A Census of Orphans and Vulnerable Children in Two Zimbabwean Districts pptx

Foreword viiAuthors ixC ontributors xAcknowledgements xiList of Tables and Figures xiiiAcronyms and abbreviations xvExecutive Summary xvi Chapter 1: Introduction 1 1.0 Background 11.1 De

children in two Zimbabwean districts

Shungu Munyati, Simbarashe Rusakaniko, Pakuromhunu F Mupambireyi, Stanford T Mahati, Peter Chibatamoto, & Brian Chandiwana

Edited by Shungu Munyati

NATIONAL INSTITUTE

OF HEALTH RESEARCH,

BIOMEDICAL RESEARCH

Welfare, Harare, Zimbabwe.

Published by HSRC PressPrivate Bag X9182, Cape Town, 8000, South Africawww.hsrcpress.ac.za

© 2006 HSRC, BRTI, NIHR & FACTFirst published 2006

All rights reserved No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, including photocopying and recording, or in any information storage or retrieval system, without permission

in writing from the publishers

ISBN 0 7969 2146 6Print management by comPressDistributed in Africa by Blue Weaver

PO Box 30370, Tokai, Cape Town, 7966, South AfricaTel: +27 (0) 21 701 4477

Fax: +27 (0) 21 701 7302email: orders@blueweaver.co.zawww.oneworldbooks.comDistributed in Europe and the United Kingdom by Eurospan Distribution Services (EDS)

3 Henrietta Street, Covent Garden, London, WC2E 8LU, United KingdomTel: +44 (0) 20 7240 0856

Fax: +44 (0) 20 7379 0609email: orders@edspubs.co.ukwww.eurospanonline.comDistributed in North America by Independent Publishers Group (IPG)Order Department, 814 North Franklin Street, Chicago, IL 60610, USACall toll-free: (800) 888 4741

All other enquiries: +1 (312) 337 0747Fax: +1 (312) 337 5985

email: frontdesk@ipgbook.comwww.ipgbook.com

iv

Foreword viiAuthors ix

C ontributors xAcknowledgements xiList of Tables and Figures xiiiAcronyms and abbreviations xvExecutive Summary xvi

Chapter 1: Introduction 1

1.0 Background 11.1 Definition and prevalence of orphanhood and vulnerability 11.2 Rationale and aims of the study 4

1.3 Conceptual framework 5

Chapter 2: Methodology 7

2.1 Operational definitions 72.2 Description of the study sites 72.3 Geographical frame for the OVC Census 92.4 Instrument 9

2.4.1 Questionnaire design and its translation 92.4.2 Pre-testing of Census questionnaire 92.5 Ethical issues 10

2.6 Data collection 10

2.6.1 Pre-enumeration activities 102.7 Deployment of enumerators and supervisors 122.8 Quality control 12

2.9 Assessment of vulnerability 132.10 Data management and analysis 13

Chapter 3: Results 15

3.1 Bulilimamangwe District 15

3.1.1 Demographic data 153.1.2 Magnitude of orphanhood 18

3.1.4 Household level data 203.1.5 Main household vulnerability indicators 233.1.6 Other income indicators 26

3.2 Chimanimani District 27

3.2.1 Demographic data 273.2.2 Household level data 313.2.3 Main household vulnerability indicators 353.2.4 Other income indicators 37

Chapter 4: Discussion 39

Chapter 5: Conclusion and recommendations 43

5.1 Challenges faced during the study 44

Appendix 2 Reactions to the Pilot Research Procedures 52Appendix 3 Fieldworker’s Introductory Letter 53

Appendix 4A Census Shona and English Questionnaire 54 Appendix 4B Code Sheet – Shona 56

Appendix 5A Census Ndebele and English Questionnaire 58 Appendix 5B Code Sheet – Ndebele 60

Appendix 6 Quality Control of Questionnaire Checklist 62Appendix 7 List of Supervisors 63

Appendix 8 Ward Analysis of Census Results by District 64Appendix 9 OVC 2003 Census Operational Structure 125Appendix 10 Vulnerability Score Assessment 126

In this era of the HIV and AIDS epidemic in sub-Saharan Africa and economic challenges

in Zimbabwe, there is possibly no other subject that has received as much media attention

of late, than that of the plight of orphans and vulnerable children, hence it is difficult to conceive of a more pertinent and perfectly timed publication than this one

Of the 3 million AIDS-related deaths globally, 2.2 million are from the sub-Saharan region

One major impact is that the disease has orphaned vast numbers of children, because those dying from AIDS are mainly in the prime of their lives and are parents These children endure overwhelming losses; living in societies already weakened by under development, poverty, the AIDS pandemic itself and whose traditional support structures, like the extended family system, have been eroded by, among other factors, urbanisation

Faced with this unprecedented crisis, the Zimbabwe Government in 1999 introduced the compulsory AIDS levy from taxable incomes to raise money which is channelled to alleviate the suffering of people living with AIDS and their dependents, who are mostly orphans It further complemented this effort by developing a National Orphan Care Policy to underpin the mobilisation of resources and ensure that orphans get, at least, minimal basic services The Orphan Care Policy combines institutional, fostering, and community-based care A plethora of interventions has been initiated by the government, Non-Governmental Organisations (NGOs), Faith-Based Organisations (FBOs) and

Community-Based Organisations (CBOs), aimed at assisting Orphans and Vulnerable Children (OVC) and their caregivers However, the work of intervention agencies is often undermined or made difficult by their use of data which is scant, outdated and vague, covering small geographical areas like wards, and is generally less objective in terms of how it identifies needy children and areas Furthermore, the process of how they identify the needs of OVC in different households and localities is not well defined Consequently, this at times leads to the implementation of badly focused interventions or selection of areas with fewer households which are vulnerable

The 2003 OVC Population Census in Bulilimamangwe and Chimanimani districts was the first of its kind to be undertaken in Zimbabwe In the past, national censuses captured data on orphanhood status without obtaining data on vulnerable children This census was done a year after the 2002 National Census Thus the framework of operation and mapping of the districts was done within the framework of the 2002 National Census

However, while the 2002 National Census was taken on a de facto basis, the OVC Census used the de jure method The additional information in this census was the household and individual vulnerability indicators, which were identified to assist in future interventions tailored for these districts Vulnerability was assessed using the Vulnerability Indicator Score, which touched on broad aspects of children’s lives such as access to food, health care and protection

On behalf of the Ministry of Health and Child Welfare and the Government of Zimbabwe,

I would like to express my gratitude to the funders of the Project, The W.K Kellogg Foundation for championing and supporting the plight of OVC, and the Human Sciences Research Council (HSRC), in particular its new President and Chief Executive Officer, Consultant and Project Champion of the OVC Project, Dr Olive Shisana, who worked tirelessly to make the projects happen in the three selected countries In addition, I acknowledge all persons who participated in the census exercise, especially the people

of Bulilimamangwe (Bulilima, Mangwe and Plumtree districts) and Chimanimani, for their

under the leadership of the Project Director, Mrs Shungu Munyati, Project Manager Mr Brian Chandiwana and the Chief Consultant, Professor Simbarashe Rusakaniko.

