Understanding Poverty and Human Resources in Zimbabwe: Changes in the 1990s and Directions for the Future December 1996 /\ ~U~'.IuJ~~7T~ f~j 4, i .a 7 r"- IS rt7 O7 Fi< I Vj;t(-;7-lt -itli-lfJ - ' |@ | I I Understanding Poverty and Human Resources in Zimbabwe: Changes in the 1990s and Directions for the Future TABLE OF CONTENTS EXECUTIVE SUMMARY ............................................................i INTRODUCTION .............................................................1 Objectives .............................................................1 Context .............................................................2 Sources of information .............................................................2 Scope of the paper .............................................................3 2. INCOME AND CONSUMPTION .............................................................4 2.1 How Prevalent is Poverty in Zimbabwe? .............................................................4 Estimates of prevalence vary, but household surveys reveal patterns and characteristics of poverty ............................................................. 4 Poverty in Zimbabwe is not as high as in many other countries ..........................6 2.2 What are the characteristics of poverty? .............................................................8 Poverty is predominantly rural .............................................................8 Poverty is most prevalent in communal and resettlement areas ...........................8 Most of Zimbabwe 's poor live in communal areas, but only an estimated one third of the population in these areas is poor .................. 10 The prevalence ofpoverty amongfarm workers on large-scale commer- cialfarms is high even though the overall prevalence on these farms is low ............................................................ 12 In resettlement areas, poverty prevalence is high but declining ............... 13 In urban areas, prevalence ofpoverty is low but probably growing ......... 14 The driest provinces tend to be the poorest ......................................................... 15 Female-headed households in rural areas are no more likely to be poor than male-headed households ............................................................ 17 Heads of poor households tend to have less formal education than do heads of non-poor households ....................................................... 17 2.3 What are the Sources of Income for the Poor? .................................................... 19 Rural cash incomes are predominantly from non-crop sources .......................... 19 Agriculture is the most frequent source of cash income, but not the largest ...... 21 Non-poor households tend to derive more cash income from formal employment than do poor households ....................................................... 22 Remittance income is substantial, especially in rural areas ................................ 22 Rural households derive half of their total income from in-kind sources ........... 24 2.4 Summary ........................................................... 24 3. STATUS OF HUMAN RESOURCES ........................................................... 26 3.1 The Population is Young ........................................................... 26 3.2 What is the status of health? ........................................................... 27 Zimbabwe has invested successfully in the health of its population .................. 27 Health status is good by international standards ................................................. 28 Prevalence of HIV/AIDS is among the highest in the world .............................. 28 Health outcomes are worse in rural than in urban areas ..................................... 29 Ill health is linked to low education levels .......................................................... 30 3.3 What is the status of education? ........ 31 Zimbabwe's achievements in education are impressive ........ 31 Education levels are lowest in the large-scale commercial farm sector .............. 31 Girls receive less education than boys, but the gap is closing ............................ 32 3.4 Summary ........................................................ 34 4. FACTORS ASSOCIATED WITH POVERTY ...................................... 35 4.1 Poverty and Inequality ........................................................ 35 Income is unequally distributed ........................................................ 35 Land distribution is highly skewed ........................................................ 36 The poor have little access to land ...................................................... 36 Rural poverty is associated with poor agricultural land ........................... 38 Land markets function poorly ...................................................... 39 Unequal access to water exacerbates the land problem ............................. 40 Agricultural services have not focused on issues appropriate to drought-prone areas ...................................................... 41 Environmental problems affect both good and poor lands ................................. 42 4.2 Why has the economy not generated employment? .............................................. 43 The formal sector historically favored capital over labour ................................. 43 Manufacturing has not generated sufficient employment ......................................................... 43 Commercial agriculture has used neither land nor labour optimally ....... 44 The informal sector is relatively small, and used to be tightly regulated ........... 44 Large numbers of school-leavers cannot find jobs ............................................. 45 4.3 Summary ......................................................... 46 5. CHANGES DURING THE 1990s ................................................. 47 5.1 What changes have affected the economy? ......................................................... 47 Structural adjustment aimed for growth and employment creation .................... 47 The reform program has only been partially implemented ................................. 50 Markets have been liberalized ......................................................... 50 The government has not managed to meet its fiscal targets ...................... 50 The Social Dimensions ofAdjustment Programme was poorly designed and insufficiently funded .............................................. 51 Over-regulation remains in some sectors ................................................... 52 Public resources could have had greater benefits for the poor ............................ 52 Drought had devastating effects, which better preparation could have reduced ......................................................... 53 5.2 What were the effects on incomes and consumption ............................................ 54 Incomes from employment have fallen ......................................................... 55 Formal sector wages have fallen sharply .................................................. 55 The informal sector has grown, mainly in low-profit activities ................. 56 Liberalization of maize markets has benefited consumers and producers .......... 58 Drought hit rural incomes and probably increased inequality ............................ 60 Effects on average incomes have been negative ................................................. 60 5.3 What were the effects on health and education9 . ..................................... 61 The health sector is under strain because of funding cuts and AIDS .61 Health spending has declined .62 Health services have deteriorated ........................................... 63 AIDS is affecting health outcomes, straining service levels and affecting the whole economy .64 Education funding cuts have not affected outcomes substantially .65 Education outcomes have worsened slightly .66 Future health and education outcomes are likely to deteriorate .67 5.4 Summary: What has happened to poverty in the 1990s? ................................. 67 6. TOPICS FOR FURTHER STUDY ............................................................ 69 Increasing understanding of the living conditions of the poor ............................ 69 Understanding formal and informal labour markets ........................................... 70 Deepening understanding of the dynamics of remittance income ...................... 71 Understanding the impact on poverty of changing agricultural markets ............ 72 Understanding how households cope with crises such as drought and AIDS .... 72 Addressing the role of institutions ............................................................ 73 7. ACTIONS FOR THE FUTURE ............................................................ 74 Implementing a poverty reduction strategy ........................................................ 74 Creating an environment conducive to economic growth ......................... 74 Investing directly in poverty reduction ...................................................... 75 Moving forward ............................................................ 76 SUMMARY OF SEMINAR PROCEEDINGS ............................................................ 78 BIBLIOGRAPHY ............................................................ 83 PARTICIPANTS IN POVERTY WORKSHOPS ........................................................ 93 TABLES Table 2.1 Poverty Prevalence by Land-Use Group, 1990/91 ..................................... 10 Table 2.2 Provincial Breakdown of Poverty, 1990/91 ............................................. 16 Table 3.1 Health Indicators in Zimbabwe Compared to those of Neighboring Countries ............................................. 28 Table 3.2 School Enrollment Status by Land-use, 1994 ............................................ 32 Table 4.1 Major Features of the Farm Sub-sectors in Zimbabwe, 1993 .................... 38 Table 4.2 Distribution of Communal Lands among Natural Regions, 1990 ............. 39 Table 5.1 Education Transition Rates ............................................. 66 . .*.*.*.*...*.*.....*.*.*.*~~~ ~~~~~~.*.*.*.... ...*.... ........... Figure 2.1 Percentage of the Population Living on Less than US$1 per Day ...............7 Figure 2.2 Distribution of the Population and of the Poor, 1990/91 .............................9 Figure 2.3 Education Levels of Household Heads, 1990/91 ........................................ 18 Figure 2.4 Sources of Cash Income, 1990/1991 ........................................... 20 Figure 2.5 Sources of In-Kind Income, 1990/91, All Households .............................. 25 Figure 3.1 Relationship between the Mother's Education and the Health of her Child, 1994 .......................... 30 Figure 4.1 Distribution of Land by Natiral Region .......................... 37 _~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ - CA Communal Farming Areas CPI Consumer Price Index ESAP Economic Structural Adjustment Programme GDP Gross Domestic Product GMB Grain Marketing Board ICES Income, Consumption and Expenditure Survey LSCF Large-scale Commercial Farm MMR Maternal Mortality Rate MSE Micro and Small Industry (in the informal sector) MPSLSW Ministry of Public Services, Labour and Social Welfare MoHCW Ministry of Health and Child Welfare NGO Non-governmental Organization NR Natural Region (also known as Agro-Ecological Zone) RA Resettlement Area SSA Sub-Saharan Africa SSCF Small-scale Commercial Farm SSS Sentinel Surveillance Survey ZDHS Zimbabwe Demographic and Health Survey I *, a 403moD 3 SZ et uN US pus sa > PCP1 -$p u pu aMwi JaScn iwd9ed sep Suedsd U! amumssw jUIAO poppAod laAra epv pus sq3a'B fU 'UOpio0 UOUlUA 'nul WJqO aBCmumw S sLp p'Bs rpeuvqWEmdqs) &qztnmnxysi '!4814N kuf!S 'AuNop pLojl '"Asuj 1g 'ioodwX pdsx tjw'i tsqn'aajr UOAa)S 'Uw"0 PMLu!M `4O0D P!AB(l WUOg JJSJLUWUO pUs SUO!IrIRu( jo0 In lfPaW oza adsd ap pnrpa qso;uppeN suotj 'ndud aqipjo -ep jazsoa us paatar *mp PWONA a tp aJO s 1 p UsA anbtyrtuoG Isruj o3iPumS Jo 9.bOJl agaC Pue AS!SnAIrn pzojxojo r HIMA . 'Aueg Pp0X aPA jo sJO enUs5so0jsoaX uVWaq Vw y UMjx JO SUfpt1% UWa)ON 'wafH-d93[NqfJo oJUljAj Usfqnl Euo suoCsa2nC pui sluauxtoa paljMap Uwog paiauaq os[p .Wdd aqL *amqrqtuwz tn sum crqm Ausap 48T pus Mrm O sd.n ppyg wgoy pus (sicw4ap.rdjo w,I aas) aisuH tq sdo4spom pus suorsnasrp IoJa eaM ml pop sauat a.ijo AuyW >lhfl Aq pajanpuoa utaq smow!ppua weos jWx S)AodJo SISAB ait pps 'a3MIqwJ0 h4!SA!n aioqmo vs aAru r ps saM qo4d Sq pusm 0P°MJM J sn4wtun atpjo uosIppv AuojL Aq iood atp ro suausnfpe rennasjo puduq ati uo qasas SujoS-uo aqs 'uasus-H tpw B j,rw 'ra auv Rq paiztnpuoo nan.puadxa a:qndjo ssl/sue uo 1aAUSUOJXO SMVlp .add aELL a4s1( awV pug uostppy Suoj° 'aWeleR LP!ON UHW uuJ 'Qogns uuw) iisuianu wqng,o pa ssuoo pCle 'oqrj sLczjaH Sq P0I SuM um aqLJ 0amqeqwUZ m suorPu o PM SpplAlp AScew 431M uopminsuoo atsolp u! 'a PpoM aqi jo uo!Sal mpyu w acpnoa pue un aQiJo dno suOwdolJAOQ uumuH atp wog Luem e £(q uauan uaaq seq iadud sja! IBRD 28092 26 28- 3'0 take Caboro Basso 32' ZIMBABWE '3j -2 - .16- PROVINCES - MOZAMBIQUE 6- *I NATIONAL CAPITAL Z A M B I A j ; * PROVINCE CAPITALS , , MASHONALAND -- PROVINCE BOUNDARIES MA'SH AN CE NTA -. INTERNATIONAL BOUNDARIES WEL k MASHONATAND CENTRAL 7 <' 0 lo0 200 KILOMETERS ar E ,, o ; Cbinhoyi0 NAMIBIA j<, ,Harare ( 18- 1 .(MASHONALAN,D S 18 ,r > g >> ~~~ ~ ~ ~~~~~~~~~~~~~EAST \, -) 2 )-/ < */~~~~~~~~~'MINGALAND eo MATABALELAND j -INGjLANDI, BOTSWAN A ( NORTH M I DL A N D S MJworeL \ '> *G-eru - . This map mcs produced by thie l -' - -'--\ Map Design Unit of The World Bonk. BuloMsyovingo The boundaries, colors, denominations * ond anyother information shoknon *J / > this mop do not mply, on the part of I The World Bank Group, any judgment > MASVINGO on the legolstos of any territory. or ,' any endorsementt or acceptance of fG'monclc/ such boundaries. I C - - 26. ' MATABALELAND t ,' 1 Z A AIRE TANZANLA SOUTH r ) ANGOLA _ -' 7-22- / 2 / j ~~~ZAMMI; _,t )f- . ;\ MO 28 SOUTH AFRICA ; 32- \ZIMBABWEj , 28 k B ,,,!XNAhMBIA , r1t>Ny + Np_1XN IBRBR 28093 SOUlH ,) j I- ~ ~ ,/ '' \-~' J OCEAN i8- io X tokeCohro80sso 32' ATLANTIC Z A M B I A _0 oceN- . %, .7 -. ............... ,6 r- _, LMOZAMBIQUE , 26- ; 00 l ' < ' 105 { NAMIBIA NATURAL REGIONS A NATURAL REGIONS- X BlwyoM vn ow/| UMI lib ( (_ / > I I V 1 c I NATIONAL CAPITAL * PROVINCE CAPITALS 22, 22' -. INTERNATIONAL BOUNDARIES B O T S W A N A , 0 100 200 KILOMETERS f 2 26- 28' _,_' SOUTH AFRICA - 32- JUNE 996 Urombo Poverty (in Shona) Nhai nhamo wakabva nepika Poverty, where did you come from? Uno shaisa mufaro vakuru nevdiki You deny happiness to women and children Usiku nesikati kuno ngova kuchema Day and night, they all cry, Kuchemera kugara hupenyu hwakanaka Yearning for a better life Urombo wakaipa - wakabvepiko Poverty, you are cruel - where did you come from? Peku gara pacho anongove matumba Our houses are just shacks Kudyakwedu - kupara - para sehuku For food we are scrounging like chickens Chokupfeka hapana kupona nokupiwa For clothing we are li-ving on donations Haiwa nharno isu tatambura Oh! Poverty, we are suffering tJrombo wakaipa - wakabvepiko Poverty, you are cruel - where did you come from? Dai iwe usipo taifarawo muhupenyu Without you we would live happily Dai iwe usipo kungadai kusina kuchema Without you there would be no suffering Dai iwe usipo taigarawo zvakanaka Without you we would have, better living conditions Dai iwe usipo taiwanawo zvoku pfeka Without you we would have adequate clothing Urombo wakaipa - wakabvepiko Poverty, you are cruel - where did you come from? Kuno kumaruwa hatina minda Out here in rural areas we are landless and nenzimbo dzokugara homeless Kumadhorobhawo varume nevakadzi In cities we are now forced to live on the strets vogara mumnigwagwa Ichingova nhamo kushayi chokubata All this because of poverty - we are helpless Haiwa nhamno isu tatambura Oh, poverty - we are suffering Urombo wakaipa-wakabvepiko Poverty, you are cruel - where did you come from? Israel E. Katsamba Denys village, Mbondoro, Chegutu District (Presented to the team during afield-vish: organized by Mrs. Mvura and Ms. Bond-Stewart of fica Community Publishing and Develcpment Trust, Zimbabwe.) Poverty and Human Resources in Zimbabwe Executive Summary EXECUTIVE SUMMARY Background and Process The study was conceived in the context of a highly charged debate about the effects of Zimbabwe's structural adjustment program. It is a synthesis of existing information on poverty in Zimbabwe. It also analyzes recent data to assess how economic reforms have affected poverty in Zimbabwe during the 1990s. The aims of the paper are to build consensus on the characteristics and causes of poverty, to inform debate on policy priorities, and to guide discussion of the agenda for further studies of poverty. The information presented in the paper has benefited greatly from the broad dialogue that the Bank conducted with non-governmental organizations (NGOs), community groups, academics, national and local government officials, and other members of the donor community. Representatives of these groups contributed to the development of the paper both by participating in meetings and workshops to discuss the substance of the paper and by providing the Bank team of authors with specific information and comments. Prior to finalising the paper, it was discussed at a seminar in Zimbabwe organized jointly by Shanduko, The Center for Agrarian and Environmental Research, and the World Bank, and presided over by the Chair of the Poverty Forum. The seminar discussed the main themes of the paper and the policy priorities, and identified the research agenda (see summary of proceedings below and at the end of this volume). The analysis contained in the paper is constrained by the availability of, and access to, reliable data. The Status of Poverty and Human Resources Poverty in Zimbabwe is predominantly rural, and is most prevalent in the country's communal farming and resettlement areas. Prevalence is lower on large-scale commercial farms and in urban areas, although there are nevertheless substantial numbers of poor people in these categories. In large-scale commercial farms, inequality is very high and poverty is concentrated among farm workers. Poverty is not as prevalent in Zimbabwe as in many other Sub-Saharan African countries. Precise estimates of prevalence vary substantially, and cannot be compared across time or among different studies because of major methodological differences in data collection and analysis. However, it is possible to make robust inferences about the patterns and characteristics associated with poverty by analyzing the available survey data. These patterns are one of the principal focuses of this paper. Poverty and Human Resources in Zimbabwe ii Executive Summary Zimbabwe has better social indicators than many other countries with similar income levels, largely because of successful government interventions in the past. Infant mortality, for example, is 53 per thousand live births, compared to an average of 93 for Sub-Saharan Africa as a whole. Primary education is now almost universal, compared to an average for the continent of 66 percent. Outcomes have improved enormously since Independence, but some positive trends have stalled or been reversed during the 1990s. Zimbabwe has one of the worst rates of HIV infection in the world, with 30 percent of the sexually active population thought to be HIV positive. This is beginning to have far- reaching, although largely unmeasured, effects on the whole society. The Factors Associated with Poverty Poverty in Zimbabwe is linked mainly to two structural factors. First, the majority of shareholders do not have sufficient access to good quality land and/or reliable water. Land is extremely unequally distributed, with the poor usually living on small plots in drought-prone areas that have poor soil. Principally for this reason, few smallholders are able to increase their agricultural production substantially. Land markets are inflexible, which constrains economic growth and leads to population pressure in ecologically fragile areas. Second, the country's manufacturing and commercial farming sectors have not historically generated sufficient employment. This is the result of a series of policy biases that favored capital over labour and emphasized import substitution at the expense of export-led manufacturing. Well-intentioned but inefficient government intervention in the labour market also stifled job creation, and unnecessary controls constrained incomes in the informal sector. Thus, while the Zimbabwean government has made successful investments in human resources in the past, it has had less success in improving the sources of livelihood for most of the poor. Changes during the 1990s The Zimbabwean economy has undergone major changes during the 1990s intended to stimulate growth in the economy. The government introduced a structural adjustment program in 1991 in an attempt to remove the incentives that favored capital intensity and to stimulate export-led growth. The program contained measures to liberalize markets, devalue the exchange rate, and reduce the fiscal deficit. This adjustment program has only been partially implemented. The government has met many of its objectives relating to trade and market liberalization, which appears to have put the economy on a better footing for growth on the future. The bias in favor of capital- intensity has been reduced, and many industries are now more competitive. Labour- intensive sectors such as horticulture and tourism have grown strongly, so growth in the Poverty and Human Resources in Zimbabwe iii Executive Summary future is likely to have higher employment elasticity than in the past. The economy appears to be in a good position to sustain growth in the future if the reforms continue to be implemented. However, the government has not managed to meet its fiscal targets. Persistent deficits have kept real interest rates high, which has deterred private investment. Moreover, interest payments on the government's debt stock took 22 percent of the total budget in 1995/96 compared to 12 percent in 1990/91, which meant that fewer funds were available in real terms for the government to make necessary social sector investments. Two droughts in four years have compounded the economic effects of policy changes and persistent deficits. The droughts reduced agricultural incomes and assets and suppressed domestic demand. GDP grew by only 1.1 percent per annum between 1990 and 1995. This is equivalent to an annual per capita decline of 1.4 percent. Household incomes appear to have declined during the 1990s. Real wages from formal employment declined by 30 percent between 1990 and 1994, while the number of jobs has not increased dramatically. Indeed, the number of jobs dipped sharply at the beginning of the decade. By the end of 1994 the situation started to improve, and there were 7 percent more jobs in the formal sector than in 1990. Although this is encouraging, it represents average employment growth of only 1.6 percent per annum -- far less than the 2.8 percent per annum increase in Zimbabwe's population of working age. The growth of employment in the informal sector has been mostly in low-profit, survivalist activities. Some of the adjustment reforms have had positive effects on household income. The liberalization of maize markets has benefited consumers and increased competition among traders, and the deregulation of millers has made low-cost maize meal far more widely available. Preliminary evidence also indicates that maize marketing margins have decreased. Meanwhile, the deregulation of the informal sector has allowed it to grow rapidly and to generate employment albeit at low income levels. On balance, it is unlikely that these positive changes have been sufficient to outweigh the effects of falling real wages and low employment growth, although this remains to be determined empirically. It can, therefore, be inferred that poverty prevalence has increased during the 1 990s. Over the same period, there have been real per capita declines in public spending on social services, principally due to the fact that the total public expenditure budget (net of debt service) has shrunk. Between 1990/91 and 1995/96, per capita spending on health fell by around 40 percent in real terms, and real government spending per pupil on education fell by 33 percent. Nevertheless, the government has protected the social sectors as much as possible. The numbers of teachers and health workers have changed little, and the share of the non-interest portion of government spending allocated to the social sectors has risen. Poverty and Human Resources in Zimbabwe iv Executive Summary Outcomes in health and education have not deteriorated at the same pace as spending has been reduced, although, at best, they have stagnated in both sectors. In the health sector, infant mortality remained constant between 1988 and 1994, but under-five mortality increased slightly. In the education sector, rates of transition from primary to secondary school have remained constant in the 1 990s, but completion rates for primary education have dropped somewhat. Chronic malnutrition fell by nearly one-third, but acute malnutrition tripled. These outcomes may have been partly due to spending cuts, but they have also been greatly affected by HIV/AIDS, drought, and other factors. Nevertheless, recent declines in social sector spending are likely to threaten future health and education outcomes, as staff morale falls, supplies dry up, and infrastructure begins to deteriorate. It is clear that there have been severe economic problems during the 1990s in Zimbabwe and that there is some justification for the widespread dissatisfaction that prevails among the population. Some of these problems are the result of policy changes associated with structural adjustment, while others are the result of the effects of past policies, fiscal overruns, drought, and possibly also AIDS. In the eyes of the Zimbabwean public, however, adjustment is the principal cause of all of the difficulties of the 1990s. Adjustment programs are typically controversial, but the program in Zimbabwe has generated particularly heated debate. There appear to be three reasons for this. First, the economy was not in a state of collapse before the program began, which reduced the extent to which the need for reform was generally accepted. Second, the government has managed to implement only part of the program. In particular, the failure to reduce the fiscal deficit is inhibiting the growth that liberalization measures would otherwise be stimulating. Third, the government did not discuss the proposed reforms sufficiently with interest groups before the program began and failed to acknowledge publicly the fact that adjustment would involve some inevitable costs. Therefore, it is difficult for the public to know where to attribute responsibility for particular effects, whether to the adjustment program as it was designed, to the program as it was implemented, or to other changes. Many groups blamed adjustment both for the costs that it did impose (such as reductions in formal sector wages and civil service retrenchments) and for outcomes it was expressly designed to avoid. Adjustment was also blamed for changes (such as declines in some health indicators) that are the result of a complex interaction of social, economic, and environmental factors. Meanwhile, the government tended to attribute a disproportionate share of the blame for negative outcomes to drought and consequent fiscal overruns. However, the debate has now become more constructive and involves a relatively open dialogue among all interested parties with the goal of reaching a consensus on future policies and reforms. A consensus already exists on the need for a fundamental change away from policies that involve the state in providing transfers to the disadvantaged, to policies that stimulate the kind of growth that creates employment and use labour and land efficiently, and to programs that target investments (not transfers) to poor communities. Poverty and Human Resources in Zimbabwe v Executive Summary The Government also undertook a national Poverty Assessment Study Survey (PASS) in 1995 to obtain current information to understand better the patterns and determinants of poverty in Zimbabwe. The PASS will provide rich information on the socio-economic characteristics of households and provide results which are valid at the district level. The final report is expected to become available in 1997. Actions for the Future The main conclusion of the paper is that after a period of stagnant growth, a painful adjustment period and several droughts, there are now fairly good prospects for reducing poverty in Zimbabwe. The characteristics and factors that are associated with poverty in Zimbabwe suggest that a strategy to reduce poverty should have two fronts: broad-based economic growth and direct investments targeted to the poor. Further studies are necessary to help us to better define the elements within these two areas, but the existing knowledge base already indicates the importance of a poverty reduction strategy that has these two main arms. First, broad-based economic growth will be essential for reducing poverty on a large scale, building on the reforms implementedi in the 1990s and concentrating on the following areas: * Reducing the fiscal deficit to bring down real interest rates and encourage private investment. * Encouraging investments in sectors with high employment elasticities. * Increasing the access of the poor to productive assets, especially land and water, by implementing the Land Tenure Commission's recommendations. Second, direct interventions will be necessary to enable the poor to take advantage of economic opportunities and to protect vulnerable members of society during the transition period between the implementation of the reforms and the growth that they are expected to stimulate. These activities will be just as necessary, although less affordable, if the economy does not grow. Priority areas include: * Restoring spending in the social sectors to previous levels in real terms and improving the quality of public spending. * Making well-targeted investments in poor rural areas to protect land quality, to mitigate the impact of droughts, to manage water resources, and to prevent the deterioration of social infrastructure. * Investing in processes that empower poor communities. * Better targeting of emergency grain distribution to improve cost- effectiveness of these interventions and reduce the likelihood of dependency. Poverty and Human Resources in Zimbabwe vi Executive Summary Working aggressively and across all sectors to reduce future rates of HIV/AIDS infection and preparing to the extent possible for the disease's devastating social and economic impact. Shanduko seminar: priorities for further study In July 1996, the main findings of this paper were critically reviewed at a day-long seminar in Zimbabwe jointly hosted by Shanduko, the Center for Agrarian and Environmental Research, and the World Bank (please see the Summary of Seminar Proceedings at the end of this volume) and was chaired by Dr. Brian Raftopoulos, the Chair of the Poverty Forum. The discussants at the seminar endorsed the main analysis and conclusions of the paper and agreed that the paper and the seminar were milestones in the dialogue between the World Bank and the public in Zimbabwe. Two major gaps were identified in the paper: land reforms and the role of institutions. There was consensus that there had been need for reform and that the economy does appear to be healthier as a result of the policy changes. There was also a view that implementing a poverty reduction strategy which promotes growth and simultaneously targets investments in the poor, is feasible and should yield positive results in the second half of the decade. The seminar also endorsed