuallearning process is necessary.Policymakers need to know how innumerablesmall farmers make decisions about what to grow, and what the effects of policy changesare likelyto be; communitiesneed informationabout prices, markets, improved health and nutrition behaviors, etc. One dimensionof learning systems, therefore, concernstwo-way (or multiple-way)information channels.

8.36 Policymakerstherefore have to learn: (i) what informationthey need; (ii) who has it; and (iii) what incentivesneed to be offered to the people who have the informationto share it. NGOs often have useful informationwhich is underutilized.Two typical obstacles to informationsharing arise. Sometimesadministrators do not recognizethat they could use such information,and that the NGOs have it. At other times, the Governmentand NGOs cannot agree on terms under which they could share information;the Governmentmay want the informationNGOs have, without offering anything in return. The institutionalchain has to get informationback to the communitiesthat need it, in ways that make sense to them. The RC systemhas been designedto be such a two-way system. While they have made much progress, more could be done usingmass media, especiallyradio, as well as face-to- face communications.

8.37 In addition to developing channels for sharing information, a learning system has content requirements.Given the gravity of Uganda's problems, and the sense of urgency, there is a tendency for those trying to alleviate poverty to carry out as many activities as possible at full speed; unfortunately, often without learning much in the process about the impact of these activities on individualsand on sets of people. Successtends to be measured in terms such as numbers of people Chapter 8 131 trained or crops planted, rather than enquiringwhat behavioralchanges have occurred as a result of the intervention.

8.38 Without knowing much about impact of particular activities, it is impossibleto cumulate learning through a system. Thus, while there are apparently numerous coordination meetings and mechanismsin Uganda,there is apparentlytoo little learningas a result. In some cases, waves of fads sweep through the development communityand are adopted with little examination; in others, a multiplicityof conflictingapproaches leads to divisiveness.The result is the same in both cases: little learning about what works.

8.39 The activityfrenzy has also led to a willingnessto fund projects involvingcommunity-based organizations or NGOs with little investigation of the legitimacy of such organizations. Accountability,to either funders or client groups, is not very advanced, with the exceptionagain of the RC system (three-yearterms, with communityrecall possible).

8.40 The result has been too little cumulativelearning about what works. The informationthat is available quantifies activities and inputs, but not usually outcomes. The entire system-donors, Govermment,and NGOs-must begin to ask more questionsabout, and require more evidence of, impact. Monitoring and evaluation systems should be devised to help individual organizations (whether community-based,NGO, government office, or donor) learn whether they are making progress. Recent participatory evaluation designs developed by several Ugandan NGOs are an encouragingsign, and should be discussed in the NGO forum and among the issue-basednetworks.

8.41 As important as the content of the informationis the form in which it is presented. The monthly and quarterly reports from field offices (Governmentor NGO) are rarely convertedinto a form which is interesting to read. Central ministries may collate these into reports for the Prime Minister's office, but they still need to be transformed in a way that highlights innovationsand lessonslearned. A number of NGO consortiahave started sectoralnewsletters which are readableand offer useful information.At the level of the developmentcommunity in Uganda, there is also much more potential for organizationsto learn from each other, if information-sharingsystems can be constructed.These do require the investmentof resources in staff who can write readable articles geared to particular audiences and the communicationsequipment to support publication and dissemination.

8.42 The developmentof a learningculture would mean that Government,donors, communities, and NGOs would have to build a sense of shared purpose. None of these actors can control the others; they have to work from mutual appreciationand influence. Each set of actors has its own interestsand constraints;nevertheless they also share a core commitmentto buildinga socio-economiic climateof developmentand povertyreduction. At present, there is little (positive)appreciation among some of the actors, and a lot of name-calling.Yet no one set of actors-Govermnent, NGOs, donors, or communities-can achieve developmenton its own. The institutionalframework has to build trust and interdependence.

8.43 ExpandingNGO Capacity.In the current donor climate, there has been relatively abundant fundingfor NGO projects. Financialresources do not translatedirectly into good programs, however. In some instances, abundant funds have tempted NGOs to expand beyond their capacity to manage and program control quality. Sometimes they branch into new areas in which they have little experience, which also makes demands on managementcapacity. At one of the large international NGOs, a staff member declared that they have reached their current limit and will need a period of 132 Chapter8

consolidationbefore taking on new activitiesor more clients. Thus, even with abundant financing, NGOs may not be able to expand capacity quickly.

8.44 While official donors have soughtto fund NGO projects, they have been less eager to fund institutionaldevelopment for local NGOs. There are two basic reasons for this:

(i) NGOs are valued because they are seen as cheap service providers, and supporting better managementsystems and staff training raises costs;

(ii) donors want their money to be used for the poor, not to support overhead.

However, investmentin institutionaldevelopment is necessary to help improve the quality of NGO programs as well as being a prerequisiteto expansion.Monitoring and evaluationsystems are scarce in NGOs (as they are in governmentprograms). They take time and some specializedknowledge to set up. Receivingdonor fundingnotoriously requires NGOsto set up specialaccounting and financial managementsystems. And NGO staff, no less than staff of official development agencies, need occasionaltraining to keep up with the innovationsin their fields.

8.45 Abundanceof donor funding carries another danger for NGOs. In addition to mobilizing voluntary contributions,by definitionNGOs are supposedto represent independentcitizen initiatives. Donors (NGOs as well as official) have their own interests and priorities. NGOs seeking funding from such sources quickly learn to cast their proposals in terms of these donor priorities if they seriously expect to be funded. The result is strikingly parallel to government-to-government assistance:the donors, not the Governmentor the UgandanNGOs, make the decisionsabout which programs are implementedand which are not. Nationals are shut out of one of the basic institutional developmentfunctions: settingpriorities and strategies,unless they are prepared to risk losing donor funding.

8.46 ImprovingNGO competenceis another complexprocess. In Uganda, many people have the most respect for the work of small, localizedNGOs which mobilize committedlocal people to work for communitybenefit. Typically, such people can only spend part of their time on NGO work, because NGOs do not have the resources to pay a living salary. While their commitmentand local understandinggive them one set of competencies,such NGOs often plunge into activity to relieve a problem without a very broad analysisof the contextand therefore without a strategy. They may end up intervening in the wrong time or place-for example, getting street children to return to their familiesrather than working with vulnerablefamilies to prevent run-aways.Often such NGOs do not have monitoringand evaluationsystems in place that would allow them to learn systematicallyfrom their experience. Their managementsystems are weak, limitingtheir scope to a very restricted scale.

8.47 Improving quality is often seen as professionalizingNGOs-hiring staff who have skills at project analysis, feasibility studies, proposal writing, monitoring and evaluation, accounting and management.Hiring professionals,however, is expensive, and often demotivatesthe volunteer staff who were the organizations'strength. It also tendsto move the organizationstoward a Kampalabase, which adds further expenseand distancebetween decisionmakers in the NGO and the clients. Opening a Kampalaoffice to mobilize donor resources also opens the NGO to the temptationto tailor what it does to donor interests of the moment.

8.48 Another solution which is practiced in the NGO community is training. Typically, this training is offered to local NGOs by internationalones, or by Ugandan professional intermediary Chapter 8 133

NGOs that specializein services to the more localizedones. Staff can be trained through courses in accounting, proposal writing, analytic techniques, and so on. Staff can also be developed through working with larger (often international)NGOs on their projects. A small number of UgandanNGOs are in a positionto train other NGOs. Unlikeother developingcountries, in Ugandathe international NGOs appear to be more oriented toward implementingtheir own projects than to strengthening capacity of Ugandan NGOs, though there are a few exceptions such as OXFAM (which acts in Uganda purely as a funder).

8.49 A third option would be to developa corps of consultantsor consultantNGOs who work with these small, local NGOs as needed. Such consultantswould perform the tasks for which individual NGOs do not need full-time, long-term staff. They could help NGOs turn their project ideas into proposals which will stand up to funder analysis; set up new managementsystems to meet funder requirements; develop indicators for monitoringand evaluation. The use of these consultantsneed not be expensive:many NGO-orientedconsultants offer their services at concessionalrates to NGOs.

8.50 Networks or consortia of NGOs, or which include NGOs with other actors, can also help improve quality. The ones which have the most impact appear to be those organized on a subsector or issue basis.'3' At their best, networks and consortia structure the exchange of experience (informing members of innovationsand lessons of experience) and information (about policies, research findings, etc.). Uganda also has an NGO-specificnetwork, DENIVA, which attempts to raise quality by offeringtraining, issue discussionsessions with donors and Government,and a code of conduct (adopted by the General Assembly but so far un-implementablebecause of lack of sanctions).

8.51 Discussionssuch as these also help NGOs, Governmentand donors become more familiar with each other. They identify possible allies and sources of interesting new approaches. These discussionscan also help organizationsjudge each others' quality, althoughfield visit must also be part of a system of choosingreliable partners.

8.52 Buing Goverunent Capacity.The Governmenthas begun to articulatea vision of poverty alleviation through institutional pluralism (combining government initiatives with those of the voluntary sector in areas where market mechanismswill not suffice). The model is an attractiveone built on recent understandings of limitations of the state and the value of independent citizen development action. The model is one of facilitation rather than control, and conforms closely to lessons learned from 30 years' experienceof developmentinterventions: where there is an economic base of millionsof small farmers, Governmentcan influencebut not control socioeconomicactivities. The same is true of the relationship between Government and NGOs: Government can seek to influenceNGOs; it can control some areas of behaviorregarding taxes and legal standing;but it risks losing their support altogether if it tries to turn them into mere instrumentsof Government.

8.53 As a consequenceof its newness, this model is not yet widely understood or shared within Government. With much outside funding already going through NGOs, due in part to lack of accountabilityin Government,Government will not, in the short term, even be responsiblefor paying

In$ In Uganda,there ae existingnetworks aid conorti in the fieldsof: comunky-baed healthcam (UCBHCA), comunurity-bhsedchild welfare(UCOBAC), disabled (NUDIPU), women (ACFODE), and natral resources(Natual ResourceManagemcnt Project mn out of Experimentin IntemationalLiving, EIL). Themeame also coordination meetingson AIDS. 134 Chapter8

for all public services."'3 is gives the Governmentsome time to develop a commonunderstanding with the public of the role of the publicsector, and to begin reorientingcivil servantstoward the new developmentmodel which emerges from the dialogue.

8.54 If the agreed upon system is one in which Government retains control of identifying, designing and producing services, the government needs to prepare one set of staff to plan and manageprojects and to strengthentechnical skills of extension staff in health and agricultureso that they can act competently as service producers, in addition to strengtheningpolicy analysis and formulationskills. If an objective is to encouragecommunity participation, staff will need some kind of training or reorientation. Supporting incentives will have to be designed to reward staff for working collaborativelywith communitiesrather than executingdirectives from their superiorswithin government.

8.55 Certain government documents have recently proposed that Government will provide for services, but not produce them directly. There are two alternativespossible that conform to this model, and there is a crucial difference between the alternatives as far as both managementof governmentstaff and working with NGOs are concerned.

8.56 One alternative is for Government to hire contractors to carry out government-designed programs. In this case the Governmentwill need personnelto identify,design and set standards.They will still need staff who are technicallycompetent in communityhealth, agriculture, and water and sanitation,but they will use these skills in judging how competentlyservices are being provided by contractors,and they will need skills to write criteria, judge responses, and managecontracts. Many NGOs may decline to work with the Government on these terms, which would make them little different from contractors in the commercialsector. NGOs and many outside observers value the independentsolutions and perspectiveswhich NGOs bring to the developmentprocess, which would be in jeopardy under this system.

8.57 The other alternativeis that Governmentand NGOs wouldjointly work out the nature of the programsto be implemented,possibly in consultationwith the affectedcommunities. For this option, governmentofficials will need yet another set of skills, in negotiating,listening, and facilitating, as well as some familiaritywith the technicalfields.

8.58 Whichever option is chosen, Governmentwill need to allow lead time for the development of training and personnelmanagement systems congruent with the option they choose. Many of these skills, such as decentralized financial management, are discussed in more detail in the District ManagementStudy."'° Training and managementsystems can be tested in the pilot decentralization districts. Others are impliedby the Government'sdecision to move toward integratedservice delivery in health and agriculture;the Ministryof Health in particular needs to design incentivesto reorient senior staff away from hospital-basedhealth care to communityhealth.

IN9 It i not, in th long un, desimableto have basicservicec funded through contributions of thevoluntary sector. When enrviceare tax-funded, eitizae have momlevermge and can be asured tht serviceswill be extendedto the whobl country.

No See Uganda:Dubict ManagementSidy, ImfratructureOperation. Division, Eastem Africa Depailmeat,World Bank,1992, for furtherdeti. Chapter 8 135

Decentralization

8.59 The decentralization process offers an opportunity to simplify procedures for project preparation and financial management(including procurement). A simpler system will support community and NGO initiative, while complex regulations raise transaction cost, providing disincentivesto collaboratewith Government.Public informationmay be more helpful than complex procedures in bringing transparencyto use of financial resources.14 '

8.60 For projects designedoutside the communitiesthey are intendedto benefit, client assessment (preferablycarried out in a participatoryway) should become a regular part of identificationand design. Uganda does not have money to waste on projects that do not meet client needs.

8.61 Donor as well as governmentdiscipline is necessary in this process. Ideally, Government, donors and NGOs can agree on the priority programs. Public expenditurereviews can be used as tools to hold both Government and donors to these agreements, which would include recentering government ministries around poverty-reducing programs and reducing the number of project implementationunits. At the district level, NGOs could also be includedin these reviews, permitting a more comprehensivepicture of developmentinterventions.