I am hopeful that this database can be used by my Ministry and others to mobilise additional resources that will be channelled towards improving the lives of orphaned and vulnerable children, their caregivers, and communities at large I also fervently hope that this database, which shows the magnitude of the Orphans and Vulnerable Children’s problem across the districts at household and ward level, will lead to the formulation and implementation of both relevant national policies and evidence-based interventions by CBOs, FBOs, NGOs and government structures aimed at tackling the pertinent challenges confronted by the OVC

Dr David Parirenyatwa (MP)Minister of Health and Child Welfare, Zimbabwe May 2005

The Editor: Shungu Munyati – MSc, BSc (Hons) Applied Biology, PhD Candidate; OVC Research Project Director and Acting Director, National Institute of Health Research (NIHR), Ministry of Health & Child Welfare, Harare (Zimbabwe).

Simbarashe Rusakaniko – PhD; OVC Research Project Chief Consultant Biostatistician, University of Zimbabwe, College of Health Sciences, Harare (Zimbabwe)

Pakuromunhu Freddie Mupambireyi – MSc Demography, BSc (Hons) Econs Statistician, University of Zimbabwe, Deputy Dean, Faculty of Commerce, Harare (Zimbabwe)

Stanford Taonatose Mahati – MPhil, BSc (Hons) Sociology & Anthropology, Social Scientist, National Institute of Health Research (NIHR), Ministry of Health & Child Welfare, Harare (Zimbabwe)

Peter P Chibatamoto – MBA, MSc Infectious Diseases, Biological Sciences; HIV/AIDS (Mainstreaming) Technical Advisor UNDP, Windhoek (Namibia)

Brian Chandiwana – BSc Econs & MBA, OVC Research Project Manager; Health Economist, Biomedical Research & Training Institute, Harare (Zimbabwe)

George Chitiyo – MSc & BSc Econs; PhD (Cand), USA.

Wilson Mashange – Dip Med Lab Tech; National Institute of Health Research (NIHR), Ministry of Health & Child Welfare, Harare (Zimbabwe)

Junior Mutsvangwa – BSc Medical Laboratory Technology and MPhil (Cand), Biomedical Research & Training Institute, Harare (Zimbabwe)

Natsayi Chimbindi – BSc HEP, (Health Education), Biomedical Research & Training Institute, Harare (Zimbabwe)

• We would like to thank the Biomedical Research and Training Institute, in particular the Centre for International Health and Policy, the National Institute for Health Research (NIHR – formerly the Blair Research Institute) of the Ministry of Health, and Child Welfare for their participation and unwavering support of the OVC project.

• We also thank Professor Exnevia Gomo, Dr M B Serbit from the College of Health Sciences, University of Zimbabwe, Mr White Soko from the National Institute of Health Research, Mr Rogers Sango, Mr Manase Viriri, Mr Patrick Bvitira, Mr Henry Semwayo from the Central Statistics Office (CSO) and Mr Tichaona Chirimanyemba (formerly with the CSO) for their valuable technical input and participation during tool development, mapping, training of fieldworkers and fieldwork

• The Grant Maker FACT Mutare, in particular the Executive Director, Mr Jephias Mundondo, is acknowledged for support during the entire OVC census

• The District Administrator for Bulilimamangwe Mr Mzingaye Sithole, the Rural District Council Chairmen of Bulilima (Mr Christopher Ndlovu), Plumtree (Mr Patrick Mabuza) and Mangwe (Mr Grey Ncube), Traditional Chiefs and Councillors are saluted for their co-operation as well as their facilitating role in ensuring successful mobilisation of the community, which led to the successful implementation of the field data collection exercise In Chimanimani, we are also grateful to the then District Administrator of Chimanimani Mr Edgar Nyagwaya, Chimanimani Rural District Council Chairman of Chimanimani Mr Joseph Harahwa, Traditional Chiefs and councillors who facilitated the census process

• The active participation of the Bulilima, Mangwe and Plumtree OVC Local Liaison team: Mr Irvine Ncube, Mrs Melta Moyo, Mr Frank Ngwenya, Mr Lincoln Ncube,

Ms Sifiso Dube and Mr Alois Sibanda during the entire exercise is very much appreciated Thanks are also extended to Mr Andrew Nleya and Mr Khumbulani Tshuma whilst the role played by headmasters, teachers and nurses in providing the entire team with free training venues, accommodation and logistical support

is gratefully acknowledged We also acknowledge the support received from the Chimanimani OVC Local Liaison team members: Mr Jobes Jaibesi, Mr Brian Muchinapo, District Nursing Officer the late Sister Mistress Ndhlovu and Sister Sifovo during the whole exercise

• The field supervisors and youthful enumerators who diligently and enthusiastically performed their tasks in Bulilima, Mangwe and Plumtree districts are acknowledged

• Gratitude is extended to the Catholic Development Commission (CADEC), Integrated Rural Development Programme (IRDP), Bulilima District Council and Mangwe District Council who provided us with vehicles during the fieldwork World Vision assisted in the recruitment of research assistants

• Data entry was performed by a large team of Data Entry Clerks under the supervision of Mr Tendai Madiro and Mr Lowence Gomo Their role is acknowledged, together with Mr Stephen S Buzuzi, Mr Timothy Mutsvari, Mr Teramai Moyana, Miss Stella Gwini, Miss Chenjerai K Mutambanengwe and Mr Maxwell Chirehwa who assisted incorporating comments from reviewers

• Appreciation is extended to Mr Norest Mapisaunga, Mr Kuziva Chatindo, Ms Tinashe Maoneke and Mr Clemence Gatsi for provision of logistical support

• We also extend our gratitude to Dr Sheila Tlou and Dr Nunu Tsheko of the University of Botswana for sharing with us their experiences of the OVC project

leadership and direction during the entire exercise

• Professor SK Chandiwana, now Head of the School of Postgraduate Studies, Faculty

of Health Sciences, University of Witwatersrand is acknowledged for the role he played in successfully securing the grant for the Zimbabwean component of the project from Kellogg through HSRC, and for the advisory role that he continues to play to the Zimbabwean team

• Last but not least, this work would not have been possible without the support and the co-operation of the people of Bulilima, Mangwe and Plumtree, in particular the orphans and vulnerable children themselves We salute them!