the elements of the poverty reduction agenda for the future emerging from the study, and identified the following research priorities to further inform the poverty reduction strategy of Zimbabwe: * the impact of economic policy on employment * the impact of agricultural marketing reform on the rural poor * the agricultural resource base of communal areas * understanding the economy and diversity of communal areas * the role of institutions Section 1 1 Introduction I 1. INTRODUCTION Objective This paper is a synthesis of existing information about poverty in Zimbabwe. It is intended to build consensus on the characteristics and causes of poverty, to inform debate on policy priorities, and to guide discussion of the agenda for research on poverty. The paper also uses recent data to assess how economic reforn has affected poverty in Zimbabwe during the 1990s The study was conceived in the context of a highly charged debate concerning the economic difficulties and deterioration of public services experienced in Zimbabwe during the 1990s. The initial objective of the report was to provide a synthesis of current knowledge about poverty in Zimbabwe and to identify key areas where further information was needed. As the study continued and as new data became available, the objective evolved to include new analysis and options for the future. Some issues, such as the role and effectiveness of public, private, and community institutions in efforts to reduce poverty, are not included in this paper. Although they are important, their coverage in existing literature on Zimbabwe is limited. The role of institutions in poverty reduction is being studied in detail in the context of design of targeted investments and community mobilization under the Government of Zimbabwe's Poverty Alleviation Action Plan (PAAP). The information presented here is the result of a broad dialogue that the World Bank team had with non-governmental organizations (NGOs), community groups, academics, national and local government officials, and members of the donor community. This consultation process began during the team's mission to Harare in July 1995 and was repeated in workshops in November 1995 and in April 1996. These meetings and workshops were intended to share information and ideas, to increase understanding among all of the parties, and to reach a consensus about future directions. Participants at the workshops contributed to the evolution of this paper both by participating in the discussions themselves and by providing specific information. The first draft of this paper was discussed in detail at a workshop with NGOs, academics, and the donor community in April 1996. A later draft was discussed at a workshop of Zimbabwean and international researchers in July 1996 (see summary of workshop proceedings in this volume). The study relies primarily on published or processed data. In some cases, such as the Income Consumption and Expenditure Survey of 1990-91, the lack of access to the primary data constrained the study team in the scope and depth of the analysis that it could conduct in the area of household incomes and consumption patterns. Section 1 2 Introduction Context Zimbabwe is located in central southern Africa and has a population of 10.4 million people, predominantly from the Shona and Ndebele groups. Administratively, the country is divided into 57 Districts in eight Provinces plus two cities, Harare and Bulawayo. Since Zimbabwe attained Independence from Britain in 1980, it has made enormous progress in providing basic social services (health, education, water and sanitation, and emergency relief) but less progress in creating employment for its population, despite having one of the largest manufacturing sectors in Sub-Saharan Africa. This was at least in part because of policies retained from the pre-Independence period that protected many industries from international competition and subsidized capital at the expense of labor. The country is characterized by a highly unequal distribution of consumption and income, a legacy of the colonial period. A dualistic landholding pattern contributes to this income inequality. Some 4,400 large-scale commercial farms occupy one-third of the country's arable land area -- nearly two-thirds of the country's high potential land -- whereas around 1 million small farms support half of the country's population on half of the country's agricultural land, mostly in semi-arid areas. In the first half of the 1990s, Zimbabwe experienced major economic and climatic upheavals. In 1991, the Zimbabwean government introduced a program of structural adjustment designed to overcome some of the factors hindering economic growth and to remove the biases against labor. The following year, the worst drought of the century hit the country, decimating agricultural production and livestock. Zimbabwe has experienced serious economic difficulties during the present decade. The economy contracted by 9 percent in real terms between 1991 and 1992 and has grown at an average ofjust 1.1 percent per annum between 1990 and 1995. This translates into per capita growth of -1.4 percent per annum (World Bank, 1996). Public spending has been squeezed. Real spending on health and education has declined by 30 percent in per capita terms between 1990 and 1995. Although there was room for increasing the cost- effectiveness of service delivery in some sectors, progress in doing so has stalled, and some outcomes are already beginning to decline. The extremely high prevalence of AIDS (30 percent of the sexually active population) is also beginning to take its toll on families, on the economy, and on aggregate health outcomes. Sources of information The paper is based on published papers and existing survey tabulations that include both quantitative survey information and qualitative research. The paper uses published material from five surveys: * The Income, Consumption, and Expenditures Survey (ICES) from 1990/91 (14,000 households) Section 1 3 Introduction * The Demographic and Health Surveys (DHS) from 1988 (4,000 households) and 1994 (6,000 households) * The 1992 population census * The biennial Sentinel Surveys from 1992-1994 (4,000 to 6,000 households), which were developed to assess the impact of drought and structural adjustment on households * The GEMINI surveys of small-scale enterprises conducted in 1991 and 1993 (14,000 and 12,000 households respectively). Preliminary results of the 1995 Poverty Assessment Study Survey (PASS) became available in June 1996 (MPSLSW, 1996). We draw on these results to some extent, especially where they show the patterns of poverty and characteristics of poor households, but we base our analysis primarily on the ICES from 1990/91 as this is the most recent complete published data set. The paper draws on several published reports. These include UNICEF's "Children and Women in Zimbabwe: a Situation Analysis Update 1994," the World Bank's Country Economic Memorandum from April 1995 entitled "Zimbabwe: Achieving Shared Growth," and drafts of the government's ongoing Public Expenditure Review of 1996. The paper also draws on a large number of studies by the government, academics, NGOs, and others, which cover topics ranging from legal issues through studies of household food security to anthropological studies (see bibliography). Scope of the paper For the purposes of this paper, poverty is measured in terms of consumption and in terms of human resource outcomes, particularly in health and education. Section 2 of the paper discusses the status of the poor in terms of their levels of consumption and of their sources of income. Section 3 discusses the status of human resources in Zimbabwe. These sections compare the situation in Zimbabwe with that in other countries and identify the characteristics of different types of poor households using the most recent available data. Section 4 discusses the fundamental factors that are associated with or that affect poverty. Section 5 then discusses actual or probable changes in both incomes and human resources during the 1 990s. This section is inherently difficult as it attempts to untangle the different effects of drought, economic contraction, structural adjustment, and AIDS on different groups of the poor. Much survey information is currently being compiled or analyzed but is not yet available, which compounds these difficulties. Where further analysis or basic information is needed, the paper points this out. Section 6 suggests priorities for further study. The fmial section concludes the paper by giving an outline of a potential poverty reduction strategy for Zimbabwe -- policies and actions that, if adopted, would immediately begin to promote economic growth and increase the access of the poor to economic opportunities. Section 2 4 Income and Consumption 2. INCOME AND CONSUMPTION This section discusses the prevalence of poverty in Zimbabwe, using household incomes and per capita consumption to indicate poverty. In this paper, we rely predominantly on published tabulations of household surveys and on secondary sources for the consumption and income data that we use as well as on information from government sources for some human resource indicators. The poor themselves often use a range of indicators to define their poverty. As more participatory studies are carried out, it will be necessary to ensure that results from these studies are taken into account to inform Zimbabwe's poverty reduction strategy. Estimates of the extent of poverty in Zimbabwe vary substantially, and major methodological differences make it problematic to compare them. The sampling techniques used, the choice of consumption versus incomes as the measure of living standards, the data collection methods, and the techniques for calculating poverty lines all vary considerably. Despite this difficulty, the surveys do yield robust results about the patterns and characteristics of poverty, knowledge of which is useful for guiding policymaking. 12.1 HOW PREVALENT IS POVERTY IN ZIMABWE? Estimates of prevalence vary, but household surveys reveal patterns and characteristics of poverty Researchers measure poverty prevalence by calculating poverty lines and then using household survey data to estimate the proportion of the population whose income or consumption falls below those lines. I Two national household surveys for Zimbabwe have been analyzed in this way -- the 1990/91 Income, Consumption, and Expenditure Survey (ICES) and the preliminary results of the 1995 Poverty Assessment Study Survey (PASS). They should not, however, be compared, because different methodologies were Consumption/expenditure data can differ quite substantially from income data, particularly at lean times of the years, such as just before the harvest. Consumption data include what the household consumes from its own stores, from government programs, and by consuming small livestock. These data also tend to be more accurately reported by survey respondents. For these reasons, consumption/expenditure data tend to be a better measure of welfare than incomes. Section 2 5 Income and Consumption used both in carrying out the actual surveys themselves and in analyzing their data. Results of the 1995/96 ICES will be available next year. This survey used a similar methodology for collecting data to the one used in the 1990/91 ICES, which will make it possible to compare the two ICES surveys to analyze changes in living standards during the 1990s. The estimate of poverty from analysis of the 1990/91 ICES is based on the purchasing power of households to meet their basic consumption needs (World ]Bank, 1995a). In this analysis, per capita consumption (cash expenditures, goods from own production, and other in-kind sources) was used to measure living standards. This is based on four poverty lines: an upper and a lower line for urban areas and an upper and a lower line for rural areas. The two upper poverty lines distinguished the poor from the nonpoor, as they were estimates of the minimum per capita expenditure necessary to meet a household's basic needs. Households with per capita consumption below the upper poverty line were classified as poor. The study set the two lower poverty lines at a level equivalent to the expenditure necessary to meet only per capita basic nutritional needs. Households with per capita consumption below the lower poverty line were therefore classified as very poor or as living in extreme poverty.2 In this analysis, it was estimated that about 25 percent of Zimbabwe's population (2.6 million people) were poor at the beginning of the decade. A sub-set of this group, approximately 7 percent of the population (slightly more than 700,000 people), was estimated to be very poor.3 The preliminary results of the 1995 Poverty Assessment Study Survey (PASS) give a much higher estimate of the prevalence of poverty in the middle of the decade. This study calculated upper and lower poverty lines in urban and rural areas for each of Zimbabwe's 58 districts, using local prices.4' The study then took the income (cash and in kind) data from the PASS household survey (as contrasted to expenditure data from the 1990-91 ICES) and estimated that 62 percent of Zimbabwe's population were poor and 46 percent were very poor in 1995 (MPSLSW, 1996). Great care is needed when interpreting the 1990/91 ICES and the 1995 PASS results. They should not be used to draw inferences about changes in poverty prevalence over 2 All four poverty lines were based on local prices and actual consumption preferences in urban and rural areas. The upper poverty lines took the basic food basket and added consumption of other basic needs, such as housing, clothing, education, health, and transport. The upper lines were Z$550 (US$162) for urban areas and Z$340 (US$100) for rural areas, and the lower lines were Z$325 (US$96) for urban and Z$209 (US$61) for rural areas, using the average 1991 exchange rate of US$1=Z$3.4 (World Bank, 1995a). 3 The population figures come from the 1992 population census (CSO, 1992). The PASS estimates were calculated on the basis of local poverty lines, but the study also used a national upper poverty line of Z$2,554.89 (US$297) in urban areas and Z$1,924.2 (US$224) in rural areas. The national lower poverty lines were Z$1,511.77 (US$176) for urban areas and Z$1,180.49 (US$137) using a 1995 average exchange rate of US$ 1=Z$8.6. Section 2 6 Income and Consumption time because of differences in the methodologies used to collect and analyze the data and in the ways in which the welfare measures were constructed. However, both data sets yield useful insights into the patterns and characteristics of poverty, which are useful for policymaking. At the time of writing, the preliminary PASS results are being reviewed, and the complete results are not yet available. This paper concentrates on analyzing the patterns revealed by the 1990/91 ICES, with a view thet they can be compared with those from the forthcoming 1995/96 ICES survey.5 In addition to analysis of quantitative household survey data, it is useful to obtain people's perceptions of the dimensions of poverty in building our understanding of poverty. People who are poor often consider themselves to be poor in terms other than just income. They often consider lack of "wealth" to be the most important factor defining their poverty. There is currently little information on how the poor perceive poverty or wealth, except for one study of a communal farming community in southern Zimbabwe (Scoones, 1995), which investigates how people in that community measure wealth (mupfumi). The study found that the community uses a range of criteria, including ownership of livestock and other assets, crop yields, education of children, age, and social status as well as cash income. The community's evaluation of the relative wealth of households sometimes differs from the results of a household survey. Within the community, different groups have different perceptions of wealth or about how a given asset adds to their security. Women, for example, tend to link wealth with access to remittance income whereas men put more emphasis on livestock. As more information like this becomes available from different communities in Zimbabwe, it can be used to enrich and complement the quantitative survey information. Poverty in Zimbabwe is not as high as in many other countries Comparing poverty across countries is inherently problematic because definitions of poverty and the availability and quality of data vary widely. Recent World Bank research (Ravallion, 1996) compares poverty across countries by estimating the proportion of the 5 The methodology used by the ICES nevertheless has certain problems that need to be taken into account when interpreting its results. The survey, for example, undersamples in certain areas, which might introduce a geographic bias. There are also risks of reporting bias, as respondents may alter their answers in order to qualify for government programs or for other reasons. The use of per capita rather than adult equivalent weightings will over-report poverty for households with large proportion of children. Finally, the use of poverty lines alone is problematic, particularly when analyzing the situation of the nonpoor, as it is impossible to differentiate between those households just above the poverty line and the very rich. Analysis by income or consumption quintiles would be preferable. Section 2 7 Income and Consumption population with consumption of less than US$1 per person per day in 1993.6 This research estimates that 39 percent of Zimbabwe's population consumed less than $1 per day, which places Zimbabwe around the middle of the list of countries in Sub-Saharan Africa, as Figure 2.1 shows. Figure 2.1 Percentage of the Population Living on Less Than US$1 per Day 90% 850 80%,1 70%~~~~~~~~~~~~~~~2 60% ~~~~~~~~~~~4 50%' 3 50% . . 20% t * S;qi7yoX1#; I*.',i,', l';".u . ;i 40% 20%~~~~~~~~ Source: Ravallion, 1996 Whatever the prevalence of poverty relative to that in other countries, the number of poor people in Zimbabwe is large. In 1990/9 1, some 2.6 million people were estimated to be unable to meet their basic needs, and, as we shall show later, that number is likely to have grown since then. The next section focuses on the characteristics of the poor and on the factors associated with poverty. Non-consumption measures of human welfare have also been used to compare standards of living across countries. As we shall see in later sections, health and education outcomes in Zimbabwe are superior to those in other countries with similar income levels. 6 This research makes comparisons across countries using a benchmark of equal value (US$I per person per day), and adjusts consumption estimates to indicate the same purchasing power across countries. It calculates purchasing power parity using Penn World Table 5.6 and 1985 prices, and makes projections for 1993 for each country using the most recent household survey available. Data for Zimbabwe are based on the 1990/91 ICES. Section 2 8 Income and Consumption 12.2 WHAT ARE THiE CHARACTERMISCS OF POVERTY? Poverty is predominantly rural Table 2.1 shows that poverty was far more prevalent in rural than in urban Zimbabwe in 1990/91. An estimated 31 percent of the rural population were poor in 1990/91 compared to 25 percent of the population as a whole. The share of the poor is also higher -- 71 percent of the total population is rural, but 88 percent of the country's poor and 92 percent of the very poor live in rural areas. The predominance of rural poverty does not imply that Zimbabwe has no urban poverty problems. On the contrary, although the urban poor are fewer in number than the rural poor, they face serious difficulties. They usually live in overcrowded conditions, and tend to be dependent on only one source of income (formal sector employment). There is also a troubling likelihood that their numbers have been growing faster during the 1990s than those in rural areas (see Section 2.2 on urban poverty and Section 5 on changes since 1990/91). Poverty is most prevalent in communal and resettlement areas Zimbabwe has historically divided its population into four groups based on residence or land use.7 These are: urban areas (agglomerations with more than 2,500 inhabitants), communal areas (where some land is held in common by the community while other land allocated to households on the basis of usufructory rights), large-scale commercialfarms or LSCF areas (which are owned freehold by individuals or corporations), and resettlement areas (where households are given long-term rights to land that the government acquired under the resettlement program).8 The prevalence of poverty in 1990/91 was far higher in communal and resettlement areas than in urban and LSCF areas. Figure 2.2 shows the distribution of the poor and very poor in different land-use groups, and Table 2.1 shows the poverty prevalence in each group. The overwhelming majority of the country's poor people live in Zimbabwe's communal areas. These areas were home to half of the country's population but to more These categories are often more administrative than socio-economic and tend to conflate households with different characteristics, constraints, and behavior patterns. Additional analysis of raw data using different criteria for disaggregation will be necessary to gain a greater understanding of socio- economic groups. 8 The census also notes other rural sub-groups numbering a total of 250,000 or 2 percent of the total population. These include small-scale commercial farmers and farmers of government land among others. Other surveys do not distinguish these groups and they are not considered separately here. Section 2 9 Income and Consumption Figure 2.2 Distribution of the Total Population and of the Poor Distribution of the Population, 1992 Distribution of the Poor, 1990/91 RASOther rural RAs 4% 2% Urban 12%- lb ~~~~~~~~~~LSCFs Urban I 6% 31% Communal Areas 52% Communal LSCFs Areas 11% 77% Distribution of the Very Poor, 1990/91 RAs Urban 6% 8% LSCFs 4% A Connunal Areas 82% Source: CSO, 1992 and World Bank, 1995a Section 2 10 Income and Consumption than three-quarters of the poor and to 82 percent of the very poor, and had the second highest poverty prevalence in 1990/91. Analysis of the ICES showed resettlement areas to have had the highest poverty prevalence in 1990/91, as an estimated 41 percent of households in these areas had per capita consumption levels below the upper poverty line. The data for resettlement areas, however, may not be reliable, because the sample size is small and not representative of certain areas, such as those that produce cash crops. Table 2.1 Poverty Prevalence by Land-use Group, 1990/91 Prvalence Prevalence Shar of Shareof Sare Of of Poverty' of Extreme Poor2 Very Poor Popuadon (%) Poverty (%) Rural" 31 9 88 92 69 Communal Areas 33 10 76 82 51 LSCF 16 3 6 4 11 Resettlement Areas 41 14 5 6 4 Urban 10 2 12 8 31 Total 25 7 100 100 100 Source: World Bank 1995a:27 I Prevalence: Number of poor households in the category/ number of households in the category. 2 Share: Number of poor households in the category/ total number of poor households. 3 The 1992 census includes other rural sub-groups, which account for 2% of the population. Most of Zimbabwe's poor live in communal areas, but only an estimated one-third of the population in these areas is poor The other principal rural land-use group, large-scale commercial farms, had a smaller share of the poor and the very poor in 1990/91, but also a small share of total population. Poverty prevalence on large-scale commercial farms was 16 percent. This relatively low prevalence is deceptive, however, because income inequality is particularly high in these areas. Workers on large-scale commercial farms are believed to have living conditions and social indicators that are among the worst in the country.'0 Unfortunately, published sources do not separate data for these workers from those in the sector overall. 10 Unfortunately, published sources do not separate data for these workers from those in the sector overall. Section 2 11 Income and Consumption Poverty was least prevalent in 1990/91 in urban areas, where 10 percent of the population were estimated to have consumption levels below the poverty line. Poverty prevalence in urban areas is believed to have increased during the 1 990s. Communal areas are a legacy of the Native Reserves established by the colonial administration at the end of the 19th century that became Tribal Trust Lands from 1965 until Independence. In 1990/91, they were home to 1.8 million people living below the poverty line, more than half a million of whom were unable to meet even their basic nutritional needs. As these people constitute most of the country's poor and very poor and they are the groups with the highest prevalence of poverty for which we have reliable estimates, much of the discussion in the following sections relates to them. Later sections discuss the factors associated with being poor in communal areas, such as source of income, household size, household headship, and health and education levels. Although the largest share of the country's poor live in communal areas, it is important to note that two-thirds of communal area households in 1991 had incomes above the poverty line. Variation of income in and among communal lands is high; indeed, in 1990/91, communal areas had the same high level of consumption inequality as the country as a whole (Jenkins and Prinsloo, 1995). Analysis of different income quintiles would be useful for understanding differences between groups of the nonpoor. Nonpoor households in the communal areas are more likely than poor households to be headed by someone who has had some education and to receive incomes from non- agricultural sources. They are likely to receive a lower proportion of their cash income from remittances than poor households. The ICES indicates that nonpoor households living in communal areas do not tend to be associated with larger landholdings than poor households. Nonpoor households in rural areas had only slightly more land per capita than the poor, which implies that the nonpoor were earning their livelihoods outside agriculture. These landholding data, however, distinguish neither among land of different quality nor among different kinds of households, such as farming or non-farming (World Bank, 1995a:32). Other sources indicate that farm size, household size, and cattle ownership all increase with income categories. These issues will be discussed in more detail in later sections. The availability and quality of land are major factors constraining own-farm incomes and food security in communal areas. In fact, many communal area households are net Section 2 12 Income and Consumption purchasers of grain, particularly in dry areas. 1I Pressure on land is increasing as populations grow yet have no access to additional land. Problems of inheritance, particularly when land is divided among family members, are leading to serious land conflicts in communal areas. The government's land tenure commission found "an untenable state of affairs which was approaching anarchy in some areas" (Rukuni, 1994:24). The majority of communal farmers live in low rainfall areas, which are increasingly vulnerable to drought. Population pressure combined with unsustainable farming, woodland, and livestock management practices appears to be leading to increasing levels of erosion (although many of the more alarming studies of erosion are thought to be unreliable). Nevertheless soil fertility does appear to be declining in many areas, and water supplies, already scarce, are becoming more precarious. The prevalence of poverty among farm workers on large-scale commercialfarms is high, even though the overall prevalence on these farms is low The poverty prevalence in the large-scale commercial farm (LSCF) sector overall appears from the ICES to be relatively low (16 percent in 1990/91). The population living in these areas is, however, highly diverse as it includes households headed both by farm owners or managers and by farm laborers. These two groups have vastly different income levels. Although few published sources of household data separate the two groups, there is little doubt that these households are among the country's poorest. One study (Jenkins and Prinsloo, 1995) shows that the LSCF sector has an index of inequality far higher than that of any other group in the country. Gathering better data on the typical conditions experienced by farm workers should be a high priority for the future. Farm workers tend to live in small communities, either employed permanently on one farm or, increasingly, working on contract and moving in response to demand in the labour market. In 1993, almost half of the country's LSCF workers were casual employees (World Bank, 1995a:94). Their living conditions vary from farm to farm because what housing and other infrastructure is available depends on the landowner. Two studies cited by UNICEF (1994:67) describe LSCF laborers as living in poor housing with poor sanitary conditions and often with unsafe water sources. Most communal area households in dry areas are net purchasers of grain (UNICEF, 1994:34; Muir, 1993:ix). One study (Jayne et al., 1990) reported estimates of the proportion of households that were net purchasers of grain in an average year ranging from 15 percent in Natural Region II to almost 100 percent in Natural Region V. Another study (Moyo, 1995:151) found that 55 percent of farmers in Natural Region IV and 82 percent of those in Natural Region V reported food shortages in 1989, which was a year of reasonably good harvests. Section 2 13 Income and Consumption Access to public infrastructure is also a problem for the families of I,SCF workers. Few schools and health posts are located close to these areas (UNICEF, 1994). As a result, forty-one percent of children between the ages of 6 and 17 living in the LSCF areas were not in school in 1994, compared to 17 percent in communal areas, and 18 percent in small-scale commercial farm and resettlement areas (MPSLSW, 1995:39). Large-scale commercial farm workers and their families have also typically lacked a political voice. Until 1993, they had no representation on rural councils as they were neither owners nor renters of land and, thus, were technically not residents (LNICEF, 1994:21). The Minister of Local Government, Rural and Urban Development now appoints members to the rural district councils to speak for commercial farm workers. In resettlement areas, poverty prevalence is high but declining Resettlement areas exist as a result of the government's attempts to redress the problem of unequal access to land in Zimbabwe. The government started resettlement schemes in 1981 immediately after Independence. It bought commercial farmland from willing sellers and took over abandoned farms, and redistributed this land to families who met eligibility conditions. By 1992, around 400,000 people had benefited from this program. Resettlement areas are typically located on better land than the communal areas. Zimbabwe has five categories of agricultural land, based on natural regions (NRs). NRs I, II, and III have good soil and relatively reliable rainfall, whereas NRs IV and V are suitable only for extensive agriculture and livestock husbandry because of poor soil and unreliable rainfall. In 1993, 56 percent of the resettlement land, supporting 80 percent of the resettlement population, was in NRs I, II, and III (Moyo, 1995:1721; Mhishi, 1995:14). Despite being sited on better land, resettlement areas were estimated to have had a higher prevalence of poverty than communal areas. In 1990/91, 41 percent of the population of resettlement areas were estimated to be poor. Difficulties with small sample sizes and inherent reporting biases in the ICES make these rates less reliable than the rates for the larger population groups (World Bank, 1995a). Yet, despite these data difficulties, few people doubt that