8.62 Donors should provide modest support for informationsharing at central and district levels. District DevelopmentCommittees should convene quarterly sectoral meetings including NGO and governmentstaff. At these meetings, special attention should be given to NGO projects which are developingmonitoring and evaluationsystems and/or who have promisingnew approachesto share.

8.63 Making districts the primary level for registrationof Ugandan NGOs is a logical extension of the decentralization policy and should be tested in the pilot districts for the decentralized management.District officials can forward informationto the designatedoffice in Kampala. With increasingsecurity in the country, it may be time to reconsider the institutionallocation of the NGO RegistrationBoard; its placementin the Ministry of Internal Affairs sends a powerful message of suspicion. More of the burden of coordinationshould fall on Governmentrather than NGOs (i.e., getting informationto the necessary officials in different central ministries).

The District Community Action Trust (DCAT)

8.64 To address poverty on a broad scale, the DCAT should be designed as a social investment fund which disburses funds quickly for projects generated by self-help groups, with technical assistance being provided from NGOs or possiblylocal private firms as needed (and funded by the DCAT). The staff of the DCAT would be relatively small and based at district level; the main function would be publicizingthe existenceof the Trust, project appraisaland monitoring/evaluation. The district-levelproject approval committeeshould also be small and should include at least one representativeof local NGOs, preferablychosen by NGOs within the district. Governmentor NGO technicalpeople would not be permanent members, but would be invited to have a say on proposals in their sector. Some local resources should be mobilized for the projects in addition to outside financing.

'41 Donorsand NGOs alike recognized that it is eay to get around 'paperworkaccountability,' whereas they could cite intanes in whichpublicizing mim of meoure and conupt practicesby radioled to rapidimprovement. 136 Chapter8

8.65 The process of designing the DCAT will be critical to its success. Few people in Uganda (whether Government, donor, or NGO) now understand what is meant by a social investmentfund (SIF). However, the Governmentneeds to establish a design process which will have the greatest possible chanceof gaining commitmentfrom donors, staff of the Governmentitself, and NGOs. In this process, focus group discussions(a well-understoodmethodology in Uganda) will be useful in generating ideas and concernsabout the DCAT. Action-planningworkshops could be used at a later stage to involve differentgroups of Government/donor/NGOstakeholders.

8.66 The Government might start by forming a small task force of no more than 6-8 people, includingone Ugandan NGO, one internationalNGO, and one representativeof the small funders network in Uganda to design the process for designing the DCAT. This group would draw out lessons from experienceof the small funders and the NURP with such mechanisms,and would visit other countries (such as Guinea, Sao Tome and Principe, and Ethiopia) which either have or are farther ahead in designing SIFs. Part of their terms of reference would be to share their knowledge with others. This Task Force would not be charged with actuallydesigning the DCAT (althoughthey mightultimately do so); they would be in charge of seeking out informationand ensuringthat critical stakeholdersparticipate in the DCAT design process, so that maximumownership is generated.

Conclusion

8.67 The Governmentof Ugandapresently lacks the capacityto deliver key social services to the poor, thereby adversely affectingtheir capacity to participate in the economicgrowth process. The decentralizationprogram adoptedby the Government,together with the interestthat Governmenthas shown in supportingcommunity initiative, offers a unique opportunityto put in place an institutional framework capable of reachingand delivering these services to the poor. However, this new model is not widely understood or shared within Government. To take maximum advantage of this opportunity,the Governmentshould devote some time and effort to develop a commonunderstanding of what the new system will entail and to begin reorienting civil servants towards this new developmentmodel.

8.68 In order to effectively involve community based organizations and NGOs into the new process, governmentofficials will need to learn to listen to their problems and to play a facilitating role. The excessively stringent procedures that presently exist for NGO registration, e.g., the verificationby the Internal Security Organization,are not warranted and only serve to confuse and isolate these organizations. The Government should establish a simplified procedure for NGO registration, preferably at the district level, and there should only be one governmentbody vested with the authorityto coordinateNGO activity. As has been discussedabove, such coordinationshould only involve the disseminationof valuable informationand should not be interpreted as permission to regulate. The informationsystem that is developedfor the purpose should be a two-way system which provides policymakerswith data of what is happening at the grassroots level and also feeds informationback to the NGOs and the communities.With the right design, the DCAT can be used effectivelyas a tool to develop a frameworkof partnership among the Governmentand the NGOs and communitygroups. Learningfrom the experiencethat other countrieshave had with social funds, the Government should give high priority to designing a system which will gain the maximum commitmentfrom the donor community,the NGOsand the cadre of civil servantswho will ultimately be responsiblefor the day-to-dayfunctioning of the system. 9

A Strategy for ReducingPoverty

With appropriWepolicies, the poor canpartcipate in growthand contribute to it, andwAen they do, rapiddeclns anpoverty are consistet with staied growth. (WorldDevelopment Report, World Bank, 1990, pp. 51-52.)

9.1 As has been discussedearlier in this report, the decade and a half of civil strife whichplagued Ugandaduring the 1970s and the early 1980sshattered its economicbase, depleted its stockof human capital and, despite its rich soil and fertile climate, reduced it to one of the poorest countries in the world. While peace and security have now returned to most parts of the country and the economy has started to pick up again, the quality of life in Ugandatoday is much lower than it was when the country attained independence in 1962. As if this historical legacy were not enough, the AIDS pandemichas now gripped the nation in a vise and threatensto rob it of its most productivesegments of society, i.e., those falling between 15 and 45 years of age. Against this backdrop, and the seeminglyoverwhelming problems facing them, the NRM Governmentmust be complimentedfor their tenacity and determinationto improve the lot of Uganda's citizens. During the past five years, the Government has implemented a far-reaching economic reform agenda which has today transformed Uganda into one of the most liberal economies in Sub-Saharan Africa. With the liberalizationof the exchange and trade regime, the abolition of the Industrial Licensing Act, the promulgation of a new investmentcode, and the gradual liberalizationof agricultural pricing and marketing,the Governmenthas succeededin establishingsome of the fundamentalpre-conditions that are essentialfor sustainablegrowth. However, the unfinishedagenda is large and there is little cause for complacency.If the Government'soverarching objective is to make a serious dent in poverty over the next decade, it will have to focus its attention on two sets of policies: (i) policies which will accelerate economic growth; and (ii) policies which will deliver key services to the poor and, by investingin human capital, ensure that the poor are able to participateequitably in that growth. Both legs of this strategy are consistent with, and serve to reinforce, the recommendationsof the 1990 World DevelopmentReport. While there is broad ownership of these policies within Government, the principal challenge ahead lies in building the capacity to ensure that policy changes can be implementedefficiently. Furthermore, there is an urgent need to strengthenthe databaseand develop a poverty monitoringsystem which can provide policymakerswith regular informationon the impact of economicand social policies on the lives of the poor.

A. Polides for Accelerating Growth

9.2 As pointed out in Chapter 5, experiencefrom other countrieshas shownthat macroeconomic stability, a high rate of investment backed by domestic savings, and high rates of literacy and numeracyare vital for rapid economicgrowth. Ugandacontinues to do poorly in each of these three areas. Inflation continues to be high and savings and investmentcontinue to be low as do rates of literacy and numeracy. The key to achievingmacroeconomic stability is to get firm control over the governmentbudget. Despite the rationalizationof the tax and tariff regime and the establishmentof the Uganda Revenue Authority, the Government's revenue effort remains one of the lowest in the world. In the coming years, it will be imperativeto further tightentax administrationand to broaden the tax base. The Governmenthas made remarkableprogress in keepingexpenditures in line with the budgeted amounts and such disciplinehas to be maintained.However, given the low revenues, the Governmentwill probably continueto be a dissaver for several years to come thereby necessitating that the private sector generate the large increases in savings that will be necessary to finance the 138 Chapter 9 investmentneeded for rapid economicgrowth. Experienceof countries like China, India, and Kenya has shown that poor countries are capable of saving 20 percent of GDP or more. This places a premium on the developmentof the financial sector because the key to higher rates of saving is an efficient financial system capable of mobilizing small savings from a large proportion of the population.Uganda, at present, is saddledwith an extremelyunderdeveloped and thin financialsector where the ratio of broad moneyto GDP has been of the order of 7 percent comparedto 38 percent in Kenya. Ugandahad, in fact, achieveda ratio of 30 percent in 1974.

Policies for Accelerating Agricultural Growth

9.3 At the heart of its growth strategy will have to be the transformationof agriculture. The agricultural sector has the potential to feed the country, to supply food to the regional market, to export horticultural products in addition to the traditional export crops, to produce industrial raw materials (sugarcane,cotton, etc.) and generally to act as the engine of growth. The key to realizing this potential is increasing yields by raising the productivityof the farmer. That means security of land tenure, investment in research and extension, control of plant and animal diseases and rural feeder roads. For some crops high-yieldingvarieties are available from local research stations or foreign stations. What remains is to adapt them as necessary, and propagate and disseminatethem. An effective extension service is needed to ensure rapid adoption of improved varieties by the farmers. Many countries in Asia have shown that rapid and sustained improvementsin yields are possible. The key question is one of timing, i.e., how quickly can Uganda put in place the infrastructure needed to raise yields. The answer partly lies in targeting incremental national and donor resources towards agriculture, particularly research and extension and rural feeder roads. Raisingagricultural productivity, along with increasedaccess to basic healthand education,must take priority in the Government'sspending program. Markets, withthe associatedinfrastructure of storage and refrigeration, also matter a great deal. This is a matter for private investors rather than Governmentexcept with regard to governmentsupport for the collectionand disseminationof market intelligence.

9.4 In additionto raising yields, Ugandamust pay more attentionto the working of the domestic markets in agriculturalproducts. This is not an invitationto Governmentto intervene in the affairs of producers, processors and traders. Rather, it is to draw attention to the need to use public policy and public expenditureto facilitate the proper functioningof markets in agriculturalproducts. Poor rural roads, for example, can place producersof certain crops, or producers in a particular region, at a disadvantagevis-a-vis producers in other regions or producers of the same or other crops. Or the way truck operators are licensed or regulated may have adverse effects on the markets for agriculturalcommodities. In Ugandathere is anecdotalevidence of very low farmgateprices for some crops relative to final consumerprices. It is the legitimateresponsibility of Governmentto find out whether the gap between farmgate and final prices is broadly reflective of costs or whether it is indicativeof a malfunctioningmarket.

9.5 Rural feeder roads constitute an essential element of Uganda's strategy for accelerated agriculturalgrowth. Roads are needed to bring in agricultural inputs and implements,facilitate the work of extension staff, bring access to manufacturedgoods, create access to basic social services such as educationand health and, most importantof all, provide access to markets for farm goods. During the early phase of the ERP the Governmentplaced emphasison the rehabilitationof the major trunk roads and much progress has been achievedin this area. The Governmenthas recently shifted attentionto rural roads and the donors have responded enthusiasticallywith funds, equipmentand technicalassistance. However, progresswith the rehabilitationof rural roads has been slow. The main Chapter9 139

reason for this is the weak implementationcapacity of the Ministry of Local Governmentand the district authoritiesupon which falls the responsibilityfor the construction,repair and maintenance of rural roads. This is a task which will be gradually shifted to the districts, in parallel with the necessary financial and trained manpower resources. The decentralizationof responsibilitieswill probably start with feeder roads maintenance.

9.6 In the drive towards greater reliance on marketforces and on the private sector, Ugandahas moved quickly to dismantle the monopoliesin the agricultural sector. These monopolieswere the Coffee Marketing Board for coffee exports, the Uganda Tea Authority for tea exports, and the Produce Marketing Board for the export of food crops. In all three areas the emergence of competitionis acting as a spur to further market development.In food crops, for example, private traders have been the driving force behind the penetrationof the European and Middle East markets in simsim. Unfortunately,Uganda is not entirely rid of monopolies.The Lint Marketing Board still has the right to purchase all cotton for export. Ginneries are free to sell lint to the domestic cotton- processing mills. However traders in cotton lint and seed must be licensed by the Lint Marketing Board. The Cotton Act, last revised in 1964,provides for the zoning of cotton production,the setting of fixed seed and cotton lint prices, restrictionson the importationof, or trade in cotton and for the licensing of ginneries and restrictionson the siting of the same. As a result the cooperativeunions have a de facto monopolyof cotton ginning. There has been much talk of amendingthe Cotton Act and the Lint Marketing Board Act with a view to introducing more competitioninto the cotton industry but little action so far. Given the potential for fast growth of a once-thriving industry, Governmentmust set itself monitorabletargets for removingthe institutionalfactors constrainingthe revival of cotton.

9.7 Ineffectiveand inefficientfinancial intermediationhurts all sectors of the economy,including agriculture. At present agriculture does not depend much on purchased inputs and implements, whether locally made or imported; nor is the bigger proportion of agricultural output marketed or processed. The heaviest demand for credit in the agricultural sector has come from the agencies responsiblefor the procurement,processing and marketingof the traditionalexport crops, particularly coffee. Apart from the Rural Farmers Scheme that has been operated by the Uganda Commercial Bank, institutionalcredit has generallynot been availableto the smallholder.As a result not enough is known about the capacity of the smallholderto absorb and repay loans. The usual presumptionis that credit is a constrainton production.This may well be true in Uganda but it does not follow that rushing credit to the farmer wouldtranslate into increasedproduction. The first priority shouldrather be to get a better understandingof the different factors limitingthe smallholder's ability to expand output.