Table 1 Distribution of wards and enumeration areas (EAs) by district 11Table 2 Demographics of population by district 15

Table 3 Comparison between National Census 2002 and OVC Census 2003 by

District 16Table 4 Age distribution in Bulilimamangwe 16Table 5 Distribution of population aged 18 and below by district 17Table 6 Population distribution by level of education attained in districts 17Table 7 Reasons for children who had never gone to school in the districts 18Table 8 Extent of orphanhood by district 18

Table 9 Population distribution of type of orphan by district 18Table 10 Population distribution of disabled people by type of disability by

district 19Table 11 Prevalence of disability in households with children by district 19Table 12 Household distribution by main water source in the districts 20Table 13 Distribution of main water sources in households with children by districts

20Table 14 Household distribution by distance from main water source in the

districts 21Table 15 Household distribution by main source of energy for cooking in the

districts 21Table 16 Household distribution by main source of energy for lighting in the

districts 22Table 17 Household distribution by toilet facility and district 22Table 18 Distribution of toilet facilities in households with children by district 23Table 19 Summary of vulnerability indicators by household and district 23Table 20 Summary of vulnerability indicators in households with children by district

25Table 21 Household vulnerability status using the Total Vulnerability Score 25Table 22 Household vulnerability status of households with children using the Total

Vulnerability Score 26Table 23 Household distribution by other income indicators and district 26Table 24 Distribution of other income indicators in households with children by

district 27Table 25 Demographics of population in Chimanimani District 27Table 26 Comparison between National Census 2002 and OVC Census 2003 in

Chimanimani District 28Table 27 Age distribution in Chimanimani District 28Table 28 Distribution of population aged 18 and below and above 18 in

Chimanimani District 29Table 29 Population distribution by ever attended school in Chimanimani District 29

Table 31 Reasons for children who had never gone to school in the districts 30Table 32 Population distribution by orphanhood status in Chimanimani District 30Table 33 Population distribution of disabled people by type of disability 31Table 34 Prevalence of disability in households with children in Chimanimani 31Table 35 Household distribution by main water source 32

Table 36 Distribution of main water sources in households with children in

Chimanimani 32Table 37 Household distribution by distance from main water source 33Table 38 Household distribution by main source of energy for cooking 33Table 39 Household distribution by main source of energy for lighting 34Table 40 Household distribution by toilet facility in Chimanimani District 34Table 41 Distribution of toilet facilities in households with children in

Chimanimani 35Table 42 Summary of vulnerability indicators by household 35Table 43 Summary of vulnerability indicators in households with children in

Chimanimani 36Table 44 Household distribution of vulnerability score in Chimanimani District 37Table 45 Household vulnerability status of households with children using the Total

Vulnerability Score 37Table 46 Household distribution by other income indicators in Chimanimani District

AIDS Acquired Immune Deficiency SyndromeBRTI Biomedical Research and Training InstituteBSS Behavioural Sero Status Survey

CADEC Catholic Development CommissionCBO Community-Based Organisation

CIHP Centre for International Health and Policy CMED Central Mechanical Equipment DepartmentCSO Central Statistics Office

DNO District Nursing Officer

FACT Family AIDS Caring TrustFBO Faith-Based Organisation

HSRC Human Sciences Research CouncilIRD Integrated Rural Development MRCZ Medical Research Council of ZimbabweNGO Non-Governmental OrganisationOVC Orphans and Vulnerable Children

Within the southern African region, the problem of orphans and vulnerable children (OVC) is on the increase, especially in the face of the HIV and AIDS epidemic Zimbabwe

is one of the countries which has been hard hit by the OVC problem and yet there are currently no reliable national statistics on OVC, let alone for the various districts of the country Reliable information on OVC is required by researchers in order to understand better the nature of the problem and thereby provide community-based organisations (CBOs), non-governmental organisations and government departments working with OVC with data for effective and efficient programme planning, especially in a resource-poor environment such as Zimbabwe

The Human Sciences Research Council (HSRC) together with its partners within the Southern African Development Community (SADC) region have been commissioned by The WK Kellogg Foundation (WKKF) to develop and implement a five-year intervention project on orphans and vulnerable children (OVC) as well as families and households coping with an increased burden of care for affected children in Botswana, South Africa and Zimbabwe

As part of this broader research and intervention project on OVC across three countries

in the SADC region, an OVC census was conducted in selected sites across Zimbabwe in order to quantify the extent of the OVC problem regarding their numbers, and, further, to use the information collected to determine a sampling frame for the other component of the broader project, which was the OVC Psycho-social Survey (PSS)

Bulilimamangwe area (comprising Bulilima, Mangwe and Plumtree districts) and Chimanimani district in Matabeleland South province and Manicaland province of Zimbabwe respectively, were the selected sites where the census was conducted in November 2003 Using the de jure approach, all households in the sites were enumerated

by a team of 527 enumerators under the supervision of 67 supervisors A two-paged census questionnaire obtained data from the head of the household or someone in a knowledgeable position in that household about all occupants and their demographic profile, living conditions, as well as their vulnerability status, using a Total Vulnerability Indicator Score (TVIS) index Nine main indicators of vulnerability were used, which were household-based rather than individual-based; that is, looking at the number of meals per day, some days they go without food, children of school-going age not going to school, ability to pay for medical fees, anyone ill in the household in the last month, adequacy

of clothing and school uniforms, children without caretakers and, lastly, those without anyone with whom to discuss their problems

The total population of Bulilimamangwe according to this census was 153 320, specifically 81 984 for Bulilima, 62 324 for Mangwe and 9 012 for Plumtree, with an average household size of 5.1, 5.4 and 4.0 respectively The total population was 107

120 in Chimanimani District, with 24 495 households and an average household size of 4.4 Over 50% of the population in Chimanimani and Bulilimamangwe were female and had either gone to primary or secondary school, and a similar proportion were aged 18 years and above (56.3% for Bulilima, 69% in Mangwe, 47.9% in Plumtree, and 51.8% in Chimanimani) Prevalence of orphanhood (only in those 18 years and below) was 30.5%

in Chimanimani, 27.6% in Bulilima, 24.5% in Plumtree and 23.7% in Mangwe The most prevalent type of orphan across the four districts were those whose father had died (paternal orphans): Chimanimani 19.3%, Bulilima 17.7%, Mangwe 15.6% and Plumtree 14.6%, and then those who had lost both parents: Chimanimani 6.9%, Bulilima 5.7%,

Regarding vulnerability assessment, between 49% and 80% of the households across Bulilimamangwe reported having only one meal a day with almost half of these indicating that there were actually some days that the household went without any food Plumtree,

in the urban area, had the highest proportion for both scenarios The majority of the households in Chimanimani district reported having at least two meals a day (89.8%), although as many as 65.4% still indicated that there were some days that they went without any food Inadequate clothing for children was another major vulnerability indicator reported by a large number of the households (Chimanimani 71.6%, 71% in both Bulilima and Mangwe and 46% in Plumtree) More than a third of the households across Chimanimani and Bulilimamangwe districts indicated that they were not in a position to pay for medical fees if their children fell ill Of note was also the proportion

of children who were heading households in Bulilimamangwe, that is, the child-headed households (average of 5% across the site) who then reported not having any caretakers (55.5%) and not having anyone with whom to discuss their problems (29.7%) as children

In Chimanimani, child-headed households were 3.1% but with almost half of them (46.9%) reporting not having any caretakers, as well as 37% indicating that they had no-one with whom to discuss their problems Further analysis using the TVIS showed that Mangwe (3.0%) and Bulilima (2.1%) had higher proportions of highly vulnerable households as compared to 0.5% for both Plumtree and Chimanimani Income indicators were also assessed as a measure of vulnerability and the census showed that most of the households in all the districts owned farm animals and farm equipment

A similar assessment as above but specifically looking at the households with children