resettlement areas have a major poverty problem. Some reports indicate that children in resettlement areas have the highest rates of both chronic and acute malnutrition of all the land-use groups (cited by UNICEF, 1994:63). Recent studies link poverty in resettlement areas to three factors. Fitst, the conditions of eligibility for the early schemes were such that beneficiaries were often too poor and ill- equipped to take full advantage of the land's potential. The early schemes selected landless people, the homeless, refugees, and ex-combatants as beneficiaries. These people often had little farming experience or assets and were, therefore, unlikely to be able to make full use of the land they were allocated, at least at first. After 1984, new beneficiaries had to be experienced farmers, but even after this change, the beneficiaries tended not to have sufficient assets, education, access to government services, or experience on different types of land to bear the risks of innovation and to take full advantage of the land they were given (Mhishi, 1995:7). Section 2 14 Income and Consumption Second, the resettlement schemes imposed rules on the beneficiaries that constrained their options. Once they had the land, the farmers had to give up any claims they had to communal land, and were not allowed to earn off-farm incomes. Both of these restrictions removed key coping mechanisms for resettlement households, making them less able to adapt in times of crisis, such as when there were bad harvests, droughts and sudden unfavorable price changes. Resettlement farmers are known not to adhere strictly to either of these rules, and the restriction on off-farm incomes has been partially relaxed, but the basic situation does not appear to have changed. Third, many of the beneficiaries of the schemes moved to their land before the government had completed the social infrastructure such as schools, roads, health posts, water and sanitation, and other basic services. This reduced the extent to which the beneficiary families could learn farming practices, market their crops, and keep their families healthy and educated (Mhishi, 1995). Recent studies indicate that farmers in resettlement areas are beginning to raise themselves out of poverty as they learn farming techniques and as basic services become available. The Land Tenure Commission found that farming standards were high and that productivity had apparently increased since the early days of the scheme (Rukuni, 1994:66). The production of cash crops such as sunflowers, groundnuts, and cotton by resettlement farmers has increased, although yields are still lower than in the LSCF sector. Cattle ownership is also increasing. The farms tend to be on better land and to be larger than most communal farms. Once basic conditions are in place, therefore, the resettlement farmers are more likely than communal farmers to be able to earn sufficient income to cover at least their basic needs (Mhishi, 1995: 24). Moyo (1995:124) cites various studies that report resettlement households generating incomes at least 20 percent higher than those of communal households by the early 1 990s. In urban areas, the prevalence of poverty is low but is probably growing Poverty prevalence in urban Zimbabwe was relatively low in 1990/91, consisting of 10 percent of the urban population. Harare is the country's largest city, but its poverty prevalence -- 7 percent -- was lower than that for other urban centers (12 percent) (World Bank, 1995a). The urban poor in Zimbabwe live mostly in "high density suburbs," to which the pre- Independence government confined the urban black population. These areas are separated by great distances from the more affluent "low density" areas, and many have only one or two access roads. They have changed little since Independence. Housing construction standards are high and still strictly enforced, which makes single family housing unaffordable for many urban dwellers. Owners often, therefore, erect second or third dwellings on their property or rent space in their house to other families. Squatter settlements are relatively rare (UNICEF, 1994), but may be increasing. Section 2 15 Income and Consumption Although overall poverty prevalence in urban areas appears to be low, there are indications that the proportion of households below the poverty line is growing. Data are not yet available so nothing conclusive can be said, but we believe that economic changes during the 1 990s may have had a disproportionate effect on urban households and caused urban poverty prevalence to rise faster than that in rural areas. Section 5 covers changes in the 1990s in more detail. Health data also give some indication of the increasing vulnerability of the urban population. Acute (weight for age) malnutrition in children under the age of three in urban areas increased four-fold between 1988 and 1994, compared to a three-fold increase in rural areas (ZDHS, 1994). This startling outcome probably reflects a number of factors. First, health centers have come under increased strain, particularly in urban high-density areas. Second, household incomes are thought to have declined, which will have especially affected urban children because they are not eligible for many government supplementary feeding prograrms. Third, HlV/AIDS is associated with acute malnutrition and is thought to be more prevalent in urban areas. We discuss health status in more detail in Section 3, and cover changes in the 1990s in Section 5. The driest provinces tend to be the poorest Disaggregating poverty data for Zimbabwe's eight provinces reveals a complex story. The 1990/91 ICES data shown in Table 2.2 do not reveal major differences in shares of the poor across provinces. Nevertheless, some provinces have higher poverty prevalences, that is, higher proportions of their population living below the upper poverty line than others. Matabeleland South has by far the highest prevalence, followed by Masvingo and Matabeleland North.'3 When we turn to the proportion of the provincial population living below the lower poverty line, we see a similar pattern. Again, Matabeleland South has by far the highest prevalence of extreme poverty, followed by Masvingo and Matabeleland North. Regional rainfall patterns provide one explanation for the variation in poverty prevalence among Zimbabwe's provinces. Large areas of the country (mostly in natural regions IV and V) have low and unreliable rainfall and support only extensive agriculture and livestock husbandry. Farming households in these areas are more likely to be poor than those in areas of reliable rainfall. Each of the country's provinces contains some low rainfall areas, but Masvingo and Matabeleland North and South have the highest 3 Another study (Stenflo, 1993:24) uses relative poverty lines and analyses the conditions of the poorest households in the country, defined as those with consumption levels less than half of the median annual consumption across the country. This method shows the same three provinces to be the poorest, albeit ranked differently (Matabeleland North has the highest prevalence, followed by Matabeleland South and Masvingo). Preliminary data from the 1995 PASS give quite different provincial rankings (MPSLSW, 1996). Section 2 16 Income and Consumption proportion of land in natural regions IV and V (see map). The land quality in these provinces is probably linked to their higher prevalence of extreme poverty as well. (See Section 4 for a discussion of land quality and poverty.) Table 2.2 Provincial Breakdown of Poverty, 1990/91 Prevalence Share Prevalence of Share Share of of Poverty of Poor Extreme Poverty of Very Populato (%) (%) Poor Manicaland 24 15 6 13 15 Mashonaland 19 6 4 5 8 Central Mashonaland East 18 18 5 17 24 (incl. Harare) Mashonaland West 22 10 5 8 11 Matabeleland North 27 13 10 17 12 (incl. Bulawayo) Matabeleland South 41 10 17 14 6 Midlands 25 13 6 1 1 13 Masvingo 32 15 12 20 12 Total 25 100 7 100 100 Source: Based on World Bank, 1995a. To complicate the provincial picture, when we turn to health and education indicators, the provinces that appear to fare worst do not overlap completely with those that were estimated to have the largest prevalence of consumption poverty in 1990/91. Matabeleland North, Mashonaland Central, and Mashonaland West appear to have the worst rates of chronic and acute malnutrition and of diarrhea among children under the age of three. 14 When the results of the 1994 DHS and the 1992 population census are combined, these three provinces appear to have the highest proportions of people with no education. The information that is currently available does not disaggregate health and education data by natural region or by land-use area.'5 14 This was calculated by ranking the provinces for each indicator (taking <2 standard devations for the malnutrition indicators) and then adding the rankings. The indicators chosen were those that tend to be related to poverty where provincial-level data were available. 5 The final report of the 1995 Poverty Assessment Survey will provide poverty information disaggregated by natural region. Recent analysis (Gleeson and Laincz, 1996) disaggregates district- level data from the 1992 population census by natural region, using pro-rata estimates of population in different NRs. This shows clearly that, for almost all health and education indicators, Natural Regions I and V fare the worst. Section 2 17 Income and Consumption Female-headed households in rural areas are no more likely to be poor than male-headed households Rural households headed by women were no more likely to be poor than those headed by men in 1990/91. Women headed 42 percent of nonpoor rural households and 39 percent of poor rural households (World Bank, 1995a). The incomes of nonpoo rural households headed by men and those headed by women are similar, yet the incomes of por rural households differ greatly according to the gender of the household head. Poor rural female-headed households have average incomes that are about one-third higher than those of poor rural male-headed households. For very poor households, the difference is less stark (12 percent higher), but still apparent (World Bank, 1995a:34). This is because the ICES does not distinguish between households that are defacto female-headed (those who have husbands or other family members who live elsewhere and send remittances to the household) and those that are dejure female- headed (with no male support). The remittance income for de facto female-headed households thus probably explains the result. We have very little disaggregated data for dejure female-headed households, but we do know that widows make up the largest group of unsupported women household heads. Eight percent of the total number of household heads in the 1990/91 ICES were widows. We have very little information about this group, but we assume that households in this category are more likely to be poor than those in many other groups. Further research about these households is needed. Female headship appears to be associated with economic hardship in urban areas. In 1990/91, female-headed households in urban areas had incomes on average 20 percent lower than those of households headed by men. Women headed 13 percent of nonpoor households and 18 percent of poor households, but only 13 percent of very poor households at that time (World Bank, 1 995a). The ICES, however, only reflects poverty as measured by per capita consumption. It would be interesting to study the differences in, for example, workload, crop yields, health outcomes, and child survival between households headed by women and those headed by men. Heads of poor households tend to have less formal education than do heads of nonpoor households People who head poor households tend to have less education than those who head nonpoor households. As Figure 2.3 shows, 29 percent of the heads of poor households and 32 percent of the heads of very poor households had no formal education, whereas this was true for only 14 percent of the heads of nonpoor households (World Bank, 1995a). The education of the household head is particularly importanit as a factor Section 2 18 Income and Consumption associated with poverty in urban areas. Although household heads with low levels of education are associated with poverty in all land-use groups, Figure 2.3 shows that the difference is particularly stark in urban areas. Figure 2.3 Education Levels of Household Heads, 1990/91 No Formal Education 50% _ _ _ _ _ _ _ _ _ _ _ _ 40% ___ 30% { _ Non-Pbor Households P .or Households 20% 0 _Very Poor Households 010%/ 0% __ All Rural Urban Secondary Education 50% r 40% 30% * ' . " Non-Poor Households .Poor Households 20Very Pbor Households 0% L E,M LIa All Rural Urban Source: World Bank, 1995a Section 2 19 Income and Consumption 123 WHAT ARE THE SOURCES OF INCOME OF 7HE PboR? | This section relies on information from household surveys, principally the 1990/91 ICES, to analyze the sources of income of different kinds of household. Interpreting data on household incomes requires particular care as it tends to be difficult to collect reliable income data from households.'6 Household income can be either cash or in-.kind. In-kind income includes food that the household has produced itself, fuelwood that household members have collected, and imputed values for rent when the household owns its house or when the employer provides housing. In rural areas, the average household depended almost equally on income in cash and in kind in 1990/91, whereas cash income made up nine-tenths of the income of the average urban household (CSO, 1994). Available studies have disaggregated the ICES income data by poor and nonpoor households only in the case of cash incomes. Therefore, in this section, we discuss the sources of cash income for urban and rural households in poor and non-poor groups. We then discuss the most important sources of cash income for the poor in more detail. Finally, we consider sources of in-kind income for all urban and rural households, as the information does not allow us to distinguish between poor and nonpoor households in this case. Rural cash incomes predominantly come from non-crop sources Rural households in all income groups have more sources of cash income than their urban counterparts as Figure 2.4 shows. The average rural household receives far more of its cash income from agriculture and remittances than does the average urban household. The average urban household, in contrast, depends almost exclusively on employment for its cash income.'7 16 For example, the respondents in the sample of the 1990/91 ICES are known to have underreported certain income categories, such as income from cotton. 17 Employment is the category used in the ICES. It is defned as "people at work" and does not include communal farmers, who are classified as self-employed, but does include agricultural wage labour. It is not clear how informal sector workers are categorized. The number of jobs in 1990/91 was disaggregated into the following sectors: production workers 26 percent; agricultural workers 20 percent; service workers 20 percent; professional workers 15 percent; clerical workers 9 percent; sales workers 8 percent; and managerial workers 2 percent. Section 2 20 Income and Consumption Figure 2.4 Sources of Cash Income, 1990/91 All Rural Households Poor Rural Households Employment Other Other 1E1p% 16% ~~~ ~ ~~~~~~~~~~~17% 1 Employment -I 7 Remittances 22% FRemittances Agriculture 30% 42% Agnculture 30% All Urban Households Poor Urban Households Other Remittances 3% Other 7% 14% -~~~~ \\\ .i. N -Remittances X .__ . . '. ~~~~~~~~~12%i-.; - X . Employment _ - - o = : 74% Employment 90% Source: World Bank, 1995a Section 2 21 Income and Consumption Poor households in both rural and urban areas tend to depend more on remittances than do their nonpoor counterparts as a proportion of their total household income. In both rural and urban areas, poor households receive a smaller share of their income from employment than do nonpoor households. Agriculture is the most frequent source of cash income but not the largest Agriculture is the source from which the largest number of households report receiving cash income. Almost all households in rural areas and over half in urban areas reported receiving some income from farming in 1991 (Jenkins and Prinsloo 1995). We take this to be income from sales of agricultural products, because the ICES classifies income from agricultural wage employment as part of employment income. Agriculture's relative importance to the overall household cash budget, however, varies with the location and the wealth of the household. Figure 2.4 shows clearly, and unsurprisingly, that in 1990/91 rural households derived a far larger proportion of their cash incomes from agriculture than did urban households. It also shows that poor households relied on agriculture for cash income to a greater extent than households that were not poor. In rural areas, nonpoor households derived 28 percent of their cash income from agriculture, whereas the poor and very poor dependedl on it for 42 and 43 percent of their incomes respectively. Agriculture was particularly important for poor households headed by males, who depended on agriculture for almost half of their household cash income (World Bank, 1995a:32). Most households depend on agriculture not only for cash income but also for home consumption. In communal areas in 1990/91, nonpoor households reported that 50 percent of their total in-kind income came from agricultural sources (own-produced food and firewood gathered by the household, Figure 2.5). Maize is the most important crop both for home consumption and for generating cash income. Understanding the way in which it is produced and consumed by different types of household is, therefore, crucial for doing poverty-related analysis and for policymaking. The importance of maize in overall household food consumption is higher for poor households than for nonpoor households in both rural and urban areas. In rural areas, maize accounted for 25 percent of the value of the food consumption of poor households in 1990/91, compared to 16 percent for nonpoor rural households. In urban areas, the proportions were 12 percent for poor households and 7 percent for nonpoor households (background papers for World Bank, 1995a). However, we do not have information on what proportion of that maize was own-produced or purchased nor on how the proportions vary by year depending on the level of rainfall. Section 2 22 Income and Consumption Maize was the largest single source of marketed agricultural produce for poor households in communal areas in 1990/91; it accounted for 34 percent of their income from marketed crops. As mentioned above, however, data on income from cash crops are somewhat unreliable. Their other sources of income were vegetables and "industrial crops." Very poor households sold less maize as a proportion of their total agricultural sales than did poor households, probably because they are net purchasers of grain. Very poor households relied more on vegetables (35 percent of their total income from sales of agricultural products in 1990/91) and on other grains (background papers for World Bank, 1995a:32). Nonpoor households tend to derive more cash income from formal employment than do poor households Employment in the formal sector (which includes agricultural wages) is very important in Zimbabwe. Figure 2.4 shows, unsurprisingly, that this source was particularly important for all urban households. However, it was also important for nonpoor rural households, which presumably reflects formal employment on farms. In both urban and rural areas, poor households derive a smaller proportion of their cash income from employment than do nonpoor households. While 30 percent of rural households were living in poverty in 1990/91, only 10 percent of rural households with a family member employed outside the own farm were poor (World Bank, 1995a:32). The importance of formal sector employment (agricultural, non-agricultural, trading, and services) in enabling households to avoid poverty may be one explanation for why the ICES showed nonpoor households as having only slightly more land per capita than poor households on average (World Bank, 1995a:32). Further study of the importance and nature of the formal sector labour market (agricultural and non-agricultural) is vital to understand the determinants of the income of households in Zimbabwe. What, for example, are the most common activities of different income groups and in different parts of the country? What is the importance of agriculture for households that receive income from the formal sector employment of a household member (as distinct from remittance income)? Wage data from formal employment would be necessary to compare the returns to formal sector employment with those to other income-generating activities. This would demonstrate empirically the importance of formal sector employment in enabling households to raise themselves above the poverty line. Remittance income is substantial, especially in rural areas Households that receive remittances from one or more family member are common, and the income sent is extremely important to the receiving household. Over 60 percent of communal area households have at least one family member working away from the farm who contributes to household income (Rukuni, 1994:36). Scoones (1995) reports that, Section 2 23 Income and Consumption whereas communal area households used to define wealth exclusively in terms of livestock, in recent years they have clearly begun to incorporate the availability of remittance income into that definition. Remittances from family members are a crucial source of income for many households, particularly in rural areas. Around 80 percent of rural households (except those in the LSCF sector) reported receiving some income from remittances in 1991, and this source accounted for 22 percent of the cash income of all rural households. For poor and very poor households in rural areas, the share of remittances in overall household cash income was 30 percent (see Figure 2.4). Female-headed households also relied on remittances. In 1990/91, they depended on money sent: from family members for more than 40 percent of their cash income, so much so that the share of poor rural households who were female-headed was less than the share of all rural households who were female-headed (World Bank, 1995a). The role of remittance income in helping households move out of poverty is probably important but cannot be determined from available analyses of the ICES data. Available information aggregates the nonpoor together, making it impossible to analyze the importance of remittances for households with incomes just above the poverty line as distinct from those well above the line. Analysis of the income sources of households by income or expenditure quintiles would be helpful in this context. The level of dependence on remittance income implies that formal sector employment may be even more important to Zimbabwean households than the household-specific data in Figure 2.4 would suggest. It is not clear what proportion of the remittance income received by different households comes fiom formal sector emplo:yment. We believe the proportion to be large, given that the informal sector in 1990/91 was small and that the majority of informal sector workers make only low profits. Remittances also play an important role in poor urban households, 12 percent of whose cash income comes from that source, compared to 7 percent for nonpoor urban households. As in rural areas, remittances are particularly important for female-headed households; they make up 28 percent of the income of all urban female-headed households compared to 5 percent for all urban households headecl by men. The role of remittance income and extended family relationships are areas that need to be given particular attention in future research. The issue is clearly of major importance to poor households, yet we have little information to address many fundamental questions. How do Zimbabweans define a "household"? Do remittances usually come from one remitting household or family member or several? What is the relationship of the remitting household to the receiving household? What proportion of their income do the remitters typically send? How do the remitters generate their income? What happens when the remitter's income level falls? Is the extended family breaking down or strengthening, and in either case, in what ways? How is AIDS affecting remittances? Section 2 24 Income and Consumption Rural households derive half of their total income from in-kind sources In-kind income accounted for almost half of the total net household income of rural households in 1990/91 (CSO, 1994). Figure 2.5 shows the importance of own-produced food in rural areas, which accounts for 38 percent of income from in-kind sources. This is 18 percent of total household income for the average rural household. Imputed rent is a major source of in-kind income, accounting for 25 percent of in-kind income in rural areas and almost three-quarters of in-kind income for urban households. However, this is sensitive to the assumptions made in calculating the value of the housing.'8 Gifts received in kind were a major income source for both rural and urban areas. Other sources included wages received in kind and imputed medical care. This is clearly a crucial source of income for the poor, therefore more information about in-kind income sources should be a key research priority in the future. 2.4 SUMMARY Although analyzing household survey data presents many problems, particularly for generating reliable estimates of poverty prevalence, survey data available in Zimbabwe does reveal robust patterns about the characteristics of poverty. The data show that the majority of the poor live in communal areas but by no means are all of the country's communal farmers poor. There are wide variations in consumption levels within and among communal areas and within and among provinces. It is a key priority to learn more about the causes of this variation and about how poor households escape from poverty. The rural population has diverse sources of income, with formal sector employment and remittances apparently being key factors for escaping poverty. Both of these factors merit further investigation. The poor, however, do not always define their situation just in terms of their incomes. Therefore, in the future, studies that use participatory methods to understand the views of the poor themselves would greatly enrich the profile of poverty in Zimbabwe. In the next section, we move from analysis based on income and consumption to examine the status of human resources, particularly the health and education levels of the people of Zimbabwe. 18 Rent is imputed for owner-occupied dwellings by using the gross rent usually paid for similar accommodation of the specified kind minus maintenance expenditures or, when the area contains no rented dwellings, the value is imputed by the value of construction multiplied by an interest rate on savings plus depreciation. When the dwellings are provided by the employer free of charge, the imputation is counted as an addition to wages paid in kind and valued at the cost to the employer. Section 2 25 Income and Consumption Figure 2.5 Sources of In-Kind Income, 1990/91 Rural Areas Income from in-kind sources is 46% of total income. Other 7% Firewood - Imputed Rent 12% 25% Gifts in kind 18% \ Own Produced Food 38% Urlban Areas Income from in-kind sources is 10% of total income. Other Firewood 6% 1% Gifts in kind 15% Owna Produced Food 5% - Imputed Rent 73% Source: ICES, 1990/91 Section 3 26 Human Resources 3. STATUS OF HUMAN RESOURCES This section gives a brief description of the population of Zimbabwe and then describes the status of human resources in Zimbabwe. It uses the most recent available information to look at outcomes and service levels in the health and education sectors and overall trends in health and education indicators between Independence and the mid-1990s. It does not discuss any changes that may have taken place in the 1 990s, as these issues are covered to the extent possible in Section 5. This section is based principally on the 1994 Demographic and Health Survey and the 1992 population census, as well as on early drafts of the government's Public Expenditure Review (GOZ, 1996). 