Policies for Accelerating Industrial Growth

9.8 Althoughimport substitution, certainly the old-fashioned,state-directed kind, is out of favor, it must be said that Uganda has substantial scope for replacing imports, provided this is done efficiently. A wide range of basic products are still imported, includingcement, paints, biscuits, processed milk, garments, blankets, tomato paste, and tinned fruit juices. Import substitutionis more likely to succeed if it is part of an outward-orienteddevelopment strategy than one focused entirely on the domestic market. The Republic of Korea, China, and other fast-growing countries have demonstratedthe superiority of outward orientation. Thus, Uganda must exploit, to the full, the opportunitiesfor entering export markets. As far as manufacturedexports are concerned, there are no obvious winners on the horizon but these cannot be ruled out. It is private investors, not governments,who pick winners (or losers). The Government's role is to ensure that the investment 140 Chapter 9

climateis conduciveto attractinginvestors who can produce a wide range of manufacturedgoods for the local and export market. In this regard, the great success achievedby Mauritiusin identifyingand exploiting an export market niche (in this case, wearing apparel) provides an indication of the opportunitiesthat Uganda can seize.

9.9 Given the dearth of long-termfinance, moderntechnology, knowledge of foreign markets and managementskills in Uganda, foreign investmenthas a crucial role to play as the catalyst for the transformationof the economy. In some countries minoritiesfrom a particular foreign country or region have performedthis role. In Ugandathe Asians fit the bill. A significantnumber of the Asians expelled by Amin in 1972 have become successful entrepreneurs in Britain, Canada and other countries. The courageousdecision of the NRM Govermmentto return the expropriatedproperties to the owners opens the way for these entrepreneursto invest in Uganda. Havingbeen dispossessed once, they will most likely exercise maximum caution in committing resources to Uganda. Nevertheless,the early signs are encouragingand a number of Asianshave embarkedupon a major rehabilitation of their properties once they got the properties back. Uganda needs to be forthright about the catalytic role that investors can play in the economy and it should mount investmnent promotionexercises directed at Asian and other foreign investors.

Employment and Labor Market Policies

9.10 As has been pointed out in Chapter 6, the central objective of employmentand labor market policies for the reduction of poverty in Uganda should be to increase the earnings of labor in agriculture. From the objectives of the land reform legislationproposed by the Agriculture Policy Committee,it appears that the Governmentrightly wishes to pursue a strategy of rural development which promotes a system of smallholderagriculture by further consolidatingthe de facto universal access to land. For this strategy to work, the proposed land reform legislation must find ways of guaranteeingtenancy rights to all the existing users of land and by improving the land endowment of the very small farmers. To encouragethe small farmers to overcomethe absenceof the advantage of economiesof scale, steps should be taken to facilitate their access to an appropriatetechnological package. As has been stated throughout the report, Ugandan agriculture operates under primitive conditionsand the knowledgeof farming systems and crop husbandrypractices is rudimentary.

9.11 Labor Mobility. The operationof the rural labor market will continueto be constraineddue to the absence of a large and mobile rural labor force. This is not by itself a problem as long as the concentration of labor in smallholder agriculture is not artificially promoted by inappropriate incentives. Improved systems of information and infrastructural facilities, along with the steady improvementof productivityin smallholderagriculture, is likely to make as much labor availableto the rest of the economyas it can efficientlyemploy.

9.12 GovernmentRole in the Formal Sector. Regulationof employmentand wages in the formal sector has traditionallybeen kept low in Uganda and the Governmenthas rightly resisted pressures to regulateformal sector wages, through such measuresas the legislationof minimumwages. As long as there is a healthy growth of earnings in the primary sector of the economy, the market might be trusted to ensure that the formal sector of the economy pays a price for labor that is adequate for living above the poverty threshold. Creatingtoo great a differentialbetween earnings in the primary sector and the earnings in the formal sector often leads to a lower than optimal rate of industrialization,an unwarranted influx to urban areas leading to an overcrowded informal sector, and a rising differentialin the earnings between the formal and informal sectors. Chapter 9 141

9.13 Educationand 7rining. Educationand training facilitiesmust adjust to make the allocation of labor more efficient. Improvedproductivity of agriculturallabor hingescritically on the expansion of primary education. Urban educationalservices would be better advised to reduce the focus on general higher education and instead increasethe emphasison technicaltraining. Clearly the entire system of pricing of educationalservices is in need of a basic reappraisal. As has been discussed in Chapter 7, at present a primary school student in Uganda is charged a tuition which is often a significantproportion of the cash incomeof an average rural householdwhereas university education continues to be free of tuition, often with access to additional subsidies.

9.14 AIDS. The effects of the AIDS pandemic on employmentand the labor market in Uganda cannot be determined with any precision. It is however clear that AIDS, which disproportionately strikes people in their prime, will increase the dependency ratio and make the task of poverty reductionmore difficult. The pessimisticprognosis is heightenedby the fact that measuresto prevent the spread of HIV infection-critically important though they are-will have little effect on the incidence of AIDS during the next decade. The Governmentof Uganda deserves credit for dealing with the AIDS issue in an open manner. Besidesconcentrating efforts on containingthe further spread of the HIV infection, e.g., by promotingthe use of condoms, the Government, working with the NGO community, should explore the need to make targeted interventionsin order to reduce the extreme effects on householdswhose labor endowmentshave been depleted by AIDS.

9.15 Public Sector Employment. Expansion of public sector employment is not an effective method of poverty reduction.Indeed an expansionin public employmentalmost certainly hurts the cause of poverty reduction by appropriating resources that might be used to expand employment more productivelyelsewhere in the economy. The number of persons employedin the public sector has grown too rapidly over the years from the standpointof the actual expansionof public services and the capacity of the Governmentto protect real wages of public employeesfrom serious erosion. The result is widespreadresort to corrupt and fraudulentpractices (e.g., bribery, keeping "ghost"workers on the payroll) and moonlightingand second informaljobs. On the whole, the effective labor time spent in public employmentmay have declined at the same time that there has been a steady rise in the number of persons on payroll. Recentlythe Governmenthas succeededin reversingthe trend and reducingthe numberof persons on payroll. The task is incompleteand should continue.To the extent that the reduction in governmentemployment comes through the eliminationof ghost workers and attrition (i.e., not replacing the low priority retirees)the process does not impose any burden on the current budget. A further opportunitymay become available if the increase in capacity utilizationin formal sector enterprises provides a leeway for productiveexpansion of employment.In that event positions in these enterprises might be filled by the redundant government workers. However, a reduction in public employmentis by itself a rather minimal measure. A central objective of the presently underway civil service reform program must be to introduce a culture of improved job performance,matched by higher real earningscommensurate with skill levels. This in turn means that a way must be found to prevent employmentin public sector to be determinedby the supply of high school and university graduates in a system of highly irrational relative costs of different forms of education.

Policies for Gender-ResponsiveGrowth

9.16 As argued in Chapter 3, the strategiesadopted by the UgandanGovernment to reduce poverty and to foster sustainableeconomic growth need to take explicit accountof the gender dimension.In this respect, they must recognize and seek to address the asymmetriesin the respective rights and obligations of men and women, and pay particular attention to the gender division of labor, and 142 Chapter9

differential incentives and opportunities facing men and women, as the country embarks on monetization,diversification, and productivityenhancement in the critically important agricultural sector. It is, in particular, necessarythat gender-responsiveactions be undertakenas an interconnected packageof measureswhich are mutuallyreinforcing. The priority areas requiring attentionif Uganda is to enable both men and womento break out of poverty, and to contributemore fully to economic and social development,are:

(i) to promote, through literacy and education, and in conjunctionwith the vigorous pursuit of the gender-responsivelegal (and customary) reform efforts underway, the legal rights and protections enablingwomen to benefit from their own labor and to have greater access to and control of economicallyproductive resources, includingcapital and land, thereby raising the status of women to enable more equal participation in household-level,community, and national decisionmaking;

(ii) to raise the productivityof women's economic (paid) labor through investmentin education aimed at overcomingsocial, financial, and culturalbarriers to female participation,including at the post-primarylevel; through investmentin basic, accessible,and affordablehealth care responsiveto the wide range of women's health needs; and through targeted actions aimed at raising women's accessto information,technology, inputs, credit, and extensionservices;

(iii) to alleviate the domestic labor constraint through substantially increased attention to and investmentin labor-savingtechnologies, in infrastructure(especially transport, feeder roads, and markets), and in water supply and woodlots, that take explicitaccount of female users' needs in design and implementation;

(iv) to provide maximumpolitical and financialsupport to the efforts, spearheadedby UNICEF, to reduce AIDS risk among young girls; and to protect the rights of children, including through institutionalmeasures in the RC system to ensure appropriate representation and articulationof children's needs.

B. Policies for the Delivery of Sodal Services

The Public Expenditure Agenda

9.17 The Governmentrecognizes that, since the single most important asset owned by the poor is their labor, the central element of its poverty reducing public expenditure strategy should be to accord highestpriority to developingtheir human capital. Human capital, more than any other factor, increasesthe incomeearning opportunitiesof the poor and contributesboth to individualand national productivity.Accordingly, during the past two years, the Governmenthas been attemptingto foster such developmentin human capital by restructuringgovernment expenditures in favor of the social sectors and rural infrastructure. Notwithstandingthe severe resource constraints, the Government must continue to ensure that these priority programs remain protected. Not only should more resources be channelled towards primary education and primary health care, but the efficiency of these expendituresshould also be improvedby ensuringthat money is spent on high impactprograms and that the combinationof expenditureswithin and across sectors are optimal. The Government needs to critically review its portfolio of investmentprojects in order to ensure that it is responsive to the country's changing needs. In other words, Governmentneeds to ensure that expendituresare made on a rational basis rather than the allocatedon the basis of historical levels. The combination of capital and recurrent expenditurealso needs to be improved. Chapter9 143

9.18 As indicatedin Chapter 7, boththe quantityand qualityof desired social services is adversely affectedby a lack of sufficientfunds. In the medium-to-longrun, mobilizationof domesticresources, through improvedefficiency in tax collectionand also throughthe judiciousadoption of cost recovery schemes in appropriatesectors, will be essential. This will need to be augmentedby shifting funds from relativelyunproductive areas such as defense, state farms, teachertraining colleges,universities, curative health, etc. to areas where the economic and social returns are the highest, i.e., primary health, primary education, agriculturalresearch and extension, rural feeder roads and rural water supply.

9.19 As has been demonstratedin Chapter 5, Uganda needs to slow down populationgrowth in order to reduce poverty withinthe shortestpossible time. At present, the Governmentof Ugandadoes not have an explicit populationand family planning policy and lags far behind other SSA countries in encouraging its citizens to have small, manageablefamilies and informing them of methods of doing this. Whereasmost women in Ugandawho know aboutfamily planning have favorableattitudes towards it, contraceptiveuse is constrainedbecause of lack of access and generallyunfavorable male attitudes. In addition to aggravating poverty, the growing population has adversely affected the environmentby increasingthe encroachmenton forests and by intensifyingfarming, resulting in soil erosion and stagnatingyields. It is thereforeimperative that the Governmentdevelop a nationalfamily planning program and closely monitorprogress towards decreasingpresently high fertility rates.

Institutional Issues and the Provision of Safety Nets

9.20 Althougha strong case could be made for this in the Ugandan context, targeting the poorest and most vulnerable is far too expensive an option for the Governmentof Uganda to consider at present, given the extremely low revenue effort and the weak administrativecapacity. Instead, the focus should be on providing fundamentalservices in rural areas: primary education, primary and preventive health care, rural feeder roads, safe, easily accessiblewater, agriculturalextension, and marketing assistance. As has been discussed in Chapter 8, communitiescan help provide some of these themnselves,with assistancefrom Governmentand NGOs. Government,donors and NGOs need to develop a tripartite systemto support self-help projects, and to ensure that priority investmentsare undertaken. If Government wishes NGOs to follow governmentleadership in a scenario in which Government formulates policy and sets standards, NGOs, including Ugandan ones, should be involved in policy formulation.For the tripartite systemto work, it must be "owned"by a substantial portion of the three sets of actors (Government,NGOs and donors). The policy developmentprocess in the areas of AIDS and child welfare could become a model for other sectors.

9.21 The program of decentralizationbeing adoptedby the Governmentoffers a unique opportunity to support communityand NGO initiatives,particularly in light of the fact that the Governmentlacks the capacity to deliver much needed services to most of its citizens. For this new model to work, it will be imperativefor the Governmentto educate the population and to reorient and educate civil servantstowards its changingrole. The DCAT, whichshould be designedas a social investmentfund which disburses funds quickly for projects generated by self-help groups, could emerge as an important instrumentto address poverty on a broad scale.

Selected Bibliography

Selected Biblography

Action Aid. 1992. Analysis of Results of Particpatory Poverty Assessment In Mityan CountY, MubendeDistrict, mimeograph.

Akiri-Kajeru, Christine. 1990. Prelminary Asssessmnt of Mulago HospitalStatistics.

Alden, John, Abby Nalwanga-Sebinaand James Sengendo. 1987. Orphanedand Disabled Children in Luwero and Kabale Districts, and In Uganda Chi Care Instlions.

Alden, John, Gerald Salole and John Williamson.1991. ManagingUganda's OrphansCXsls. Report Prepared for USAID.

Bwanika-Bbaale,Hon. A. Undated. The Role of the Ugandan Gowrnment In National Strategy Formulation.Paper presented to the Working Group on Children Orphanedby Aids in Africa.

Evans, Alison. 1992. A Revkw of the Rural LabourMarket In Uganda.School of African and Asian Studies, Universityof Sussex, mimeograph.

Fleuret, Anne, et. al. 1992. Girls' Persistenceand Teacher Incentives In Primary Education In Uganda. USAID, Kampala, Uganda.