18 years and below, which was inclusive of the households with OVC, indicated that around 50% of these households in Bulilima and Mangwe reported that they were having only one meal a day, whilst it was 20% in Plumtree, which is urban Over 70% of the households with children in Mangwe and Bulilima, 47.9% of those in Plumtree and 66.3 % in Chimanimani went for some days without food Again, inadequate clothing for children was another major vulnerability indicator reported by a large number of these households (more than three quarters in both Bulilima and Mangwe, 80% and 79%

respectively), nearly two thirds in Plumtree (60.8%) and 87.5% in Chimanimani Except for Plumtree, over half reported that their children had inadequate school uniforms (55.3%

Bulilima, 55.8% Mangwe, and 61.6% in Chimanimani) Mangwe had just over half (54.9%), whilst Bulilima (47.2%) and Plumtree (43.2%) had nearly half of the households reporting that they were not in a position to pay for medical fees if their children were sick In Chimanimani this was 42.4% Of note is that all these figures depicting vulnerability were slightly higher in the households with ‘children’ as compared to the households with ‘adults’

On other personal and household conditions, it was found that around one in every five households with children in Bulilima, Mangwe and Chimanimani districts had at least one person who had a disability, whilst in Plumtree it was about ten per cent The most common forms of disability were ‘difficulty in moving’ and ‘seeing’ In Bulilima, Mangwe and Chimanimani the main water source was a protected well or borehole, except Plumtree with indoor piped water In all the districts, with the exception of Plumtree, wood was commonly used for cooking while paraffin was commonly used for lighting,

It is therefore concluded from this study that census data are a useful source of scientific data This census has managed to extrapolate the population profile of Bulilimamangwe and Chimanimani districts The census data has also highlighted the extent of the problem

of OVC in the districts and the conditions in which these children live The study therefore recommends that interventions tailor-made for OVC should use the statistics generated to help in the sampling frame of their target groups It is interesting to note that the national demographic profile parameters as defined by the OVC census of 2003 are generally in agreement with the National Census of 2002 The burden of other problems like disability, sanitary facilities and sources of energy has been well documented The results have also identified gaps at population level that need to be addressed, like the need for toilet facilities and clean water Around a quarter of the children were orphans across the two sites and there were more paternal orphans than maternal orphans Shortage of food and lack of adequate clothing seemed to be the major problems facing the communities in Bulilimamangwe and Chimanimani

Regarding the key findings in assessing vulnerability, it is apparent that in all sites the households had high levels of poverty as evidenced by the lack of adequate meals It is therefore essential that interventions be directed at issues to do with food security, for example, equipping the community and the caretakers with skills and resources to come

up with income-generating activities, targeting mainly those households with children, especially OVC

Intervention agencies should, when donating household items, prioritise and make provision for clothing and school uniforms for the children, who in this assessment included OVC Being inadequately clothed or not having school uniforms singles you out as a child in Bulilimamangwe and Chimanimani, and has a bearing on issues of stigmatisation within the communities Furthermore, Social Welfare and other intervention agencies should assist in lobbying for subsidies for medical fees/medicines for those households with vulnerable children, including orphans

Background

Parents are the most central and enduring influence in children’s lives The loss of parents has an indelible impact on the living conditions of children and can have a strong bearing on their future quality of life The circumstances under which orphans and vulnerable children are growing up might not auger well for the development of any country Children are bearing the brunt of many of the socio-economic problems bedevilling most developing countries; especially in sub-Saharan Africa, as the supportive structures at household, community and national level have, over the years, been eroded

or overwhelmed by the magnitude of the burden, set against a background of scant resources For intervention agencies it is critical to determine the numbers of orphans and vulnerable children if adequate and appropriate strategies are to be adopted and implemented successfully

As one way of fulfilling its mandate, the WK Kellogg Foundation (WKKF), in 2001, funded the Human Sciences Research Council’s (HSRC) Social Aspects of HIV/AIDS and Health Programme (SAHA) to prepare a policy document reviewing social and economic problems linked directly or indirectly to the HIV/AIDS problem in southern Africa

This required that an analysis of problems related to orphans and vulnerable children (OVC) be prepared, together with recommendations on potential interventions in rural development programming On completion and submission of the report to WKKF, HSRC was then asked to produce a draft strategy for the care of orphans and vulnerable children in Botswana, South Africa and Zimbabwe, a task that was undertaken jointly by SAHA and the Child, Youth and Family Development (CYFD) programme of the HSRC

This was accepted by WKKF and led to the signing of a Memorandum of Understanding (MOU) between them and the HSRC, which required that the latter prepare an

Operational Framework for Research-Driven Interventions for Orphans and Vulnerable Children, including performance targets and indicators The framework was submitted to WKKF and also approved The MOU in addition required that the HSRC develop systems

to implement and monitor the HIV/AIDS OVC Operational Framework and provide research to support innovative and sustainable models that target orphans and vulnerable children, as well as families and households coping with an increased burden of care for affected children The OVC Census was one of such research studies

1.1 Definition and prevalence of orphanhood and vulnerability

The definition of an ‘orphan’ varies In 2002, a study was carried out to find out the local definition of an orphan in the Bulilimamangwe and Chimanimani districts of Zimbabwe (Skinner et al., 2004) Similar studies were concurrently conducted in Botswana and South Africa1 (ibid) In Zimbabwe, members from service providers, community leaders, OVC and caretakers of OVC participated in the study The communities in both Bulilimamangwe and Chimanimani districts overwhelmingly defined an orphan (nherera/

intandane) as a child who has lost one or both parents They broadly categorise orphans into two groups, that is, those without guardians and those with guardians

The communities further indicated that there is no age limitation to orphanhood and that they would remain under the banner of being called an orphan until they were able to look after themselves or when they got married They, however, considered childhood as lasting up until the age of 18 years

Skinner et al (2004) noted that the threat posed by the HIV/AIDS pandemic has resulted

in the mushrooming of numerous social support institutions in our communities, all aimed at giving care to orphaned and vulnerable children left as victims of the pandemic

It is further pointed out that the proliferation of the related family disintegration creates a situation in which those affected automatically become victims of stigma, which becomes apparent in the way they are treated, and also in the terminology used to describe them It therefore becomes necessary to take due care in defining what an orphan is In addition, social implications brought about by this pandemic, including the surrounding poverty and increased numbers of vulnerable children creates a situation where in dealing with the problem, organisations find themselves having difficulty defining OVC and drawing parameters for their work, especially in resource-poor environments where there

is a lack of adequate caring mechanisms and service structures to support them

UNAIDS defines an orphan as a child under 15 years of age who has lost their mother (‘maternal orphan’) or both parents (‘double orphan’) to AIDS (UNICEF/UNAIDS, 1999) Many researchers and intervention groups increase the age to 18 years, but most appear

to use the UNAIDS definition It is also being more generally accepted that the loss of the father would also classify the child as an orphan (RAISA, 2002)

Again Skinner et al (2004), note that community definitions of an orphan and vulnerable child are also often different from the government definitions For instance, assistance

to children by the government is directed by particular age limits – any child who falls outside those limits may be excluded There was general consensus during the research interviews conducted by Smart (2003) on definitions of vulnerability in Botswana, Rwanda, Zambia and South Africa that the government should adopt a ‘bottom up’ approach, so taking guidance from community level, when setting parameters for assistance This provokes a debate about addressing the specificity of needs versus what

is bureaucratically feasible To get a real sense of where to introduce interventions or support, a clear understanding of the community’s perspective is required Time has to

be spent in the community listening to people who are doing work there already, and particularly to the caretakers and the vulnerable children themselves

