A Multiple Overlapping Deprivation Analysis

July 2016

Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis

July 2016

All rights reserved. No part of this report may be reproduced or utilized in any form or by any means, electronic or mechanical, including photocopying, recording or by any information storage or retrieval system, without permission in writing from the publishers.

This publication is available on the following websites: www.unicef.org/zimbabwe and www.zimstat.co.zw/

The Child Poverty in Zimbabwe is based on A Multiple Overlapping Deprivation Analysis of data collected in the Multiple Indicator Cluster Survey 2014.

Recommended citation:

UNICEF Zimbabwe (2016) The Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis of Multiple Indicator Cluster Survey 2014.

Harare, UNICEF Foreword

Poverty reduction is a national priority for Zimbabwe. The Ministry of Finance and Economic Development coordinated the development of an Interim Poverty Reduction Strategy Paper (IPRSP) (2016-2018) for Zimbabwe. The purpose of the IPRSP is to enhance Government efforts in fighting poverty in Zimbabwe and ensure inclusive growth, guided by the country’s national development plan, Zimbabwe Agenda for Sustainable Socio-Economic Transformation (ZimASSET).

This report presents the multidimensional child deprivation analysis for Zimbabwe, applying the Multiple Overlapping Deprivation Analysis (MODA) methodology that measures various aspects of child poverty. Child poverty is defined as non- fulfilment of children’s rights to survival, development, protection and participation, anchored in the Convention on the Rights of the Child. The 2014 Multiple Indicator Cluster Survey (MICS) data was used, taking the child as a unit of analysis and applying a life-cycle approach when selecting dimensions and indicators to capture the different deprivations children experience at different stages of their life. The main objective of the report is to present Child Poverty in Zimbabwe using a direct method of child poverty measurement which analyses deprivations experienced by the child. The report goes beyond mere deprivation rates and identifies the depth of child poverty by analysing the extent to which the different deprivations are experienced simultaneously.

Focus on the key determinants affecting children today creates an enabling environment to support evidence-based advocacy around investment in critical basic services, including where high future returns can be achieved (e.g. education, nutrition, water and sanitation and early childhood development). These will help to accelerate the unfinished business as well as sustain the gains on the Millennium Development Goals (MDGs) and pave the way for achievement of the Sustainable Development Goals (SDGs).

We thank Swedish International Development Cooperation Agency (SIDA) for their financial support and acknowledge the great technical support and contribution of Prof. Dr. Chris De Neubourg and Nesha Ramful from the Economic Policy Research Institute (EPRI) and Marlous de Milliano from the UNICEF Office of Research – Innocenti. Special thanks to the Zimbabwe National Statistics Agency (ZIMSTAT) team, namely Taizivei Mungate, Evelyn Marima, Handrick Chigiji, Lloyd Mahere, Lovemore Sungano Ziswa, Jacob Chitiyo, Tinashe Mwadiwa and Tidings Matangira (Zimbabwe 2014 MICS Lead Team), the UNICEF Zimbabwe Country Office team i.e. Chief of Social Policy and Research, Samson Muradzikwa, the Social Protection Specialist, Chrystelle Tsafack and the Social and Economic Analysis Officer Rumbidza Tizora. We also acknowledge with much appreciation all those who participated in the N-MODA workshops held by UNICEF Zimbabwe in 2015 and 2016 and provided their invaluable inputs.

Dr. Mohamed Ayoya M. Dzinotizei Representative, UNICEF Harare, Zimbabwe Director-General, ZIMSTAT

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis i Table of Contents

Foreword ...... i

List of Abbreviations ...... vii

Executive Summary ...... ix

1. Introduction ...... 1 1.1 Background ...... 1

2. Methodology ...... 3 2.1 MODA methodology ...... 3 2.2 Indicator and dimension selection ...... 4

3. Results ...... 8 3.1 Deprivation among children aged 0–17 years (aggregation of results for all age groups of children) ...... 8 3.1.1 Multidimensional child poverty ...... 8 3.1.2 Overlap between child poverty based on wealth status of the household and multidimensional poverty ...... 18 3.2 Deprivation among children aged 0–23 months ...... 22 3.2.1 Single deprivation analysis ...... 22 3.2.2 Multiple Overlapping Deprivation Analysis ...... 33 3.3 Deprivation among children aged 24–59 months ...... 45 3.3.1 Single deprivation analysis ...... 45 3.3.2 Multiple Overlapping Deprivation Analysis ...... 55 3.4 Deprivation among children aged 5–14 years ...... 65 3.4.1 Single deprivation analysis ...... 65 3.4.2 Multiple Overlapping Deprivation Analysis ...... 71 3.5 Deprivation among children aged 15–17 years ...... 80 3.5.1 Single deprivation analysis ...... 80 3.5.2 Multiple Overlapping Deprivation Analysis ...... 87 3.6 Analysis of multidimensional poverty according to the wealth status of children ...... 96 3.6.1 Single deprivation analysis ...... 96 3.6.2 Multiple deprivation analysis ...... 99 3.6.3 Overlap analysis between monetary poverty and multidimensional deprivation ...... 104

References ...... 106

Annexes ...... 107 ii Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Table of Contents

Index of Tables

Table 1: Dimensions, indicators and age groups for N-MODA Zimbabwe using MICS5 ...... 5

Table 2: Multidimensional deprivation indices for children aged 0–17 years at the national level ...... 9

Table 3: Percentage distribution of children aged 0–17 years by province . . . .10

Table 4: Deprivation distribution for children aged 0–17 years by child’s characteristics ...... 10

Table 5: Overlap between child poverty for children aged 0–17 years based on wealth status of households and multidimensional poverty by area ...... 19

Table 6: Deprivation rates for children aged 0–23 months by dimension and child’s characteristics ...... 32

Table 7: Percentage distribution of children aged 0–23 months deprived on each number of deprivations by province ...... 38

Table 8: Deprivation distribution for children aged 0–23 months by child’s characteristics ...... 39

Table 9: Multidimensional deprivation indices for children aged 0–23 months at the national level ...... 40

Table 10: Deprivation rates for children aged 24–59 months by dimension and child’s characteristics ...... 55

Table 11: Deprivation distribution for children aged 24–59 months by child’s characteristics ...... 60

Table 12: Multidimensional deprivation indices for children aged 24–59 months at national level ...... 61

Table 13: Deprivation rates for children aged 5–14 years by dimensions and child’s characteristics ...... 70

Table 14: Deprivation distribution of children aged 5–14 years by child’s characteristics ...... 75

Table 15: Multidimensional deprivation indices at national level for children aged 5–14 years ...... 76

Table 16: Deprivation rates for children aged 15–17 years by dimension and child’s characteristics ...... 86

Table 17: Deprivation distribution for children aged 15–17 years by child’s characteristics ...... 91

Table 18: Multidimensional deprivation indices for children aged 15–17 years at national level ...... 92