3.1 THE POPULATION LS YOUNG | Between 1980 and 1991, Zimbabwe's population grew by an average of 3.14 percent per year, one of the world's fastest growth rates. As a result, 45 percent of the population is under 15 years of age (CSO, 1992). Fertility rates have dropped considerably, although the population continues to grow fast at least in part because child mortality declined sharply after Independence. The 1994 DHS survey showed that, on average, a woman will bear 4.3 children, down 22 percent from the figure of 5.5 shown in the 1988 survey. Because of the large proportion of young people in the population, the country's dependency ratio is high. This ratio is calculated by taking the number of family members who are either younger than 15 or older than 65 and dividing the total by the number of adults aged 16 to 64 years. The 1992 census gives a dependency ratio of 0.94 for the country as a whole, down from the 1982 figure of 1.03 (CSO, 1992:103). This rate is high by international standards, but is somewhat lower than for neighboring countries that have also experienced fast population growth. The rate in both Zambia and Malawi, for example, is 1.06. Poor households tend to have higher dependency ratios (World Bank, 1995a:35), and, as AIDS takes its toll on Zimbabwe's adult population and the number of orphans increases, these ratios are likely to increase. The National AIDS Control Programme predicts that there will be 600,000 orphans in Zimbabwe by the year 2000, compared to 67,000 in 1992 (UNICEF, 1994:85 and 110). Section 3 27 Human Resources 12 WHAT IS THE STATUS OF HEALTH? | In this section, we use data from 1994 to present a picture of the status of health outcomes and services in Zimbabwe, while trends during the 1990s are discussed in Section 5. Zimbabwe has good health information relative to other countries in Africa. However, the data available are disaggregated neither by income levels nor by land-use groupings nor by natural regions. It is, therefore, possible that the trends in the aggregate figures mask changes in health outcomes or services for particular groups. Thus, it is particularly important to ensure that health data collected in the future are disaggregated if the profile of poverty in Zimbabwe is to be full and accurate. Zimbabwe has invested successfully in the health of its population For nearly a decade after Independence, Zimbabwe made impressive and consistent progress in the areas of health, population,, and nutrition. The government increased expenditures on public health and shifted their emphasis from urban and high-income areas towards rural areas. Between 1979/80 and 1987/88, government expenditures on health rose by 48 percent in real per capita terms, and rose from 2.2 percent of GDP to 3.1 percent in 1990/91 (GOZ, 1996a). Public spending focused on preventive and primary health care, family planning, and nutrition programs, and Zimbabwe pioneered many of the cost-effective primary and community-based health interventions that are today considered to be examples of best practice. The gox c,rnment has also increased access to health services throughout the country. Over the past 15 years, it has built, rehabilitated, or upgraded more than 600 rural health facilities so that by 1988, 85 percent of Zimbabweans lived within 8 km of a health facility (ZDHS, 1988) and it also greatly increased the flow of drugs to these facilities (GOZ, 1996). In addition, the government has invested in extensive nutrition and family planning interventions. These efforts have yielded results. During the 1980s, Zimbabwe invested slightly more than average for a country of its income level, yet it achieved much better results than average for similar nations. Life expectancy increased from 56 in 1980 to an estimated 61 in 1990 (GOZ, 1996 and CSO, 1992). Between 1980 and 1994, the infant mortality rate fell from nearly 100 deaths per 1,000 live births to about 50, weight-for-age malnutrition for children under the age of three fell from 22 to 16 percent, and child immunization rates increased from 25 to 80 percent (GOZ, 1996 anld ZDHS, 1994). Despite this progress, Zimbabwe's health problems are still predominantly those related to poverty. These are communicable, infectious, and parasitic diseases, respiratory infections, and maternal and peri-natal conditions. Acute respiratory infection, malaria, skin diseases, STDs, and diarrhea account for half of all out-patient attendance in Zimbabwe's health facilities (GOZ, 1996). Section 3 28 Human Resources Health status is good by international standards Many indicators of health outcomes are good relative to other countries with comparable income levels, even though the positive trends since Independence have reversed for some indicators in recent years (see Section 5). Infant mortality is 53 per 1,000 live births, compared to an average of 93 for Sub-Saharan Africa, and under-five mortality is around 77 per 1,000 live births, compared to 170 for the continent as a whole. Acute child malnutrition was 6 percent in 1994, one of the lowest rates in Africa (UNICEF, 1994). Table 3.1 shows that Zimbabwe is performing well in terms of child nutrition and under-five mortality relative to other Sub-Saharan African countries. Table 3.1 Health Indicators in Zimbabwe Compared to those of Neighboring Countries Percentage of Children Under-five Mortality PPP estimate of under Five: (0-59 months) G/Pcepita per 1,000 Uve births Underweight Stunted Zimbabwe (1994) 16% 21% 77 2,000 Botswana (1988) 55.9 5,160 Kenya (1993) 22% 26% 93.2 1,290 Malawi (1992) 27% 49% 239.5 690 Namibia (1992) 26% 28% 91.6 3,790 Tanzania (1992) 29% 47% 153.6 580 Zambia (1992) 25% 40% 177.6 1,040 Source: Demographic and Health Surveys for each country by Macro International Inc. The prevalence of HIV/AIDS in Zimbabwe is among the highest in the world HIV/AIDS is causing a major health crisis in Zimbabwe, and its effects also portend a major social and economic crisis. From the late 1980s onwards, the disease began to take its toll. This situation will continue to increase for at least the foreseeable future. The most reliable recent estimates put HIV prevalence at around 30 percent of the sexually active population (Latif, forthcoming). AIDS is now the leading cause of death for children between one and four years of age in Harare where every ninth child is born HIV-positive. This is one of the highest rates in the world, and evidence suggests that it will continue to increase at least in the foreseeable future. AIDS has a major impact even on basic health indicators. The US Bureau of Census has predicted that life expectancy in Zimbabwe, which peaked at 61 in 1990, could drop to 40 by the year 2010 because of the disease, although other projections show less dramatic but still substantial falls. AIDS has already sparked a dramatic rise in cases of tuberculosis and is having a noticeable impact on child mortality and acute malnutrition. Section 3 29 Human Resources Using widely accepted assumptions, we can illustrate the effects of AIDS on child health on a national level. It is believed that at least 30 percent of pregnanit women in Zimbabwe are HIV-positive. If we assume that 25 percent (which is considered a conservative assumption) of them pass the infection on to their children, then 7.5 percent of all Zimbabwe children born are HIV-positive. Almost all of these children will die before the age of five, meaning that HIV alone will soon account for 75 childhood deaths per 1,000 live births -- approximately the same as the current under-five mortality rate (UNICEF, 1994:85). It is, therefore, certain that child mortality will increase markedly in years to come, irrespective of what standards of care are available. In addition to its emotional and social effects, AIDS has severe economic effects on households. An adult usually has to care for the patient, thus reducing the time that this person has available to work or to care for the children of the household. If the breadwinner is infected, there is the additional effect of the loss of his or her major contribution to the household income. Treatment and funeral costs also reduce any savings the household may have and increases its economic vulnerability. Health outcomes are worse in rural than in urban areas Most indicators show that health outcomes in Zimbabwe are worse in rural than in urban areas. The under-five mortality rate in 1994 was 80 in rural areas compared to 63 for urban areas. Rural children were twice as likely to have suffered a recent episode of acute respiratory infection. Malnutrition rates, however, were not vastly different, being 17 percent for rural children versus 13 percent for urban children. Some sources report that child malnutrition rates are highest in resettlement areas whereas others indicate that the worst rates are among large-scale commercial farm worker households (cited in UNICEF, 1994:63). Health services tend to be less accessible to rural populations, although the differences are less stark than in many other countries. Children in rural areas are four times more likely to have been born at home than urban children, as 38 percent of births in rural areas are at home, compared to 9 percent in urban areas (ZDHS, 1994). Differences in the attendance of a qualified assistant at the birth are somewhat less stark, as the birth of 58 percent of rural children was attended by a doctor or nurse versus 96 percent of urban children (MPSLW, 1995:13). Vaccination coverage is not markedly different between urban and rural areas (ZDHS, 1994). Section 3 30 Human Resources III health is linked to low education levels Education of mothers is the principal variable associated with poor health outcomes of children. The mother's education has a positive association on the health and nutrition of her children as Figure 3.1 indicates. The children of women who have no formal education have more than 2.5 times the rates of overall malnutrition as those with mothers who have at least some secondary education. They are also more than three times as likely to have received no vaccinations. When the mothers have no education, the children are 30 percent more likely to have diarrhea and the mothers are less likely to know about simple treatment such as oral dehydration therapy. Low education of mothers and the poor health status of their children may be linked to lower incomes of these households as both causes and effects of low incomes, although the DHS data cannot be analysed from an income perspective. Figure 3.1 Relationship between the Mother's Education and the Health of Her Child, 1994 100 90 80 70 60 50 40 | - No Education g Pnimary 30 E3 Secondary or more 20-I 1 0 0 Under-5 % Who % of % of Malnutrition Mortality Deliver at Children Children rate Rate Home with ARI with (<36m) Diahrrea Source: ZDHS, 1994 Section 3 31 Human Resources 13.3 WHAT IS TIE STATUS OF EDUCATION? Zimbabwe's achievements in education are impressive Since Independence, Zimbabwe has invested heavily and successfully in education. Between 1979 and 1995, the number of children enrolled in primary school increased by 213 percent and the number enrolled in secondary school by a remarkable 1,000 percent. Primary education is now almost universal, with 86 percent of the appropriate age group enrolled in 1994, while around 60 percent of the appropriate age group is enrolled in secondary school. This is reflected in the rise in the literacy rate -- from 62 percent of the population over 15 years of age in 1982 to the current 80 percent (World Bank, 1995a and UNICEF, 1995). Education outcomes are outstanding in comparison with Sub-Saharan Africa. The adult literacy rate in Zimbabwe (80 percent) is far ahead of the continent's average of 50 percent. Zimbabwe has one of the highest gross primary school enrollment rates (the number of children of any age enrolled as a percentage of the relevant age group); at 108 percent, it is much higher than the average of 66 percent for the continent as a whole (MPSLSW, 1995 and ADAE, 1995). Zimbabwe's educational achievements also compare favorably with those of countries in other continents with similar levels of income. Of eight countries with comparable income levels measured in terms of purchasing power parity, only one (Honduras) has a higher rate of feffmaie literacy (World Bank, 1995d). Zimbabwe's figures may, however, be unrealistically high, as their definition o:f literacy differs from that of many other countries.' Education levels are lowest in the large-scale commercial farm sector Outcomes are reasonably consistent among land-use groups, with one exception. Education coverage in the large-scale comrnercial farming (LSCF) areas is startlingly lower than in the other groups. Table 3.2 shows that 41 percent of 6 to 17 year olds in LSCF areas were not attending school in September 1994 compared to 17 percent in communal areas and 11 percent in urban areas. Commercial farming areas also have lower enrollment rates for primary school-aged children -- 38 percent of eight year olds living in LSCF areas were not attending school compared to 5 percent in communal areas Any child who has completed three years of primary school is considered to be literate in Zimbabwe, in contrast to UNESCO's definition of five years of primary school. A literacy survey would be useful to assess actual functional literacy levels in different areas of the country and for each gender. Section 3 32 Human Resources and 3 percent in urban areas (MPSLSW, 1995:42). Sources agree that the low enrollment rates in LSCF areas are concentrated among children of farm worker households. Table 3.2 School Enrollment Status by Land Use, 1994 Percentage of Children Aged.6.17 Enrolled i. Enrolled In N*t im Toal Primary School Secondary School School Rural Areas 70 1 1 19 100 CommunalAreas 71 12 17 100 LSCFAreas 55 4 41 100 Resettlement and SSCF Areas 71 11 18 100 Urban Areas 63 26 11 100 Total 68 15 17 100 Source: MPSLSW, 1995 A 1994 study of education in LSCF areas estimated that there is only one school for every six farms on average, each of which is likely to contain a large number of school- aged children. LSCF children have an average distance to school of 8 km. Some farms provide schools for the children of their workers. These schools are often not registered, so the government has little control over their educational standards. Some of these unregistered schools do not offer classes for the later grades of primary education. Some farm owners have expressed interest in providing registered schools but find the conditions governing registration excessive, given the relatively small numbers of children in each age group on their farms (for example, the regulations do not allow multi-grade teaching and the rules governing construction are rigid and expensive). Other farm owners are reported to have said that they have no interest in registering their farm schools because government teachers have a destabilizing effect, education makes workers more mobile, and Ministry rules might prevent the farm owners from using schoolchildren to work on the farm during labour-intensive periods (GOZ, 1996 citing Nyagura and Mupawaenda, 1994). Girls receive less education than boys but the gap is closing Gender biases in Zimbabwe, as in most other countries, manifest themselves in many ways. According to some indications of biases easily picked up by surveys, the differences between girls' and boy's education certainly exist but appear to be diminishing. Section 3 33 Human Resources Girls are somewhat less likely than boys to complete education at various levels, and this difference increases at higher levels. In 1994, girls dropped out of primary school at almost the same rate as boys (76 percent of girls who enrolled finished primary school compared to 78 percent of boys). Once they had graduated from primary school, girls were more likely than boys to make the transition to secondary school (70 percent of girls went on to secondary school versus 68 percent of boys). However, once in secondary school, girls were less likely to finish than were boys, as 66 percent of girls finished secondary school (form IV) compared to 75 percent of boys (UNICEF, 1995). These are average figures, and may mask greater inequality between the genders in different income groups or in different kinds of communities. In many subjects, boys pass more 'O' levels than girls. In 1994, 30 percent of boys sitting the mathematics 'O' level passed versus 18 percent of girls. The corresponding pass rates for the science 'O' level were 38 percent for boys and 20 percent for girls. Gender differences were greater in those schools that serve less affluent communities (in other words, farm and mine schools, rural council schools that serve LSCF communities, and district council schools that serve communal farm communities). Boys do not get better exam results in all subjects, however. In the 'O' levels for English, Shona, and Ndebele, for exarnple, girls performed better than boys (GOZ, 1996). The gaps between the genders appear to be closing. Primary school dropout rates for girls were almost 32 percent higher than for boys between 1986 and 1988, but the difference fell to less than 7 percent in 1990-92 (GOZ, 1996). This trend also holds for 'O' level passes. For all subjects and in all types of school, there were more and faster improvements in girls' indicators than boys' between 1984 and 1994. Girls' pass rates for mathematics 'O' level, for example, increased by 71 percent cornpared to an increase of 16 percent for boys (GOZ, 1996). Although the gender differences are worrying, education in Zimbabwe is remarkably equitable by international standards (as measured by rates of enrollment and exam passes alone). Gender biases in Zimbabwean education are far lower than in other parts of the continent. For example, the literacy rate for women in Zimbabwe is almost 90 percent that of men, whereas, on average across Sub-Saharan Africa, women's rates are only two- thirds those of men (ADAE, 1995). Section 3 34 Human Resources 13.4 SUMMARY Despite the enormous progress that has been made since Independence in health and education, problems related to poverty remain. The diseases associated with poverty dominate outpatient clinics. In addition, AIDS is likely to have a major, if as yet unmeasured, effect on poverty. Rural communities have less access to healthcare and worse health indicators than urban communities. Education is one of Zimbabwe's great successes, with impressive levels of enrollment and completion. Children in large-scale commercial farming areas have least access to education. There are some indications that gender differentials in educational outcomes are diminishing. In the following section, we look at the factors that are associated with poverty in Zimbabwe and that comprise some of its underlying causes. Section 4 35 Factors Associated with Poverty 4[ FACTORS ASSOCULTED WITH POVERTY 14.1 POVERTY AND INEQUALITY Opportunities for generating income have not grown along with Zimbabwe's impressive gains in human development. The poor have been constrained from moving into more productive activities by a combination of interlinked factors. Income is extremely unequally distributed. In smallholder agriculture, the poor have been constrained by a lack of access to good land, the unequal distribution of which is linked to the country's skewed income distribution. While large-scale commercial farms are generally located in areas with good quality soil and adequate rainfall, most mallholders farm small plots of land in the country's driest regions which are becoming increasingly environmentally degraded and vulnerable to drought. Large farms do not use land or capital efficiently and, thus, have not generated as much employment as might be expected. The chances for the poor to generate non-agricultural income have also been limited. The formal sector has been biased in favor of capital at the expense of labour, while excessive regulation and lack of access to credit have historically suppressed the growth of the country's informal sector and indigenous businesses. This section discusses each of these constraining factors. Income is unequally distributed The unequal distribution of income is striking. In 1991, 50 percent of the population received less than 15 percent of total annual incomes and accounted for about 15 percent of total consumption, while the richest 3 percent of the population received 30 percent of total incomes and were responsible for 30 percent of total consumption (Stenflo, 1993). One measure of inequality, the Gini coefficient, which measures rising inequality from 0 (absolute equality) to 1 (absolute inequality), gives a figure of 0.57 for Zimbabwe, which is among the highest in the world. It compares to 0.61 for South Africa and 0.62 for Malawi and is far higher than the rates of 0.44 for Zambia and 0.41 for Uganda. (Chen et al., 1993; Republic of South Africa, 1995; World Bank, 1995d). Inequality varies across the country's socio-economic groups. The 1990/91 ICES data showed income inequality to be highest in commercial farming and urban areas, although when white, Asian, and colored households were excluded from the commercial farming data, a picture of considerable equality emerges. The members of the black households Section 4 36 Factors Associated with Poverty tend to be farm workers and to be uniformly poor. Similarly, there was a reasonably equal income distribution among households in resettlement areas. Communal area households, however, were more heterogeneous. They had the same high level of inequality as the country as a whole (Jenkins and Prinsloo, 1995). Unequal access to land, water, other natural resources and to employment, all contribute to this income inequality and are covered in this section. Other important factors are access to and ownership of livestock and the availability of infrastructure (especially roads and transportation to markets). Land distribution is highly skewed The poor have little access to land Some 4,400 large-scale farms and ranches occupy one-third of the country's arable land, including 60 percent of the high quality land in Natural Regions (NRs) I and II. These farms produce more than 85 percent of the value of the country's marketed agricultural output (World Bank, 1994:11). The average commercial farm is 2,200 hectares in size, although this average is distorted somewhat by a few huge ranches in southern Zimbabwe. In contrast, an estimated 2 million farms in communal areas occupy half of the country's agricultural land. Three-quarters of the communal farm areas are in NRs IV and V where the quality of the land is poor. The average communal farm is 23 hectares, but the amount of arable land per farm is typically only three to five hectares. A 1989 survey of 750 communal households found that 70 percent had access to less than 2.5 hectares of arable land, while one-third had access to less than 1.5 hectares (Moyo, 1995:137). Communal farming lands are more densely populated and tend to be used more intensively than commercial farm land. Livestock density is higher, and, as Table 4.1 shows, communal farms have three times the population density and 3.5 times the cropping intensity of commercial farms. Maize yields from communal farms in 1989 were estimated to be 3.4 tonnes per hectare compared to the average of 5 tonnes per hectare on large-scale commercial farms (Moyo, 1995:152). Section 4 37 Factors Associated with Poverty Figure 4.1 Distribution of Land by Natural Region Distribution of Land in NRs I, II, & III RAs 14% - ~ Co rriunal 32% LSOF 6% 48% Distribution of Lasnd in NRs IV & V RAs 7% LSCF 26% ConTmnal SSCF 64% 3% " Source: Rukuni, 1994:13 Section 4 38 Factors Associated with Poverty Table 4.1 Major Features of the Farm Sub-sectors in Zimbabwe, 1993 Communal Resettlement Large-scale Otbere Area Farms Area Farms Commercial Farms Number of Farms 1,000 65.8 4.4 8.6 (thousands) Share of Total National 51% 10% 33% 6% Agricultural Land Population Density 32 13 11 varies (persons/kin2) Cropping Intensity (planted 14% 6% 4% 2-4% area/total area) I Other includes small-scale commercial farms and large-scale government farms Source: World Bank, 1995a:83. The figure for the number of LSCFs is from Rukuni, 1995. Ruralpoverty is associated with agricultural land of poor quality The problems of land distribution are not only related to the size of the plot but also to the quality of the land and the reliability of rainfall. The majority of communal farmers live in areas with the poorest land and the least reliable rainfall as Figure 4.1 shows. Less than a tenth of Zimbabwe's communal land is located in good farming areas, whereas almost half is in NR IV and 26 percent is in NR V (Moyo, 1995:130). The different quality of land is associated with great differences in productivity between communal farms in NRs I, II, and III and those in the two less favorable regions. Table 4.2 shows that communal lands in those regions that have the lowest rainfall produce far lower grain yields and virtually none of the country's marketed surplus. It also shows the dominance of farms in NRs I and II in terms of marketed surplus. The expansion in smallholder farming after Independence came from a very limited number of farms, principally in NRs I, II, and III. On average since 1983, 10 of the country's 170 smallholder farming areas have accounted for 75 percent of the Grain Marketing Board's (GMB) intake of grain from smallholders (Jayne, 1991) Section 4 39 Factors Associated with Poverty Table 4.2 Distribution of Communal Lands among Natural Regions, 1990 Natural Percentage of Percentage of Percentage of Percentage of Rein Total Communal Communal Smallholder Smaliholder Maize Land Area Land Grain sold to GMB Population Production I and II 9% 20% 50% 80% III and IVi 62% 63% 40% 17% V2 29% 17% 10% 3% ' It would be more appropriate to combine information for NR IV with NR V, but the data are not available in that form. 2 Data for NR V depend on some definitions of smallbolder/communal farmers. Source: Based on UNICEF, 1994:34 using data from R(ohrbach, 1990. The problems associated with land quality in NRs IV and V are exacerbated because the areas are physically marginalized. Communal areas are located in 56 of the country's 58 District Council Areas, 20 of which are in remote areas of the country, often near the borders, with few roads, railways, and urban centers.' These differences among communal farms in different zones underscore the need to analyze poverty by household in different natural regions. Land markets do notfunction well Excessive controls and/or unclear practices for buying and selling land are causing problems in all of the farming sub-sectors. In large-scale commercial farm areas, owners have freehold title to their land but need perminssion from the Department of Physical Planning before they can subdivide the farm or change its use to non-agricultural purposes. Procedures for obtaining this permission are slow and biased against sub- division. This has two results. First, many farms are too large for efficient land use, especially as there is no cost for under-using the land. This reduces the LSCF sector's efficiency, creates farming for mechanised farming and thus is likely to inhibit employment creation. Second, the large size and, therefore, cost of the farms makes raising enough capital to buy the whole farm prohibitive for the vast majority of potential buyers (Rukuni, 1994:89). Land conflicts are growing in communal areas and often center on inheritance and subdivision of plots. Unclear laws and varying practices are leading to serious problems, particularly as the majority of farmers die intestate. Authority over land rights in communal areas is muddled. It is formally the responsibility of Rural District Councils, 1 Communal areas also sometimes consist of fragmentary plots located around large-scale farms and state land (Moyo, 1995:13 1). Section 4 40 Factors Associated with Poverty but in practice, residents have tended to look towards traditional leaders on issues of land and natural resource management (Rukuni, 1994:25). The Land Tenure Commission found that customary laws and practices are becoming ever more ambiguous, and, as they have not been codified, they are becoming open to misinterpretation, abuse, and misapplication (Rukuni, 1994:32). More information is needed on how these conflicts are linked to poverty and how large-scale land titling programs are likely to affect different groups of the poor (for example, those who do not get title to land). Unclear procedures for inheritance are a particular problem for women in communal, resettlement, and small-scale commercial farming areas. Women rarely become heirs, as it is customary for land rights to pass to the eldest son. Widows, including minor wives in polygamous marriages, customarily continue to use their husbands' land and to live in the marital home. The Land Tenure Commission found, however, that the social checks that ensure that this takes place are diminishing, and widows are increasingly losing their rights to use land and livestock. Divorcees and widows who wish to remarry outside their former husband's family lose all rights to their former land and homestead, as the women typically have to go to their natal or new husband's home (Rukuni, 1994:3 1). Unequal access to water exacerbates the land problem The Land Tenure Commission found that there are great inequalities in the present distribution of water rights and that racial conflicts about access to water are increasing (Rukuni, 1994:91). Although new water legislation is currently being written, the existing water laws in Zimbabwe gave rights of ownership of surface water in perpetuity to the owners of the land at the time the law was passed in 1976. Most people in communal areas had no formal title and thus had no rights to surface water. Large and small-scale commercial farmers did generally take up their water rights (World Bank, 1994a). Underground water is not currently controlled at all, and anyone has the right to explore and use it, although wells are supposed to be registered with the local authority. New framework legislation has been passed that changes many of these rules, but the details are not yet available. Eighty percent of water use in Zimbabwe is for irrigation purposes, the infrastructure for which is highly unequally distributed. The commercial farming sector accounts for nearly 90 percent of the country's irrigated land, with the infrastructure typically paid for by the farm corporation or owner on a purely private basis (World Bank, 1995a). Because the farm owns the water rights and the infrastructure, it pays nothing at all for each unit of water used (unless withdrawals are downstream from a government dam, in which case the farm is charged a small fee). Thus, farm managers have little incentive to use water efficiently, so wasteful practices and inappropriate technologies, such as daytime spray irrigation, are common (World Bank, 1994). Section 4 41 Factors Associated with Poverty The Land Tenure Commission suggests several potential solutions to the land problem. It emphasizes four ways to increase smallholders' access to land. These involve either improving markets or bringing the costs of holding land to full economic value by measures such as: * Making legislative changes to facilitate land sub-division and to clarify procedures for inheritance (particularly for women) * Raising water prices for large-scale farmers * Introducing a land tax * Helping potentially successful smallholders to purchase or lease land. Agricultural services have notfocused on issues appropriate to drought-prone areas Extension, research, and marketing services have proved to be inappropriate for many smallholder farmers, particularly those in NRs IV and V. Before Independence, government support for agriculture was specifically designed to protect the interests of large-scale commercial farmers. After Independence, the government extended services such as veterinary, extension, tsetse-eradication, and research and development to smallholders but focused them overwhelmingly on smallholders in high-potential areas (Rukuni, 1994:41; World Bank, 1995a:67). The technologies that extension agents disseminated, such as cash crops and better seeds with fertilizers, tended to be appropriate only for high-potential areas. In addition, the structure and content of extension services was more appropriate to male farmers, as female farmers tend to favor subsistence over cash crops and tend not to go to the large, formal meetings where extension workers typically meeet with farmers. Meanwhile, drought relief has tended to reinforce unsustainable patterns of consumption and production in NRs IV and V. White maize is not a drought-resistant crop, yet it has been the principal foodstuff and seed stock distributed by extension and relief programs. White maize is popular because it produces high yields in non-drought years, is easy to process, and the husks can be used as a supplement for cattle feed. Programs are necessary to promote other, more drought-tolerant grains (e.g. sorghum), to challenge the benefits of the drought-vulnerable white maize. Many households in the country's driest areas receive drought relief more or less permanently, which has tended to reinforce these households' dependency on government largesse and to act as a disincentive for them to prepare for future droughts (GOZ, 1996). Section 4 42 Factors Associated with Poverty Environmental problems affect both good and poor lands Problems of environmental degradation in many areas of Zimbabwe limit the potential for smallholder farmers to increase their agricultural incomes. The problems are principally related to soil erosion. This reduces soil fertility and, therefore, crop yields. It also causes siltation of dams and rivers and adds to problems of water scarcity. One study (cited in Davies and Ratts0, 1996) estimates the cost of replacing soil nitrogen and phosphorous lost by erosion at around 35 percent of GDP. The highest rates of soil erosion are found in the smallholder areas in Zimbabwe's high- potential agricultural land, which led Zimbabwe's expansion of smallholder agriculture after Independence. A disproportionate share of Zimbabwe's severely eroded land is in NRs I and II, and all of the communal area districts that are the source of the expansion in smallholder production since Independence are classified as experiencing moderate to severe levels of erosion. This is one factor (among others) behind declining soil fertility and stagnating or declining yields since the mid-1980s in these areas (World Bank, 1994:26). It is nevertheless important to note that estimates of land erosion tend to be unreliable as they are typically based on trials at field stations rather than on farmers' plots. Environmental degradation also threatens the livelihoods of people living on poor quality lands. Increasing population density has certainly put pressure on the land. Over the years, plots have undergone continued sub-division, bringing the proportion of land that is cropped to between 2.5 and 3.5 times that of other major agricultural areas. This pushes cultivation onto marginal and unsuitable land, which, along with increased cutting of firewood, overstocking of livestock, and neglect of the fallow period, can contribute to erosion, declining crop yields, and a decline in the reliability of water sources. For poor households, the threat to their already precarious incomes and food security increases their vulnerability to both natural and price shocks. The household is then likely to reduce the time over which it plans to use its available resources and is less likely to be able to make long-term investments in the land, which feeds into the negative cycle of poverty and environmental degradation. One study (Whitlow and Campbell, 1989) found population density to be the best single explanation of land degradation. The effects of fuelwood shortages and the degradation of water sources typically have a particularly large impact on women. As women and children have the primary responsibility for collecting water and firewood, when these resources become scarce, this increases the time that they must spend on these activities and reduces the time they have available to spend on other activities. One study estimated that, in the early 1990s, rural women in Zimbabwe spent around two and a half hours per day fetching water at the height of the dry season (cited in World Bank, 1993). ENDA/ZERO (1992) report that one-fifth of communal areas reported shortages of firewood in 1980, but by the early 1990s, the proportion had risen to two-thirds. The study reports that in the 1990s, women needed to travel 2 kilometers to collect firewood when previously they had been able to Section 4 43 Factors Associated with Poverty find sufficient wood within one-fifth of a kilometer from their villages. Another study estimates that 37.5 percent of communal areas reported extreme shortages of fuelwood in 1989, while 28.4 percent had moderate shortages and only 32.7 percent reported an adequate supply (Davies and Ratts0, 1996:399). 