‘Vulnerability’ is much more difficult to define than defining an orphan World Vision (2002) listed some identifiers, such as children who live in a household in which one person or more is ill, dying or deceased; children who live in households who receive orphans; children whose caregivers are too ill to continue looking after them; and children living with very old and frail caregivers These categories focus on factors related to HIV (Skinner et al., 2004) There is an entire set of variables that needs to be considered that relates to more general aspects of the child’s context, such as poverty; access to shelter, education and other basic services; disability; impact of drought or extreme weather conditions; stigma and political repression – all factors that could influence vulnerability A consultative meeting in Kenya defined as vulnerable, children

in households with a chronically ill parent or caregiver, but later focused on vulnerability

in terms of access to key resources such as food, shelter, education, psychosocial and emotional support and love (NACC Taskforce on OVC, 2002)

Vulnerability for this OVC census is defined using nine vulnerability indicators, which included access to food, protection, health care and clothing among other indicators (see section 2.9) However, these indicators differ slightly from those used by World Vision in assessing vulnerability World Vision indicators of vulnerability focused on HIV-related factors like children who live in a household where one person or more is ill, dying or deceased, children whose caretakers are too ill to continue to look after them (Skinner

et al., 2004) On the other hand, this census focused on general factors such as poverty and education

Zimbabwe, along with Zambia, Malawi, Lesotho and Rwanda, has the world’s highest proportion of orphaned children December 2001 estimates by UNAIDS (2002) suggest that in these five countries, 17–18% of all children under the age of 15 have lost one

or both parents In Zimbabwe, this amounts to just over 1 million orphans, out of a child population of around 5.8 million in a total population of 12 million It is estimated that by the year 2010 one out of every three or four children in Botswana, Malawi, Tanzania, Uganda, Zambia and Zimbabwe will be an orphan A generation of orphans are being cared for by grandparents, family members or through self-care in child-headed households (UNAIDS, 2002) According to UNICEF (2003), there were 240 000 Zimbabwean children (0–14 years) living with HIV and AIDS by the end of 2002 Since the beginning of the epidemic 900 000 children have lost one or both parents to the disease (SAfAIDS/Panos, 2001) It is therefore projected that in Zimbabwe, Zambia and South Africa, 20–30% of all children younger than 15 years may be orphans by 2015 (FHI

2003 in ZHDR, 2003)

The HIV/AIDS epidemic is resulting in increasing numbers of OVC and child-headed households, and is creating a strain upon extended families – particularly grandparents – and communities, who are desperate for resources and support to deal with ever-increasing demands However, Skinner et al., (2004) noted the existence of many community-based initiatives supporting OVC, through identifying and monitoring vulnerable children, visits, creating community foster homes, providing material and psychosocial support, and disseminating information on HIV/AIDS prevention

The National Plan of Action for OVC for Zimbabwe seeks to prioritise and address the urgent issues facing OVC, their families and communities (Government of Zimbabwe, 2003) Although communities and community-based organisations have responded to the plight of OVC with a proliferation of local initiatives, lack of co-ordination means that their impact is fragmented and existing resources cannot be utilised in a manner that fully benefits children The harsh socio-economic situation in Zimbabwe has resulted in alarming increases in children with the basic survival needs for food and health services

In Zimbabwe generally, national policies and laws establishing the legal infrastructure for the co-ordination of OVC programmes and services have not been implemented, for lack of financial, material and human resources Anecodotal evidence shows that due to the lack of full policy implementation protecting children’s rights, OVC service providers report an alarming increase in cases of child abuse Furthermore, the widespread lack of birth certificates prevents children from accessing basic services and rights and children, particularly OVC, lack the ability to participate in decisions that affect their lives

Nowhere is the lack of children’s participation more evident than in the legal and policy issues that affect them in Zimbabwe The extent of the orphan problem and the new

Legal issues pertinent to children in Zimbabwe are subject to interpretation according

to a dual legal system, comprising customary law and codified, general law found in the Constitution and statutes The existence of these two systems can result in situations where one system can pre-empt or contradict the other Moreover, existing legislation aimed at protecting children against abuse is often fragmented, generally requiring access

to legal advocacy to ensure that the law is carried out in favour of the child

Generally, the GOZ, NGOs and the international community (donors) are now recognising that approaching community initiatives collectively results in significant gains for OVC Escalating numbers of OVC mean that line ministries, local organisations and communities need to expand coverage with service provision and care, and that local efforts are co-ordinated to maximize resources and avoid duplication of effort

1.2 Rationale and aims of the study

Conducting population censuses in Zimbabwe began as early as 1901 but was confined

to the non-African population The first enumeration of the whole population was done

in 1961–62 and then again in 1969 However, the two races were enumerated separately

In 1980 Zimbabwe attained political independence from the United Kingdom, and in

1982 the first census with a single reference period was conducted using the same questionnaire for the whole population

After the 1992 census, quite a number of population-related or demographic surveys were undertaken within the framework of the Zimbabwe National Household Survey Capability Programme The surveys include the 1994 and the 1999 Demographic and Health Surveys and the 1997 Inter-censual Demographic Survey The last population-based survey was performed in 2002

The HIV and AIDS epidemic, which has had serious negative impacts on households and communities in Zimbabwe since the early 1990s, has orphaned thousands of children, and the economic problems bedevilling the country have exacerbated the plight of children in general Life expectancy has dropped from 61 years in 1990 to 43 years in

2003 (Zimbabwe Human Development Report, 2003) As of 1999, an estimated 4.7 million children in the Southern African Development Community (SADC) region had been orphaned due to AIDS

In response to the AIDS epidemic and economic challenges seriously affecting the lives of orphans and vulnerable children (OVC) and threatening to ruin their future socio-economic well-being, various intervention agencies like the government, non-governmental organisations (NGOs) and community-based organisations (CBOs) have been very active in trying to identify those children in need of various forms of assistance, and directly or indirectly assisting them and their families In Zimbabwe, there are no population figures available for OVC at national level The data on the OVC population is fragmented, scant and is found at organisational level such as NGOs, government departments, schools and in different districts, wards, etc

This enumeration of OVC is therefore an endeavour to make sure that all the affected children are identified and the burden of the OVC problem in households and communities assessed, so that reliable information is available to assist the local municipalities and CBOs in planning their interventions

The main objectives of the OVC Census were therefore:

• To determine the numbers of OVC in the selected sites and the extent of their vulnerability

• To describe and characterise the challenges faced by OVC and their households

• To provide a sampling framework for the Psychosocial Survey (PSS) to be carried out among OVC

The goals of the project are to:

• Improve the social conditions, health, development and quality of life of vulnerable children and orphans

• Support families and households coping with an increased burden of care for affected and vulnerable children

• Strengthen community-based support systems as an indirect means to assist vulnerable children

• Build capacity in community-based systems for sustaining care and support to vulnerable children and households, over the long term