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis iii Table of Contents

Index of Figures

Figure 1: Selected dimensions for each age group ...... 5 Figure 2: Deprivation distribution for children aged 0–17 years ...... 8 Figure 3: Percentage distribution of deprived children aged 0–17 years by area ...... 9 Figure 4a: Multidimensional child poverty (K=2) for children aged 0–17 years by province ...... 12 Figure 4b: Multidimensional child poverty (K=2) for children aged 0–17 years by area ...... 12 Figure 5: Multidimensional headcount ratios (K=2) for children aged 0–17 years by child’s characteristics ...... 13 Figure 6: Deprivation overlap for children aged 0–17 years by dimension . . . .14 Figure 7: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation ...... 15 Figure 8: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation by area ...... 16 Figure 9: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation by poverty status of households ...... 17 Figure 10: Deprivation overlap for children aged 0–17 years based on wealth status of households and multidimensional poverty ...... 18 Figure 11a: Child poverty for children aged 0–17 years based on wealth status of households by province ...... 20 Figure 11b: Child poverty of children aged 0–17 years based on multidimensional poverty (K=2) by province ...... 21 Figure 12a: Percentage distribution of deprivation rates for children aged 0–23 months by indicator and dimension ...... 22 Figure 12b: Percentage distribution of deprivation rates for children aged 0–23 months by dimension ...... 23 Figure 13: Deprivation rates for children aged 0–23 months by area ...... 24 Figure 14a: Deprivation rates for children aged 0–23 months for Nutrition by province ...... 26 Figure 14b: Deprivation rates for children aged 0–23 months for Health by province ...... 27 Figure 14c: Deprivation rates for children aged 0–23 months for Physical development by province ...... 28 Figure 14d: Deprivation rates for children aged 0–23 months for Water by province ...... 29 Figure 14e: Deprivation rates for children aged 0–23 months for Sanitation by province ...... 30 Figure 15a: Percentage distribution of deprivation rates for children aged 0–23 months with/without a health card ...... 31 Figure 15b: Percentage distribution of deprivation rates for children aged 0–23 months with/without a birth certificate ...... 31 Figure 16: Deprivation overlap for children aged 0–23 months for each dimension ...... 33 Figure 17: Deprivation overlap for children aged 0–23 months ...... 35 Figure 18: Deprivation overlap for children aged 0–23 months by area ...... 36 iv Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Table of Contents

Figure 19a: Percentage distribution of children aged 0–23 months deprived on each number of deprivations ...... 37 Figure 19b: Percentage distribution of children aged 0–23 months deprived on each number of deprivations by area ...... 38 Figure 20: Multidimensional deprivation headcount for children aged 0–23 months by various cut-off points ...... 41 Figure 21: Multidimensional deprivation indices (K=2) for children aged 0–23 months by province ...... 42 Figure 22: Multidimensional deprivation headcount (K=2) for children aged 0–23 months by child’s characteristics ...... 43 Figure 23: Adjusted deprivation headcount rate (K=2) for children aged 0–23 months by area ...... 44 Figure 24: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 0–23 months by area ...... 44 Figure 25a: Deprivation rates for children aged 24–59 months by indicators and dimensions ...... 47 Figure 25b: Deprivation rates for children aged 24–59 months by dimensions . .47 Figure 26a: Deprivation rates for children aged 24–59 months by indicators and gender ...... 48 Figure 26b: Deprivation rates for children aged 24–59 months by indicators and possession of a birth certificate ...... 49 Figure 27: Deprivation rates for children aged 24–59 months by dimension and area ...... 50 Figure 28a: Deprivation rates for children aged 24–59 months for Water by province ...... 51 Figure 28b: Deprivation rates for children aged 24–59 months for Physical development by province ...... 52 Figure 28c: Deprivation rates for children aged 24–59 months for Health by province ...... 53 Figure 29: Deprivation rates for children aged 24–59 months by the presence of infant mortality cases in the five years preceding MICS5 ...... 54 Figure 30: Deprivation overlap for each dimension for children aged 24–59 months ...... 56 Figure 31: Deprivation overlap for children aged 24–59 months ...... 57 Figure 32: Deprivation overlap for children aged 24–59 months by area ...... 58 Figure 33a: Percentage distribution of children aged 24–59 months by number of deprivations at national level ...... 59 Figure 33b: Percentage distribution of children aged 24–59 months by number of deprivations by area ...... 60 Figure 34: Multidimensional deprivation headcount for children aged 24–59 months by various cut-off points ...... 62 Figure 35: Multidimensional deprivation indices (K=2) for children aged 24–59 months by region ...... 62 Figure 36: Multidimensional deprivation headcount (K=2) for children aged 24–59 months ...... 63 Figure 37: Adjusted deprivation headcount rate (K=2) for children aged 24–59 months by area ...... 64 Figure 38: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 24–59 months by area ...... 64 Figure 39a: Deprivation rates for children aged 5–14 years by indicator ...... 66

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis v Table of Contents

Figure 39b: Deprivation rates for children aged 5–14 years by dimension ...... 66 Figure 40: Deprivation rates for indicators under the dimension Education, children aged 5–14 years ...... 67 Figure 41: Deprivation rates by urban and rural area, children aged 5–14 years ...... 67 Figure 42a: Deprivation rates for children aged 5–14 years for Education by province ...... 68 Figure 42b: Deprivation rates for children aged 5–14 years for Information by province ...... 69 Figure 43: Deprivation overlap for children aged 5–14 years for each dimension ...... 71 Figure 44: Deprivation overlap for children aged 5–14 years ...... 72 Figure 45: Deprivation overlap for children aged 5-14 years by area ...... 73 Figure 46a: Percentage distribution of number of deprivations experienced by children at national level ...... 74 Figure 46b: Percentage distribution of number of deprivations experienced by children by area ...... 74 Figure 47: Multidimensional deprivation headcount for children aged 5–14 years by various cut-off points and area ...... 76 Figure 48: Multidimensional deprivation indices (K=2) for children aged 5–14 years by province ...... 77 Figure 49: Multidimensional deprivation headcount (K=2) for children aged 5–14 years by child’s characteristics ...... 78 Figure 50: Adjusted deprivation headcount rate (K=2) for children aged 5–14 years by area ...... 79 Figure 51: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 5–14 years ...... 79 Figure 52a: Deprivation rates for children aged 15–17 years by indicator ...... 81 Figure 52b: Deprivation rates for children aged 15–17 years by dimension . . . . .81 Figure 53: Percentage distribution of Education indicators for children aged 15–17 years by sex ...... 82 Figure 54: Deprivation rates for children aged 15–17 years by urban and rural areas ...... 83 Figure 55a: Deprivation rates for children aged 15–17 years for Education by province ...... 84 Figure 55b: Deprivation rates for children aged 15–17 years for Information by province ...... 85 Figure 56: Percentage distribution of children aged 15–17 years by deprivation overlaps for each dimension ...... 87 Figure 57: Deprivation overlaps for children aged 15–17 years ...... 88 Figure 58: Deprivation overlaps for children aged 15–17 years by area ...... 89 Figure 59a: Percentage distribution of children aged 15–17 years by number of deprivations at national level ...... 90 Figure 59b: Percentage distribution of children aged 15–17 years by number of deprivations by area ...... 90 Figure 60: Multidimensional deprivation headcount for children aged 15–17 years by various cut-off points ...... 92 Figure 61: Multidimensional deprivation indices (K=2) for children aged 15–17 years by province ...... 93 vi Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Table of Contents