Governmentof Uganda. 1988. NatdonalAgricultural Research Strategy and Plan. Uganda, Vol. II: Prioritiesand Programs.

Hesse, Mary Chinery. 1989. EngineeringAdjustnentfor the 1990s. Report of a Commonwealth Expert Group on Women and Structural Adjustment.Commonwealth Secretariat, London.

Holt, Sharon. 1991. "The Role of Institutionsin Poverty Reduction: A Focus on the Productive Sectors". PRE WorkingPaper No. 627. World Bank, Washington,D.C.

Husain, Ishrat. 1992. Adjustment and the Impact on the Poor, Ihe Case of Africa. Mimeograph. World Bank, Washington,D.C.

Jarawan, Eva. 1991. Women in Development:Current Issues and Agendafor Fkther Research. Populationand Human ResourcesDivision, WorkingPaper, World Bank, Washington, D.C.

Kaijika, Emmanuel, Edward Kaija, Anne Cross and Edilberto Loaiza. 1989. UgandaDemographic and Health Surey 1988/89. Ministryof Health, Entebbe, Uganda.

Khan, Q.U. Employment Siuation In GovernmentSector. Manpower Planning Department. ILO, Geneva.

Korten, David. 1980. 'Community Organization and Rural Development: A Leaming Process Approach". Public AdministrationReview. 148

Krueger, A.O., M. Schiff and A. Valdes. 1988. "AgriculturalIncentives in DevelopingCountries: Measuringthe Effect of Sectoral and EconomywidePolicies". The World Bank Economic Review. Vol. 2, No. 3.

Lochhead,Alison. 1991. Genderand Developmentin Dodoth County,Karamoja. OXFAM, Kampala, Uganda.

ManpowerPlanning Department, Ministry of Planning and EconomicDepartment. 1989. Manpower and Employmentin Uganda,Report of the NationalManpower Survey. Kampala, Uganda.

Martin, Doris and Fatuma Hashi. Law as an Institutional Barrier to the Economic Progress of Women.

. Sub-Saharan Africa: Gender, Evolution of Legal Institutions, and Economic Development.

Ministry of Finance and EconomicPlanning. 1992. UgandaNational Plan of Actionfor Children, The Way Forward IV: Prioritiesfor Social Services Sector Developmentin the 1990s. Kampala, Uganda.

Moser, CarolineO.N. 1989. "GenderPlanning in The Third World: Meeting Practical and Strategic Gender Needs". WorldDevelopment, Vol. 17, No. 11.

Muyayisa,David. 1992. Street Childrenin Kampala,unpublished manuscript.

Muyinda, Herbert and Tom Barton. 1992. Socio-Economic Influences on the Health and Rehabilitationof Physically HandicappedRural Children: A case study in Kayunga Sub-County, MukonoDistrict, Uganda.Final Report. Child Health and DevelopmentCentre, MakerereUniversity, Kampala,Uganda.

Muyinda, Herbert and Tom Barton. 1992. Socio-Economic Influences on the Health and Rehabilitationof PhysicallyHandicapped Rural Children, unpublishedmanuscript. Child Health and DevelopmentCenter, Makerere University.

Obbo, Christine. "Women, Children and a 'Living Wage'", in Holger Bernt Hansen and Michael Twaddel, eds., ChangingUganda. FountainPress, Kampala, Uganda.

Palmer, Ingrid. 1991. Genderand Populationin the Adjustmentof African Economies:Planningfor Change. Women, Work and DevelopmentSeries No. 19. (Geneva,InternationalLabor Office).

Saito, Katrine. 1992. Raising the Productivityof Women Farmers in Sub-SaharanAfrica, Vol I. OverviewReport, PHRWD. World Bank, WashingtonD.C.

Salmen, Larry. 1992. "ReducingPoverty: An InstitutionalPerspective". Poverty and Social Policy Series, Paper No. 1, ProgramDesign and Implementation.World Bank, Washington, D.C.

Sserunjogi, Louise. 1992. VitaminA Intake Among Children Below Six - Kamuli District, Child Health DevelopmentCenter, Mulago Hospital, Kampala,Uganda.

Summers,Lawrence. 1992. Investing in All the People: EducatingWomen in DevelopingCountries. Remarks prepared for a DevelopmentEconomic Seminar, World Bank, Washington, D.C. 149

Tadria, H.M.K. 1987. 'Changes and Continuitiesin the Position of Women in Uganda." Beyond Crisis: DevelopmentIssues in Uganda, eds. Paul Wiebe and Cole Dodge, Makerere InstituteSocial Research, Kampala, Uganda.

Tuhaise, Charles. 1992. 'Need for Foster Care in Uganda." The Vulnerable Child. UCOBAC Newsletter.

UNDP. 1992. PreparatoryStudy on Poverty Alleviationand Rural DevelopmentIn Uganda. Draft Report (prepared by ManagementSystem and EconomicConsultants Ltd.), Kampala.

UNICEF. 1989. Childrenand Women in Uganda: A SituationAnalysis.

. 1992. New Phase of UNICEF Support for AIDS Control in Uganda. Summary Document. Kampala, Uganda.

UNICEF/ACFODE. 1988. Uganda:Women's Needs AssessmentSurvey.

von Braun, Joachim. 1991. 'Social Security in Sub-Saharan Africa: Reflections on Policy Challenges",in E. Ahmadet al, Social Security in Developing Countries.Oxford UniversityPress.

The World Bank. 1993. Uganda:Social Sector Strategy. Populationand HumanResources Division, Eastern Africa Department,Washington, D.C.

____. 1993. Uganda: Agriculture Sector Memorandum. Agriculture and Environment Operations Division, Eastern African Department,Washington, D.C.

_ 1992. Women in Development: The Legal Issues in Sub-Saharan Africa Today. AFTSP Division Working Papers, Nos. 2, 3 and 4, Washington,D.C.

1990. Primary Education: A World Bank Policy Paper, Washington, D.C.

- 1991. Public Choicesfor Private Initiatives: Prioritizing Public Expendituresfor Sustainable and Equitable Growth in Uganda, Eastern Africa Department,Washington, D.C.

_.__ 1990. World Development Report. Washington, DC.

.1992. UgandaDistrict Management Study. InfrastructureOperations Division, Eastern Africa Department, Washington,D.C.

The Office of the Commissionerof Data Processing. 1987. Census of Cvil Service.

1989. Non-GovernmentalOrganizations Registration Statute, Governmentof Uganda.

1990. StatutoryInstruments, Governmentof Uganda.

1990. Ihe Non-GovernmentalOrganizations Regulations, Governmentof Uganda.

1992.Report of the VicePresident's Select Commitee on Non GovernmentalOrganization, Kampala, Uganda. . 1992. World Development Report. Washington, DC.

Annex I

StatisticalAppendix

- 153 -

Annex I

Statistical Appendix

Contents

Table 1.1 Country Data Sheet ...... 155 Table 1.2 The 1991 Populationand Housing Census ...... 156 Table 1.3 PopulationDensity by Region and District ...... 157 Table 1.4 Age Structure of Population ...... 158 Table 1.5 Populationby Age and PopulationType by Sex ...... 159 Table 1.6 Populationand Literacy by Rural-UrbanDistribution ...... 160 Table 1.7 Populationand Literacy by Sex Distribution ...... 161 Table 11.1 GDP by Sector at Current Factor Cost ...... 162 Table 11.2 GDP by Sector at Constant 1987 Prices ...... 163 Table 11.3 GDP by Expenditurein Constant Prices ...... 164 Table Ill.1 Balanceof Payments ...... 165 Table 111.2 Compositionof Exports (FOB) ...... 166 Table 111.3 Compositionof Recorded Imports (CIF) ...... 167 Table 111.4 Service Imports ...... 168 Table IV.1 External Debt ...... 169 Table V.1 Central GovernmentOperations ...... 172 Table V.2 Central GovernmentRevenue ...... 173 Table V.3 Central GovernmentExpenditure ...... 174 Table VI.1 Monetary Survey ...... 175 Table VI.2 Interest Rate Structure ...... 176 Table VI.3 ExchangeRate Movements ...... 177 Table VI.4 Foreign ExchangeAuctions ...... 178 Table VII.1 Productionof PrincipalManufactured Commodities ...... 179 Table VII.2 Index of Industrial Production (AnnualSummary) ...... 180 Table VII.3 Index of Industrial Production (MonthlySummary) ...... 181 Table VII.4 Production and Exports of Principal AgriculturalProducts ...... 182 Table VIII.1 New ConsumerPrice Index: Kampala ...... 183 Table VIII.2 New ConsumerPrice Index: Kampala ...... 184 Table IX.1 Orphan Census Derived from the 1991 National Census ...... 185 Table IX.2 Pattern of Food Expenditureby Region ...... 186 Table IX.3 Percentage Shares Spent on SelectedExpenditure Categories ...... 187

UGANDA TABLE I.1

COUNTRY DATA SHEET

Ganeral Area, land q km 197,096 Population (1991) 16,671,705 Growth rate % pernnum 2.5 Density (1991) per sq km 85 Socio-Economic hidicators Total hours worked per daty,women 15-18 Total hours worked per day, m 8-10 Female ownership of land percent 7 Share of credit to women percent <1 Social Indicators Population Characteristics (1990) Crude birth rate per 1,000 51 Crude death rate per 1,000 22 Life expectancy at birth, male yoe 47 Total fertility rate 7.3 Contraceptiveprevalence rate percent 5 Female-headedhoueholds percent 30 Health Infant mortality (1990) per 1,000 live births 117 Maternal mortality rate per 100,000births 500 Female morbidity rate percent 76 Population per physician (1991) 24,700 Population per hospital bed (1991) 1,200 First birth in 12-15 age range percent 29 First birth after age 19 percent 22 Income Distribution, (1989-90) 1/ Highest quintile of national income % of population 8 Lowest quintile of national income % of population 30 Access to Electricity (1989-90) % of Urban population percent 40.1 % of Rural population percent 1.9 Nutrition (1991) Calories per capita 2800 Education (1990) Literacy rate, 1991 percent, age 10 and over 54 Female literacy, 1991 percent, age 10 and over 45 Adult literacy rate, 1991 percent, age 20 and over 52 Female adult literacy, 1991 percent, age 20 and over 39 Primary school enroLIment- total % of relevant population 72 Primary school enrollment - femle % of relevant population 63 Sources: Houshold BudgatSurvey, 1939-90;Statitics Departmat. MFEP; Uganda:Social Sector Straogy, 1993. 1/ Basedon expendtr dat fromthe Houshld BudgetSurvey.

155 UGANDA TABLE 1.2

THE 1991 POPULATION AND HOUSING CENSUS Final Results

Percentage Region & District Total Male Female Urban Rural Urban

CENTRAL Kalangala V6.37 6,4 214 1.95 .. Kampala Sl,16 7441 .. lO...... Kiboga S 6....9 547 136,330 ...... 3 Luwero 2)9 29.4±9 ... $3..43.60.64 Masaka . .... 41 4... 746 4t'*9 Mpigi 45,ET ...... 46 1346 7±41 Mubende I00,....254Am34,54 44,895 464356 Mukono 82 .O4 43,8 4102 ...... J8 ... 75I6 . 2 Totals No*,9,1+8 A 05 11982 36.0 4 EASTERN Iganga #443 4149 440 40 04*1 Jinja 14.3.16,401%...... 0 ..$ ...... Kamuli 4865*242 37.....513k. 2411 8.6 7692 Kapchorwa 167.0 xx6477~ 5642%..I4* Q4 Kumi 11 ...... ~31247 1(4M1*35 Mbale 7190 $513 5,7.O2f $AZ6 Pallisa 3746 1316 1310 11 5190 Soroti 4039 2950 206 024 341 0 Tororo $544 2l20 2824 4,5 1471 Totals44649 2243 Z AS 3260 ,01 7 NORTHERN Apac 4..,.4 2....5...... 1 ..... 1 Arua 63.4 4,2 3 .2 ...... +24 4 Gulu 336427 ...... 0..697 3043 I KCitgium M5,* ...... 5 4 t~.7~ Kotido 1~O 241 1355 971 16344 Lira $245 WG 545 Z~6 7,9 . Moroto 741 061 -9445 1...1....3 - 7 Moyo 1,4 504 051 5? 61S'O Nebbi 3616 1205 1413 2±4 9,2 . Totals 31195 S5677 1656 00 .1,25 WESTERN Bundibugyo 1056 5±1 37017317 Bushenyi 7431 347 1,9 f45I Hoima 7,5 994 n,o ::::4:V$l:j3423 Kabale 471 9.9 1 2 Kabarole74±0 3911 37±1 3094 3±444 Kasese 330 642 15Ol 3,9 0±0 1 Kibaale 221 W9$ $5. 4;j. 74 j1 Kisoro .1.... +. 007 4* 1 164 Masindi 24,0±740I#$ 12±6 431> 2644 Mbarara 43±7414? 4255 4464 *4%5 Rukungiri 9,0 115' 225 2*% $7753 Totals 446? 215 ,±44 ~ 7S 4±134 UGANDA Totals 14l70 11,77 1415 953 18,2 LI 5 1

Source: Statisticts Department, MFEP.

156 ...... gm mg.: gg ..... ai: N ...... X:O...... I...... R :X.N.1 .1-K ...... > ...... PQ Z4. .t.4 ......

......