The need by intervention agencies to have accurate, reliable, up-to-date statistics and broad-based information in order to efficiently execute their work cannot be over-emphasised Population censuses are a principal means of collecting basic population statistics They form part of an integrated programme of data collection and compilation aimed at providing a comprehensive source of statistical information for economic and social development planning, for administrative purposes, for assessing conditions

in human settlements, for research and for commercial and other uses The value of each census is increased if the results can be used together with those from other investigations

This OVC Census was therefore conducted in order to document the extent of the problem in the selected districts by conducting a house-to-house enumeration exercise It also located households and communities with children in need of assistance without bias

or prejudice

The use of census data as a base or benchmark for current statistics can furnish information needed for conducting other statistical investigations This was the secondary aim of the census – that is, to provide a good basis for a sampling frame for other scientific studies The statistics generated usually provide good estimates of prevalence and sample size determination

2.1 Operational definitions

In the context of the OVC project in Zimbabwe the following definitions were used:

Caregiver: a person who regularly voluntarily assists a household, whose members are related or not related to him/her, in doing household chores, providing advice and giving spiritual, psycho-social and material support

Child: a person under the age of 18

Child-headed household: a household in which a person aged 18 years and below

is responsible for making day-to-day decisions for a group of persons who stay or who usually reside together and share food from the same pot, whether or not they are related

by blood

District: a district is made up of several wards

Enumeration area: the smallest demarcation of a district that is a cluster of about 100 households

Head of Household: a person, regardless of age, who is responsible for making day decisions for a group of persons who stay or who usually reside together and share food from the same pot, whether or not they are related by blood

day-to-Household: a group of persons who stay or who usually reside together and share food from the same pot, whether or not they are related by blood

Orphan: a person under the age of 18 who has lost either one or both parents

School-going age: in Zimbabwe, school-going age starts at seven years (grade 1) and goes up to form six (grade 13), although often if someone completes form four and does not proceed to form 5-6 he or she is still regarded as having finished school

Vulnerable Child: a child who is living in a household having one meal a day, receiving inadequate caregiving (child-headed households), with a sick household member who has been seriously ill for a month, households that are not able to pay for medical fees, and children with inadequate clothing Of note is that there is no direct relationship between orphanhood and vulnerability One can be an orphan and yet not be vulnerable

or one can be vulnerable without necessarily being an orphan

Ward: a ward is made up of 500 to 600 households

2.2 Description of the study sites

The 2003 OVC Census was conducted in the two districts of Bulilimamangwe and Chimanimani, which were then the WK Kellogg Foundation’s Integrated Rural Development Programme (IRDP) sites The WK Kellogg Foundation has been funding various intervention programmes through the IRDP in these two sites Thus, for logistical reasons and the need for continuity of intervention programmes, Bulilimamangwe and Chimanimani were selected as research sites.2

In 2002 the government recognised that Bulilimamangwe was too large in terms of administration and decided to split into three districts, namely Bulilima, Mangwe and Plumtree Bulilima and Mangwe districts are considered rural whilst Plumtree is urban

However, for the purpose of this study, the three districts will be treated under the name Bulilimamangwe, although at times the districts will be specifically distinguished from

2 In Zimbabwe the OVC intervention programmes are also being implemented in the following areas: Nyanga, Mutasa

each other Bulilimamangwe, located in Matabeleland South province and bordering with Botswana to the west, is in Region 5, which is prone to severe droughts (Figure 1) There are diverse economic activities in Bulilimamangwe, but the most common is cattle rearing

It has a total of 35 wards, specifically Bulilima in the north with 19 wards, Mangwe in the south with 12 wards, while Plumtree district, which is between Mangwe and Bulilima districts, has four wards

According to CSO (2002) the population of Bulilimamangwe is 172 788 (Bulilima 94

361 and Mangwe South, 78 427), with 54% being female The Ndebele and Kalanga are the dominant ethnic groups in the districts The average household size for Bulilima is 5.1 and Mangwe is 5.3 There is one district hospital, one mission hospital, three rural hospitals, seven rural health centres and 11 clinics for the three districts

Chimanimani district is located in Manicaland province in the Eastern Highlands of Zimbabwe (Figure 2.1) The district, which is predominantly rural, is 155 km south east of the provincial capital Mutare and borders Mozambique to the east, Buhera district to the west, Chipinge district to the south and Mutare district to the north The district is divided into 23 wards that are further divided into enumeration areas (Appendix 1) The district has a small urban area that is named after the district The current population estimate is

115 250 with 52% being female (CSO 2002) The average household size in the district is 4.4 and the population is predominantly Ndau, which is a Shona sub-ethnic group The district represents all five agro-ecological zones found in Zimbabwe The high and rugged terrain in the eastern side receives high rainfall, while the low-lying flat lands in the western part, at 1 600m above sea level, is characterised by a very erratic rainfall pattern The main economic activity in the district is agriculture The district has two mission hospitals, one of which is the acting district hospital, and three other hospitals, with one

of them being a rural facility, five rural health centres, thirteen clinics and two sub-clinics.Figure 1: Location of Bulilimamangwe and Chimanimani Districts

2.3 Geographical frame for the OVC Census

The geographical framework for the 2003 OVC Census followed the spatially defined structure of the districts and wards In an effort to cover all the selected districts, the exercise followed the smallest demarcation of each district, known as the enumeration areas (EAs) Consultants who work for Zimbabwe’s Central Statistics Office (CSO), using the 2002 National Census (Zimbabwe Census 2002 Preliminary Report) border definitions, mapped the boundaries of each ward and its enumeration areas On average, each enumeration area was assigned to one enumerator with an average of six EAs constituting

a ward, which was assigned a supervisor

2.4 Instrument

2.4.1 Questionnaire design and its translationThe research team adopted and modified the OVC Project’s Generic Census Questionnaire from South Africa’s Human Sciences Research Council (HSRC) and the census

questionnaire used by the Zimbabwe’s Central Statistics Office (CSO, 2002) The OVC Project’s Generic Census Questionnaire did not have the following information: geo-mapping, number of visits by the interviewer, disability, type of disability, highest level

of education completed, reasons why never been to school (those who never attended school), type of housing, detailed information about water and sanitation, other income indicators and, most importantly, indicators on vulnerability Zimbabwe’s Central Statistics Office’s census tool did not have a section on indicators of vulnerability

The questionnaires were pre-coded (Appendices 4a and 5a) and divided into 4 sections:

A) Geo-coding – identification of enumeration area and householdB) Household listing and socio-demographic characteristics of household members including disability

C) Socio-economic characteristics of the household (for example, type of dwelling unit, availability of water and sanitation, etc.)