Figure 62: Multidimensional deprivation headcount (K=2) for children aged 15–17 years ...... 94 Figure 63a: Adjusted deprivation headcount rate for children aged 15–17 years by area ...... 95 Figure 63b: Adjusted deprivation headcount rate and its decomposition (K=2) for children aged 15–17 years by area ...... 95 Figure 64: Deprivation rates for children aged 0–23 months by wealth quintile .97 Figure 65: Deprivation rates for children aged 24–59 months by wealth quintile 97 Figure 66: Deprivation rates for children aged 5–14 years by wealth quintile . . .98 Figure 67: Deprivation rates for children aged 15–17 years by wealth quintile . .98 Figure 68: Number of simultaneous deprivations for children aged 0–23 months by wealth index ...... 99 Figure 69: Number of simultaneous deprivations for children aged 24–59 months by wealth index ...... 99 Figure 70: Number of simultaneous deprivations for children aged 5–14 years by wealth index ...... 100 Figure 71: Number of simultaneous deprivations for children aged 15–17 years by wealth index ...... 100 Figure 72: Multidimensional indices (K=2) for children aged 0–23 months by wealth quintile ...... 101 Figure 73: Multidimensional indices (K=2) for children aged 24–59 months by wealth quintile ...... 101 Figure 74: Multidimensional indices (K=2) for children aged 5–14 years by wealth quintile ...... 101 Figure 75: Multidimensional indices (K=2) for children aged 15–17 years by wealth quintile ...... 102 Figure 76: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 0–23 months ...... 102 Figure 77: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 24–59 months ...... 103 Figure 78: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 5–14 years ...... 103 Figure 79: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 15–17 years ...... 103 Figure 80: Overlap between poverty and deprivation for children aged 0–23 months ...... 104 Figure 81: Overlap between poverty and deprivation for children aged 24–59 months ...... 105 Figure 82: Overlap between poverty and deprivation for children aged 5–14 years ...... 105 Figure 83: Deprivation overlap between poverty and deprivation for children aged 15–17 years ...... 105

Index of Annexes

Annex 1: Selected dimensions, indicators and thresholds for each age group .107 Annex 2: Indicator ‘Wish-list’ ...... 109 Annex 3: Assets used to build the wealth index ...... 110

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis vii List of Abbreviations

A Average intensity across the deprived

BEAM Basic Education Assistance Module

CC-MODA Cross Country Multiple Overlapping Deprivation Analysis

ECD Early Childhood Development

EPRI Economic Policy Research Institute

H Headcount Ratio

Hh Household

HIV/AIDS Human Immunodeficiency Virus/Acquired Immune Deficiency Syndrome

IPRSP Interim Poverty Reduction Strategy Paper (IPRSP)

K Cut-off point

M0 Adjusted deprivation headcount

MCA Multiple Correspondence Analysis

MDG Millennium Development Goal

MICS Multiple Indicator Cluster Survey

MODA Multiple Overlapping Deprivation Analysis

MTCT Mother-to-Child Transmission

N-MODA National Multiple Overlapping Deprivation Analysis

OPHI Oxford Poverty and Human Development Initiative

PICES Poverty Income and Consumption Expenditure Survey

Q Quintile

SDG Sustainable Development Goal

SIDA Swedish International Development Cooperation Agency

TV Television

UN United Nations

UNICEF United Nations Children’s Fund

WFP World Food Programme

WHO World Health Organization

ZimASSET Zimbabwe Agenda for Sustainable Socio-Economic Transformation viii Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Executive Summary

The case for investment in children in Zimbabwe

Zimbabwe has a very young population, with 48 per cent (6.3 million) of its people below 18 years of age. These children will be responsible for the social, demographic, environmental, economic and political dynamics that shape the future of the country. Investing in human capital development today ensures that these children contribute to the country’s growth path, as well as to poverty reduction. Tapping and exploiting this demographic dividend can help to boost the economy of Zimbabwe. Creating the best possible environment for children to develop into healthy, productive and balanced adults results in high-quality human capital. While much research has confirmed that investments in children are of high value and yield high returns for entire societies, it is crucial that these investments are made as early as possible in a child’s life cycle. Any circumstances which inhibit a child’s ability to fully realize its capabilities and functioning are especially dire in the child’s earliest stages of life, as many such deficiencies can often not be regained or reversed as the child grows older.1

The methodology

In line with the above, a multidimensional poverty analysis for children was carried out based on the Multiple Overlapping Deprivation Analysis (MODA) methodology. The MODA methodology has been developed by the United Nations Children’s Fund (UNICEF) to provide a framework in which children’s (monetary) poverty and deprivation can be measured, quantified and identified. The methodology concentrates on the measurement of child deprivation, and has the following characteristics: 1. It takes the child, rather than the household, as the unit of analysis. 2. It underlines the use of individual-level data where possible, so that any differences across sex, age or within households may be observed. 3. It makes use of the life-cycle approach, changing indicators according to the changing needs of children at different life stages. 4. It broadens the scope of sector-based approaches through overlapping deprivation analysis. 5. It includes the prevalence and the depth of deprivation for each child, revealing the most vulnerable children with a high number of simultaneous deprivations. 6. It generates profiles in terms of the geographical and socio-economic characteristics of the (multiply) deprived, allowing for better targeted, more effective policy responses and interventions.

In order to alleviate poverty, it is important to concentrate on the most vulnerable children, especially those simultaneously deprived across many dimensions of well-being, because several deprivations at a time during childhood and even adolescence can have irreversible effects on the productivity and social inclusion of these children.

1 A young child who is exposed to health risks in unsanitary living conditions, unsafe water, and inadequate feeding may incur severe effects on in its physical and mental development. As the child enters school age, he/she may already be at a disadvantage in his/her ability to learn and retain information, and these factors could intersect with further disadvantages the child will face in other areas of life as he/she grows older and eventually enters the workforce.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis ix Executive Summary

The MODA analysis utilized the 2014 Multiple Indicator Cluster Survey (MICS) data, which was part of the Round 5 survey programme. Age groups, dimensions, indicators and thresholds used in the MODA analysis were decided2 at a workshop with key stakeholders, taking into account internationally agreed definitions of the essential rights and needs of the child (see Annex 2). The results for each of the following age groups are presented throughout the report: 0–23 months, 24–59 months, 5–14 years, 15–17 years and 0–17 years.