...... 00 z

...... %; ...... :- LLI ...... X::,.ft Cd ...... 0 ...... Xm...... '5 ...... U 5 Cd 4 0 Cd M 0 0. cci .4 0 14o 0 .0 & 1.2 - .) I W W,, - &2J. 0 %79 E.o 0.2 CsSM >e CA rA En Cd .0 cin M M 4 4 4 ;SE0 0 z Eo- Cs 9,03 F-0. Eo- -t4 0z 2 3 Eo- 31,04,2 UGANDA TABLE I.4

AGE STRUCTURE OF POPULATION

<--- RURAL -> <- URBAN > <- TOTAL -- > '000 'loW 1 Age Group Male Female Totl Malo Foale ToUl Male Female Toftl

......

20-24 2 4 25 -29

30 -34 )4~ . 35 -39 z. 40 - 44 ~I

45 -49 -~~

50-54

55 -59 60 - 64 65 - 69

70-74 -~

75 79

Not Stated .. ~ '

Source: StatisticsDcp.rtment, MFEP.

158 UGANDA TABLE 1.5

POPULATION BY AGE AND POPULATIONTYPE BY SEX

HOUSEHOLD POPULATION INSTMTUTONAL& FLOATING TOTAL POPULATION POPULATION Age Group male Female TOta Male Peale Total Male Female Total

0 4 :::ts>6 ~ ~ ~ b8 . 59~~~~~~~~~~~2s 64512046Z%9 10 -14 A2IASgkz,a...... 15 -19 5+2 uemo P.4 tns 20 - 24 4.6 H W 1AS *567~ 1M 25 -29 ji'j ...

30 -34 4$ t 4O2$ EhS 28,7 35-39 ~ ~ ~ ~ ~ ~ ~ S6 ~4A3 S,0 +1 40 -44 2 . 2.P 2.f 4,4 45 49------50 - .54 20,4 # 7.1. 3,46 42,2

65 - 69 MAVSwiBEU 3.65 17,4 70 -74 7f4 824 8,3 6.5 75 - 79 45:,i*I~ M)4,51 4,91 1,4 80 + 6 2~~%~ t~6 NotStated .*. .>...... ,10

Total 8074..B

Source: Statistics Department, MFEP.

159 UGANDA TABLE 1.6

POPULATION AND LrrERACY BY RURAL-URBAN DISTRIBUTION Age 10 and Over

RURAL URBAN TOTAIL Percent Percent Percent District Lkerate Total Literate Literatz ToW Literate Literate ToW LitrraW ...... Apac ...... i...... " *. ,: Arua iMR.:% iffi:::.:ft-::Mii Bundibugy...... Bushenyi Gulu h Hoim Iganga ...... Jinja ...4 Kabale ...... Kabarole Kalangala Kampala ...... Kamuli Kapehorw ...... Kasese MM'i w ...... M: Kibale Kiboga M- ...... Kisoro M Kitgum . ...!

Kotido ...... gg ...... Kumi ...... Lira Luwero Masaka k'..... Masindi ...... J Ubale ...... Mbarara Moroto Moyo ...... '04...... - ...... mpigi ::*:...... ::A Mubende ...... Mukuno ...... Nebbi M ...... Pallisa ...... Rakai W...... Rukungiri Soroti ...... Tororo ...... Total:

Source: The 1991Population and HousingCensus, Statistics Department, MFEP.

160 > ] X :g:..l S FEmEElE* F.~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... X ':' .' ' :'':: : :: : ::§ E ' ':: g 'E': ::: 1 ' ' '' '' ' '' -."'.E'E..'g:'.'' :': ''.' ''.'..:' ''. @~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~X S:~~~~ ~ ~~~~~~~~~~~~~~~...... ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~a~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.... A'' ::§::::':E :' .'::'.:::' ' DE:::.Et '.:E:E':" ' .:' ' : :: a: : .:.E.....; §...... | ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... g...... ~~~~~~~~~~~..--a gBBBKgg"' ...... g...... ":'g"B...... ,B,B:0g ...,.,.,.,.,..,.,.,.,...... ,,... .g .. ., ., .,...... * g * ... , , * . . .. - g . ; . g , .~~~~~~~......

..3: '''"':' -: - .. :.- '.'. " '...... :':"'''D:"""g''' : . L - ::- -::- :, '~~~~~~~.:...... , .. .,...... ,;...... ,.''-''':.'''§'.'§:'

p J :S~~~~~~~~~~~~~~~~~~~~~~~~.g...:... . s.{. ....N...... -... f-SE

; < - n . > ffi -- n --- E-~~~~~ ..- ~~~~.e...... -: : 0. : :