D) Vulnerability indicators – to determine the extent of their vulnerability using a set

of indicators, for example, number of meals eaten per day, number of school-going children who were not attending school, number of children without adequate clothing, etc

The questionnaires were translated into the two main vernacular languages (Shona and Ndebele) for easy use in the districts The questionnaires were not back translated, but to ensure accuracy in translation, the translated versions were reviewed by an independent group of experts speaking the same language, to compare translated versions with original English texts

2.4.2 Pre-testing of Census questionnaire

A total of 40 questionnaires (20 in each district) were administered during the pre-test, which was conducted from 5-8 October 2003 The list of aspects that were evaluated during the pre-test are summarised in Appendix 2 After the pre-test, the census tool was slightly amended by deleting repetitions, adding some skip instructions, adding some codes (for example, either N/A or 99 for response that does not apply) and clarifying some questions which were vague Both the physical address and the name of the household were added for easy identification of households

2.5 Ethical issues

Permission to carry out the OVC Census was granted by the Medical Research Council

of Zimbabwe (MRCZ), which is the institution responsible for sanctioning all research

on humans in Zimbabwe In addition, supervisors and enumerators were trained on the following aspects of research ethics:

• Gaining access into research sites and households

• Getting ordinary and sensitive information

• The importance of respecting the respondents

• Risks, harm and benefits of participating in the exercise

• Confidentiality

• What to do when they met vulnerable groups including children, people with terminal illness, pregnant women and people who were stressed economically and socially

• The generally acceptable dress code

• Culturally acceptable ways of approaching dwelling units and how to introduce themselves and the organisation they were representing (Appendix 3)

• Obtaining informed consent from participants

2.6 Data collection

A population census can be taken either on a de jure or de facto basis Using the de facto approach, only those people who spend the nights of the census exercise in the district are counted, whilst the de jure involves enumerating people at their place of usual residence The OVC population census was done on a de jure basis This was done through enumerating all households (those with and without OVC) irrespective of their age, sex, ethnicity, citizenship and nationality or residence status in the three districts A head of household or key informant for the household was identified and interviewed to provide information on the household composition and other information relating to the household members’ living conditions

The main activities included:

• pre-enumeration, which involved mapping the districts, questionnaire design, preparation of supervisors and enumerator manuals, pre-tests sensitisation meetings, etc

• the actual enumeration

• post-enumeration, involving data processing, analysis, releasing preliminary results, editing report, dissemination of results

2.6.1 Pre-enumeration activities2.6.1.1 Publicity and sensitisationPublicity was properly and effectively done at all levels of authority

At the provincial level – this was done by sending letters stating the objectives of the OVC Census to the following: Provincial Medical Director, Police Chiefs and other influential people

At the district level – letters stating the objectives of the OVC Census were sent to the following: District Administrator, District Medical Officer, Police Chiefs and other influential people

Sensitisation meetings – The research team held sensitisation meetings in all the three districts before the census exercise was conducted These meetings involved all district and community leaders and key people in government departments

The main objectives of holding sensitisation meetings were:

• To ask for the co-operation of the communities during the census and the subsequent Psychosocial Survey (PSS) in terms of allowing the research team to conduct the studies in their areas and providing honest responses

• To ask the community leaders to assist in the selection of supervisors and enumerators from their wards

At the local level – core research team members and field supervisors contacted the local leadership (Ward Councillor, Chief and Village Head) and informed them about the census before the onset of data collection This was done in all enumeration areas

2.6.1.2 Mobilisation and recruitment

In each district, the research team selected and recruited local people to serve as members of the OVC Project Liaison Committee The Committee assisted in mobilising the local community to support the OVC project programmes, organising community meetings and logistical issues They further assisted in the mobilising of both the

census supervisors and enumerators who were then screened for final selection by the research team

The basic educational requirements for eligibility to participate as an enumerator were three Ordinary Level passes For supervisors, preference was given to people who had more than five Ordinary Level passes, university graduates and those who had participated in other surveys, for example, researchers, teachers, nurses, etc

A total of 67 fieldwork supervisors (Bulilimamangwe 44 and Chimanimani 23) and 527 enumerators (Bulilimamangwe 315 and Chimanimani 212) who were mostly youth were recruited (Appendix 7)

2.6.1.3 Training of fieldworkers The supervisors and enumerators were trained for three days in Plumtree and Chimanimani towns Due to logistical reasons both Bulilimamangwe and Chimanimani districts were divided into five clusters for the purpose of training and supervision In each district the research team was accordingly split into five teams and each team was assigned a cluster to train and supervise

The supervisors assisted in the training of their enumerators A summary table for the total number of enumerators and supervisors trained in the enumeration areas is given in Table 1 below

Table 1: Distribution of wards and enumeration areas (EAs) by district

District Total number of wards Total number of enumeration areas

Supervisors Training: – This was based on the Training Manual developed by the research team Objectives for the supervisors training were to:

1 define the role of the supervisor

2 define the supervisor’s expected activities before and during enumeration

3 define the supervisor’s role in quality control procedures for enumeration, highlighting the objectives of quality control from the supervisor’s perspective

4 equip the supervisors with knowledge on research ethics

Enumerators Training: – The training was based on the Training Manual developed by the research team The objectives of this training were to:

1 standardise the way enumerators were going to ask questions

2 define the parameters within which enumerators should operate

3 introduce the enumerators to the tools

4 go through the census tool and make sure the enumerators understood the objectives of each question in the tool

5 impart skills on rapport creation with the interviewee

6 impart skills on how to tackle difficult interviews

7 familiarise the enumerators with mapping procedures and definitions of the ward and EA boundaries in which they were going to work

8 equip the enumerators with knowledge on research ethics

2.7 Deployment of enumerators and supervisors

Each supervisor was in charge of a group of enumerators corresponding to the number of EAs in the ward Certain wards, which were large in terms of the number of households to

be covered, were given more than one team and supervisor Each team was expected to have a vehicle assigned to it for field activities, though some teams had to share vehicles

2.8 Quality control

There was an overall team that was responsible for co-ordinating the fieldwork and ensuring quality control by monitoring the work of the supervisors and enumerators In the field, the completed questionnaires were checked for completeness, accuracy and consistency and were also checked again by the enumeration area supervisor Corrections were made on the spot and call-backs were done to collect missing information The supervisors also made sure that the enumerators had covered all the households in their area of jurisdiction After the enumeration exercise, all the completed questionnaires were checked at the OVC Census’s command centres (Plumtree town in Bulilimamangwe and Chimanimani urban) by the core research team members

A quality control of questionnaire checklist was then developed (see Appendix 6) It checked on completeness of questionnaires and adherence to protocol

The head of household was the targeted respondent even for those households that were headed by children (see operational definitions) Where there was no-one at home

an enumerator would re-visit the household If the head of household or a member knowledgeable about the household was not in a position to respond (for example, due

to a busy work schedule), an enumerator would make an appointment to re-visit There were no call-backs for those who refused to be interviewed

2.9 Assessment of vulnerability

The Vulnerability Indicator Score was defined using the nine vulnerability indicators listed below:

1 The number of meals the household usually had a day

2 Households which indicated that there were some days they would go without food

3 Households with children of school-going age (7–18 years) who were not attending school

4 Households that were not able to pay medical fees if children were sick/ill

5 Households with children who did not have adequate clothing

6 Households with a household member who had been ill during the month preceding the census exercise

7 Households with school-going children who had no adequate school uniform

8 Child-headed households that had no caretaker

9 Child-headed households that had no-one with whom to discuss problems

Each indicator was coded 1 if the household was commensurate with vulnerable status and coded 0 if otherwise The nine vulnerability indicator scores were then summed up

to come up with a Total Vulnerable Indicator Score (TVIS), which was then expressed

as a percentage The maximum possible score was nine, indicating a state of being highly vulnerable