Results

The MODA results for children in Zimbabwe provide both broad and specific insights into the situation of children. They reveal in which dimensions of well- being children were deprived, how these deprivations might be interrelated (overlap) and how to identify the deprived children. Furthermore, the analysis reveals how overall poverty reduction should not rely solely on the reduction of monetary poverty (proxied by asset poverty in this report) – especially in the case of children. Children living in households that were monetarily poor were not necessarily the same children who were affected by the analysed deprivations. Policies addressing the needs of children, therefore, require a distinct sensitivity to understanding how ‘poverty’ manifests itself among children, and how ‘deprivation’ of access to necessary goods and services adds a critical level of complexity to the poverty in the country. The main findings are summarized below.

Very few children were deprived in one dimension at a time; most of them experienced overlapping deprivations

About one child in ten was not deprived on any of the dimensions of child well- being used in this analysis. This implies that 90.1 per cent of the children experienced at least one deprivation. It was noted that 59.6 per cent of children experienced at least two deprivations, while 27.6 per cent experienced between three and five.

The Sanitation dimension yields the highest proportion of deprived children

It was observed throughout the analysis that there was a very high proportion of children deprived on the Sanitation dimension across all the age groups, but more so for children from 24 months to 17 years, with more than 7 children out of 10 being deprived. The high deprivation rates were driven by the indicators toilet type and sharing of toilet facility. Disaggregating the results at the provincial level revealed that there was almost the same proportion of deprived children in all the provinces, implying that sanitation appears to be more of a national problem than specific to some provinces. Children are disproportionately affected by sanitation provisions, carrying the highest burden of diarrhoeal diseases and worm infections. Better sanitation and hygiene practices are very closely linked to good health, and can significantly help reduce illness and prevent unnecessary deaths.

Provincial differences for the Water dimension

The Water dimension had the highest level of disparity among the provinces. The level of deprivations on the Water dimension varied significantly across the

2 The final selection reflects the opinions of the key country stakeholders, national standards, research interests and data availability. x Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Executive Summary

provinces for all age groups. More than 50 per cent of children in all age groups in Matabeleland North and Masvingo provinces were deprived on the Water dimension. However, in the metropolitan provinces of Bulawayo and Harare, deprivation rates were quite low, ranging from 1 per cent to 10 per cent, depending on the age group. Deprivation on the Water dimension was often accompanied by deprivations in Health and Sanitation. In the past few years, access to safe water has been a serious problem in Zimbabwe. In 2013, there was an outbreak of water-borne diseases, such as typhoid, dysentery and diarrhoea, which claimed many casualties.3 Matabeleland North and Masvingo were among the provinces with the highest number of sick people and deaths.

One child in four below the age of 5 was stunted, and deprivation on the dimension Physical development seemed to overlap with Nutrition, Health, Sanitation and Water

About 25 per cent of children aged below 5 years were stunted. Several studies have demonstrated that a stunted growth can have a sustainable impact on the physical and intellectual development of a child. Stunted children have the tendency to drop out of school earlier. Even if they do not drop out, their cognitive capacity is lower and inferior to those of their non-stunted counterparts. This overlap analysis revealed that many of the children deprived in Physical development tended to be deprived in other dimensions, such as Nutrition, Health, Water and Sanitation. For example, of the 30.3 per cent of children deprived on the Physical development dimension in the age group 24–59 months, 3.4 per cent were deprived in that dimension only.

Notable disparities between deprivation rates, based on geographical location

Throughout the study, it was noted that for all dimensions and age groups, the deprivation rates were higher in rural areas for both simple and multiple deprivation analysis. The deprivation headcount rate adjusted for intensity index (M0), was high for children living in rural areas, implying that overall rate, depth and intensity of deprivation were higher in rural areas. Nonetheless, despite low percentages, the prevalence of deprivation in urban provinces was inherent.

Bulawayo Province had the lowest deprivation indices, while Matabeleland North and Masvingo provinces had the most deprived children. Notably, although the proportion of deprived children varied significantly across the provinces, the average intensity of deprivation across the deprived children4 was not different. This implies that the deprived children in all provinces had, on average, the same number of deprivations.

By decomposing the deprivation index, M0, it was found that the dimensions used in the analysis contributed to the index in different ways, based on the geographical location and poverty status of the child

In general, the Water and Information dimensions contributed more to deprivation levels in rural areas, while Education contributed more in urban areas. Similar findings were noted with respect to poverty status of the household; while the

3 Zimbabwe Situation (2013) [Online]. Water-borne diseases resurface. Available at: http://www.zimbabwesituation.com/news/zimsit_water-borne-diseases-resurface/ 4 In this study, a deprived child was defined as one who faces at least two deprivations simultaneously.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis xi Executive Summary

Water and Information dimensions contributed relatively more to deprivation rates for children living in poor households, Education was the highest contributor in non-poor households. It was important to note that each of the contributions was relative. The actual deprivation in Education might not have been more severe for children living in non-poor households, but its relative importance was higher.

Overlap between poverty based on wealth5 and deprivations

Some 45 per cent of children were both poor and deprived; 18 per cent were poor but non-deprived; while 15 per cent were non-poor but deprived.

Deprivation affects the children in all wealth quintiles especially for children below 5 years; 19 per cent of children aged 0–23 months in households in the richest wealth quintile were deprived in two or more dimensions simultaneously. This revealed that a relatively high standard of living of households does not necessarily protect children from highly adverse conditions (especially for children below 5 years of age). For children in households in the poorest wealth quintile, more than one child in three was deprived on at least three dimensions. Joint policies targeting both poor and non-poor households should be implemented to tackle both poverty and deprivation.

The MODA analysis explored the profile of the most vulnerable children by locating them both geographically and socially. It facilitates understanding of how the different deprivations by sector overlap, to show which deprivations may need to be addressed simultaneously and to inform equity-based public policy responses to child deprivation. It also highlights the deprivation coincidences that need further theoretical and empirical elaboration. The study, however, was limited by the availability of data, so that not all dimensions of children’s well-being were studied. To guide the collection of data in future in Zimbabwe, the local stakeholders came up with a ‘wish-list’6 comprising more dimensions and indicators that they wish to be collected in forthcoming surveys or censuses.