4i2 WHY HAS THE ECONOMY NOT GENERATED EMPLOYMENT? The formal sector has historically favored capital over labour Government policies have historically created incentives for both manufacturers and commercial farmers to utilize capital rather than labour, which has meant that economic growth in Zimbabwe has not created jobs. The manufacturing sector has not generated sufficient employment Zimbabwe has a large manufacturing sector that has historically been very protected. The economy is one of the most diversified in Sub-Saharan Africa. Industry, both before and after Independence, has been protected from imports and has thus depended almost exclusively on the domestic market, which is driven by agricultural output and government consumption. This inward-looking policy has made the manufacturing sector vulnerable to contractions in local demand, which has exacerbated the negative economic effects of droughts. A combination of labour, interest rate, and foreign exchange policies created incentives for manufacturing companies to invest in capital equipment rather than to hire workers. Labour market policies, such as setting minimum wages and creating administrative and financial penalties for laying workers off, that were introduced with the intention of protecting workers, actually acted as disincentives for companies to hire new employees. The Employment Act in 1980 raised the costs of hiring workers and is reported to have reduced long-run demand for labour from the manufacturing sector by 25 percent (cited in Knight, 1996). This ensured high wages and security for those who had jobs but tended to work against the creation of new employment. As a result, earnings from the urban formal sector were 36 percent above average earnings in 1990, although this had fallen from 68 percent in 1979 (Knight, 1996:21) The bias against employment inherent in the high costs of hiring and firing workers was compounded by the reduced costs of capital, which, through the foreign exchange licensing system, made capital imports artificially cheap. Negative real interest rates provided manufacturers with an incentive to take out loans to invest in capital equipment. In addition, high depreciation allowances, an over-valued exchange rate, and high marginal tax rates gave companies an incentive to invest all of their surplus as quickly as Section 4 44 Factors Associated with Poverty possible. These economic incentives probably added to the inclinations of the foreign- born factory owners and managers to buy the imported capital equipment with which they were familiar. As a result of these two concurrent biases, capital-intensive rather than labour-intensive sectors and practices have grown, and the manufacturing sector has generated less employment than could have been expected. In 1982, Zimbabwe had a capital to gross output ratio of 0.95, far higher than South Africa's 0.35, which is already high by international standards (World Bank, 1995a: 121). In fact, the share of labour in output declined from around 18 percent in the mid-1980s to 13 percent in 1991. Unemployment in 1992 was 22 percent of the economically active population (CSO, 1992). Commercial agriculture has used neither land nor labour optimally A similar pattern can be seen in commercial agriculture. Former government policies protected the sector's access to capital, technology, foreign currency, and commodity markets. This led to an over-capitalized sector that did not take full advantage of available labour and, therefore, did not generate the expected amount of employment. The patterns of employment have changed in recent years. The overall number of employees dropped steadily until the late 1980s but began to pick up again in the early 1 990s as the production of tobacco and horticultural products increased. An increasing proportion of new jobs were casual or short-term. The reduction in permanent employment is partly the result of a reduction in the planted area within the sector, but it also reflects the costly employment regulations of the 1980s (World Bank, 1995a:93). The relative efficiency of land use in the large-scale commercial farming sector compared to other farming sectors has been the source of much controversy. Yields per planted hectare are higher in the LSCF sector (on average between 1982 and 1989, yields were between three and seven times more than those of the communal farm sector depending on the crop), but better soils and rainfall account for much of the difference. Productivity in the sector has also been helped by decades of state support, including agricultural research, extension, subsidized credit, electricity and diesel fuel, and investments in irrigation and marketing infrastructure (Moyo, 1995:98). There is a consensus that a proportion of the high-potential land in the LSCF sector is underutilized, although there is little consensus on how large that proportion is. The under-utilized land is either not used at all for agricultural purposes or is devoted to low-intensity production that has lower social returns and generates less employment than would be possible under crop cultivation (World Bank, 1 995a: 103). The informal sector is relatively small and has been tightly regulated The informal sector in Zimbabwe is not as large as in other countries in Africa, as strict government controls have limited its operation. Recently, however, some controls have been relaxed and the sector appears to have grown quickly, especially in urban areas (see Section 5). As in other countries, the informal sector has two principal groups -- the Section 4 45 Factors Associated with Poverty informal survivalists who engage in low-value activities to supplement their incomes in the absence of higher-paying alternatives and dynamic young firms with some growth potential. The informal sector is an important source of employment, providing five times as many jobs as the formal sector between 1988 and 1992 (Daniels, 1994:23). Informal sector jobs in 1993 were split between rural and urban areas in roughly the same proportions as the population (Daniels, 1994:23). The majority of these jobs are low-skill and low-profit activities -- two-thirds of all informal operators make less profit than the minimum wage for domestic workers, which is considered the lowest rung on the formal employment ladder in Zimbabwe (Daniels, 1994:23). As in most southern African countries, women represent the majority of proprietors of informal sector firms and are concentrated in low-profit, low-growth activities. In Zimbabwe, women head 71 percent of the businesses and 86 percent of the country's one- person enterprises. Two-thirds of female-owned enterprises are in five sectors (crocheting, grass/cane or bamboo processing, knitting, vending farm products, or vending other products), which are all in the lowest two profit quintiles for informal sector activities. Eighty-nine percent of women proprietors reported spending their profits exclusively on their households (Daniels, 1994:30). In general, growth rates for informal sector activities are slow, and graduation into the formal sector appears to be less common in Zimbabwe than in other Sub-Saharan African countries. A 1994 report based on a survey of micro- and small enterprises (MSEs) found that only 18 percent of all informal firms employing one to four people experienced growth of any sort. Only 4 percent of these firms expanded to employ more than 20 people. This is a lower percentage than in other Sub-Saharan African countries. Graduation into the formal sector is also less common in Zimbabwe than in neighboring countries. For example, fewer formal sector firms in Zimbabwe report having started out in the inforrnal sector than in Kenya, Botswana, and Swaziland (Fidler, 1996:9). This is to be expected given the prior narrow scope of Zimbabwe's informal sector. Large numbers of school-leavers cannot find jobs The lack of jobs in the formal sector and rigidity in the informal sector had major implications for the increasingly large number of young people graduating from Zimbabwe's schools in the 1980s. The number of people leaving secondary school increased ten-fold between 1980 and 1987. These young, relatively well-qualified people expected to find non-agricultural jobs just as their predecessors had done. Formal sector jobs, however, were being created at an average rate of around only 15 percent of the number of young high-school graduates entering the job market. The number of 17 year olds leaving secondary school in the late 1980s averaged more than 122,000, whereas the number of new jobs averaged 18,000. Clearly, the formal economy was not absorbing or making use of the newly acquired skills of the country's young people (Knight, 1996:45). Section 4 46 Factors Associated with Poverty 43 SUMMARY Many fundamental factors are associated with poverty in Zimbabwe. Some, such as the highly unequal distribution of income and of good agricultural land, are a legacy of the country's colonial past. Others, primarily the inability of manufacturing industry and commercial agriculture to generate sufficient employment and growth, result from a combination of former colonial policies and more recent interventionist labour market policies that were well-intentioned but had unforeseen negative consequences. In the early 1990s, the government began a series of policy changes that were intended to remedy some of the structural factors impeding economic and employment growth. In the next section, we review these policy changes, how they were implemented, and other factors that have affected the economy over the same period. We then do a preliminary analysis of how incomes and poverty are likely to have changed during the 1990s. Section 5 47 Changes during the 1990s 5. CHANGES DURING THE 199Os In the previous section, we discussed some of the structural factors associated with poverty in Zimbabwe. These factors restrict the ability of the poor to respond to, and thus benefit from, economic growth. In the early 1990s, as a means of counteracting some of these structural factors, the government introduced a program of economic reforms that primarily involved measures to eliminate the bias against job creation and to stimulate export-led growth. In this section, we first discuss the economic changes that have taken place in Zimbabwe during the 1990s and assess how these changes may have affected poverty. First, we discuss the factors that have affected the economy. These include those parts of the reform program that have been implemented and those that have not. We also discuss the effects of the major drought that hit the country in 1992 with enormous economic consequences. We then analyze changes in incomes and in the status of human resources to the extent possible, given the limited information currently available. The results of two major household surveys will shortly be available, which will enable future researchers to analyze the changes in poverty and household welfare during the 1990s in more detail. Other in-depth studies are also currently underway that will provide more information on the impact of adjustment on the poor. 15.1 WHAT CHANGES HAVE AFFECTED THE ECONOMY? Structural adjustment intended to stimulate growth and create employment By the late 1980s, it became clear that major economic policy reforrns were necessary in Zimbabwe. Industrial protection and heavy regulation were blocking economic growth, and an overvalued dollar discouraged the production of export goods. Inefficient parastatals and internationally uncompetitive industries dominated the manufacturing sector. Between 1980 and 1989, gross domestic product (GDP) grew by 2.7 percent per year, while the population grew by 3.1 percent per year, and the labour force grew much faster than employment in the formal economy (Knight, 1996; Skalnes, 1995). Section 5 48 Changes during the 1990s It had also become clear that public efforts at redistribution of incomes were unsustainable (Davies and Ratts0, 1996). The government was consistently spending more than it was collecting in revenues. Public expenditures had risen to 44 percent of GDP in 1989 compared to revenues of only 34 percent of GDP. The result was an average annual fiscal deficit equivalent to 10 percent of GDP between 1986 and 1989 (GOZ, 1996). In response to this situation, in 1990 the government began a five-year program of economic policy reforms, known as the Economic Structural Adjustment Programme (ESAP). This program was intended to increase economic growth and stimulate job creation by reformning economic incentives in favor of export-led sectors and by liberalizing some markets. The reform package included targets for reducing the fiscal deficit, devaluing the currency and floating the exchange rate, liberalizing imports, and reducing government intervention in the economy, particularly in crop marketing. Although the ESAP contained targets for reducing public spending and some measures to protect health workers and teachers from retrenchment, it did not specify measures to protect the kind of public spending that most benefited the poor. The ESAP also did not include any element of land reform, although the government did appoint the Land Tenure Commission in 1993 to investigate options for addressing the land question. At the time the ESAP was introduced, some stakeholders supported and some opposed the program, depending on how they expected to fare from it (to the extent that they had relevant information). Groups that had something to gain from market liberalization and trade reform supported adjustment. They recognized that their situation would stagnate or deteriorate without such fundamental changes. Large-scale commercial farmers and some industrialists, for example, recognized that their sectors would grow only if the economy opened up to external markets and/or if their products were made more competitive by a devalued exchange rate. Organized labour probably had the most to lose from the ESAP. They opposed the program, mainly because they saw it as threatening job security and formal sector employment. The government had designed the Social Dimensions of Adjustment (SDA) Programme to shield people from the expected transitional costs of adjustment, but as we shall see later, the SDA turned out to be neither well-designed nor sufficiently funded. In practice, the ESAP has not yielded the expected benefits. In fact, during the 1990s, Zimbabwe has experienced severe economic and social difficulties, caused partly by policy changes implemented under the ESAP, partly by the government's failure to meet fiscal targets, partly by drought, and partly by policy distortions that remained. What the situation would have been in the absence of the ESAP is, of course, not known. In the eyes of the Zimbabwean public, however, the ESAP alone is the principal cause of all of the difficulties of the 1990s. Highly charged debates are very often associated with structural adjustment programs, but even so, the ESAP in Zimbabwe has generated particularly heated arguments. There appear to be three reasons for this. Section 5 49 Changes during the 1990s First, in contrast with most countries that have undergone structural adjustment programs, Zimbabwe did not experience a major economic collapse before the ESAP was introduced. Therefore, there was little general acceptance or recognition of the need for policy reforns at the outset. Second, the government has not implemented certain key parts of the program. While it has made major changes to its exchange rate and trade policies and has taken measures to control public spending and the deficit, it has done neither of these things adequately. As a result, fiscal overruns are currently suppressing the growth that liberalization would otherwise be stimulating. Third, the program was designed and implemented without sufficient discussion with interested groups. The government did not communicate the objectives and scope of the program to the public and made little effort to foster debate about relorm options. Therefore, it has been difficult for the public to distinguish between the ESAP as it was intended to be, the ESAP as it was implemented, and other factors affecting the economy. The government also did not sufficiently acknowledge in public the expected costs of adjustment, which would have been impossible to avoid even if the ESAP had achieved all that it was intended to and within the intended timeframe. By not acknowledging the expected costs and by overselling the likely benefits of programs such as the SDA, the government inadvertently encouraged the public to expect rapid growth with little pain. Thus, by the middle of the 1990s, Zimbabweans have experienced the pain of adjustment and have seen only few gains. This has generated great distrust of the ESAP, of the government, and of the World Bank and IMF who advised the government on the reforms. Many people and organizations blame the ESAP for the human costs it did impose, such as retrenchments in the civil service and in manufacturing and the compression of wages. However, they often do not recognize the benefits that have accrued to both producers and consumers by maize market reforms, which were also part of the ESAP. Many groups also blame the program for things it was explicitly designed to avoid, such as the fiscal deficits that have contributed to inflation and high real interest rates. The ESAP also seems to bear the brunt of public recrimination against changes, such as declining health outcomes, that are not the result of the program but of a complex interaction of economic, social, epidemiological, and environmental factors. The controversy over adjustment has meant that no-one involved has been willing to engage in a constructive and open discussion about how the reforms could have been designed or implemented differently to minimize some of the negative changes experienced in the 1990s. Perhaps even more unfortunately, the controversy has also shifted the focus of the debate away from a major factor behind declines in some key health indicators -- the spread of HIV/AIDS. Developing a strategy to prevent further infection and to prepare for the inevitable social and economic consequences of the disease needs to be given at least as much attention as the reform of economic policies. Section 5 50 Changes during the 1990s The reform program has only been partially implemented Five years after the start of the adjustment program, progress on implementing the reforms has been mixed. Some reforms are in place, while others are not. This section discusses each element of the program in turn. Markets have been liberalized The reforms that have been implemented are mostly those that involve reducing controls on markets: * The foreign exchange allocation system has been dismantled and current account transactions freed from control. The market now determines the exchange rate. The real exchange rate (expressed as the ratio of the price of tradable goods to the price of non-tradable goods) rose by approximately 52 percent between 1990 and 1994. * Price controls have been abolished and consumer subsidies reduced. * Investment licensing has been abolished for all projects under US$40 million. * Controls have been removed on the domestic marketing of agricultural commodities, most importantly maize. The marketing boards have been, or are being, turned into commercial organizations, and domestic trading is now allowed. * Controls on agricultural processing have also been removed. Small-scale milling, primarily of maize, has been allowed to develop. * Some of the regulations inhibiting small-scale activities (such as zoning, hawker and vendor licensing, and transport controls) have been reduced and/or their enforcement relaxed. * Dividend remittances have been freed from controls for all foreign investment after 1980, and the repatriation of capital is now guaranteed for all investments after April 1993. * Labour regulations have been amended to reduce the government's involvement in determining wages and working conditions. Procedures for recruitment and retrenchment are also now more flexible. * The civil service was reduced by 16 percent between 1990 and 1995 (Knight, 1996:38). The government has not managed to meet its fiscal targets The original fiscal objectives of the ESAP have not been met. The budget deficit has stayed at around 10 percent of GDP, rather than falling to the targeted 4.6 percent. This has had a serious effect on the ability of the economy to respond to market reforms. The government borrowed on the domestic market to finance the deficit, which increased the demand for funds, thus pushing up their price. Thus, real interest rates have been in Section 5 51 Changes during the 1990s double digits throughout the 1990s, which in turn has deterred private investment and so suppressed economic growth. The fiscal overruns have two primary causes. On the expenditure side, actual spending has consistently exceeded planned budgets. There are many reasons for this, including poor planning and budgeting, insufficient expenditure control, ad-hoc expenditure items and insufficient intra- and inter-sectoral prioritization (see the following section). Meeting fiscal targets is made all the more difficult because the overall amount of money available to the public sector is shrinking -- a circular phenomenon caused, in turn, by fiscal overruns. As the government's debt grows, financing charges are taking an increasing amount of government expenditure, reaching 22 percent of total public expenditures in 1994/95 up from 12 percent in 1990/91. High debt service payments are, therefore, reducing the amount of funds available to finance essential public services, such as health and education. Even where the government has protected or increased the share of discretionary expenditures allocated to the sector, as it has in the case of health, the real amounts have been shrinking. In fact, total real public spending on all sectors net of interest payments declined by around 10 percent between 1990/91 and 1995/96. Total public expenditure including interest payments, however, grew by 4 percent over the same period. As a share of GDP, total public expenditure fell slightly from 43 percent in 1990/91 to 41 percent in 1995/96. However, again, total expenditure net of interest has declined from 37 percent of GDP in 1990/91 to 30 percent in 1995/96 (GOZ, 1996:8). On the revenue side, revenues fell from 34 percent of GDP in 1990/91 to 29 percent in 1994/95. This is the result of lower tax rates and of weakening tax collection and administration capacity. Also, the slow growth of the economy and the compression of wages have reduced the overall amount of income on which the government can levy taxes, which has further constrained revenues. The Social Dimensions of Adjustment program was poorly designed and insufficiently funded The SDA has not achieved even its limited objectives. The SDA had two elements, employment and training and social welfare, and the government established a Social Development Fund (SDF) to implement the SDA. However, the SDF did not become operational until more than two years after the ESAP had begun, and problems emerged in both elements soon after the SDF began operating. The employment and training program (intended to help retrenched workers to establish small companies) has suffered delays. Very few retrenched workers have submitted applications, and the fund has approved even fewer. The social welfare program (intended to exempt poor families from health and education fees and to provide the urban poor with money for maize) has similarly reached only a small fraction of its intended beneficiaries. The school fees exemption, for example, has reached only 20 percent of the target population and the food money program only 4 percent (Kaseke and Ndaradzi, 1993). Section 5 52 Changes during the 1990s There are two basic reasons why the program has been so unsuccessful. First, cumbersome, time-consuming procedures were needed to enforce the eligibility criteria, which placed the burden of proof of eligibility on the applicant. Second, the government managed the program without any additional staff or budget, meaning that the workload rapidly overwhelmed staff and led to great delays (Kaseke and Ndaradzi, 1993). The government has since moved away from the original objectives of the SDF and has designed a broader Poverty Alleviation Action Plan (PAAP), the aim of which is to provide investments in infrastructure and income-generating opportunities rather than transfers to people in poor communities. Over-regulation and ambiguous regulation remains in some sectors Although the adjustment program has removed much unnecessary regulation of markets, regulations on some activities are ambiguous. In some cases, the government has announced its intention to change legislation but has not yet completed the necessary legislative reforms, so entrepreneurs and local officials are often not clear about the status of the regulation. This constrains the ability of the private sector to respond to economic opportunities. In the formal sector, the excessive regulation of certain markets, such as telecommunications, deters entrepreneurial activity and improvements in service quality. In the informal sector, unnecessary controls restrict the growth of many small businesses. Permits, for example, still require a cumbersome and lengthy application process. Public resources could have provided more benefits for the poor Zimbabwe's public sector receives a considerably higher share of national income than in many other countries. In 1995/96, public expenditures were equivalent to an estimated 41 percent of GDP (down from a peak of 47 percent in 1988/89). This is far higher than the 26 percent average for Sub-Saharan Africa, the 23 percent average for East Asia, and the level of 33 percent in South Africa (GOZ, 1996). In real terms, however, the government's budget has been shrinking throughout the 1990s. The decline in available funds, caused by shrinking revenues and the burden of debt service, makes it all the more critical that public spending be allocated in ways that maximize the returns to the poor. The government could have made its spending more pro-poor in two ways. Section 5 53 Changes during the 1990s First, allocations among sectors could have given a higher priority to activities that benefit the poor. The country's 60 state enterprises received a subsidy of US$46 million in 1994/95. This was equivalent to almost one-third of the Ministry of Health's budget for that year, and represents a transfer from the taxpayers to the managers and workers of these firms and/or to the users of their subsidized products who are rarely the poor. Other non-essential expenditures included unplanned ad hoc items such as the All Africa Games. Second, allocations within sectors could have been more directed towards the services most likely to reduce poverty. The following section will show how the share of spending on basic health and primary education has fallen relative to that of higher-level services. Untargeted drought relief programs are another example of inefficient -- although in this case effective -- public spending. The government successfully manages to provide emergency food to vulnerable communities, sometimes on an enormous scale. At the height of the 1992 drought, for example, the government distributed free food to over half of the population (World Bank, 1995a:72). The government purchased three- quarters of the almost 2 million tonnes of grain imported into the country on commercial terms at that time (Muir, 1993:6). These programs are, however, largely untargeted and, hence, are unnecessarily costly. They have large numbers of non-poor beneficiaries and involve high financial and administrative costs.' Droughts have had devastating effects that could have been reduced with better preparation Drought is increasingly becoming a fact of life in Zimbabwe. The country has suffered increasingly serious reductions in its rainfall every seven to nine years since 1947 (UNICEF 1994:35), which have repercussions for the whole economy. In 1992, the country suffered the worst drought of the century, followed by another less severe drought (in terms of rainfall) in 1995. Agricultural production fell by 25 percent in 1992 and by 12 percent in 1995, largely due to declines in grain production of 75 percent and 40 percent (World Bank, 1 996a). Although massive inflows of donor aid and effective government distribution of food prevented widespread starvation, many households lost livestock and other assets and became more vulnerable to subsequenit droughts. At the macro level, the reduction in household income from agriculture reduced demand for processed goods. Industry, already suffering because of high interest rates, was hit hard, partly because of its high dependence on the domestic market. In addition, the cost Food relief programs have also become something of an entitlement with around 15 percent of the population receiving emergency relief on a permanent basis. By handing out white maize, these relief programs have encouraged the habit of consuming white maize, which is not a suitable crop for growing in drought-prone areas. Section 5 54 Changes during the 1990s of the massive relief effort was an unplanned expenditure that increased the fiscal deficit and, therefore, contributed to the rise in interest rates. Some of these far-reaching negative effects could have been lessened if the country had systematically prepared for droughts. Given the regularity of droughts and the very real possibility that the country is getting drier (the downward trend of ten-year mean rainfall is becoming steeper according to UNICEF, 1994), it is increasingly important to take every measure to prevent droughts where possible and to prepare for their occurrence. Some measures that would mitigate the impact of drought include promoting millet and sorghum (which are more drought-resistant than white maize) as staples, promoting conservation tillage techniques to save water and reduce erosion, and conserving water for irrigation on large farms. 