The TVIS was developed taking into account some of the basic Childs’ Rights (access to food, education, health and protection), which are enshrined in the United Nations (UN) Convention on the Rights of the Child (1989) However, these rights were not weighted in order to signify their relative importance in the overall score

The TVIS was then defined into the following three categories:

1 Less Vulnerable: – a TVIS below 50% of the total score

2 Moderately Vulnerable: – a TVIS of 50% to 74% of the total score

3 Highly Vulnerable: – a TVIS of 75% and above of the total score

The cut-off of 50% was chosen on the basis that anyone who scored more than half the total expected score was more likely to be at higher risk of being vulnerable

2.10 Data management and analysis

Soon after the completed questionnaires were quality-checked, by both the supervisor and the research team who were co-ordinating the fieldwork, they were ferried to Harare for data entry Fifteen data-entry clerks were recruited and trained for one week

on how to use the EPI-Info 6 statistical package and to understand the census tools (questionnaire and code sheet) As part of familiarising them with the questionnaire, trial runs on entering data were conducted Double entry of data took about one and half months Data was then exported to STATA from EPI-Info 6 for analysis After data entry the research team did data cleaning, which involved performing consistency checks

For analysis, basic frequencies and cross tabulations were done Descriptive statistics (for example, means, proportions, etc.) were computed for selected variables After data analysis, the data was stored electronically and the questionnaires were kept in a room with access limited to only selected research-team members

The following section describes the census results analysed at district level, whilst the results at ward level for each district are presented as appendices (see Appendix 8) For each district, results describe what is found in the communities at person level and also

at household level, with an attempt to pick out any differences that may be apparent between households with and without children 0–18 years, the latter which includes orphans and vulnerable children The first section (3.1) presents the findings from Bulilimamangwe followed by the section (3.2) on the findings of Chimanimani district

3.1 Bulilimamangwe District

3.1.1 Demographic data The populations of Bulilima, Mangwe and Plumtree districts were 81 984, 62 324 and

9 012 respectively, with the average household size for these districts being 4.8 persons

There were more females than males across all the three districts, with a sex ratio in Plumtree of 85 males per every 100 females, 84 and 83 males per 100 females in Mangwe and Bulilima respectively Over 70% of the households in the three districts had children

Table 2 below illustrates these findings in the three districts in more detail

Table 2: Demographics of population by district

The population profile for the three districts demonstrates consistency with the national profile in terms of sex ratio (male:female), which shows that there are more females than

males as well as the average household size The population census in the three districts has also shown that most of the people are below the age of 18 years.

Table 3: Comparison between National Census 2002 and OVC Census 2003 by district

National Census 2002

OVC Census 2003

Variation National

Census 2002

OVC Census 2003

Variation

Sex ratio (Male:

Female)

Total population

94 361 81 984 -12 377 (13.1%) 78 427 71 336 -7 091 (9.0%)

Total number of households

Average household size

Table 4: Age distribution in Bulilimamangwe

When further broken down, using the cut-off point of 18 years of age to separate the

‘children’ from the ‘adults’ as defined for this census, you find that the majority of the

population in Bulilima and Mangwe was 18 years and below (56% and 58% respectively) whilst Plumtree had slightly less children (48%) compared to adults (52%), (see Table 5).

Table 5: Distribution of population aged 18 and below by district

The reported levels of education attained among those who indicated that they had gone to school are summarised in Table 6 Of those who indicated that they had attended school, only 59 900, 47 068 and 7 185 from Bulilima, Mangwe and Plumtree respectively indicated the level of education they had attained The table shows that the largest group in Bulilima and Mangwe were those who had reached primary school level (71% and 70.3%) with only about 21% having attended secondary school, whilst for Plumtree, which was the urban area, there were slightly more (46%) who had reached secondary school

Table 6: Population distribution by level of education attained in districts

Education level Bulilima

3 In Zimbabwe primary education covers grades 1 to 7

Table 7: Reasons for children who had never gone to school in the districts

Reason for no education Bulilima

* Zimbabwe’s Ministry of Education, Sports and Culture policy states that a child must start going to school at the age of

7 years However, for this exercise determination of the numbers of those who were too young to attend school were not analysed according to school-going and non-school-going age groups

3.1.2 Magnitude of orphanhoodAccording to our definition of ‘child’ the proportion of orphans in the districts was determined based on the total population of those aged 18 years and below – a total

of 86 692 in the three districts As shown in Table 8, approximately one quarter of all children in each district were orphans (28%, 23% and 25% for Bulilima, Mangwe and Plumtree respectively)

Table 8: Extent of orphanhood by district

Table 9: Population distribution of type of orphan by district

Orphanhood status Bulilima

3.1.3 DisabilityThe prevalence of disability across the three districts was 5.5% for Bulilima, 5.2% for Mangwe and 2% for Plumtree (Table 10) with ‘difficulty in moving’ being the most common type of disability (26%) in both Bulilima and Mangwe followed by ‘difficulty in seeing’ (Table 10) In Plumtree, the urban district, difficulty in seeing (24.9%) was more common than difficulty in moving (20.9%) Table 10 also shows that ‘multiple disabilities’

were also prevalent across all the three districts (12% in Bulilima, 14.1% in Mangwe and 13.6% in Plumtree) It is worth noting that there was a high prevalence of ‘other’

disabilities, which were not specified across the three districts Further research needs to

be done to identify these

Table 10: Population distribution of disabled people by type of disability by district

Table 11: Prevalence of disability in households with children by district

Total number of households with children Bulilima

2 313 (22.9)

3.1.4 Household level data3.1.4.1 Sources of water

In the general population as shown in Table 12 below, households in Bulilima and Mangwe had protected wells or boreholes as their main water source (58.7% and 53% respectively) whilst Plumtree, being an urban setting, had mostly households with piped water, either inside the home (57.4%) or outside (29.5%) About a third of the households

in Bulilima and Mangwe (30.7% and 34.8%) fetched water from rivers, streams or dams.Table 12: Household distribution by main water source in the districts

Main source of water Bulilima

Well/borehole protectedWell/borehole unprotectedRiver/stream/dam

Table 13: Distribution of main water sources in households with children by districts

Main source of water Bulilima

Well/borehole protectedWell/borehole unprotectedRiver/stream/dam

premises or at least within a radius of 500m of their households, whilst the majority of the Bulilima and Mangwe population had water sources either outside a radius of 500m (35.3% and 36.6%), or more than a kilometre away (36.7% and 30.4% respectively).

Table 14: Household distribution by distance from main water source in the districts

Main source of water Bulilima

an urban setting Here the main source of energy was electricity for both cooking and lighting purposes

Table 15: Household distribution by main source of energy for cooking in the districts

Source of energy for cooking

Table 16: Household distribution by main source of energy for lighting in the districts

Source of energy for lighting

Table 17: Household distribution by toilet facility and district

Type of toilet facility Bulilima

Pit latrineCommunal

The status of sanitation facilities in the households with children across the three districts

is summarised in Table 18 Again, the majority of these households in Bulilima and Mangwe had no toilet facilities