= Z . :- . : .: . . :.:.: ~~~~~~...... :.:::::.:...... ¢

9 .~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

...... X:.-~~~~~~~~~~~~~~~~~~~~~~. Z ,-,: ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.:-:... ::::...... -...-::-

~~~~~~~~~~~~~~~~~~~~~~. > gg *...... g. g.g~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~..*..¢...... -a--E--r.uggggE.2.agBBga...... 2::B:.~~~~~~~...... }| > ......

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0~~~~~~~~~~~~~~...... mm . .. :::::: ...... - ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ ...... ~~~~~~~~~~~~~~~~~~~~.cc ...... 2<~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~'-~A0 o;

ffi ...... 3°.3..A ...... x ... 2o-° n~~~~ . 3. 0% 0 . . 0 ...... UGANDA TABLE 11.1

GDP BY SECTOR AT CURRENT FACTOR COST In millions of U Sb

1985 1986 1987 1988 1989 1990 1991

Agriculture 1503 3015 1fl417 43$3Q4 677044 112459 1012044 g&. >$$ W463 ww nuu uwo# Cash Crops . 57*546 725470 Food Crops .

Livestock ,. ,. .P uzn. Forestry & Fishing >; IUI 42559 #4011

Industry t6tD 4 S* 15S714 Manufacturing tnt> 2474w Mining & Quarrying 47 121 »¼ 4532 *491 a. a flY 4o$ 607w 19541 Coffee, Cotton, Sugar t;4 171 5 j444 ... S$0 ZIQ$ Food Products .4 4Th 4t t#<. W1 45852 Miscelleneous P1* 64W Z9P1' $4004 4WZ7 '58164 Public Utilities >10 t . Construction 106 182$ >> 40041 *IflS 134465

I .A,> >> Services d SflP0 4i $> 1O17 17434

Trade 'UW > <> $4Q9 197907 25*817 Transport & Communication 9 $'C ' flU 903 103001 Road $7 . flK3 fl514 ivroa Rail ,.i.4 ?$)6 26 "2360 3856 Air .:....*...... 16< '7$ tIC 52132 7146 Conununication \t4EZ $' 3%2 12303 General Government ""'00.*4 > . 4fl$ 4%$... fl$ J5M 67264 Education "#fl 1$fl> 4% US) 1$44> 4gj4 458524 Health 1$, > 5 J%1, 4$ 7*$ 15119 2560* Rents 47G 1* 1* 1$4W 27142 34041 79S%* Owner-occupied Dwellings 01 140 4977 4t# 25101 4000 66029 Miscelleneous >494 43 *0- $flhC '33136 43241

> >> GDP - Factor Cost M74V> S14< tttflt At IXEMI) 1495260 2103892

01w Non-Monetary ...... >ZW4 659094.

.> .... ' ....... ,...... -...... -... Indirect Taxes 26779:.:.>:.:.>:.:.:.:.::::c:::::::::k.:::.. !i5?4. 957

GDP - Market Prices 4E01 620* t47* 41W0? 116U17 4590918 2236Y

Source: Statistics Department, MEEP.

162 UGANDA TABLE H. 2

GDP BY SECTOR AT CONSTANT 1987 PRICES In millions of U Sh

1995 1996 1987 1988 1989 1990 1991 W.P:K.. Agriculwe Apm El ...... 110 Cash Crops Food Crops .. Livestock Forestry & Fishing gk

Industry

X0,04-K Mhiing & Quarrying Manufacwring K. Coffee, Cottcm, Sugar Food Products Misceflancous 04

Public Utilities K' i4g Constmction

Services

Trade Tramport & Communication Road X. Rail Air Communication General Govemment Education Health Rents xx Owner-occupied Dwellings Miscellencous ...... K

GDP - Factor Cost

o/w Non-Monetary X

Memorandum Item:

GDP Deflator (1997=100)

Source: Stedstics Dq)artment, MFEP.

163 UGANDA TABLE 11.3

GDP BY EXPENDlTURE IN CONSTANT PRICES In millions of U Sh

1985 1986 1987 1988 1989 1990 1991

Consumption 2 4 _ Private1841 195 190 t57294 2337 237 Public 173 105 140-t t h 53 14

Gross domestic investment11$ 1#7.0$ 214 Z41 37 17.

Fixed investment 1-3 1 1 2 Private610 66 734 kS~100 I29 101 Public 13$ U~ $t aI S4 11 15$

Gross domestic expenditure 1'

Exports (g+nfs)195 1W 143 I0 1S( 124 194 Terms of trade adjustment 7 j 0 Import capacity 204 153 133 103 155 <82 79 Imports (g+nfs) 20 56 t~'$2 99 P 2l

Resource balance 192 1 .

Net factor service income e -2 j Net current transfers 47 94 41 f' 415 33 03 Current account balance 4_ .

GDP at market prices2123 002 2270277 2646 217 213.

GNP D 220236

Gross domesticsavings 116 119 80 f4 77 26 63 Gross national savings 1 1

Gross domestic income2226 2512240251 2243 63(5 711 Gross national income 272 146204 556292 696 203

Source: Statistics Department, IMF and staff estimates.

164 UGANDA TABLE Ill.1I

BALANCE OF PAYMEENTS In millions of US$ Est 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92

Exports (g+nfs) 40m.4~ 0 -246 0 9 Merchandise (fob) 1,7. 7 o/w Coffee .0.. .2..1.. 9 Non-factor services 2 22 Imports (g+nfs) ...... :.:.:..:... >:~65 Merchandise (cif)54 55 IS0 42 o/w Petrol 6 6~75 Non-factor services ...... Resource balance 44043 Net factor income 4 76 76 4 o/w Net interest .4 77 Current private transfers 010 147S0 12 C/A balance (excl grants) 11 2$ 31 49 -4 4 C/A balance (inci grants) l~ (3 20 ~ 7 29 -~ Official transfers4031 120 7 o/w Importsupport0 342 W Net M< loans 4 11 15 25 11I Disbursements 14 21U 4 Project loans 15 11 1 ~ 1$7 Import support loans 0. 4$...... Repayments 9 t 67 5 12 ForeignLinvestment/Kenya comp. 2 01 Short-term, net317 721 Errors/value adjustment 2 07001 Overall balance196 O44 111 Financing:70644 1 0 Monetary authorities 2 41 04 Gross reserve changes 3 41 11 3 IMF, net 17712 SAF/ESAF and purchases574 944005 Other, net7 000 Short term/comnmercial 1 .. I- External arrears 4 71 ~6 2 Exceptionalfinancing 24614I1 Rescheduling92 4 7000 Debt cancellation 0321 Residual finance gap0001 Memoradum item: Gross reserves (EOP) 1/ in months of imports. Source:Bank of Uganda, IMF, and staff estimates. I/ Includesreserves of the Bank of Uganda held as tied import support.

165 UGANDA TABLE m1I.2

COMPOSMTIONOF EXPORTS (FOB)

1980 1981 1982 1983 1984 1985 1986 1987 19 1989 1990 1991

Volume: '000 Tons

Coffee 1 1 4 4 _ . Cotton ~ . A .A~ ~1~ Tea * ~ U U4~. Tobacco .w Maize & b

Unit Value: US$mig

Coffee . Cotton g Tea 04 O'4 Tobacco fl . Maize &beans

Value (FOB): US$ imilions

Coffee

......

Maize & beans .. .,.'

Other exports ~ *..

Total value 4A M

Source: Research Department, Bank of Uganda.

166 UGANDA TABLE III.3

COMPOSMON OF RECORDEDIMPORTS (CIF) In millions of US$ First Semester 1986 1987 1988 1989 1990 1991 1991

Imports not requiring foreign exchange

Food and beverages n b Other consumer goods 73 Intermediate goods K% Z4 27 Petroleum and petroleum by-products 1 X . Capital goods ~~ 4

Totalof w6 1

Imports req 19ringforeign exchange

Food and beverageson a c y 4s Other consumer goods it does n c to the Intermediate goods * 9 t4 $ M Petroleum and petroleum by-products I/ flI4R 1 Capital goods 44 46 , & ZS 2.

Total 4 M6 2$ W$ *A 4.

Total of recorded importsPam.t 1111. 2445

Project-related imports 2/1~ ~ 1~4, ~2~d ~m t6 Total imports 3L ~ ~3~& ~~~9

Source: Reseach Department, Bank of Uganda

I/ For 1986, this category also includes chemicals. 2/ Source is IMF for years prior to 1990.

Note: Details of projoectimports are not recorded. This table in reported on a calendar year basis and does not include barter imiports. Consequently, it does not rcorrespondto the Balance of Payments (Table III. 1).

167 UGANDA TABLE III.4

SERVICE IMORTS In millions of US$

1987 1988 1989 1990 1991

-...... ,2o<,,. Overdue bills 8.1 22.8 4. 0.5 Medical treatment 00 0 Ft Education iS 21 : 09 05 Public . ... . ,...+...... +. 0

Public 5 15 79 Private 41 6 6 . Diplomatic missions -- :113 -t$5 -11 10 4 Savings and maintenance 1.1- 1 2 1 0 1 Dividendsand reinsurance .+ I 3

External debt payments 01... 36 S 3

Totalservices it'.'',..''.,tsA,li. 14'

Source: Research Department, Bank of Uganda.

i/ Includes clearing/freight charges, fees, commissions and services related to imports and exports.

168 UGANDA TABLE IV. 1

EXTERNAL DEBT In millions of US$ 1986 1987 1988 1989 1990 1991

Total external debt 1 1 4 2-31 2 2

Debt outstanding& disbursed 2O$ 1f 61 . 21, 2Z Lang-term~debt Public and publicly guaranteeddebt % 1 Official creditors S3 X8 if 6~ 14 0 Multilaterals m IDA 30 54 5$ A$ 9$ 17 IBRD414 f Bilaterals26 . 5 $ 61.. 44 64

Private creditors * 8 7 29 Zf 23 Commercial banks . :: ...S>. 7 Bands 0 1 0 0 0 0 Other private 1.7...... 15 Private nonguaranteed 0 ...... Use of IMF credit 24~ 27$ 252 % Cnu 282 Short-term debt 67 66 ~ 1 1~ 143 ~76 oiw Interest arrears on LDOD 3 . 3 3 5

Long-term debt 441 5 ...... Public and publicly guaranteeddebt 4177 9f Official creditors 4 51 66 60 Multilaterals35.:r,r.,,~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 43 ...... 55 57 70 73 IDA 2 3 ~ ~ . IBRD .... 00 Bilaterals 6 5 8 4 ~ ~ Private creditors $ 5487 43 4 Commitments10 47 1 6 46 43 Long-term debt 4 1 1 16 46 43 Public and publicly guaranteeddebt10 41 1 36 46 43 Official creditors75 275 43 Multilaterals15 25 4 37 11 IDA 0 16 1Z 9 S IBRD0 0000 Private creditors 3 7 i~

Continued

169 UGANDA TABLE IV. 1

EXTERNAL DEBT - Continued

1986 1987 1988 1989 1990 1991

Disbursements , , . , , , 180 , , , ,,,3,, 2,3 Long-term debt <8 0 8 1 0 7 Public and publicly guaranteed debt 0 4 185 3 0 Official creditors 1$ 2 4 6 6 7 Multilaterals .3$ 2 117 146 233 1- IDA 6 1 3 9 1 3 IBRD0 I RD § g fl ~~~~~~~~~~~~~~~~~~~~~~~~.0 .0...... 0 Bilaterals 7 31 3 236$ Private creditors 32 j11 7.., 3 4j Commercial banks 2. ...41.1.0 Bonds ... Other private 3B ... 4 Private nonguaranteed ...... 0.0...... IMF purchases 0 6 25 47

Principalrepayments 1; 1 1 7 Wi Long-term debt 1 4 .. 549 6 Public and publicly guaranteeddebt 31 E: 74 58 48..E Official creditors Bilaterals itiX ~~~~~~~~~~~~~~~~~~~~~~.22 Al 2 ...... 23...... {f_EEEEEE 29 Multilaterals114 19 2181 IDA 1..1... .. IBRD2 33 45 Bilaterals 2 ...... Private creditors 2.0 3. 2 , 2,., ,,, Commnercialbanks 14..... Bonds ' ' 0 i 0 0 Other private14 1326 13 2 Private nonguaranteed 00 IMF repurchases 7 77 6463 Lnterestpayments49 441 4333 Long-term debt 22 23 22 23 15 24 Public and publicly guaranteed debt 2 2 2 52 Official creditors 18 21 19 1914 2 Multilaterals 5 18 111 13,... °ngUaranteed , ''WL Ej 2 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~...... , IDA $$4 $ IBRD 3 22 Bilaterals 3.....2 Private Creditors4 Commercialbanks21 0103 Bonds 0...... Other private 1...3.1.. . 1.. Privatenonguaranteed ...... IMF charges25 U6 38 Short-term debt 12 3 4 Continued

170 UGANDA TABLE IV. 1

EXTERNAL DEBT - Continued

1986 1987 1988 1989 1990 1991 Total debt service 167. 1:4 170 128 13X Long-term debt . 0 9- 1 :64. - Public and publicly guaranteeddebt . 65 6 81 6* 9 Official creditors32 4 0 437. 4 Multilaterals 26 S 6 .. , 1 . 36 IDA 6657 10: IBRD 66. Bilaterals 2 28$ 10 Private creditors ...... i 4- Commercialbanks 421 1 3 1 Bonds ...0 ... 0. Other private 16 2 4: Private nonguaranteed 0 0 0 0 0 0 IMF repurchases and charges 95 - 5 - 41 Short-term debt (interest only)- - . Net flows 7 4 1 4 O 5 Long-term debt 11 24.~3~ ...... Publicand publiclyguaranteed debt l4g | 111 2 ! 25 Official creditors . 4 7 3 24- 4 Multilaterals 1 124 124 215 133 IDA 114...... 211 . 1. IBRD 4: 5ii t<->- Bilaterals. 13 30 Private creditors -- s 11. . 16.. -3 Commercial banks 2 5 .1l8A-1 -1 Bonds 0 0 0 0 0 0 Other private 18 10 17 -10 2 . -21 Private nonguaranteed 0 0 0.0 .. .. . - : IMF credit P 15 7: 1 38 45 Short-term debt 13 -5..... 6 .. . Net transfers ~ iuw(26~ 2 6 Long-termdebt 1 1 483 Public and publicly guaranteeddebt 2 $ iSE 231 1 Si Officialcreditors 116 ZW 21 ,821 2: Multilaterals 06 206 0 17 2022 1:14: IDA 61 10 7 92 06 12 IBRD 6 - 6 - Bilaterals $1 1-712 28: $~ Private creditors211 14 1 -0 Commercial banks 44 2 . 1 1 Bonds0 0000 Otherprivate 16 1 2 -2. Private nonguaranteed 0 0 0 0 IMF credit 4 257 Short-term debt (interest only) -: _: .:.-:--. .4...-

171 UGANDA TABLE V. 1

CENTRAL GOVERZNMENT OPERATIONS In millionsof U Sh

Est. Budget 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93

Total revenue 2S4...... ~ ...... $ 069 Tax revenue, fees & licenses 4q4qL Q7,U36~17)$ 2f0 olw Coffee 9 37 41 170 250 Non-tax revenue ~~~ 47 Totalexpenditures ~4~I l ~ ~ M ~ 1P 7~ Current expenditure ~~9Z ~ ~ ~ P0 Wages& salaries 70 97 0 243 476 600 Interestpayments ~~Q $7 1~ OPI 1~) Other 20 10 420 47I 75 S7I 975 Net lending/investment 00 12 0 10 70 Capitalexpenditure 44 90 '10 4w 1~1 491 276 External W J40 i07 46$ V S Domesticcounterpart 12 80 40 16 95 28 20 Localcapital 006PZ 4L1 27 . 00 Current cccountbalance 2 ~ 10f 37 Z$ 17 80 Overall deficit(comnmitment)-83 t 4~4 640 124 381 Changein arrears, net $ Z -4$ 16$~~ 21

Overall deficit(cash) -- 434 -X0

Financing:$2. 30 404 344 444 .:.:9.:$437 Budgetarygrants $ $P 140~ 26 14~144.3~9

External,net 4N)KQ ~0 11 *%8 439 261 Borrowing 1 *0 20 54 62 3~S 261 Repayment ~~Q 40 $) Other 1/ 0': .40...... 61. 49 )61 4 Domestic ..... 7.0...3..... Bank 49 10 130 134 36 19 6 Non-bank "0 10 .91 50 4 2 0

Memorandum item: ...... Revenueand graLnts07 94 24 .41 2693 809 Overalldeficit plus grants 449 '110 -. 44 -M8 -46 -916-247

Source: Ministry of Finance and Ec-onomicPlanning and IMF.

I/ Includes debt rescheduling, moratorium interest, etc.

172 UGANDA TABLE V.2

CENTRAL GOVERNMENT REVENUE In millions of U Sh

Est Budget 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93

Taxes on incomeand profits

PAYE(Pay-as-you-earn)344$ 1$ MI 2 27 Company/individualprofits and Pai a 4Z1$

Taxeson goodsand services $~O ~7E1 ~1S7

Salestax 4O Local'0 44 0# 9 1$ Imports ~~* ~ ~ ~ 0 ~ 40 Exciseduty '4 0$ 7O h1 10W . Commxercial1transaction levy' 1~'~4 .

Taxes on international trade ~ j M$~~

Importduties f0 P~~ Petroleum ( 10 3t 66 6~~71 I Other P t 44 2Z $4 *4 Exportduties 1 4 44 $23 2$0 Coffee 0 ~ 1 ~ $S 0 Other 3 00

Other tax revenue70 Fees and licenses 4a ~ 0 41 ~ 0 ~ 4

Totaltax revenue,fees & licenwse 4 O1 M 9 16 tIS m

Non-tax revenue16 * 12 S1 44 4$i R1

Totalrevenue $4. 4$4 46 6S 19I

Grants 0 0$ 464 )30

Totalrevenue & grants f3 Z t l04 34 G69 O$ M

Source: Ministry of Finance and Economic Planning and IMF.

173 UGANDA TABLE V.3

CENTRAL GOVERNMhENTEXPENDllIURE In millions of U Sh

Est. Budget 1986/87 1987/88 1988/89 1989/90 1990/91 1991/92 1992/93

Total currentexpenditure 490 2.0417 Z~ 1O8 3Z3017.~9

Economicclassificdation and csfao4o00 this tablc WagesaLnd salaries O 0 f 4l 446 O0 Interestpayments nerI/ be cnse withText Table 71a7. Other 1~10 ~ ~6%6 a~rn £~7a t~7 Functionalclassification .9...... ~ S9~0 Generalpublic services 17. 34 ~4 ~3 ~9 Defenseand security 3 44 M Z174 Education 3 $$ W5 US .M495 Health ~40 '41 40 Other social services ~55 Economicservices ~0 ~ ~ ~ 84 Interestpayments 0 30 50 55? 14 09 150 Total capitalexpenditure *2 90~~R0.~9 30$ 24~8 374 Economicclassification415 1 308 2491 349 Domestic 4$ O0 4O 4 95 138 External 33 $10 0 420 805 237 W* Local capital $ 11 42 Z± 10 Functionalclassification 44 40 5~~10~~S*,**.****~* Generalpublic services X1 94 543 Defenseand security ~ 4' S Education *7 '1 '~40"2 ~ 93 Health 41 38 Other social services21 53 '7' 357 26 Economicservices 84 t08 145 091 81 461

Net Lending/Investment 00 * ~680 .0 Total Expenditure. 14M

Source: IMF.

Note: Because of different sources and classifications,this table may not necessarily be consistent withText Tables 7.1 and 7.2.

174 UGANDA TABLE VI. 1

MONETARY SURVEY In billionsof U Sh

1986/87 1987/88 1988/89 1989/90 1990/91 1991/92

Foreign assets, net ; .9 1- 1 :- - 7..

Domesticcredit 2 2 S5 7, 2. 0. Claims on Government, net 4 1 11 9 26 92 12 64,0- o/w Ban of Uganda 4...2.2..... 10 14.3 6.32 Claimson privatesctor 4 1lllgi 1x| 42 1 70. 10. l| .,,,,,,>,,.00 ,...E l00 ,lS, ...... Ote r 26 7t7226.46 .1 6. } '.'

Total assets $4 46 ".. 'g0 7

T otliabilities al b4 S 4.. ' ......

Broad money,M2 " '''R'': ' '0 RR 4, 1'. 222.'R3"'"""""'R

Currency in circulation 0a4 14 4-v 2- 2 3 |lg.6 i 562 4.

Otertms,-net 14000l 213' 5'6 1°tti*03 10900l4 299.2 i 1

o/wCurrency revaluation 14 1 14'l>'0"R'1tg -4'''i3 Rig2t=^<,2 ,<,, ......

Memorandum items:

Velocity ...... 1.4.6 .. 12.7

Source: Bank of Ugandaand IMF.

175 UGANDA TABLE VI.2

INTEREST RATE STRUCTURE In percent

1984 1985 1986 1987 1988 1989 1990 1991 1992 Dec. Dec. Dec. Dec. Dec. Dec. Dec. Dec. June

Bank of Uganda. Waysand means . S I 14 4 Re-discount rates. Bank rate to commercialbanks4$51 M:

Treasury bills 35 days 2~ 2 63 days $ 4 91 days 2 2 2 1 3 0 2

Government stocks Ilyear 4 $ 5 years2$ 2 40 3 0 4 4 4 lO years ~4 ~ 4 P 4 15 years.2 . 7

Commercial banks Deposit rates I/ Demanddeposits 5 57 1 0 I Savings deposits 11 2 9 2 $ Time deposits I/ 3-6 months 2 5 1 13 4 3 7-12 months 2 9 3 0 3 $ 2 $ 2 Minimum 1year 2 2 3 2 2 7 2 6 3 Lending rates 2/ Developmental 4 2 22 2$ 2-P 3 3? 4

Source: Bank of Uganda.

1/ Minimum rates. 2/ Maximum rates.

176 UGANDA TABLE VI.3

EXCHANGE RATE MOVEMENTS U Sh per US$ OfficiaU/ Bureau Real Bureau Official Period Effective Bureau Average Average Average Rate Market FY & CY Official FY & CY Official/ FY & CY Sept. Mid-Rate (Note 1) Mid-Rate (Note 1) Bureau (Note 1) 1989=100

1989 January February 461$:~~ ~ :...... ,. March in tc April 2I 0 May June 4 July August Z.1 September 0 October November74 247 December less coptiie 1990 January to Dse February 1 March44 $9 April64 $ May June ~.6l4 44 August 745 4M4 September 114 4 October ±4644 November75 140 December 79~0 1991 January17 0 February60 March .40 April 46441 May 7 S40 June 0 5 July ~119142+ August ~ . 0 ~ September 11517 October 0 4 November .:31 December ~ 14~10 f 94I 1992 January 4Jfl72 February 16 March .34 April 2 15 614 may 109, 5 June 12 10 16 52 6 4 Julyv~17 .. 4 August ~ 5 4 September ~ S 15913+ October 2211 434 November 4119962A December 1$ 24 f4 7

Source: Bank of Uganda and IMF. Note: An increase in the real effective exchange rate implies that Uganda becomes less competitive. / FY refers to fiscal year, July to June, and is read from the June line; CY refers to caendlar year, january to December, and is read from the December entry.

177 UGANDA TABLE VI.4

FOREIGN EXCHANGE AUCTIONS U Sh per US$

Spread: BureauRate, Bureau/ Total Declared Auction Clearing Highest AuctionDay Cleaing Sold Amount No. Date Rate Rate (Note I) (Note 2) US$ m US$ m

1 3 1-Jan-92 W 9%235. 2 07-Feb-92 IO1 *01..4...... 3 14-Feb-92 9919 l . 1%2.5 ...... 4 21-Feb-92 9...... 3... .. ::::::.....3iu 28-Feba-92 % 3~2~ ...... 3 1-Mar-92 4..

9 27-M aLr-92 ...... 2.Z ...... 4. 11 20-MApr-92 1.. 10....f4 . * ...... 12 21-Apr-92 . ~~~~...... 0.1* 7 .1 3.... 13 240-Apr-92 :::.i:tt%I0~12 0%0 14 04-MAyr-92 1. 100 117..3 0 1.

17 22-MAy-92 ...... 2.clS 193 05-Jun-92 101$..1...... L.... 21 190-Jun-92 021) 7 xf . . 23 03-Jul-92 12 0614 7 42 24 10-Jul-92 ~14 257 17-Jul-92 02W614 0 . 26 240-Jul-92 1021 1316 ... 25... 27 312-Jul-92 1510%041$ 29 14-Aun-92104141*75%031 31 28-Aug-92 12 051.61.%O . 232 04-Sep-921011011513%214 33 11-Sep-92 10 *X04x ZZ. . . 35 17-Jul-92 02 151...... 36 024-Oct-92 ...... 1151 10 . .4..1... 3 7 312-Oct-92 1010 1601 . .

41 06-Nov-92 3'061525 429 13-Nov-92 13 0114 06 . . 30 27-Nov-92 *05lO`11*13% . 461 11-Dec-92 103 4.11....2 . 1. . 2.0..... 347 18-Dec-92 13 491.714 . . 48 28-Dec-92 15 20 .0.0713 .. 1..0 Source:Bank of Uganda. xx...... Notes:~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~X.. : . 3 TC6 Bureau-9trvelrshekrae 2/Fo7 ti calOcultion,thceaigat..cuds..Ucareso 39 2.43utiOaction no..8.nclusve;.threaftr,.2%

178~~ ~~~~...... 4- 4.t" 'A IN '17, It "troun-,S lw Z ...... 't.c. .k " M a .. C4 , AKon N M.1m. n: ...... : ...... : . - ...... : ...... : ...... kd ...... N ...... S. S. 4v q, r4:ln.Ml Ua*.lV..-.* .... . fm Ra wl,:- :: _14 . .... Of M.M.M.. ..an ... . r--t- W g -M.g t- N 'O 00 00" kn Ttl v ...... bn A t:K 5 4.,rwt tRinm .... *tn n tinv-, 4-w kn n- . : 4 rnininin vn,knln3n:. ht Nni ...... *,., W r- r- V ;: t ...... :...... :...... ::... - ...... - - -:...... :...... m...... : ...... - ..... :...... M...... :...... mkV& tn tn kb r4o 4p ..... ba: -4 Nw _t, M.-INPIR t::n Z. ig BS ..if4 :64 ...... XS. .... r2:...... R A rq 'A m ...... X:4 :Avc &4 ...... tn ...... O*: M) C44n: W:: ...... - . ... :' ...... q ...... ::::: ...... onn'm %CMA .'t.- M. b Ar.. 40 :.66...... X, ...... :q so .... ]:::: ...... Is ...... A.. ir .:tn...... :::+: ...... O.s M X.Mm xq ... Z. PA. . ..Kin ...... ]x]: ...... F: ...... X" :...... ::.... -b -4

sm ...... - ...... LL

...... am ......

......

Z ......

U bo -5 L' 'O 00 Z rA 0 Z C4M %no t- 00 0% CP% N 00 00 oo 00 00 Co Co at) Ch CK a, a 0% e.,2., O, 0, Ch Ch sz ...... g4 ...... W...... C14 Cq 0.

......

...... I!Xw ...... RR ii...... CPO 00 ...... X ...... 0 0 ...... n . ....

...... sm ::::...... e ...... tn ...... m::::Ad ...... 00 ...... Ch

......

...... 4 ...... C ) ...... z w ...... -6' .00 -0 bo 0 40 0 bo 00 .5 49 P 0 40 C), u S-* 93 es twD -13 0 co CQ 0 o,

-ta tA

mo-:sn 0 fli-nudHo COCA u k- -o AQU 4 0 ...... A. . . O N.. ff M ...... V ... ..

R...... a, 'n A...... Pt.ilost...... mscov......

CON

...... :: ...... :::::.. ... - " :...... m: ...... :::...... c), ...... -Wm ...... ' *vi +di ...... 4 ...... C ...... QN. ;l ...... :...... Mzr. ; ...... W ...... I...... rIP4 _- ...... - ...... :. .:.... : ...... -, ... .q...... - t ...... _.:Vt 0 ... 4CI6$M ...... :!k i6 ...... tz ...... n ... 'Itt: 1 i ...... to ...... k ...... Wo...... -J , J. im Li...... 0 ...... in ga m tlvl. MM ...... gm ! ...... 1...... :::...... :...... C j ...... :...... 901lk...... -tq...... -,I... C. M C> ek n 0)04&TA._-: ...... oq nr ...... -...... 4 M=:zB ...... IS...... Xlg: 9 ...... X ...... :...... :.:: ...... W 00 ...... :::...... q.: ...... %n...... n ii 4::...... le IMR:-.4,04R, Rma: ...MN'.C "bf-An X...... x.x xx...... :'LM R ...... :: -1- ....:::...... X.m. ..x ...... - ...... x ...... > y 6,>O <2" 0- 00 Oo Q. "-W <.Ozlm A LI. cq as 0 k'Q :3 00 00 oo ao Do co 00000, Q\ 00 (71 0\ 06 zw c 0-4 --n

......

...... ii, C14 -q ii i Mx.. t Roxti . . ... M wiffilvii: z . -. 4 W -i4O ...... m ...... X ......

a ...... M O X...... X...... RA ...... 1 S.:: . .!-9 ...... ad 1 16 ...... a...... to .A. M. (O ...... -X ...... a ...... a m--m X.: ...... W A : ...... Xm:m ...... -...... so- ...... W X ...... is *k J...... X ...... 0: ...13 -1Mmi, R R .:.4.,:rW W...... ok: ...... W .. I I4.V .i 40 m : ;"V: 6. Mw: :40: ...... X -m... -100 .... :::::: ...... X :: :: ------...... %...... :.:.:.:.:.:.:.:.::::::::...... -X ...... ::...... : :: ...... ;...... NOW 4 3...... M ai.: A W ...... a a '.Wd :A W : K a : w IMI.M...... N, :...... ft,..J ..'s...... a...... a...... m ...... N ...... i9W M. T. Naigilwo. m:::,%v:R F: zj co K: An.W. -k 0::m ...... :::.7::: ...... :: - .: ...... M...... ;:;mmmmmmm ...m mmm...... > 8 9 .8 0 clooz > cd Z C) O Z0 cl 0 Z 0 Z N k C,4 cn -,t Ln O t-- oo (7, 0 C) C-4 0, (71 - ("IC, (Z Z H ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~M

w. .0a~~~~~~ Li. ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~......

N, ...... 44*4+4~~~~. . *w ... 44 4 . 4 44 4 4- - 44 ...... -.....4 t4: c .0 '-44 *~~~~~~~~~~~~~~~~~~~~~~~~~~~~- 0 ...... 0 ... ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~ .~~~~~~~~ ...... 0.~~... ~~~~~~<~~~~~~~~~~~~~~~~~~~~<~~~~~~~~~~~ OZ c j, ~ ~ ~ ~ ~ ~ ~~~~~~~......

0 ~~ ~ ~~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~~~~~~~~~0