52 WHAT WERE THE EFFECTS OF ADJUSTMENT ON INCOMES AND CONSUMPTION? In this section, we do some preliminary analysis of the effects of structural adjustment on household incomes and consumption. Our analysis is incomplete because there is as yet little systematic information on these effects, which makes it difficult to draw firm conclusions. We discuss changes in the most important components of household income, namely employment income from both the formal and informal sectors, and changes in maize prices. We then discuss the effects of the 1992 and 1995 droughts on households. We end by drawing some initial conclusions about the net effects of the various changes on average incomes. Although nothing definitive can be said at this stage, it seems likely that average household incomes have fallen during the 1990s. GDP grew by an average of just 1.1 percent per year and fell in annual per capita termns by 1.4 percent between 1991 and 1995 (World Bank, 1996). Real wages fell sharply, and formal sector employment grew by an average of only 1.6 percent per year, with sharp declines in the early 1990s. On the positive side, the informal sector has grown substantially, although mainly in low-profit activities. Another positive development has come from changes in the maize markets. The price of maize meal to consumers has fallen since 1993, when restrictions on small- scale milling were lifted, making cheaper (although lower quality) meal available. Consumers switched from industrially milled maize meal to the lower-priced, locally milled product (MPSLSW, 1994 and 1995; Chisvo and Munro, 1994). Producer prices of maize have either remained constant or increased somewhat. The gains from informal sector growth and from maize market reforms, however, are unlikely to be sufficient on aggregate to offset the declines in household incomes caused by the drop in formal sector wages. We, therefore, infer that average incomes have fallen. Section 5 55 Changes during the 1990s Incomes from employment have fallen Formal sector wages have fallen sharply The Zimbabwean economy has undergone major structural changes in the 1990s, as a result of policy reforms under the ESAP. Real wages from formal employment declined sharply overall between 1990 and 1994, although they did increase in certain sectors. Over the same period, the number of formal sector jobs has grown somewhat, although by less than the growth in the labour force. The most recent available information indicates that real wages fell by 30 percent between 1990 and 1994. This is true whether wages are deflated by the GDP deflator or by consumer prices, whether wages in formnal manufacturinf alone are considered, or whether agriculture and services are included (CSO, 1995). Wage declines are a normal consequence of structural adjustment policies, but in stable macroeconomics conditions, lower wages typically stimulate employment creation. Total employment from all sectors has grown at an annual average of 1.6 percent since 1991, although this average hides considerable yearly variation. Growth in employment is still much less than the 2.8 percent annual increase in Zimbabwe's working age population. The decline in formal sector real wages, combined with sluggish employment growth, is likely to have translated into a decline in average household incomes and an increase in the percentage of households living below the poverty line. Urban households are likely to have been directly affected as they relied on formal employment for 91 percent of their cash income in 1990/91 (see Section 2). Rural households are also expected to have been directly affected, although to a lesser extent, given their more diverse sources of income. Rural households are also likely to have been affected indirectly because the fall in wages is likely to have led to a decline in urban-to-rural transfers or remittances. However, remittances may not have declined by the same proportion as wages. In other words, the combined income of the remitting and receiving households (as defined by the ICES and other surveys) is likely to have fallen, but the remitting household may have cut back on its own consumption in order to protect the level of remittance it sends to the receiving household. Further research will be necessary to determine the nature of these changes empirically. Within this picture of overall decline in ernployment income, there are some promising signs. The composition of formal sector output and employment has changed. The terms of trade have changed in favor of tradable goods rather than non-tradable goods. 2 The decline in wages may be overestimated because employers are increasingly paying workers non-salary benefits, which are not included in the published wage data. Section 5 56 Changes during the 1990s Employment in tradables also grew by 7.2 percent between 1990 and 1994 (from 539,000 to 578,000 jobs), whereas employment in non-tradables grew by just 5.6 percent (from 654,000 to 691,000). Particularly strong sectors were finance and insurance, tourism, and construction. Growth in tradables probably contributes to reducing poverty because most of the rural poor derive their income from sources within the tradable sector such as smallholder agriculture, from wage employment on commercial farms, and from remittances from family members working in manufacturing and mining.3 It is likely that the economy is on a more competitive footing than it was before adjustment. The anti-employment bias that remained from pre-Independence days has now been largely reversed for two main reasons. First, high interest rates and the devalued Zimbabwe dollar have removed the bias in favor of imported and labour- substituting equipment. Second, real wages are lower and labour policies less restrictive, which has removed previous disincentives to hire new workers. Future growth is, therefore, likely to be more labour-intensive than in the past. There is an urgent need for concrete up-to-date information about key economic indicators. There are indications, for example, that the GDP statistics do not capture changes in Zimbabwe's economy in the mid-1990s adequately. For example, the statistics do not include informal sector activity, although it is known to have increased dramatically. In addition, the GDP figures are calculated using sectoral weights that reflect the industrial structure that existed before reform. Many sectors that have experienced major declines, such as the textile industry, still have high weightings, whereas others that have grown in recent years, such as tourism, are only given small weightings. Further investigation is necessary to assess the precise effects of these changes on the poor. What kind of workers, for example, take up the newly created jobs? Are they poor or not? Are they employees who were retrenched? What opportunities are there for young school-leavers? What are the implications for women? Are the new job opportunities absorbing urban workers or drawing in migrants from rural areas? The informal sector has grown, mainly in low-profit activities Before the ESAP, the regulatory environment strongly discouraged informal businesses. Zoning regulations, strictly enforced controls on unlicensed trading, and limits to and complicated procedures for obtaining a license all imposed high costs on microentrepreneurs. As a result, Zimbabwe had a smaller informal sector than other countries with a comparable per capita GDP level. Some remittances also come from wage-earners in non-tradable sectors. Information about the sources of remitance income is not currently available. Section 5 57 Changes during the 1990s Many regulatory restrictions have now been reduced. New operators may now enter the market, and existing operators can trade more freely. The principal changes affecting the informal sector have been the deregulation of the transport industry, relaxed enforcement of licensing and registration restrictions, and trade liberalization. The deregulation is not complete, however. In many sectors, there has been a time lag between the government announcing its intention to remove controls and when it actually changes the legislation. This has meant that some officials enforce the laws that are on the books, while others follow the government's stated policy. This leads to uncertainty for operators in the informal sector. The informal sector has grown substantially in terms of operators and workers despite the costs imposed by remaining regulatory problems. Between 1991 and 1993, the number of workers in informal enterprises grew at a rate of 14 percent, with the biggest increases in those sub-sectors that benefited most directly from the policy changes. During this period, the number of people working in transport grew by 35 percent and in trading by 23 percent (Daniels, 1994:24). Since then, the numbers are thought to have increased substantially, although data on this are not yet available. The growth in transport and trading has also had spin-off effects on other informal traders. The growth of the informal transport sector, for example, appears to have had a positive effect on many micro- and small informal enterprises. In 1991, 16 percent of all medium and small enterprises reported that transport problems were a major impediment to growth of their enterprise, whereas by 1993, only 3 percent reported problems of this nature (Daniels, 1994). It appears, however, that much of the growth in the informal sector in the 1990s has been in low-profit activities. The informal sector can grow in response to growth in consumer demand, as appears to have been the case in the informal transport sector, and it can also grow because of excess labour supply, in that people operate in the informal sector because they have no other options. The growth of informal operators in Zimbabwe in the 1990s has probably mainly been due to excess labour supply. The proportion of one- person firms in Zimbabwe's informal sector increased between 1991 and 1993, with most of the increase in rural areas (Daniels, 1994:13). Women in the informal sector appear to be involved predominantly in low-profit activities, and their participation in these activities has grown. In 1991, women headed 68 percent of all informal sector firms, which grew to 71 percent in 1993. Most of the growth, however, was in low-profit sectors. Firms owned by women also grew less (in terms of the number of jobs created) than those owned by men. In 1993, 4 percent of female-headed firms had increased the number of people they employed, compared to 10 percent of all those headed by men (Daniels, 1994:3 1). Economic contraction seems to have had two effects on informal sector workers. Some informal businesses are likely to have been hurt by it, particularly survivalist traders engaged in low-profit activities with low barriers to entry. As regulations were lifted and Section 5 58 Changes during the 1990s as unemployment increased, more and more workers entered the market, and average earnings appear to have declined. Skilled traders, however, are likely to have benefited. Demand for the products and services of the higher-skilled areas of the informal sector, such as metal work and furniture-making, is likely to have increased as purchasers have shifted their consumption from formal sector suppliers to lower-cost suppliers in the informal sector. The liberalization of maize markets has benefited consumers and producers The ESAP has lifted restrictions on trading, marketing, and processing maize and has also removed subsidies on maize meal for sale to consumers. Research into the effects of these changes is ongoing, but early findings indicate that private traders and processors have established themselves throughout the country and have generated considerable employment. The overall effects on production and consumption are positive. Consumers are benefiting from lower real prices of maize meal (albeit meal of lower quality), while producers are benefiting from more flexible markets and no apparent decline in prices. The specific effects on poor producers, who tend to live in remote areas and to sell small quantities of grain, have yet to be determined. Before the ESAP, the system of maize marketing was highly inefficient. The Grain Marketing Board (GMB) bought all of the country's maize and transported it to four industrial millers who were the only people licensed to process maize on a large scale. The large-scale millers then sold the meal at subsidized prices to both urban and rural consumers. These millers used expensive technology, had high profit margins, and generated relatively little employment. Local, small-scale millers could have produced coarser but more nutritious meal more cheaply, and would have generated more employment because the processing technology they use is more labour-intensive. The ESAP removed many of these restrictions. In June 1993, the government removed the subsidy on meal from the large-scale millers. It also removed the GMB's near- monopoly over maize purchase and its control over the movement of maize between different areas. The GMB did, however, retain its control over exports. The ESAP also removed the licensing restrictions on processing. These changes had significant effects. Removing the subsidy immediately caused the consumer price of roller meal to go up by approximately 55 percent. Several hundred small-scale mills sprang up throughout the country as did around 30 larger rural production millers. These mills were more labour-intensive than the large-scale millers, and allowed consumers to obtain and sell meal at lower prices. In late 1993, small mills (known as hammer mills) in urban areas were selling maize meal at prices lower in both nominal and real terms than those of the subsidized roller-milled meal six months earlier (MPSLSW, 1994 and 1995; Chisvo and Munro, 1994; Chipika 1995:21). Section 5 59 Changes during the 1990s Consumers switched on a large scale from roller-milled meal to hammer-milled meal. The Sentinel Surveillance Survey (SSS) of 1994 reported a dramatic shift in the composition of the country's food basket from industrially milled meal to locally milled hammer meal. In October 1992, 82 percent of households surveyed had consumed some roller meal in the previous month, while 25 percent had consumed some locally milled meal. In contrast, in December 1993, only 23 percent of households had consumed roller meal, while 71 percent had consumed locally milled meal (MPSLSW, 1994:15). The 1995 SSS showed that this change was sustained in late 1994 (MP';LSW, 1995:32). Per capita consumption of all grain increased from 10.4 kilograms per capita per month in December 1993 to 11.8 kilograms per capita per month in September 1994 due mainly to an increase in consumption of hammer-milled meal. The structural changes in this sector have probably had beneficial effects on employment, although more detailed data is needed before this can be investigated further. Employment in the hammer milling sector has increased enormously, whereas some jobs have been lost in the large-scale mills. Although no conclusive data are yet available, employment gains in the hammer mills probably outweigh retrenchments. To increase understanding of the impact of these reforms on poor consumers, more information is needed about how maize prices have changed since the reforms. This is especially important at the provincial and district levels, given that the private marketing of maize varies by province. In addition, more information is required about the consumption patterns of the rural poor in order to understand the impact of the price changes on them as consumers. Is it possible for them to substitute hammer-milled meal for roller meal? Information about how markets are developing in different areas will also be important. It is unclear how the reforms have affected poor households as producers of grain. Initial information indicates that private marketing operations of many different types and sizes have established themselves in some areas. It is not yet clear how effectively these operators serve the poorest, who typically live in remote areas and sell small quantities of grain. The closure of many GMB storage facilities could be a problem for poor farmers if private traders do not fill the gap in their part of the country on a regular and timely basis. There is a need for more information about grain markets and about the quantities of grain that poor households sell in a typical year. There is also a need to understand how much of their total income poor households derive from grain. The ICES shows that, in 1991, 16 percent of the total agricultural iricome (in-kind and cash) of poor and very poor rural households came from crop sales, compared to 22 percent for non-poor rural households (background paper to World Bank, 1995a). However, we do not know what proportion of total household income this represents; without access to the original data sets, the importance of agricultural cash and in-kind household income cannot be calculated. Further investigation of this issue is clearly a priority. Section 5 60 Changes during the 1990s Changes in the markets for other agricultural products, such as cotton, sunflower seeds, and tobacco, also affect some poor smallholder farmers. More research is also necessary on this topic to understand the nature of changes in these markets in the 1 990s and their effects on poor households. Drought reduced rural incomes and probably increased levels of inequality The shocks of two droughts in close succession had far-reaching effects on households. Many rural households lost all their livestock. Cattle losses of more than 50 percent were recorded in some provinces. The first drought left rural households with no marketable crop surplus with which to buy inputs for the following year (UNICEF 1994:35; Wilson, 1993). When the second drought came only three years later, they had fewer assets and lower incomes to cushion them from the effects. Drought may have increased inequality in rural areas, because households with insufficient agricultural and off-farm incomes would have been forced to draw on their assets and sell or consume their livestock. Families that managed to maintain their incomes were then able to buy livestock at distress sale prices. One study (Wilson, 1993) found evidence of this pattern in one communal area. Further information both on changes in assets (tools, savings, and livestock) in rural areas and on remittance income for different types of household will be necessary to determine the extent to which this pattern holds true in other regions. The effects on average household incomes have been negative The net effect of the economic changes that have taken place during the 1 990s will be different for different households depending on their sources of income and levels of consumption before adjustment. Some households will be better off, but many will be worse off. On average, we infer that most households have become worse off for several reasons. First, the country's total formal sector wage bill has fallen. Formal wages have declined sharply while the number of people employed has grown only slightly and has certainly not kept pace with the large numbers of school-leavers entering the job market each year. This situation affects urban households directly and rural households both directly and, through remittances, indirectly. These negative effects on household incomes were tempered by some growth in the informal sector and by maize market reforms, which probably had a net positive effect on average disposable incomes. It is unlikely, however, that the positive changes outweighed the major negative effects of the sharp declines in formal sector wages. Section 5 61 Changes during the 1990s Different households are likely to have been affected differently. Urban households are likely to be worse off on aggregate because of declines in real wages. Rural households that were poor before adjustment and had little access to income from employment or remittances are probably not worse off as a result of reforms, although they are likely to be worse off because of drought. The negative impact of drought is likely to have been tempered by agricultural reforms that made cheaper locally milled maize meal available to consumers and by the government's massive drought relief effort. Rural households that depended on remittance income to bring their consumption levels above the poverty line before 1990 may well have become poor since that time because of a likely decline in remittance income. As female-headed households (as defined by the surveys) are most likely to be in this group, it is likely that more female-headed households live in poverty than before. 15.3 WHAT WERE THE EFFECTS ON HEALTH AND EDUCATION? The health sector is under strain from funding cuts anid AIDS Health indicators improved steadily in Zimbabwe in the 1980s, but this trend appears to have stalled in the 1 990s. Infant and child mortality rates began to rise as early as the late 1980s. Different sources give different rates, but in all, a flat or even a slight upward trend is evident. The 1994 Demographic and Health Survey (ZDHS) gives a figure for infant mortality of 53 per 1,000 live births, the same as in 1988, and gives figures of 77 per 1,000 for under-five mortality compared to 75 in 1988. Maternal mortality rates have also increased during the 1990s, although the data on this are unreliable (UNICEF, 1994:72). There have been conflicting trends in the nutritional status of children. Wasting (weight- for-age malnutrition), which reflects acute deprivation, tripled between 1988 and 1994 from 1.3 percent to 5.5 percent of the population under the age of three. In urban areas, rates quadrupled over the period, so that wasting is now more prevalent in urban than in rural Zimbabwe. The case fatality rate for malnutrition also rose 11 percent in the early 1990s. Stunting (height-for-age malnutrition) showed the opposite pattern, falling to 21 percent in 1994 compared to almost 30 percent in 1988 (ZDHS, 1994), which amounts to major progress. The decrease in stunting rates is in part a tribute to the government's active nutrition efforts over the past decade. The reasons for the increase in wasting are more complex. Deprivation caused by the 1992 drought is probably one cause, while the fall in income levels discussed above is probably another. HIV/AIDS is also a major factor, as the disease is associated with severe wasting. This might explain the steeper increase in urban areas where AIDS is believed to be more common. Section 5 62 Changes during the 1990s These health outcomes reflect changes in the national averages, and thus do not necessarily reflect changes in the health of the poor. The health data that are currently available are not broken down for different income groups. Thus, it is possible that the health outcomes for some groups have improved and mask deterioration in outcomes for other groups. Health spending has declined Public spending on health has declined by all measures. As a proportion of the government's total budget, spending declined by more than 30 percent, from 6.4 percent in 1990/91 to 4.3 percent in 1995/96. It declined from 3.1 percent of GDP to 2.1 percent over the same period. Due to the growing population, in per capita terms the drop in spending has been even more stark. The 1995/96 budget allocation for health represented a 40 percent drop in real per capita spending from its peak in 1990/91. Consequently, real per capita public spending on health care is now at the lowest level since Independence. Not only has the overall level of spending fallen sharply, but the composition of that spending appears to have become less favorable to the poor. The proportion of the public health budget allocated to preventive and primary health care, for example, appears to have declined. Calculating the share of public resources allocated to these services is difficult, but one study (GOZ, 1996) estimated the share to be 36 percent of total Ministry of Health and Child Welfare (MoHCW) outlays in 1994/'95, a fall from 39 percent in 1992/93. These calculations were based on assumptions derived from detailed costing studies in three districts (MoHCW, 1993). Public spending on primary health care is now insufficient to provide even for basic needs. On a per capita basis in 1994/95, primary health care spending amounted to Z$51 (in 1995 Z$) or US$6. This is only two-thirds of the US$9 per person of public funding that is considered necessary to provide a basic package of health care in Zimbabwe (World Bank, 1994b). Despite this bleak picture, the government has protected the health sector to the extent possible. The MoHCW's share of the budget fell by less than that of any other ministry except Home Affairs. Health now accounts for a greater share of the government's non- interest expenditure than at any time since Independence. Health fees have proved to be a contentious issue. The MoHCW stepped up the enforcement of fees in 1991, with the objective of increasing cost recovery. The government established an exemption scheme based on the patient's level of income. In Some of the decline has probably been offset by private spending (although this is unlikely to have been relevant to the poor) and by extra-budgetary support from donors. Section 5 63 Changes during the 1990s practice, however, the scheme proved to be administratively cumbersome, and increases in the cut-off level for exemptions lagged behind fee increases. In some rural areas, health fees payable for primary care were not enforced in practice, although the public generally believed that fees were being collected and so may well have been deterred from seeking treatment nevertheless. A new fee schedule took effect in January 1994, which was designed to encourage users to enter the system at the appropriate level and not to bypass primary care. Fees for primary care did, however, rise considerably, and the exemption scheme remained so confusing and cumbersome that many of the poor were unable to obtain the necessary documentation to prove their eligibility. The MoHCW suspended fee collection at primaiy facilities in rural areas in March 1995. In urban areas, people with household incomes of less than Z$400 per month are now exempt from fees, and immunizations and communicable disease control are free to everyone. However, the exemption cut-off rate has not been adjusted for inflation since that time. Health services have deteriorated The decline in spending may have had a negative effect on the quality of care. Much of the reduction in spending has come from salaries, which have fallen by 30 percent since 1990/91, while the actual number of people employed in the health sector has remained fairly constant. The pay differential between public and private sector doctors is now huge, with negative effects on staff morale and recruitment of new staff. More than 90 percent of doctors in rural areas are now non-Zimbabweans (World Bank, 1995b; UNICEF, 1994). Non-wage recurrent spending, however, dropped even faster than spending on salaries. On a real per capita basis, this spending declined by 42 percent between 1990/91 and 1995/96. This decline is troubling because the category includes allocations for basic drugs and supplies that are needed to treat rnost common diseases. Drug shortages are becoming increasingly common. Utilization of services has remained constant on average between 1')88 and 1994. The proportion of births in facilities and attended births remained basically unchanged at around 70 percent in each case, and the proportion of women receiving antenatal care increased from 91 to 94 percent. Immunization coverage remained constant at around 80 percent, although coverage at the correct age dropped somewhat. This is the first period since Independence in which health service coverage has not increased. Problems of service levels have been particularly acute in urban areas. Probably at least in part because of increased demands for services at a time of declining resources, perinatal mortality rose from 36 per 1,000 in 1994 to 42 per 1,000 in 1995 (City Health Department, Harare, 1996). Between 1994 and 1995, the City of Harare experienced a 14 percent increase in under-five clinic attendances, and in referrals to central hospitals, and a 9 percent growth in primary care clinic attendances. Section 5 64 Changes during the 1990s HIV/AIDS is affecting health outcomes, straining service levels, and affecting the whole economy Some of the worsening of health indicators can be attributed to AIDS. This is evident from the fact that the downward trend in many indicators began when health spending was still increasing and before the economic downturn and drought of the early 1 990s, which is when the effects of AIDS first began to make themselves felt. Rapid growth in acute malnutrition and tuberculosis are both associated with AIDS. Many overall indicators of child health are also affected. For example, as mentioned in Section 3, AIDS alone will soon be responsible for 75 childhood deaths per 1,000 live births, compared to the 1994 total childhood death rate of 76. The epidemic is already putting a severe strain on the health care system. The MoHCW and the World Bank estimate that it will cost US$90 in drugs alone to treat each person with AIDS in Zimbabwe, while total estimated cost of caring for an AIDS patient in the formal health system would average US$614 (1991 estimates). Since around 1 million people in Zimbabwe are currently HIV-positive, this represents a total cost of almost four times the public expenditure allocation for health in 1994/95. HIV/AIDS poses enormous problems not only for the health system but also for the economy as a whole. With infection rates currently at around 30 percent of the sexually active population and still growing, the ramifications of the disease will soon hit the country's social and economic fabric at all levels. Preventing further infection and preparing for the inevitable economic and social consequences must be an urgent national priority. Women are likely to be particularly affected by the crisis for many reasons. They are biologically more likely to become HIV positive (the virus passes more easily from men to women than vice versa). Cultural taboos surrounding marriage also preclude them from refusing intercourse or insisting upon condom use. Also, polygamy means that one man may infect several women. Women are also affected because they are typically the primary caregivers if a member of their family becomes sick. In addition, the unclear rules that govern the inheritance of land often prevent widows from inheriting their late husband's land (see Section 4), making their widowhood financially precarious. Section 5 65 Changes during the 1990s Education funding cuts have not affected outcomes substantially Education spending has fallen steadily since 1988/89 as a share of both GDP and of total central government spending. In 1990/91 the sector received 18 percent of the central government budget, which had fallen to around 15 percent in 1994/95. As with health, the government has protected education spending as much as possible, and the sector's share of non-interest spending rose from 21 percent to 26 percent over the same period (GOZ, 1996). The government has given priority to maintaining teachers' salaries and student/teacher ratios. This policy is reflected in the figures for spending on supplies and teaching materials, which fell faster over the 1990s than did spending on salaries. Non-staff costs accounted for more than 10 percent of the primary school budget in the mid-1980s, but by 1993/94, this share had fallen to less than 5 percent. Despite these cuts, Zimbabwe still spends an unusually large proportion of GDP on education. At 8 percent, this is more than twice the proportion typically invested in education in the high-growth economies in East Asia (GOZ, 1996). This is partly because teachers' salaries in Zimbabwe are higher than is usual in other countries. Even after the cuts of the 1990s, a primary school teacher still costs around 6.5 times Zimbabwe's GDP per capita, which is unusually high by international standards. The government has managed to maintain student/teacher ratios, mainly by reducing teachers' real wages by between 25 and 40 percent between 1990/91 and 1994/94. The number of teachers has remained constant during this time. The quality of the teachers appears not to have declined, despite the fall in wages. The proportion of qualified teachers actually increased from 52 percent in primary schools, and from 48 percent in secondary schools in 1990 to 71 percent and 79 percent respectively in 1995. The adjustment in teachers' salaries was probably necessary as wages may have been unsustainably high. The composition of intra-sector spending appears to have favored higher education. Budget allocations for university education rose between 1990/91 and 1994/95, whereas those for primary education fell to their lowest level since Independence. Real government expenditure for primary education fell by 22 percent between 1988/89 and 1994/95, even though enrollment grew by 17 percent. Real resources per primary pupil thus fell by 33 percent. Section 5 66 Changes during the 1990s Education outcomes have worsened slightly Indicators of the quality of education provide mixed signals about outcomes in the sector. The principal measure of educational achievement used in Zimbabwe is passes at 'O' level. Between 1984 and 1994, pass rates increased for all subjects except for science. All pass rates increased between 1991 and 1994. 