5 A wealth index was created, based on the assets and some characteristics of the household using the multiple correspondence analysis (MCA) technique. Households were then classified in increasing order of wealth status; thus categorizing them in quintiles. According to the PICES 2011/12, 62.6 per cent of Zimbabwean households were deemed poor, this implies that households constituting the first three wealth quintiles of the wealth index were poor in absolute terms. 6 See Annex 2 for the ‘wish-list’ of dimensions and indicators that local Zimbabwean stakeholders would like to see in the next data collection. xii Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Introduction

1.1 Background

In the 1980s, Zimbabwe was one of the best performing countries in the southern African region, with a vibrant agricultural export sector. It also had the highest literacy rate in sub-Saharan Africa. In the 1990s consumption poverty rate in Zimbabwe was 25% compared to poverty rate in sub-Saharan Africa which was fluctuating around 46%.7 Over the past few decades, the country has experienced an economic meltdown, and today around 72 per cent of the Zimbabwean 1 population lives in consumption poverty, with children being the most vulnerable. Analysing the situation of children can provide crucial indicators to help understand better the nature of poverty in Zimbabwe – who the poor are, why their poverty persists and how poverty is transmitted intergenerationally. This aids in the formulation and evaluation of policies to address poverty in Zimbabwe and overcome the roadblocks that might prevent those in need being reached. Now that the Millennium Development Goals (MDGs) have been replaced by the Sustainable Development Goals (SDGs), it is also pertinent to assess the current situation of children, in order to prepare and work towards the achievement of the goals by 2030.

Building an understanding of the nature of child poverty and children’s deprivations is essential to addressing the needs of children through suitable programmes and policies. Where the majority of historical methods of poverty analysis have focused on a household’s or an individual’s monetary well-being, income or expenditure, these measures are less appropriate for measuring child poverty or well-being. As children are not typically decision-makers or primary income earners in a household, their access to household-level income does not always lead to proportional improvements in their well-being. In addition, children’s needs are unique and specific at different stages of childhood development, and these needs are not necessarily met solely by increased access to income. Multidimensional poverty analysis helps to complement existing studies on monetary-based poverty. UNICEF’s Multiple Overlapping Deprivation Analysis (MODA) methodology is one such method, aiming to holistically identify and quantify the nature of children’s deprivations to help inform the design of services and interventions that more accurately meet the needs of children.

This report presents the results of a multidimensional poverty analysis using MODA methodology to identify, locate and create a profile of children with deprivations in Zimbabwe. The analysis uses the 2014 Multiple Indicator Cluster Survey (MICS) data to identify the type, level and overlaps of deprivations faced by children in the country. Further decompositions reveal regional disparities, urban/rural cleavages, and other individual or household characteristics which may be linked to the multidimensional phenomenon of child poverty.

While the results of this analysis may corroborate those of existing studies in monetary poverty, where children living in financially poor households suffer multiple deprivations, the MODA analysis also reveals the deprivations of children

7 Source: Alwang, J. R., Mills, B. F., & Taruvinga, N. (2002). Why has poverty increased in Zimbabwe? World Bank Publications. Chien, S., & Ravallion, M. (2001). How did the world's poorest fare in the 1990s? Review of Income and wealth, 47(3), 283-300.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 1 Introduction

living in non-poor households. Furthermore, the study emphasizes the severity of multiple deprivations that a group of children may suffer. Such aspects are not covered in existing child poverty analyses.

The deprivation analysis covers the dimensions of access to basic essential services in the areas of nutrition, health, child development, physical development, water, sanitation, education and information. Children were grouped in the following age groups: 0–23 months, 24–59 months, 5–14 years, 15–17 years and 0–17 years. A summary detailing the methodology supporting the selection of dimensions, indicators, deprivation thresholds and age groups is found under the Metholodogy section below.

The next section of the report discusses the methodology and its application in the broader context of children’s deprivations in Zimbabwe. That is followed by a section describing the results of the MODA analysis of children’s deprivations.

2 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Methodology

2.1 MODA methodology

This study uses the Multiple Overlapping Deprivation Analysis methodology to measure multidimensional deprivation among children aged 0–17 years. The MODA methodology was developed at UNICEF to provide a framework in which children’s (monetary) poverty and deprivation can be measured, quantified and identified. The methodology builds on existing approaches to multidimensional poverty measurement, such as UNICEF’s Global Study on Child Poverty and 2 Disparities (see Gordon et al., 2003; UNICEF, 2007) and the Oxford Poverty and Human Development Initiative (OPHI) Multidimensional Poverty Index (see Alkire and Santos, 2010; Alkire and Foster, 2011). The methodology concentrates on the measurement of child deprivation, and comprised the following key elements: l It takes the child rather than the household as the unit of analysis. l It underlines the use of individual-level data whenever possible, so that any differences across gender, age or within households may be observed. l The method makes use of the life-cycle approach, changing indicators according to the changing needs of children at different life stages. l It broadens the scope of sector-based approaches through overlapping deprivation analysis. l It includes the prevalence and the depth of deprivation for each child, revealing the most vulnerable children with a higher number of deprivations at the same time. l It generates profiles in terms of the geographical and socio-economic characteristics of the (multiply) deprived, allowing for better targeted, more effective policy responses and interventions.

The MODA methodology was described in more detail in the Step-by-step guidelines to MODA (de Neubourg et al., 2012). The method has already been extensively applied to a series of Cross Country Multiple Overlapping Deprivation Analysis (CC-MODA) using comparable datasets, indicators and thresholds to facilitate the analysis of multidimensional child deprivation across lower- and middle-income countries. N-MODA is an application of the MODA methodology to specific national contexts, with the customized use of datasets, age groups, dimensions, indicators and thresholds, often utilizing richer information that is available from national datasets.

N-MODA’s focus was on producing country-specific analyses, with the aim of: i) capturing national values and objectives concerning child development; ii) exploring the profile of deprived children, to locate them both geographically and socially; iii) improving understanding of how the different deprivations by sector overlap to inform which deprivations may need to be addressed simultaneously; iv) informing equity-based public policy responses to child deprivation; and v) indicating deprivation coincidences that need further theoretical and empirical elaboration.

8 Profiling was the basis for the equity analysis, showing differences between genders, geographical provinces, urban and rural areas, parents’ socio-economic situation, wealth quintiles, and other variables.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 3 Methodology

2.2 Indicator and dimension selection

The analysis of multiple and overlapping deprivations was based on indicators, dimensions, deprivation thresholds and age groups, which were selected following data-driven feasibility assessments, discourse with national partners, and consideration of internationally agreed definitions of the essential rights and needs of the child. The final selection reflects the opinions of the key country stakeholders, national standards, research interests and data availability.

Children in a particular age group were assessed on the basis of a deprivation threshold in a set of indicators, which make up a set of dimensions. In each dimension, a combination of indicators and respective thresholds determines the level of deprivation of the child on that dimension. In line with the union approach, children were regarded as deprived on a dimension if they were deprived on at least one of the dimension’s indicators.

Some dimensions may not apply to the entire child population for reasons that may include empirical consistency and data constraints. For instance, the Education dimension only covers school-aged children (aged 5–17), while the Sanitation dimension covers children of all age groups. In fact, the dimensions Water and Sanitation were based on household-unit data and therefore apply to all age groups.