~~~~~~0000~~~~~~~~~~~~~~~~~~~~~~~~~ a' a.~~~~~~~~~~~~~~~~~~~~~~~~~~~0 UGANDA TABLE VIII.2

NEW CONSUMER PRICE INDEX, KAMPALA September 1989 = 100, All Households

All Items Annual Percent Index Change

Calendar Year (January - December)

1982 1983 ... 7 183...... 1984 10...... 1985 ...... 1986 ...... 198719 1988 6 1989 1 1990 12 ~ 1991 16 8 1992

Fiscal Year (July - June)

1982/83 0

1984/85 ...... 1985/86 ...... 1986/87 13 ~ ~ 1987/88 1988/89 ..... 1989/90 .1....~ 1990/91 p7

Source: Statistics Department, MFEP.

184 UGANDA TABLE IX. 1

ORPHAN CENSUS DERIVED FROM THE 1991 NATIONAL CENSUS Provisional Results

Orphans as % Total Population in District Children Orphans of Children Final Census Count

Kumi Ol3 74.%.3$ Rakai Gulu @~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~.:. S491% .ei...... S o roti ...... Gulu ,Q 19 4 3,7 Kitgum 2-5614 3n49 Mayo I$2122 % 74 Luwero 2 :.. 2 Masaka , .4% 54 Kalangala '"' "6R,, 11. .'.. K~~~~~~~~~~~~~~~~~~ bog...... -YI0 Moroto 407 2%147 Mpigi .1...... : Mnioi :,~~~~~~~~~~~.o311 - :.,':,,'...,...... R1''''''4',;a:i.yS Rukungiri .,,,.,,21,6,,.,,,,,,,,. Kabarole 1991Pic Kotido ...... % Mubende ...... ~ P1~) Lira 397 4 71% Mbarara 43rare a.Orp 15years Kampala ~%7,4M1~4 Mukawo have lost oertpes Masindi 120,156 11~406 ~% 26G~198 Kabale2044 930% 1 5 1 Nebbi14,1 9% Bushenyi $ '9.94 25973,1 Bundibugyo M9241R9 Arua 9.7 496% Apac ...... 1401...54.G Kibale 109 ~ 3 ~ 2,6 Hoima h4 %1h5 Tororo $I2 ,*1% Kisoro X 7 9 ...... Jinja 3,9$9347 Pallisa 4,f $V1 Iganga 465 ~ 437 Kamuli 22,4 4,7 6 #;1 Mbale 343 0526 1,B Kapchorwa . ~ ,193576 K~asese 6%

Source: The 1991 Population and Housing Census, Statistics Department, MFEP. Note: Children are under 15 years. Orphans are children under 15 years who have lost one or both parents.

185 e +v - - S d~~~~~~~*flONNtt~~~~~~~~~~~~~ttt..*~~~~~~~~~~~"*;5 - .. S~......

S =X O~~~~~$~~~~~ ~ ~ ~ ~ ~S~ v . r .-.....

, ...... ~~~~~...... ~~~~~~~~~...... ,...... , , ....* ,,. O...... ot.1..ntAao>~~~~~~tt~ ~~~.r)r.. g~~~~~~~~~~~~~~~~~:7ii. iii i iiii ' : :'"t' i-'-.:''i .'' .....

i~~~~~~~~~~~~~~ z z

H H U R~~~~4 a - 0 4)

| E 5 = -.gc. 3 X0 28 a g rci...... m00=gSov>>zuz6SESgL-oca4c:6H UGANDA TABLE IX.3

PERCENTAGE SHARES SPENT ON SELECTED EXPENDITURE CATEGORIES In percent

< ------Percent accounted for by ------> Urban Rural Total Non-poor Poor Poorest Total

Health -. 0 -. ;. - - -. .- . Education d 7 10: Clothing 21- 2. . 1.. Charcoal - : . 7: 1. 4 -0- Firewood 0 ...... 70.. . .. 2... . 00..

Market food .01''|- Bread and cereals 3 6 100 77 15 4 '00 Meat and poultry 7...... ~..... Fish22 7 1062 31 10 Milk, chees and eggs 4.. 1051 1 ...... 1...00 Oils and fats .. 100 .. 73 ...... Fruit and vegetables $ X 7' Matok51 4 10799 2 10 Potatoes and tubers 1..6 ..... 2....1. Sugar 4 6 077 2 10 Salt 1urve0 ...... Other food 24IS6 100 73 2 4 100

No4n-m~arket:edfood ~~ .9...... Bread and cereals 97...0.61...... Meat and poultry 6 94..... 3 .. 1.. Fish310 26 10 Milk, chees and eggs ...... 4 ...... 20....Q. Oils and fats 6 9 .00....22... Fruit and vegetables 10 .. .. 6...... 1 Matoke3 97 1071 24 10 Potatoes and tubers 3 9 0 31 0 Sugar 3..... 00.1...0... Salt3 97 001 10

Other food 100::r

Source: Household Budget Survey

187

Annex II

Poverty: A Child's View

191 Annex II

Rich people sleep on a bed (top left house) and have electricity (light bulb hangingfrom ceiling in top left house). They own many things, such as spoons (bottom row in top right house), cups (above spoons), plates and knives (above cups). Their houses have tiled roofs (bottomhouse) or iron sheets. They can afford shoes and handbags, and even have cars (they look like buses, but it is because they have to fit the wholefamily!). 192 Annex 11

, ~~~~~~~~~~~~~~

-m~~~~~~f~

Rich homes are tidy (see items arranged in boxes) and there are many things in them, including glasses, cups, dresses, chairs. They have electricity (bulb is hangingfrom the ceiling, between boxes of items) and ironsheet roofs. Here too, rich people are depicted as wearing nice dothes, with shoes and handbags, and having a car. 193 Annex II

.< S~ ~el E . ~

Th symbols of wealth are cars (with radios!), enough money to go to a nice school (two-story building with trees around it), a television (square at the bottom). This is the only child who depides a television, while many drew radios. 194 Annex 11

A poor household.