'O' level passes are, however, not necessarily the best measure of sector performance. Other measures present a more mixed picture. Although pass rates are increasing, overall success rates are very low. Fewer than 5 percent of children who entered primary school in 1981 passed five 'O' levels at grade C or better. Around 20 percent of secondary schools do not pass a single successful candidate, and around half pass only 5 percent of their entering class. This problem is particularly acute in rural areas. Furthermore, completion rates for primary school, for example, have fallen. Seventy-nine percent of the 1985-91 cohort of children completed primary education. Three years later, for the 1988-94 cohort, the primary completion rate had dropped slightly to 77 percent. The drop was greater for girls, since boys' rate of completing primary school had remained almost constant. This development is particularly worrying from a poverty perspective, as poor children are those most likely to have been affected. However, for those children who complete primary school, rates of transition to secondary school have stayed fairly constant (see Table 5.2). Rates of completion of secondary school have increased somewhat. Table 5.1 Education Transition Rates Grade 7 to Form 1 (beginning Form I to Form XV (completing secondry scbool) secwdary school) (% of au cohort) (% of age cohort) 1992 1994 1992 1994 Boys 73% 67% 73% 75% Girls 64% 70% 64% 66% Total 69% 69% 69% 71% Source: Ministry of Education and UNICEF, 1994 Section 5 67 Changes during the 1990s Future health and education outcomes are likely to deteriorate Public spending on health and education has declined in the 1990s. Although wages have borne the brunt of this compression, the government has protected the number of teachers and health workers as much as possible. Wages remain high relative to other sectors. Non-wage expenditures have suffered and indeed have declined even more rapidly than wages. Some of the cuts, particularly in education wages, may have been necessary given the high proportion of GDP dedicated to education in Zimbabwe relative to other countries. So far, in neither sector have outcomes declined commensurate with the size of the spending cuts. This may be partly explained by increases in the cost-effectiveness of service delivery, driven principally by the compression of real wages. It may also be that the adverse effects of the spending cuts on outcomes have not yet become evident. Certainly the trends are worrying, particularly for the poor. Non-wage expenditure for certain types of basic services are now too low. Staff morale in both sectors is thought to be suffering and infrastructure are not maintained adequately, threatening the quality of services to be provided in the future. Current outcomes are, therefore, fragile. It will be necessary to increase public spending on primary and preventive health care and on primary and secondary education if the gains made in both sectors since Independence are not to be eroded. 5.4 SuMmARY: WHAT HAS HAPPENED TO POVERTY IN TMHE 1990s? We have inferred that poverty has increased in Zimbabwe during the 1990s, although nothing conclusive can be said about the extent of the increase until forthcoming household data are analyzed. At this stage, it is also not possible to attribute causality. We base our inference on four types of information. First, incomes from formal employment have fallen over the 1990s.. The most recent information available indicates that real wages declined by almost a third between 1990 and 1994, whereas the number of people ernployed increased by only 7 percent over the same period. The total wage bill for the formal sector declined by about one-third during that period. Moreover, because of the importance of urban remittances for maintaining the incomes of rural households, holding other things constant, this decline in formal sector wages is likely to have increased the percentage of households in both urban and rural areas living in poverty. Section 5 68 Changes during the 1990s The changes in distributional patterns are not clear, but preliminary indications from data on the formal sector are that inequality has increased. The share of wages in profits has fallen. More information on the composition of the non-wage share of value-added will be necessary, as will more household data to increase our understanding of the impact of structural adjustment on income distribution. Second, there is clear evidence that the informal sector has grown during the 1 990s, but available information indicates that this has mostly been in low-profit activities. It remains to be determined empirically whether total earnings from informal activities have been sufficient to compensate for the decline in formal sector earnings. Third, the liberalization of maize markets has benefited both poor and non-poor consumers of maize meal by lowering consumer prices. The gains in consumption have probably not been sufficient to outweigh the drop in real incomes from formal employment and the negative impact of droughts on agricultural production. Changes in maize markets have probably also benefited producers. It is not yet clear, however, how these changes have affected poor producers, who tend to sell small quantities and to live in remote areas where private traders are less likely to be active. It is unlikely that the aggregate benefits have been high. Fourth, the poor have also been affected by the stagnant or declining provision of social services. Per capita spending on health fell by around 40 percent in real terms between 1990/91 and 1995/96, and real government spending per pupil on education fell by 33 percent during the same period. Although the government has tried to protect spending in these two areas, the available budget is severely constrained, not least by the need to service the government's debt, which now absorbs 22 percent of the total budget compared to 12 percent in 1990/91. Although outcomes have not yet declined to a similar extent, some negative trends are already evident. Reduced levels of public spending on social services, therefore, put future health and education outcomes at risk, particularly for the poor. Section 6 69 Topics for Further Study This paper highlights many areas where further information is needed to refine policies and to improve the design of interventions in Zimbabwe. This section discusses these topics and suggests the kind of new information that would be useful. Investigating some of these issues will require analysis of national household data sets, whereas others will require structured interviews and participatory techniques and, therefore, more focused studies. We note that the Government also undertook a national Poverty Assessment Study Survey (PASS) in 1995 to obtain current information to understand better the patterns and determinants of poverty in Zimbabwe. The PASS will provide rich information on the socio-economic characteristics of households and provide results which are valid at the district level. The final report is expected to become available in 1997. The issues where further analysis is deemed a priority can be groupecl into six areas: * Increasing our understanding of the living conditions of the poor, particularly how they may have changed during the 1990s * Analyzing the labour market in the formal and informal sectors * Deepening our understanding of the dynamics of remittance income * Understanding the impact that the country's changing agricultural markets have had on poor producers and consumers, including the relationship between poverty and access to land and water * Carrying out participatory studies of poverty, including questions about how households cope with crises such as drought or AIDS * Assessing the role of the country's institutions in reducing poverty. Increasing our understanding of the living conditions of the poor Zimbabwe has relatively abundant and comprehensive sources of household data, which have the potential for providing invaluable insights into the living conditions of the country's poor people. Analyzing a combination of existing data sets, the forthcoming Poverty Assessment Study Survey, and the 1995/96 ICES will allow researchers to increase our understanding of the characteristics of poverty in Zimbabwe. Methodological differences among data sets make it difficult to make comparisons over time, but building a poverty monitoring system would improve this situation. Carrying out participatory studies would also enrich the analysis. Yet even in the absence of these measures, we see five areas where analysis of existing data sets could increase the amount of information currently available to policymakers. Section 6 70 Topics for Further Study * Documenting changes in poverty, vulnerability, and health status over time: How has the incidence of poverty changed for different administrative groups, in different parts of the country, for female-headed households, and for young people? Has household size or composition changed? Are the effects of AIDS reflected in the data? How have income and consumption changed in different areas? Has income distribution changed? Once comparable longitudinal data are available, it will be possible to analyze the effects of the ESAP on poverty more systematically. How do the sources of income of different population groups differ? * Gaining a detailed understanding of the situation faced by vulnerable groups in the population: We need reliable data about large-scale commercial farm workers, households headed by single women or divorcees, households in marginal areas, mine workers, and other vulnerable groups. In the cross-tabulations that are currently available, data relating to these groups are aggregated with other data. * Understanding the factors behind the diversity of wealth and of welfare in communal areas: Seventy-six percent of Zimbabwe's poor households and 82 percent of the country's very poor households lived in communal areas in 1990/91. Policies and direct interventions aimed at reducing poverty, therefore, need to be designed to enable people in communal areas to meet their basic needs, which in turn will require a detailed understanding of the dynamics of these households and the forces acting upon them. Is poverty in communal areas associated with regional location, natural region, quality and size of available land, access to employment, gender, levels of education, distance from economic infrastructure, illness of a family member, or combinations of these and other factors? Which households move in and out of poverty as circumstances change, and under what conditions? Alternatively, which households are trapped in a cycle of poverty and why? Has the impact of adjustment been different on different types of household? What determines decisionmaking in communal farming households? What factors make the farmer more likely to invest in his or her land or willing to risk diversifying into new crops? How does the uncertainty associated with rapidly changing prices or weather affect the process? What effect does the the government's emergency relief provision have on a household's decisionmaking process? * Understanding the demands on people 's time, particularly women 's time: Women in particular tend to have many responsibilities such as fetching water and fuel, preparing and cooking food, growing crops, tending livestock, and raising children. How does this constrain their access to economic opportunities? Do government services such as extension or health care take women's timetables into account? What actions would reduce the demands on women's time? Understanding the formal and informal labour markets In Section 2 of this paper, we showed that rural households in Zimbabwe depended relatively little on own-farm agricultural production for their cash income in 1990/91. Section 6 71 Topics for Further Study Instead, they tended to rely on income from employment in the formal and informal sectors for more than half of their cash income. To gain a better understanding of the relationship between employment income and poverty in Zimbabwe, the following issues merit further investigation. * First, there is a need to know more details about formal sector employment. A greater understanding is necessary of what types of household earn income directly from what kinds of formal sector employment. IFurther investigation would also be useful into the effects of the ESAP and other changes during the 1990s. Which workers have been retrenched? Who are taking up the new jobs? What opportunities are there for young school-leavers? What are the gender implications? * Second, there is a need for more information on the informal sector. The informal sector is clearly an important area for the poor, either as a survival strategy that enables poor people to generate minimal incomes in the absence of other options, as a means of topping up household income, or because it provides an outlet for entrepreneurialism. We know that five times as many jobs were created in the informal sector than in the formal sector between 1991 and 1993 (Daniels, 1994), but the nature of these jobs is little understood. Do people engage in more than one informal activity? Do they have formal jobs as well as informal jobs? Do they engage in informal sector activities as well as farming? Do they undertake several informal sector activities at any one time or do they take on different informal activities at different times of the year? What links are there, if any, between informal sector activities and poverty? What is the gender dimension of informal sector activities? Similarly, what is the age profile of the informal sector? Are young people disproportionately concentrated in informal activities? Deepening understanding of the dynamics of remittance jincome Remittance income and the related issue of extended family relationships need to be given particular attention. This source of income is clearly of major importance to poor households, yet we have little information to address many fundamental questions. What relationship does the remitter tend to have with the head of the receiving household? From how many sources do different types of household receive remittances? What proportion of their income do the remitters typically send? How do the remitters generate their income? What happens when the remitter's income level falls? Are inter-household relationships changing? Is the extended family breaking down or strengthening, and if so, in what ways? How is AIDS affecting patterns of family ties and incomes? Section 6 72 Topics for Further Study Understanding the impact on poverty of changing agricultural markets It is necessary to gain a deeper understanding of the effects of agricultural reforms on the poor. Maize marketing reforms have affected the poor both as producers and as consumers. Most communal farming households produce some maize for sale in years of reasonable rainfall, even if they do not grow enough to meet their total household needs. There is little detailed information available about the amount of grain farmers tend to sell, relative to the amount they consume and the amount they grow. There is also insufficient informnation about the relative importance of this income source for different types of household. There appears to be little difference between the extent to which poor and non-poor households rely on own-produced food. What is the explanation for this? From which sources do different households receive their income or their food? What proportion of cash income were poor households spending on maize before and after the reforms? Now that a freer maize market is beginning to develop, it would be useful to gain greater understanding of how it functions. What, for example, is the role of private traders? What are their buying policies? Most critically, is the market competitive and how does this vary across different areas of the country? How much market power does the GMB have? How are prices determined, and how do they vary in different seasons? Has this affected household or private storage decisions or facilities? Is trade across regions developing? Have natural resource constraints (such as lack of access to land and water or environmental degradation) affected farmers' ability to respond to the policy changes? What is the situation with the markets for grains other than maize? There is also a need to address the implications of the agricultural resource base (land, water, natural resources, infrastructure) for poor communities. How, for example, do changing tenure systems affect different types of household? There is also a need to deepen and broaden our understanding of the relationship between population growth and movements and the capacity of different communal areas. Understanding how households cope with crises such as drought and AIDS Having a family member contract AIDS has a major economic effect on households. Ongoing studies are already investigating Zimbabweans' awareness of AIDS and its health care implications. The ways in which poor households cope with the infection and subsequent death of a family member, however, merits specific attention. How does news that the primary income eamer is infected affect household decisionmaking, for example? If a household member has AIDS, how much time and money does the household spend caring for the person and, subsequently, on his or her funeral? What happens to children whose parents are afflicted with or die from AIDS? How does this situation differ between urban and rural areas? Section 6 73 Topics for Further Study The gender dimension of AIDS needs to be given particular attention. Women bear a disproportionate share of the burden for several reasons. They are biologically more likely than men to become HIV positive, they are culturally inhibited from insisting upon men using condoms or from refusing to have intercourse, and women are the primary caregivers. In addition, widows face particular economic problems and difficulties in inheriting their husbands' land. Drought is another major source of shock to Zimbabwean households. It has a direct effect on poor rural households because it affects their water supply, crops, livestock and other assets (e.g. tools). Because the effects of droughts percolate throughout the economy, they also affect the cash income of the poor. How has drought affected levels of inequality in rural Zimbabwe? How have txvo severe droughts in the 1 990s affected traditional coping mechanisms, such as the extended family? Are social networks developing in newer communities such as resettlement areas? What is the link between drought, food security, and poverty? How does lack of food security link with poverty? We have little information, for example, on how households respond to deficits of own-produced food. Do they buy grain, collect wild food, borrow from neighbors, and/or receive emergency rations? If food-insecure households are forced to liquidate their assets in order to buy food, are they likely to have even less food security the following year? In addition, it would be useful to bring together the literature on food security and environmental conditions with the literature on the effects of economic reform to assess whether adjustment has affected the ability of households to maintain food supplies in times of drought. What are the effects of the food security programs in terms of providing incentives or disincentives for food-insecure farmers? Do specific population groups, such as households headed by commercial farm workers, have particular problems? These households appear to be more dependent on purchased food than are other rural communities. This would make them vulnerable to sudden increases in the prices of staple foods, especially if they do not have access to cheaper substitute foods. Addressing the role of institutions Effective institutions are vital for successful public or private poverty reduction activities. What are the responsibilities of existing public, private, and community institutions in poverty-related endeavors? How effectively do they fulfill these responsibilities and what constrains them from being more effective? What are the political economy considerations that lie behind allocations of public resources at a national and sub- national level? How could a more effective institutional environment be created? Section 7 74 Actions for the Future 7. ACTIONS FOR THE FUTURE Poverty in Zimbabwe may not be as prevalent as in many other countries in Sub-Saharan Africa, but a large number of people are nevertheless unable to meet their basic needs. The distribution of income and assets is highly unequal, and the country is not meeting its full potential in terms of economic growth, employment creation, and poverty reduction. Nor is the situation improving. Indeed, the number of people living in poverty has almost certainly increased during the 1 990s. This means that reducing poverty in Zimbabwe in the future will depend crucially on two sets of actions -- continuing to implement policy reforms to stimulate growth and create employment, and by improving the effectiveness of direct measures aimed at reaching the most vulnerable groups of the population. Implementing a poverty reduction strategy To stimulate broad-based economic growth in the future, it will be necessary for the government to achieve and maintain fiscal stability and ensure that the regulatory environment does not discourage economic growth. The whole of society can be expected to benefit from these policies, but reducing poverty requires that macroeconomic policies be supplemented by measures to enable the poor to respond to new opportunities and by an effective and efficient safety net. We now discuss the specifics of each component of the strategy. Creating an environment conducive to economic growth Economic growth is essential for reducing poverty on a large scale. A dynamic economy provides opportunities for employment, markets in which the poor can sell their goods, and a large revenue base that the government can use to invest in essential social services. Although growth alone will not be sufficient to reduce poverty, especially in a country such as Zimbabwe where the initial conditions are highly unequal, poverty cannot be sustainably reduced without it. The government has already created many of;the conditions for future economic growth. It has liberalized markets, removed some trade barriers, and freed the exchange rate from external control. As we saw in the previous sections, the structure of the economy is changing and some of the conditions for employment creation and economic growth are in place, but fiscal overruns are impeding faster growth. Therefore, measures to improve the environment for poverty-reducing economic growth include: Section 7 ,75 Actions for the Future * The first priority if poverty is to be reduced in a sustainable way is to decrease the government's deficit. Reducing the deficit would allow real interest rates to fall, which in turn would enable companies to borrow, invest, and take advantage of the opportunities created by the liberalized economy. To control the deficit, the government will have to reduce the quantity of its public spending and increase revenues. There is potential for this, even given the already highly constrained situation of the mid-1990s. However, it is crucial that the government should avoid making cuts across the board and that it should allocate its funds towards activities and services that are likely to have the greatest impact on reducing poverty. Once the fiscal deficit is under control, the conditions will be right to continue liberalizing markets, including deregulating the telecommunications market and accelerating the privatization of many state-owned enterprises. * The second priority is to remove constraints to growth in key sectors in which Zimbabwe has a comparative advantage and which have high employment elasticities of growth. In recent years, employment has been increasing most noticeably in the low-income informal sector. Yet Zimbabwe has one of the most extensive formal manufacturing and agriculture sectors in the region. Remaining constraints on growth in key sectors such as garments and textiles, where high employment potential exists, need to be removed on a priority basis. Furthermore, the conditions need to be identified which would facilitate technology transfer in high employment industries, not only through foreign direct investments but through other forms of collaborative ventures. * The third priority is to increase the access of poor farmers to good quality land and water. Current inflexibility in the land market prevents the efficient operation of the agricultural sector, and restricted access to land is preventing the poor in the communal areas from being able to raise themselves out of poverty. Experience in Zimbabwe and elsewhere shows that carefully designed land tenure and market reforms are likely to bring benefits to the poor in the long term as well as the short term. The report of the Land Tenure Commission contains recommendations that should be used as the basis for a process of consensus-building and reform. The option of improved operation of land markets in the commercial farm sector is likely to lead to greater flexibility and dynamism in resource allocation in the medium to long term. Linkages to the smallholder farm sector (e.g. through labour markets) would enable the incomes of the latter to grow too, thereby reducing the pressure on the land in communal areas. A key action is thus to build on the Land Tenure Commission's Report and the experience of successful strategies in other countries to develop consensus and and then implement reforms in land tenure and/or markets. Investing directly in poverty reduction measures There is still a pressing need to take specific actions targeted to the poor regardless of whether policy reforms are implemented or even whether growth is achieved. Even the Section 7 76 Actions for the Future current constrained level of public resources could be reallocated to have a greater impact on reducing poverty. Measures to improve the quality of public spending include: * Protecting and, if possible, restoring levels of public spending on health and protecting spending on education. Allocations within the sectors should give priority to primary and preventative health care and primary and secondary education. * Making selected, well-targeted investments in rural areas, concentrating on measures to manage land, water, and other natural resources, and activities to prepare for and mitigate the effects of drought. Investments should also aim to prevent the deterioration of existing infrastructure. Areas with good economic potential could benefit from economic infrastructure that would enhance the community's ability to respond to growth opportunities. Areas with little economic potential would probably benefit most from investments in social infrastructure. * Building the capacity of poor communities to identify and implement effective solutions to specific problems, that is, empowering communities to take advantage of economic opportunities. * Better targeting of emergency grain distribution and child feeding programs to the poor, so that the benefits are less likely to accrue to the non- poor, and so that eligible communities are less likely to become dependent on the relief provided. * Working aggressively and across all sectors to reduce future rates of HIV/AIDS infection and to prepare as much as possible for the devastating social and economic impacts the disease will have. If these interventions are made in a situation of strong economic growth, they should focus on enabling the poor to respond to new opportunities and on providing a safety net for those who are unable to do so through age or illness. If growth is slow, the interventions should focus more specifically on providing targeted safety nets, because these are even more vital, although less affordable, in this situation than when the economy is growing. Moving forward Because the interventions outlined above build upon current policies and initiatives, they could be implemented immediately. New information and insights generated by further research and study will only enhance the ability of policymakers to update, refine, and add to their poverty reduction strategy. In contrast with the situation that prevailed when the 1990 structural adjustment program was introduced, there have now been five years of real economic difficulties in Zimbabwe. As a result, the public has shown some desire for change and some recognition that the welfare state of the 1980s is financially Section 7 77 Actions for the Future unsustainable. Again in contrast with the way the ESAP was designed and implemented, the government now sees the need for and benefit of public dialogue on the next phase of reforms. Thus, there are now fairly good prospects for reducing poverty in Zimbabwe. In the first half of the 1 990s, the Zimbabwean economy underwent some painful changes but has benefited little from them because the economy has not grown. Also, the poor were not adequately protected from some of the effects of the reforms. However, the economy does appear to be healthier as a result of the policy changes, which means that implementing a poverty reduction strategy that promotes growth and simultaneously invests in the poor should be feasible and should yield positive resulls in the second half of the decade. Seminar Proceedings 78 Summary . SUMMARY OF SEMAR PROCEEDINGS I Poverty and Adjustment in Zimbabwe Co-hosted by Shanduko, The Centre for Agrarian and Environmental Research and The World Bank Introduction A day-long seminar was held at Thetford Estate, Mazowe in Zimbabwe on July 25, 1996 hosted by The World Bank's Southern Africa Department and Shanduko, the Centre for Agrarian and Environmental Research. There were two main objectives of the seminar: first, to review two draft papers, Understanding Poverty in Zimbabwe prepared by the World Bank and The Impact of Economic Reform on Livelihoods and Poverty by Mr. Tony Addison of the University of Warwick; second, to discuss priorities for further study necessary to make policies and targeted interventions more effective in reducing poverty. The participants at the seminar were members of the academic and NGO community in Zimbabwe, together with international researchers (see attached list of participants). The workshop was chaired by Mr. Brian Raftopoulos, the Acting Director of the Institute