Figure 1 below illustrates the dimensions used in the N-MODA analysis for children in Zimbabwe. Differences between the dimensions used were based on the life-cycle approach, which states that a child has different needs at different ages. The choice of dimensions was also driven by the data available in MICS5. While two age groups might share some dimensions (e.g. health), their deprivation on these dimensions was measured differently, using different sets of indicators depending on their age (see Table 1). For children aged 0–23 months, health was measured by skilled birth attendance, full immunization, mother’s knowledge of mother-to-child transmission (MTCT) and cooking fuel, while for children aged 5–17 years, the same dimension was measured only by cooking fuel. It should also be noted that MICS5 has richer data for children below 5 years.

9 The rationale for using this approach is to capture all children showing any sign of deprivation in a specific dimension. If there is more than one indicator per dimension, the selected indicators have been chosen to complement each other in the explanation of the dimension they represent and to jointly identify the children’s status in the respective dimension. This method does not account for the depth of deprivation within a given dimension because the methodology is developed to focus on the dimensions rather than separate indicators.

4 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Methodology

Figure 1: Selected dimensions for each age group 0-23 24-59 5-14 15-17 months months years years

Nutrion Health Health Health

Child Health Educaon Educaon development

Physical Physical Water Water development development

Water Water Sanitaon Sanitaon

Sanitaon Sanitaon Informaon Informaon

The dimensions and indicators used for each age group are listed in Table 1. A more detailed table, including the thresholds used for each indicator, is found in Annex 1.

Table 1: Dimensions, indicators and age groups for N-MODA Zimbabwe using MICS5 Age groups Dimension Indicator 0–23 24–59 5–14 15–17 months months years years Infant and young child feeding (incl. exclusive X Nutrition breastfeeding for 0–5 months) X (12–23 Full immunization X months) Mother’s knowledge of X Health MTCT Skilled birth attendance X Cooking fuel X X X X Early childhood X (36–59 education attendance months) Child Child–adult interaction X X development Availability of toys or X books

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 5 Methodology

Table 1: Dimensions, indicators and age groups for N-MODA Zimbabwe using MICS5 Age groups Dimension Indicator 0–23 24–59 5–14 15–17 months months years years

Stunting X X Physical development Wasting X X Underweight X X School attendance X X Primary school X Education attainment Grade-for-age X

Water source X X X X Water Distance to water X X X X source Handwashing X X X X

Sanitation Toilet type X X X Sharing of toilet facility X X X Availability of Information information/ X X communication devices

Note: x denotes where the indicator was applicable.

Children in a particular age group were assessed on the basis of a deprivation threshold in a set of indicators, which make up a set of dimensions. For each dimension, a combination of indicators and respective thresholds determines the level of deprivation of the child on that dimension. According to the union approach, children were regarded as deprived on a dimension if they were deprived on at least one of the dimension’s indicators. For example, a child aged 0–23 months is considered deprived on the Health dimension if he or she was deprived in skilled birth attendance, even if not deprived in the indicators covering full immunization, mother’s knowledge of MTCT and cooking fuel. The depth of deprivation has not been taken into account, but all indicators have been chosen on the basis that they all partly explain the realization (or not) of a child’s rights. Since each of the selected dimensions reflects a basic right, in the analysis they are considered to be of the same importance.

For each age group, the following analyses were carried out: l Single deprivation: the percentage of children deprived on each dimension (and for each indicator) was estimated to give a sector perspective. It gives a first insight into which deprivations were particularly relevant for children in the five different age groups in Zimbabwe. l The distribution of the number of dimensions on which children were deprived: the number of deprivations per child was counted to give an overview of the distribution of all deprivations among the different age

6 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Methodology

groups and according to different background characteristics (profiling variables). The deprivation count also enabled analysis of the depth of multidimensional deprivation. l Multidimensional deprivation overlaps: the analysis looked at the different deprivations that were usually experienced simultaneously. Combinations of deprivations have been highlighted and estimates of the number of children suffering from one to five deprivations at the same time have been given. l Multidimensional deprivation indices: several multidimensional deprivation indices were calculated to provide summary statistics: (i) the headcount ratio (H) to look at the incidence of multiple deprivation on the various dimensions; (ii) the average intensity (A) to look at the number of deprivations that a deprived child experiences as a percentage of all possible deprivations; (iii) the adjusted deprivation headcount (M0) was calculated to capture both the incidence and depth of deprivation.

This report captures notable results of the analysis of multiple overlapping deprivations among children in Zimbabwe. All tables and figures produced for this analysis are available in a separate document that complements this document. Where possible and meaningful, the thresholds for each indicator were defined following internationally agreed definitions, mainly by the UN, UNICEF and the World Health Organization (WHO).

Note: Chi-square tests were performed to determine whether the difference between deprivation rates for various age groups of children were significantly different at the 5 per cent level of significance. Throughout the document, the use of an asterisk (*) implies that the difference was significant.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 7 Results

This chapter presents the results of the multidimensional child deprivation analysis for Zimbabwe. The findings are first presented for all children (0–17 years). Results are then presented by age group and in increasing order of age. The analysis for each age group is divided into two sections, with the first section showing the results of the single deprivation analysis, the overlap analysis between the dimensions under study and the results of the Multiple Overlapping Deprivation Analysis. In the final section of this document, an analysis is made of 3the overlap between child poverty based on the wealth status of households and multidimensional deprivation.

3.1 Deprivation among children aged 0–17 years

This section is divided into two sub-sections. The first presents the main results for multidimensional and non-monetary child poverty for children aged 0–17 years (aggregation of results for all age groups of children), while the second focuses on the overlap between multidimensional poverty and poverty based on the wealth status (measured by an asset index) of the child.

3.1.1 Multidimensional child poverty

The Multiple Overlapping Deprivation Analysis reveals that the majority of children in Zimbabwe experience several deprivations simultaneously. Figure 2 depicts this situation graphically, showing that 9.9 per cent of children faced no deprivations, while 59.6 per cent suffered at least two deprivations simultaneously. The distribution of deprivations shows that most children faced one (30.5 per cent) or two (31.9 per cent) deprivations simultaneously. More than a quarter of children were severely deprived on three or more of the five dimensions. Table 2 further explains the situation of children by giving an indication of the depth and severity of deprivation using different thresholds. For example, 59.6 per cent of children were deprived on at least two dimensions (H). On average, children faced deprivation on 2.6 of the five dimensions (52.2 per cent).