Thepoor use a hand hoe because they cannot affort ox ploughs. The family owns a bicyde, used by the father. Houses have a thtched roof, often akang (that's why there i a person on the roof to repair it). Water Is carded in old fashioned day pots (home made) rather than in modern but expensiveplasi jenry cans. Annex 11 195

15~~~~~~~~~~~~~~~~~~~~9

Poorpeople live in sunple huts with thatched roofs. Their schools are in poor shape (note shaky IYnes)and pupils have to sit on benches rather than having proper desks. Poor households may own a few chickens, but certainly not cows. Fanners can only afford localUy made hand hoes to cultivate their fields (see two people at the bottom), and grain has to be pouded by hand because more efficient technology is not aVailable(see person with stick, on right). There are snakes near the schools and houses of the poor, because compounds are poody kept (a refletion of high dependency rates and time-consuming work practices?). 196 Annex II

Poor houses are undy and have thatched and often leaking roofs. Poor people receive donationsfrom the rdch (here they receive 100 Sh) and have to eat cassava. They are barefoot. Women can't 4ffordplasic jerry cans to car7y water. AnnexII 197

Poor people receive donationsfirom the better off (the child and the person on the dr are both receivinga donation). 7Theycarry waterin day pots ratherthan in mnodernbut expensive jerry cans. Even the pots used for cooking are homemnaderather than rnctal ones (see them inside the house).

Annex III

Adjustments to the HBS Data

Annex III

Adjustments to the Household Budget Survey Data

1. The fieldworkfor the HouseholdBudget Survey (HBS)took place over the course of the year from April 1989 to April 1990 and the enumeratorsvisited a differentset of parishes in each district on six different occasions during the year. Inflation in Kampala over this period was about 40 percent. Consequently,if we imaginetwo identicalhouseholds, with identical purchases of goods, one sampledat the beginningof the fieldworkand the other sampled at the end of the fieldwork,then the second householdwill have expenditureswhich are approximately40 percent higher because of inflation over the year. So if we had simply added up total expenditure for each household for comparison,we may have concludedthat otherwise identicalhouseholds had a 40 percent difference in welfare. Similarly price variations across the country would lead to biases in comparing households. Therefore, in order to correct for inflation and other regional price variations, we calculated a set of unit-valuesfor each parish sampled, using the information in the HBS on the quantitiesof food consumed,from which we derived a food price index. For the adjustmentof food expenditure, we simply used this food price index. For total expenditure, in the absence of any additionalinformation on non-food prices, we assumed no non-food price variation in constructing an overall price index.

2. In the HBS, when householdswere asked about the value of food from own-production,the values sought were equivalent to farmgate, or producer, prices. So the reported values of food consumedout-of-own-production are understatementsof the consumptionvalue of that food. So we have revalued the recorded values of consumptionout-of-own production at market prices, so that the consumptionvalue of all food is equallyreflected in the valuationsrecorded in the HBS.

3. It needs to be stressed that like all large survey data, the HBS data will carry statisticalnoise from unavoidablesampling and enumerationerrors. The informationfrom the HBS has proved to be robust in aggregate;but we are utilizing a level of detail, particularly in the calculationof the price adjustments,which could lead to some increases in the statistical noise, rather than unambiguous improvementsin the estimates. Further refinements are presently being made to the data by the Central StatisticalOffice and, as mentioned in Chapter 1, there is also further analysis planned to estimate "equivalencescales' for Uganda. All of these efforts, together with the ongoing SDA survey, are expectedto further develop the databaseand provide additionalinsights into the poverty situation in Uganda. The results presented in this report represent a first attemptto put together a poverty profile for the country, based on the HBS and other supplementaryinformation, and should in no way be taken as final.

Distributors of World Bank Publications

ARGENTINA The Middle EastOberver KENYA SOUTH AFRICA. BOTSWANA CaroosHirsch SRL 41, Shenf Street Africa Book Servi (E.A) Ltd. Forsingl tite: GalerhaGoemes Cairo Quaran House. Mfanpno Street Oxford Universty Pren Florida 165,4th Floor-Ofc. 453/465 P.O. Box 45245 Southern Afnca 1333Buenr Air" FINLAND Naiurobi PO Box1141 AkateerunnenKirjakauppa Cape Town 800o AUSTRALIA, PAPUA NEW GUINEA. PO. Box 128 KOREA. REPUBLICOF FIl SOLOMON ISLANDS, SF-00101Heinki 10 Pan Kore BookCorporbon For sucrnpt ordm: VANUATU, AND WESTERN SAMOA P.O. Box 101.Kwangwhamun International SubscriptionService D[A. Books & journals FRANCE Seoul P.Q Box 41095 648Whitehore Road World BankPublications Craighall Mitcham3132 66,avenued'I6na MALAYSIA Johannesburg2024 Victoria 75116Pais Univerity of MalayaCooperative Bookshop,Limnited SPAIN AUSTRIA GERMANY P.O.Box 1127,Jalan PantaiBarm Mundi-Prensa Libroe SA. Gerold and Co. UNO-Verlag 59700 Kuala Lumpur Csteilto37 Graben31 PoppelsidorterAllee 55 2001 Madrid A-1011 Wien D-5300Bonn 1 MEXICO IN~FCEC Librerta IntermacionalAEDCE BANGLADESH HONG KONG. MACAO Apartado Post]l 22-860 Consell de Cent 391 Micro Industries Development Asia 2000Ltd. 14060Tlalpanr Mexico D.F. 038D9Barcelona Assitance Socmety(MIDAS) 4648 Wyndham Street HousS, Rod 16 WinningCente NETHERLANDS SRI LANKA ANDTMEMALDIVES Dhanmondi R/Area 2nd Floor De Landeboom/InOr-Publikaties Lake House Bookshop Dhaka 1209 Central Hong Kong P.O.Box 202 P.O Box 244 7480AE Haaksbergen 130.Sir ChittampalamrA. SmBac ofltca: INDIA Cardiner Mawatha Pine View, Ist Floor Allied Publisher, Prvate Ltd. NEW ZEALAND Colormbo2 100Agrabad Commercial Are 751Mount Road EbSCONZ Ltd. Chittagong 4100 Madras- 6C 002 Private Mail Bag 99914 SWEDEN New Market Forsingietilgf: 76. K DA. Avenue B,xcl. offce: Auckland Frites Fackbokstoretaget Kuina 9100 ISJN. I-erediaMarg Regeringspgtan12 Box 16356 Ballard Estate NIGERIA S-10327 Stockholm BELGIUM Bombay- 400038 University Pref Limited JeanDe Lannoy Three Crowns BucidIngJericho Forsuseyhs ornder: Av. du Roi 202 13/14Ast Ali Road Pnvate Matl Bag5095 Wennergren-WiliLarnsAB 1060 Brusels New Delhi -110002 Ibadan P. O.Box 1305 S-17125Solna CANADA 17Chittaranpan Avenue NORWAY Le Diffuseur Calcutta- 700072 Narvesn Information Center SWfTZERLAND CP. 85 15018 rue Ampere Book Department Forsingle titr: BouchervmIleQubec JayadevaHostel Building P.O. Box 6125Ettertd Librarie Payot 14BSE6 5th Main Road,Gandhinagar N-0602 Oslo 6 CAsepostale 3212 Bangalore- 560 009 CH 1002Lausanne CHILE PAKISTAN Invertec IGT S.A. 3-S-1129Kachiguda Mirma Book Agency Fwr uss-ts mis: Amenco Vmpuio Norte 1165 Cross Road 65, Shahrah-e-Quaid-Azam Llbrane Payot Santiago Hyderabad -500027 P.O Box No. 729 Service de Abonnements Lahome54000 Cae potale 3312 CHINA Prarthanr Flats, 2nd Floor CH 1002Lausanne China Fuancial & Economic Near Thakore Baug Navrsngpura PERU Publishing House Ahmedabad - 380009 Editorial Desarrollo SA TANZANIA 8, Da FoSi DorngJie Apartado 3824 Oxford University Prin Beijing Ptiala House Lima I P.O. Box 5299 16-A Ashok Marg Maktaba Road COLOMBIA Lucknow-226W01 PHILlPPINES DaresSlaam Inoenlace Ltda Internbonal Book Center ApartadoAereo34270 Central BazaarRoad Suite 1703.Citylarud 10 THAILAND BogotaD0E. 60 Ba,ajNagar Condominium Tower I Centri DepartaentStore Nagpur 440010 Ayala Avenue. HV. dela 306Silom Road COTE D'IVO IRE Costs Extension Bangkok Centr d'Edition et de Diffusion INDONESIA Makti. Meo Manu Africaines (CEDA) PL Indira Limited TRINIDAD & TOBAGO, ANTIGUA 04 B.P.541 JalanBorobudur20 POLAND BARBUDA. BARBADOS, Abidjan 04 Plateau P.O.Box 181 IntenaubonalPublishing Service DOMINICA. GRENADA. GUYANA, Jakarta 1320 Ul. Piekna31/37 JAMAICA, MONTSERRAT, ST. CYPRUS 00-677Warzawa KlTS & NEVIS, ST. LUCIA, Center of Applied Reearch IRELAND ST. VINCENT & GRENADINES Cyprus College Govemment SupphesAgency For,suicratxn ordmer: SystemritcsStudies Unit 6 DiogenesSbeetrEngomi 4-SHarcourtRoad [PSJournals #9 WattsS et P.O.Box 2006 Dublin 2 U1.Ohurena 3 Curepe NIcosia 02-916Warszawa Trinidad, Wet Indies ISRAEL DENMARK Yozmot Literature Ltd. PORTUGAL TURKEY SamfundsLitteratur P.O. Box 56055 Livraria Portugpl Infotel Rosenoers Allk 11 Tel Aviv 61560 Rua Do Carimo70-74 Narlabahce Sok. No. 15 DK-1970 FrederiksbergC 1200Lisbon Cagaloglo ITALY Istanbul DOMINICAN REPUBLIC Licosa Commasionara Sarsoni SPA SAUDI ALRABIA QATAR Editom Taller, C par A. Via DucaDi Calabra, 1/1 Jarir BookStore UNITED KINGDOM Restsuraci6ne Isabeila Cot6fica309 Csella Postale552 P.O. Box3196 Microinfo Ltd. Apartado de Correos 2190Z-1 50125Firenze Riyadh 11471 P.O. Box3 Santo Domingo Alton, Hampshire GU34 2PG JAPAN SINGAPORE. TAiWAN, England EGYPT,ARAB REPUBLIC OF Easte Book Service MYANMARBRUNEI Al Ahram Hongo 3-Chome,Bunkyo-ku 113 Informatoas Pubcations VENEZUELA Al Gl"a Street Tokyo Privte, Ltd. Librena dedEste Cairo Golden Wheel Buading Aptdo. 60.337 41,Kallang Pudding, 00443 Caracas1060-A Singapore 1334

IBRD24881

SUDAN , -' / S ~~~~~~~~~~~~~~~UGANDA y - - > ~~~~~~~ADMINISTRATION ~t?t /Moio0 < KOTIDO

) ( ARUAsi J , ~~~KITGUM.a gS v3, MOYO * J 2 k t * ~~~~~~~~~~~~~~DISTRICTCAPITALS t RUA 2 IGMKTD 0<@ NATIONALCAPITAL

ZAIRE) s GULU -'' RIVERS ZAIRE{e V 3 g GULU 9NEBBIO INTERNATIONALBOUNDARIES

+/ ~~~~~~~~~APAC IRA LIRn,n2

-2' / l, * (_ )~~~~~~~~~~.PASOROTI vC ~~MASINDIJJ w a/ B~~MASINDI SOROT HOIMAS

,J KASESE Q } _ S MUKONMPIGOIGIL KENYA- _ (KAESE J 5>)\--. o L ¢\ / 8) ioM~~~~ASAKA/)i , :A 1> 9 ^ 2 MASAK~~~~11A KALANGALA , LaJke3 BUSHENI, MBARARA 4^g^

) ,1 2 e~~~~~.; e 9 @ < ~~KALANGALA| AgLGAN {RUKUNGilNee i_/ a RAKAI \+ ~~~~~LAKEVICTORIA t -AL 4 t ~~~at<' i~~RAKAI,>

f _;-^urr> ~~~~~TANZANIA \.

- RWANDA> 32 34 \ 3s ~~~~~~~~~~~~~~~~~~~~~~~~MAY1993

-4

4' - *#.s.

* -. 4-, , r" 'Z4

3-diome ., joG

F3)3Z1 T&3a24838 - - ''K. 'A- le - ." t

P it>

-* t>f

L4 K -/ .4.. ( ..*'- ,,-.

- fY '--v. i' i"

. --t, .

44 *5 *

j 'fl-i.

12460 CEM iQO 0 6 13-2460 6 JGANDA GROW CUT O POVE

40 QQ06i90 $12.95 BNO-8213-2460-8 -#w;' A - 5.

'.,-,'. .-.