of Development Studies (IDS) at the University of Zimbabwe and the Chair of the Poverty Forum. It was opened by Ms. Winifred Goromonzi, a senior official from the Ministry of Public Services, Labour and Social Welfare, which is charged with coordinating the Government's Poverty Alleviation Action Plan (PAAP). The World Bank gratefully acknowledges the financial support of the Dutch Trust Fund for Poverty Assessments which made the organization of the workshop possible, and the efforts of the organizer, Dr. Dale Dord of Shanduko. Main responses to the papers The discussants at the seminar regarded the papers and the seminar as milestones in the dialogue between the World Bank and the public in Zimbabwe in that they were seen to place the issue of poverty reduction firmly 'on the map'. The papers were assessed to Seminar Proceedings 79 Summary have provided a comprehensive overview of recent research on poverty in Zimbabwe, as well as a balanced analysis of the patterns of poverty, their causes and possible solutions. The seminar concluded that the focus of the papers on poverty in the communal areas was appropriate because that is where the majority of the poor live, but that this should not detract from the need to address the growth of urban poverty. Two major gaps were evident in the papers: land tenure reforms and the role of institutions. Although the studies do highlight land tenure reform as a priority in the policy agenda, they do not cover the linkages between access to land and poverty in sufficient depth. This is particularly pressing as implementation of the recommendations of the Land Tenure Commission appears to be slow. In response to the query on institutions, it was mentioned that a study on the role of institutions in poverty reduction is to be undertaken in the course of preparation of the CDP. Identifying priorities for policy-oriented research A key objective of the workshop was to discuss priorities for further study in order to inform the design of policies and targeted interventions to reduce poverty in Zimbabwe. In this regard, five key topics were drawn from the papers and identified for discussion: the impact of economic policy on employment; the impact of agricultural marketing reform on rural poverty; the agricultural resource base of communal areas; the economy and diversity of communal agriculture; and the role of institutions. Short papers were presented by Zimbabwean researchers on each of the five topics' (a full account of the proceedings, including the short discussion papers presented, is available on request from Shanduko). In the afternoon, participants were asked to join one of five groups, each of which discussed one of the main seminar topics. Each group was asked to highlight the most important poverty variables and their causal links to economic and agricultural policy and to suggest a few key questions for further research. In the late afternoon, the rapporteurs from each group presented the results of the group's discussions. On the basis of the discussions which took place, a number of policy and research priorities emerged in each of the five areas. Priority area 1. The impact of economic policy on employment: understanding the relationship between economic growth and employment generation following reform; assessing the appropriateness of labour-market institutions as a means for stimulating employment growth, especially for the young; understanding the determinants of investment and labour demand in the private sector; quantifying the flow (and direction) of remittances and the impact of remittances from the urban sector on accumulation and diversification in the communal sector. The five presenters were: Messrs./Mmes. Jessirnan Chipika, Kay Muir-Leresche, Dale Dore, Godfrey Mudimu and Bevlyne Sithole. Seminar Proceedings 80 Summary Priority area 2. The impact of agricultural marketing reform on rural poverty: the constraints affecting the formation of private markets for maize, other grains and cash crops (such as infrastructure, market information, the cost of credit, and the consistency of government policy); the competitiveness of private marketing by crop and by district; the impact of market liberalization on the consumer prices of maize and other grains, and therefore on the incentive for smallholder maize producers to diversify away from maize into more profitable cash crops; the problems of marketing in marginal areas and the appropriate public policy response for communal farmers in those areas; the linkages between commercial and communal farmers in marketing; the impact of the development of hammer milling and oil seed processing on employment in the rural micro-enterprise sector. Priority area 3. The agricultural resource base of the communal areas: assessing the implications of proposals of the report of the Land Tenure Commission for poor communities and poor people within communities; the development of individual tenure and the implications of different forms of tenure relationships in the communal areas for the access of the poor to natural resources; the relationship between population growth (and shifts) and the carrying capacity of the communal lands and the sustainability of communal agriculture; the extent to which intensification of communal agriculture is possible given a generally less favorable natural resource base than in (for example) East Africa; the socio- economic and micro-political relationships which influence the ways in which resources are used and through which policy changes and projects must work. Priority area 4. Understanding the economy and diversity of communal areas: the scale of economic differentiation (characteristics of livelihoods, role of non- agricultural income sources etc.) across the communal areas; the prospects for policy and project interventions (and the costs and benefits) in the communal areas depending on their natural resource base (e.g. NR II versus NR V); the livelihoods of women in the communal areas and their access to productive resources, including land, and credit; the pathways by which poor households can accumulate and diversify their livelihoods, and the constraints within different types of communal area on those processes. Priority area 5. The role of institutions: the responsibilities of public, private, and community institutions in the rural areas; their focus on poor communities and on the poor within better-off communities; their effectiveness in meeting those responsibilities and constraints on that effectiveness (infrastructure, management capacity, and government-community relations); improving the co-ordination between institutions including clearer definition of responsibilities, reducing bureaucratic layers, and improving communication; the political economy of allocating public resources both at the macro-political level (the cabinet, the Seminar Proceedings 81 Summary ministry of finance and the line ministries) and the micro-political level (provincial government, and political administrations at ward and district levels). Summing up At the end of the day, the World Bank's team leader, Ms. Helena Ribe, summed up the most relevant points to emerge from the day's proceedings. * Although the seminar concentrated on poverty in communal areas, this did not mean that there was not a pressing need to address the growth of urban poverty. * The two main gaps of land reform and the role of institutions are important issues for poverty reduction. The papers do identify the importance of access to high quality land but does not go into great depth given that the Land Tenure Commission has already done so. There is scope for examining further the extent to which the Commission's findings have been reflected in policy actions. The role of institutions will be given very careful consideration during the planning of the Community Development Program. * If policy interventions are to be effective, policymakers must have not only a conceptual model of poverty but also a clear understanding of how poor people perceive their own poverty and how they believe that they can contribute towards improving their own circumstances. * It is very important to find out more about the differences among those who are categorized as poor. For example, who are the chronically poor? Which households manage to escape poverty and how do they do so? It is now recognized that it is necessary to carry out household surveys and analyze the resulting data on a regular basis to ensure that informnation on the rural poor is current and fully disaggregated to reflect the great variation in the levels and sources of communal household incomes. * An issue that was raised repeatedly during the seminar was the need to know the cost effectiveness of public expenditures and their impact on the reform process. In particular, the government's priorities should be examined with reference to the level of its defense expenditure vis-ai-vis the level of its spending on social services that benefit the poor. In particular, how does over-expenditure by the government inhibit economic growth and employment? Seminar Proceedings 82 Summary * The poverty and inequality implications of the pattern of growth remain largely unquantified. Are data available to test the elasticity of employment with respect to economic growth? How could these findings be used to improve future economic reforms? * It is now recognized that macroeconomics policies that constrain employment creation and the real wage rate are likely to depress the level of remittances. Thus, additional information is needed on the growth of the formal and informal sectors, the role and levels of remittances, and the incentives that cause urban workers to send money to their relatives and friends in rural areas. * Various studies have confirmed the positive effects of the liberalization of maize markets. There is now a need to explore what kind of agricultural marketing institutions are emerging, as well as their impact on marketing margins and equity. It may be that the drive for efficiency is adversely affecting the rural poor who live in remote areas where GMB depots have been closed because they are not considered to be economically viable. There is a need to find innovative ways to ensure that households living in areas of poor potential and with low population densities can benefit from liberalized agricultural marketing. What role, for example, can the setting of floor prices play in resolving this dilemma? * Finally, it is important to find out more about the population, land, and environment nexus. We need to know more about the rate and direction of technological change in agriculture and its variation across different agro- ecological regions. * It would seem from the findings of the Land Tenure Commission that the land issue cannot be divorced from the need for proper land administration and effective institutional structures, especially at the village level. Clarifying tenure in the communal areas will involve finding a way to transform the present intricate and dysfunctional institutional set-up into a workable system. Ms. Ribe stressed that further research needs to be done on all of these issues to improve the information base for refining and/or designing policies for poverty reduction. Reliable information and data are also needed to develop indicators with which to monitor the direction and strength of changes taking place at different levels within the Zimbabwean economy. Nevertheless, enough is already known for the constitution of the key elements of Zimbabwe's poverty reduction strategy. These elements include: broad-based growth, protection of social sector spending, land reform and investments targeted to the poor. BIBLIOGRAPHY Poverty and Human Resources in Zimbabwe 85 Bibliography Africa Community Publishing and Development Trust. The Suffering are the Cornerstone in Building a Nation: Community Views on Poverty, Poverty Alleviation and Wealth Creation. Sponsored by The Government of Zimbabwe and the UNDP. January 1995 ADAE (Association for the Development of African Education). A Statistical Profile of Education in sub-Saharan Africa 1990-1993. c/o UNESCO, Paris. 1995 Amanor-Wilks, Dede Esi, and Contributors. In Search of Hope for Zimbabwe 's Farm Workers. Dateline Southern Africa and Panos London. 1995 Bidani, Benu and Martin Ravallion. Decomposing Social Indicators using Distributional Data. WB Policy Research Working Papers 1487. July 1995 Burgess, Stephen. A Bottom Up Perspective on Rural Organizations in Africa: The Case of Small-Scale Communal Farmers in Zimbabwe. African Rural and Urban Studies. Vol:l 1994. pp 141-169. Campbell, B.M., P.N. Bradley and S.E. Carter. Sustainability and Peasant Farming Systems: Some Observationsfrom Zimbabwe. Stockholm Environment Institute. 1994 Chatora, Clever. The Resource Water. Presentation to a Workshop on "Institution, Extension Interventions and Local Strategies in Agricultural Resource Management". January 1995 Chen, Shaohua, Gaurav Datt and Martin Ravallion. Is Poverty Increasing in the Developing World? World Bank Policy Research Working Paper # 1146. June 1993. Chipika, Stephen. Economic Reform and Small Scale Enterprises in Zimbabwe: An Indigenous Enterprises Empowerment Strategy. Intermediate Technology Development Group, Zimbabwe. 1995. Chisvo Munhano and L. Munro. Review of the Social Dimensions ofAdjustment in Zimbabwe. UNICEF, Harare. 1994. City of Harare. Annual Report of the City Health Department 1995. March, 1996. CSO (Central Statistical Office), Zimbabwe. Quarterly Digest of Statistics CSO (Central Statistical Office), Zimbabwe. Population Census 1992 CSO (Central Statistical Office), Zimbabwe. Income Consumption and Expenditure Survey Report, 1990/91. July 1994 Poverty and Human Resources in Zimbabwe 86 Bibliography Daniels, Lisa. Changes in the Small-Scale Enterprise Sector from 1991 to 1993: Results of a Second Nationwide Survey in Zimbabwe. Growth and Equity through Microenterprise Investments and Institutions (Gemini). Technical Report No. 171. March 1994 Davies, Rob and J0rn Ratts0. "Growth, Distribution and Environment: Macroeconomics Issues in Zimbabwe". World Development. Vol 24. No. 2. pp. 395-405. 1996 Fidler, Peter. The Impact of Economic Structural Adjustment on Poverty in Zimbabwe 's Informal Sector. Report to the World Bank. Mimeo. January 1996. Gibbon, Peter (ed.). Structural Adjustment and the Working Poor in Zimbabwe: Studies on Labour, Women Informal Sector Workers and Health. Nordiska Afrikainstitutet, Uppsala. 1995 Gleeson, Brian and Chris Laincz. Poverty in Zimbabwe 's Natural Regions. Course paper for Woodrow Wilson School of Public Policy. 1996. Mimeo. GOZ (Government of Zimbabwe, with the assistance of the World Bank). Zimbabwe. Public Expenditure Review. Mimeo. 1996a. Grootaert, Christian, Ravi Kanbur and Gi-Taik Oh. The Dynamics of Poverty: Why Some People Escape from Poverty and Others Don 't. An African Case Study. The World Bank Policy Research Working Paper 1499. August 1995. Hicks, Deborah. An Evaluation of the Zimbabwe Drought Relief Programme 1992/93: The Roles of Household Level Response and Decentralized Decision Making. World Food Programme of the United Nations. Harare, Zimbabwe. June 1993. Hongoro, Charles and S. Chandiwana. Effects of Users Fees on the Health Care System in Zimbabwe. Blair Research Laboratory, Ministry of Health and Child Welfare, Harare. 1994. Iliff, Peter. Health in Times ofAIDS, Poverty and ESAP. Silveira House Social Series No lOor 1. Draft. n.d. Jayne T.S. et al. "Unraveling Zimbabwe's food insecurity Paradox: Implications for Grain Marketing Reform" in TS Jayne et al (eds.). Integrating Food and Nutrition Policy in Zimbabwe: Proceedings of the First Annual Consultative Workshop on Food and Nutrition Policy. University of Zimbabwe/Michigan State University Food Security Research in Southern Africa Project., Department of Agricultural Economics and Extension. University of Zimbabwe, Harare. 1990. Cited in UNICEF 1994. Poverty and Human Resources in Zimbabwe 87 Bibliography Jenkins, Carolyn. Economic Policy and Investment in Zimbabwe, 1980-1990. Centre for the Study of African Economies, University of Oxford. Mimeo. November 1995 Jenkins, Carolyn. Economic Policy and Rural Welfare in Zimbabwe 1980-90. Centre for the Study of African Economies, University of Oxford. Mimeo. January 1996 Jenkins, Carolyn and Erika Prinsloo. Inequality among Households in Zimbabwe: An Assessment using the 1990/91 ICES. Centre for the Study of African Economies. University of Oxford, St. Cross Building, Manor Road, Oxford OX13UL. August 1995. Kanji, Nazneen. Gender, Poverty and Economic Adjustment in Harare, Zimbabwe. Environment and Urbanization. 7:1. April 1995. Kaseke, Edwin and Maxwell Ndaradzi. A Situation Analysis of the Social Development Fund Ministry of Public Service, Labour Emd Social Welfare and UNICEF. Harare. July, 1993. Kinsey, B.H. Parallel Markets, Private Traders, Government Policy and Food Security in Zimbabwe. Mimeo. n.d. Kinsey, B.H. The Geography of Nutrition in Rural Areas: Three Resettlement Schemes in Zimbabwe. Department of Rural and Urban Planning, University of Zimbabwe. Mimeo. n.d. Kinsey, Bill and John Hoddinott. Development Policies and Rural Well-being: Resettlement, Health and Household Welfare in Zimbabwe. Report prepared under ESCOR grant R6185. January 1996 Kinsey, B.H., Hilary McQuie and Mandivamba Rukuni. Rural Household Studies in Zimbabwe: A Review. Working Paper AEE 2/95. Department of Agricultural Economics and Extension, University of Zimbabwe. March 1995 Knight, John. Labour Market Policies and Outcomes in Zimbabwe. Centre for the Study of African Economies. Institute of Economics and Statistics, University of Oxford. Mimeo. March 1996. Laincz, Chris and Brian Gleeson. Indicators of Poverty in Zimbabwe 's Natural Regions. Mimeo. 1996 Latif, A. A Report on a Study to Determine the Aetiology and Pattern of STD Amongst Men and Women Presenting to Health Centres in Harare, Zimbabwe, and to Determine Risk Factors For Cervicitis Among Symptomatic and Asymptomatic Women. Forthcoming. Poverty and Human Resources in Zimbabwe 88 Bibliography Lennock, Jean. Paying for Health: Poverty and Structural Adjustment in Zimbabwe. Oxfam. 1994 Loewenson, Rene and Munhamo Chisvo. Transforming Social Development: The Experience of Zimbabwe. Prepared for UNICEF, Zimbabwe. July 1994. McIntosh, Alison. Population Policy and Environmental Impacts in Rural Zimbabwe. 1993. Chapter VII in Gayl Ness, William Drake and Steven Brechin, eds. Population- Environment Dynamics: Ideas and Observations. Ann Arbor: University of Michigan Press. Mead, Donald C. and Peter Kunjeku. Business Linkages and Enterprise Development in Zimbabwe. Gemini Technical Report No. 55. April 1993. Mhone, Guy C.Z. The Social Dimensions of Adjustment Programme in Zimbabwe: Critical Review and Assessment. European Journal of Development Research. 7:1 June 1995. Mhishi, S.G. A Critical Analysis of the Resettlement Programme in Zimbabwe. Mimeo. 1995 MoHCW (Ministry of Health and Child Welfare) and UNICEF-Harare. Health Service Costs, Resource Use and Financing in three districts in Zimbabwe. December 1993. Moyo, Sam. The Land Question in Zimbabwe. SAPES Book, Harare. 1995 MPSLSW (Ministry of Public Service, Labour and Social Welfare). 1995 Poverty Assessment Study Survey -- Preliminary Report. Social Development Fund Unit. April 1996 MPSLSW (Ministry of Public Service, Labour and Social Welfare). Report of the Fifth Round: Sentinal Surveillance for SDA Monitoring. Inter-Ministerial Committee on SDA Monitoring, Government of Zimbabwe, Harare. November 1995. MPSLSW (Ministry of Public Service, Labour and Social Welfare). Report of the Fourth Round: Sentinal Surveillance for SDA Monitoring. Conducted by the Government of Zimbabwe Inter-Ministerial Committee on SDA Monitoring with Support from UNICEF. June 1994. Muir, Ann. Livelihood Strategies and the Household Economy in Binga District, Zimbabwe. Save the Children Fund, UK. Harare, Zimbabwe. May 1993. Mundy, Verity S. The Urban Poverty Datum Line in Zimbabwe. Catholic Commission for Justice and Peace in Zimbabwe. 1995 Poverty and Human Resources in Zimbabwe 89 Bibliography Mushauri, J. and Associates (Pvt) Ltd. Study on the Situation Analysis of the Poverty Alleviation Action Plan Implementation. Royal Netherlands Embassy. April 1995. Nyoni, J.M. and Associates. The Determinants ofPoverty: An Overview of the Nature, Causes and Effects of Poverty in Zimbabwe. Zimbabwe Council of Churches. October 1995. Oxfam. The Oxfam Poverty Report. 1995. Rakodi, Carole. Urban Poverty in Zimbabwe: Post-Independence Efforts, Household Strategies and the Short-Term Impact of Structural Adjustment. Journal of International Development: 6, 5: 655-663. 1994 Ravallion, Martin and Shoahua Chen. What Can New Survey Data Tell us About Recent Changes in Living Standards in Developing and Transitional Economies?. Background Paper to World Development Report 1996. Poverty and Human Resources Division, Policy Research Department, the World Bank. February 1996. Republic of South Africa. Key Indicators of Poverty in South Africa. Ministry in the Office of the President: Reconstruction and Development Programme. October 1995. Ridell Commission. The Status of Poverty in Zimbabwe. 1981 Ridell, Roger and Mark Robinson. T7he Impact of NGO Poverty Alleviation Projects: Results of the Case Study Evaluations. Overseas Development Institute Working Paper 68. November 1992. Rohrbach, D. D. et. al. "Agriculture and Food Security." In Integrating Food, Nutrition and Agricultural Policy in Zimbabwe. Proceedings from the First National Consultative Workshop, Juliasdale, University of Zimbabwe, Harare. 1990. Rukuni, Mandivamba. Zimbabwe: Report on the Commission of Enquiry into Appropriate Agricultural Land Tenure Systems. October 1994 Scoones, Ian. "Investigating Difference: Applications of Wealth Ranking and Household Survey Approaches among Farming Households in Southern Zimbabwe. Development and Change. Vol 26 (1995), 67-88. Sibanda, Arnold Elson and Mary Maboreke. Changes in Economic Roles And Social Status of Women in Selected Rural Areas of Zimbabwe. Institute of I)evelopment Studies Monograph #3. University of Zimbabwe, Harare. April 1993. Skalnes, Tor. The Politics of Economic Reform in Zimbabwe: Continuity and Change in Development. Macmillan Press, Ltd. Basingstoke, Hampshire, UK. 1995. Poverty and Human Resources in Zimbabwe 90 Bibliography Stenflo, Gun Alm. Poverty in Zimbabwe 1990-91. Reportfrom a Short Term Mission to the Central Statistical Office, Harare, November 9-December 19, 1993. Stoneman, Colin. South Africa... Lessonsfrom Zimbabwe. CIIR Briefing Paper. n.d. Tichagwa, Wilfred. The Effects of Drought on the Condition of Women in Zimbabwe. Focus on Gender 2:1. Feb. 1994 UNDP (United Nations Development Programme). Human Development Report 1995. UNICEF. Children and Women in Zimbabwe: A Situation Analysis Update 1994. United Nations Children's Fund, Harare Zimbabwe. June 1994 UNICEF. Basic Fact Sheet on Education. Mimeo. 1995 USAID Family Early Warning System Project, Zimbabwe. An Assessment of Vulnerability in Zimbabwe 's Communal Lands. November 1994. Whitlow, R. and B. Campbell. "Factors influencing erosion in Zimbabwe: a statistical analysis." Journal of Environmental Management, Vol. 29. pp. 17-29. 1989. cited in Davies and Ratts0, 1996. Wilson, Ken. The Impact of Drought and Structural Adjustment on Household Livelihood and Welfare in the Mazvihwa Communal Area of Svishavane District of Southern Zimbabwe. Refugee Studies Programme, University of Oxford and Oxfam UK. September 1992. Wilson, K. B. Preliminary Findings from the Field Study of the Impact of Drought, Structural Adjustment and HIV on Livelihood, Fertility, Mortality, Health and Nutrition in the Mazvihwa Communal Area of Zvishavane District, Zimbabwe. Mimeo. 1993 World Bank. Rural Community Infrastructure: Building on the Rural Water Supply and Sanitation Program. Infrastructure Operations Division, Southern Africa Region. June 1993. World Bank. Country Environmental Strategy Paper for the Republic of Zimbabwe. Agriculture and Environment Division, Southern Africa Department. 1994a. World Bank. Better Health in Africa: Experience and Lessons Learned. 1994b World Bank. Zimbabwe: Achieving Shared Growth. Country Economic Memorandum. Macro, Industry and Finance Division, Southern Africa Department. 1995a World Bank. The Public Sector and Poverty Reduction Options. Background paper for the Country Economic Memorandum. The World Bank, Harare. Mimeo. 1995b. Poverty and Human Resources in Zimbabwe 91 Bibliography World Bank. Malawi: Human Resources and Poverty: Profile and .Priorities for Action. Southern Africa Department. 1 995c. World Bank. World Development Report 1995. 1995d World Bank. Memorandum to the President of the International Development Association to the Executive Directors on a Country Assistance Strategy of the World Bank Group for Zimbabwe. Macro, Industry and Finance Division, Southern Africa Department. Mimeo. 1996. ZDHS (Zimbabwe Demographic and Health Survey) 1994. Central Statistical Office, Harare and Macro International Inc. September 1995. ZCTU (Zimbabwe Congress of Trades Unions). Beyond the Economic Structural Adjustment Programme (ESAP): Framework for a Long-Term Development Strategy in Zimbabwe. n.d. Zimbabwe Farmers Union and the Friedrich Ebert Foundation. Future Tenure Systems in Zimbabwe Smallholder Farming Areas - Implications on Investment, Resource Conservation and Productivity. Conference Proceedings. 30 November 1995. Zimbabwe Textile Workers Union, National Union of the Clothing industry Union Shopstewards with the Zimbabwe Congress of Trade Unions Health and Social Welfare Department. Structural Adjustment and its Impact on Workers. April 1993 P iCrNT$ POVERTY WORKSHOPS Harare Novembber 1995 April 1996 July 1996 Poverty Seminars 95 List of Participants LIST OF PARTICIPANTS in alphabetical order Addison, Tony University of Warwick (UK) Alwang, Jeffrey Virginia Tech (USA) Bere, Liz National Council of Negro Women Bucknall, Julia The World Bank Campbell, Bruce University of Zimbabwe, Department of Environmental Studies Cavendish, William University of Oxford (UK) Chatitza, Sensly Ministry of National Affairs and Employment Creation Chigudu, Hope Zimbabwe Women's Resource Centre Network Chikanza, Marck A.R.C. Chikuni, Gilbert Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Chipahsura, Langton Zimbabwe Boy's Brigade Chipika, Jesimen University of Zimbabwe, Department of Economics Chipika, Steve ITDG (Intermediate Technology Development Group) Chiteiceetekel, John Glen Forest Training Centre (GFTC) Cook, David The World Bank Danha, Fadzai Oxfam U.K. & Ireland Davies, Rob Universtiy of Zimbabwe, Department of Economics Dhliwayo, Letwina Royal Netherlands Embassy Disch, Arne The World Bank Dore, Dale Shanduko Trust Francis, M. University of Zimbabwe, Institute of Development Studies Gokova, Jonah K. Ecumenical Support Services Goromonzi, W. Ministry of Public Services, Labour and Social Welfare Hamdok, Abdulh A. Poverty Assessment Study Survey (PASS) Team Hammar, Amanda Social & Institutional Development Consultant Hansen, Keith The World Bank Haque, Trina The World Bank Hartsough, Tala Catholic Commission for Justice and Peace Hatendi, Mercy Women's Action Group Henjesand, I. Norwegian School of Economics & Business Administration Hungwe, Alois Soils Incorporated (Pvt) Ltd Hwekwete, Naomi N. University of Zimbabwe, Institute of Development Studies Jamali, David Zimbabwe Human Rights Association (ZIMRIGHTS) Jassat, Ebrahim University of Zimbabwe, Institute of Development Studies Poverty Seminars 96 List of Participants Jazdowska, Niki Training & Research Support Centre Jubenkanda, S. Office of the President and Cabinet, Monitoring & Implementation Juliusdottir, Magnfridur University of Zimbabwe, Institute of Development Studies Kalawe, A.Y. World Vision International Kaliyati, Jacob University of Zimbabwe, Institute of Development Studies Kamidza, Richard Southern African Association of Political Science Trust (SAPES Trust) Kasch, Volker E.Z.E., Germany Kinsey, Bill University of Zimbabwe, Department of Rural and Urban Planning Kleppe, Ingeborg Norwegian School of Economics & Business Administration Kooijmans, Joost International Labour Organisation/SAMAT Kostermans, Kees The World Bank Kuchera, Rev. Murombedzi Zimbabwe Council of Churches Leher, Yvonne Girl's Brigade Zimbabwe Lynam, Tim J.P. World Wide Fund for Nature Malaba, J. Poverty Assessment Study Survey (PASS) Team Manenji, D. Oxfam UK & Ireland Manenji, Fridah African Development Education Network (ADEN) Maridza, Jadua GREDA, Masvingo (GUTU) Matiringe, Thecla Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Matshalaga, N. University of Zimbabwe, Institute of Development Studies Maunganidze, Langtone Catholic Commission for Justice and Peace, Civic Education Group Mavunga, Jeanlinda Zimbabwe Environment Research Organisation (ZERO) Mawanza, Simeon Zimbabwe Human Rights Association (ZIMRIGHTS) McKay, Lloyd The World Bank Meszhawidza, Phides Women In Business Mhishi, S.G. Ministry of Public Services, Labour and Social Welfare Mhlanga, Ethan OXFAM UK & Ireland Milort, Anna UNDP Mispelaar, Mike CARE International Mlambo, Emmanuel Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Mombeshora, Solomon University of Zimbabwe, Department of Sociology Moyo, Janice Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Msiyandima, Adiel Zimbabwe Women's Finance Trust (ZWFT) Muchawaya, L. Morgan Chitungwiza Integrated Youth Survival Alternative Project I (CHIYSAP) Poverty Seminars 97 List of Participants Muchechesi, Edson Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Muchena, Deprose Zimbabwe Council of Churches Muchono, Samuel World Vision International Mudehwe, Jonah NATIONAL ASSOCIATION OF NON- GOVERNMENTAL ASSOCIATIONS (NANGO) Mudimu, Godfrey University of Zimbabwe, Department of Social Economics Mugabe, Makusha Zimbabwe Human Rights Association (ZIMRIGHTS) Mugededza, C. University of Zimbabwe, Institute of Development Studies Muir-Leresche, Kaye University of Zimbabwe, Department of Agricultural Economics Mungwashu, Rube Zimbabwe United Nations Association Munro, Lauchlan UNICEF Murapata, Patience Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Mutongerwa, Jane Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Muza, Joseph Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Ndebele, Talita Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Nijholt, A. UNDP Peacock, Eng T. Halcrow Zimbabwe (Pvt) Ltd Pichon, Francisco The World Bank Plastow, C. ODA/British High Commission Pokawa, Joseph Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Posi, Angeline Chitungwiza Integrated Youth Survival Alternative Project (CHIYSAP) Pushpanath, Krishnamurthy OXFAM (U.K.) Raftopoulos, Brian University of Zimbabwe, Institute of Development Studies Ribe, Helena The World Bank Ruwona, David P. ALOZ (Adult Literacy Organisation of Zimbabwe) Samuriwo, Margaret Southern African Research & Documentation Centre (SARDC) Saunders, Chris Save the Children Fund (U.K.) Shambare, R.M. Nekor Business Consultants Sibanda, A. University of Zimbabwe, Institute of Development Studies Siddharth, Veena OXFAM International Siliamaa, Leo Lutheran World Federation Sithole, Bevlyne University of Zimbabwe, Department of Environmental Studies Tarwirei, James Chitungwiza Integrated Youth Survival Alternative Project Poverty Seminars 98 List of Participants (CHIYSAP) Verner, Dorte European University Institute Watkins, Kevin OXFAM, UK & Ireland Wirth, Peter DANCHURCHAID, Denmark Zeitlyn, Sushila ODA Zhou, Emerson Chief Economist, Zimbabwe Farmers' Union Zimudzi, Herbert Helpage Zimbabwe Zimunya, Kingstone CARE International Zitsanza, Nancy Ministry of Agriculture Zulu, Ben Media for Development Trust The World Bank 818 H Street, N.W 'Washingt'on D.C. 20433,, , U.S.A. 'A ;0; , ; -; A''StU';'' A . .. . .. . . . , . ,;, -, i .,- , ,;~~~~~~~~~~~A A~ ~~~, ,.A .- - , .............. . . .A A