Figure 2: Deprivation distribution for children aged 0–17 years 35 31.9 30.5 30

25 20 20

15 9.9 10 6.6 5 1.1 Percentage of children deprived, % deprived, ofchildren Percentage 0 012345

Number of simultaneous deprivaons experienced by the child

8 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 2: Multidimensional deprivation indices for children aged 0–17 years at the national level

Number of Deprivation Average intensity Average intensity deprivations headcount across the deprived across the (dimensions) (H) % (A) in % deprived (A)

1–5 90.1 41.3 2.1

2–5 59.6 52.2 2.6

3–5 27.6 66.3 3.3

4–5 7.6 82.8 4.1

5 1.05 100 5

Some characteristics of the children’s households were highly correlated with the number of simultaneous deprivations that they face. This was particularly true for the geographical location of the household, household size, education level of the household head and the mother, religion, number of children in the household, labour-constrained household and the asset index (used as a proxy for wealth status). Looking at the geographical location of children (Figure 3), it was observed that the majority of urban children had no deprivations (23.4 per cent) or only one (46.5 per cent), while in rural areas only a very small proportion of children (5.8 per cent) had no deprivations, and most (68.7 per cent) had at least two simultaneous deprivations.

Figure 3: Percentage distribution of deprived children aged 0–17 years by area 5* 1.4 0.0

4* 8.3 1.0

3* 24.0 7.0 2* 35.0 22.0 deprivaons 1* 25.6 46.5

Number of simultaneousof Number 0* 5.8 23.4

40 30 20 10 10 20 30 40 50 60 Percentage of deprived children

Rural Urban

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 9 Results

Table 3 shows that Bulawayo and Harare provinces have the highest proportion of non-deprived children, with 28.1 per cent and 18.4 per cent, respectively. On the other hand, Masvingo Province was the worst-off province, with 11.4 per cent of children deprived on 4–5 dimensions simultaneously.

Table 3: Percentage distribution of children aged 0–17 years by province Number of simultaneous deprivations Province 0* 1* 2* 3* 4* 5* Manicaland 7.4 32.1 30.9 19.9 7.8 1.9 Mashonaland Central 8.3 27.5 32.8 22.5 8.0 1.0 Mashonaland East 9.6 34.1 31.9 17.8 6.2 0.5 Mashonaland West 8.7 31.6 32.0 21.4 5.6 0.6 Matabeleland North 5.6 20.3 37.8 27.8 7.6 0.9 Matabeleland South 7.1 28.1 36.2 21.2 6.7 0.7 Midlands 11.8 27.9 30.3 20.3 7.7 1.9 Masvingo 4.7 23.3 35.6 25.1 9.7 1.7 Harare 18.4 44.6 25.6 9.8 1.6 0.1 Bulawayo 28.1 45.4 20.3 5.6 0.5 0.0

The deprivation distribution analysis for other characteristics of households reveals that children living in households with an uneducated household head, in big households (more than six members), in households with more than two children and in households with labour constraints tend to be simultaneously deprived on a higher number of dimensions. The same result was observed for children belonging to households that practised traditional religion, households in the poorest asset quintile and rural poor households (see Table 4).

Table 4: Deprivation distribution for children aged 0–17 years by child’s characteristics

Profiling variables 0 1 2 3 4 5 Secondary or 13.9* 36.6* 29.9* 14.8* 4.1* 0.7* Educational higher level of household Primary education 6.4* 25.7* 33.5* 24.2* 8.8* 1.4* head No education 3.5* 18.2* 36.3* 30.8* 10.1* 1.2* More than 6 9.2* 27.0* 34.2* 21.0 7.3* 1.3* members Household size 4–6 members 10.5* 32.6* 30.5* 19.6 6.0* 0.9* 1–3 members 9.1* 0.2* 32.8* 19.7 7.2* 1.0* Above median 9.1* 29.8* 32.5* 20.5* 6.9* 1.1* No. of number of children children in household Below or equal to (median=2) median number of 11.6* 32.0* 30.7* 19.0* 5.8* 0.8* children

10 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 4: Deprivation distribution for children aged 0–17 years by child’s characteristics

Profiling variables 0 1 2 3 4 5 No religion 5.7* 24.2* 33.6* 25.9* 9.2* 1.4* Other religion 11.7* 25.7* 35.6* 22.1* 4.8* 0.0* Traditional 4.4* 20.9* 36.3* 26.9* 9.8* 1.8* Other Christian 8.0* 28.0* 31.3* 23.2* 7.5* 2.0* Religion Apostolic sect 6.5* 28.7* 33.4* 22.1* 8.1* 1.3* Pentecostal 15.4* 39.5* 28.7* 13.1* 3.1* 0.1* Protestant 18.0* 35.8* 28.1* 14.2* 3.4* 0.6* Roman Catholic 14.0* 35.8* 31.7* 14.7* 3.5* 0.3*

Labour constrained: 7.3* 27.6* 33.7* 22.2* 7.8* 1.5* Labour- dependency ratio constrained greater than 2 household Not labour 11.4* 32.3* 30.9* 18.8* 5.8* 0.8* constrained Q5 (Richest) 23.4* 38.1* 26.5* 9.8* 2.0* 0.1* Q4 14.7* 43.7* 27.6* 11.4* 2.3* 0.2* Asset index (proxy for Q3 8.5* 35.2* 33.9* 17.1* 4.5* 0.7* wealth) Q2 3.7* 25.7* 36.1* 25.5* 7.9* 1.1* Q1 (Poorest) 1.3* 14.8* 34.6* 32.7* 14.0* 2.6*

Area and Rural non-poor 11.9* 33.9* 33.9* 16.0* 3.9* 0.3* poverty status Urban non-poor 26.3* 47.1* 20.6* 5.5* 0.5* 0.0* (based on asset Rural poor 3.8* 23.0* 35.3* 26.5* 9.6* 1.7* index) of household Urban poor 9.4* 43.4* 29.2* 14.7* 3.1* 0.1*

In this analysis, a child was considered multidimensionally poor if he/she was deprived on at least two of the dimensions. Based on this definition, 59.6 per cent of Zimbabwean children were multidimensionally poor (Table 2). At the provincial level (Figure 4a), the provinces with the highest level of poor children were Matabeleland North Province (74.1 per cent) and Masvingo Province (72.1 per cent).

In Figure 4b, the multidimensional poverty indices were disaggregated for children living in urban and rural areas. Using the threshold K=2,10 it was observed that 8.1 per cent of urban children were deprived, compared to 33.7 per cent of rural children. Despite this large disparity in the headcount rate, the difference in the average intensity of deprivation between children in rural and urban areas was small, though it was significant at the 5 per cent level. This implies that the average number of deprivations facing multidimensionally poor children hovers around

10 The threshold K=2 implies that a child was considered multidimensionally poor when he/she was deprived on two or more of the dimensions of child well-being used in the study.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 11 Results

60 per cent for both rural and urban children. Thus, multidimensionally poor children in both urban and rural areas were almost equally worst-off. M0 is simply a summarized index aggregating the deprivation headcount and intensity. A higher M0 of 0.224 for children in rural areas (compared to 0.05 for urban children) indicates that child poverty was more severe in rural areas. Multidimensionally poor children are profiled by other characteristics in Figure 5.

Figure 4a: Multidimensional child poverty (K=2) for children aged 0–17 years by province

Mashonaland Central 64.3

Mashonaland West 59.6

Harare 37.1 Mashonaland East 56.3

Matabeleland North Midlands 74.1 Manicaland 60.3 60.5

Bulawayo 26.5

Masvingo Matabeleland South 72.1 64.8 Total

26.5 - 56.3 56.4 - 60.5 60.6 - 74.1 Missing Value l

Source: ZIMSTAT_MICS_2014

Figure 4b: Multidimensional child poverty (K=2) for children aged 0–17 years by area 80 0.3

66.5 62.6 60 0.224 0.2

40 33.7

0.1 20 0.05

8.1 headcount deprivaon Adjusted

Deprivaon headcount and intensity,% and headcount Deprivaon 0 0.0 Urban Rural

Deprivaon headcount (H) %* Average intensity across the deprived (A) on %* Deprivaon headcount adjusted for intensity (M0)*

12 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 5: Multidimensional headcount ratios (K=2) for children aged 0–17 years by child’s characteristics

Deprivaon Headcount Rao (%) 0% 25% 50% 75%

Naonal Naonal 28% female 26% Child's gender* male 29% female 28% Gender of hh head* male 27% Secondary or higher 20% Educaonal level Primary educaon of hh head* 34% No educaon 42% hhsize >6 30% Household size* hhsize 4-6 27% hhsize 1-3 28% No. of children in Above median number of children 29% hh (media n =2)* Below or equal to median number of children 26% no religion 37% other religion 27% tradional 39% other chrisans 33% Religion* apostolic sect 31% pentecostal 16% protestant 18% roman catholic 19% Labour Labour constraint: dependency rao>2 31% constraint hh* Hld not labour constraint 25% Q5 (Richest) 12% Q4 14% Asset Index (proxy for Wealth)* Q3 22% Q2 35% Q1 (Poorest) 49% rural non-poor 20% Area and poverty urban non-poor 6% status (based on asset index) of hh* rural poor 38% urban poor 18% female headed hh (non-poor) 13% Gender of hh head and poverty status male headed hh(non-poor) 13% (based on asset female headed hh(poor) 36% index) of hh* male headed hh(poor) 37%

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 13 Results

3.1.1.1 Deprivation overlap analysis

In Zimbabwe, the majority of children experience several deprivations at one time. The deprivation overlap analysis by dimension, illustrated in Figure 6, confirms this observation. The figure shows overlap of deprivations for each of the dimensions used in the analysis for all children.11 Of the 25 per cent of children deprived on the Health dimension, 2.5 per cent were deprived on only that dimension. Most were deprived on one (8 per cent) or two (9 per cent) other dimensions. Almost identical results were found for the Water dimension, with 2 per cent of children deprived on only that dimension. Ultimately, designing programmes to tackle only health or water issues for children will make a very small percentage of children non-deprived. For example, solving the problem of water on its own would make 2 per cent of children non-deprived; 12.9 per cent would suffer only one instead of two deprivations; 13.2 per cent would face only two instead of three deprivations; and so on. The best strategy to reduce child poverty would be to design multi-sectoral policies that solve several problems that face children concurrently.

The highest deprivation was observed for the dimension Sanitation, with 75.9 per cent; it also has the highest proportion of children (22.7 per cent) who are deprived in only one dimension. Addressing the problem of Sanitation (measured by toilet type, sharing of toilet facility and handwashing) will help the majority of children in Zimbabwe.

Figure 6: Deprivation overlap for children aged 0–17 years by dimension 80

7.3

60 18.5

40 27.4

6.5 Percentage of deprived children ofdeprived Percentage

5.6 13.2 20

8.9 22.7 12.9 7.7

0 2.5 2.0 Health Water Sanitaon

Deprived only on the speciﬁed dimension Deprived on one other dimension Deprived on 2 other dimensions Deprived on 3 or more other dimensions

11 The dimensions Nutrition, Physical development, Child development, Education and Information were not analysed here because they were not used for all the age groups used in this analysis. For example, the dimension Nutrition was used only for children aged 0–23 months. Estimating the deprivation overlap results for those aged 0–17 years would lead to an underestimation of the rates.

14 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Venn diagrams were used to analyse which of the deprivations were experienced together. This will help in the design of appropriate policies, as alluded to earlier. Figure 7 shows the deprivation overlap across the dimensions of Health, Water and Sanitation for children aged 0–17 years. Almost a quarter (24.7 per cent) of children were deprived on Health, 34.7 per cent on Water and 76 per cent on Sanitation. The children who experienced deprivation on all three dimensions simultaneously accounted for 7.7 per cent. Of the 34.7 per cent of children deprived on the Water dimension, 3.5 per cent of them were deprived on that dimension only. The majority faced Health and/or Sanitation problems.

Figure 7: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation

Disaggregating the results by urban/rural and poor/non-poor households (wealth was measured by the assets owned by the households) in Figures 8 and 9, we can see that a higher proportion of children living in rural and poor households experienced all three deprivations at the same time. In urban and non-poor households, it seems that the majority of children tend to suffer only one of the deprivations analysed in the Venn diagram. Urban areas had a higher percentage of non-deprived children than did rural areas. Similarly, the percentage of children living in non-poor households who were non-deprived accounted for 25.3 per cent, compared to 6.8 per cent living in poor households.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 15 Results

Figure 8: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation by area

Urban Children

Rural Children

16 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 9: Deprivation overlap for children aged 0–17 years for the dimensions Health, Water and Sanitation by poverty status of households

Children living in non-poor households

Children living in poor households

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 17 Results

3.1.2 Overlap between child poverty based on wealth status of the household and multidimensional poverty

This section examines the relationship between child poverty based on the wealth status of the household and multidimensional deprivation. The wealth status of a household was determined by the assets possessed by the household – a list of assets which was representative of wealth in Zimbabwe was determined for urban and rural areas (see Annex 3 for the list of assets). The households were then classified into wealth quintiles. Children who belong to households in the three lower wealth quintiles were deemed poor. On the other hand, a child was multidimensionally poor if he or she was deprived on at least two dimensions of well-being. The results, presented in Figure 10 and Table 5, confirm that poverty based on wealth status and deprivation were two distinct phenomena. Child poverty based on the two concepts does not completely overlap – 45 per cent of children were poor based both on the wealth status of their household and on the deprivation level; 18 per cent were poor in terms of their wealth status but were non-deprived; while 15 per cent were non-poor and yet deprived.

Figure 10: Deprivation overlap for children aged 0–17 years based on wealth status of households and multidimensional poverty

Non-poor hh based on wealth status and muldimensional poverty (23%)

Poor hh - based only on wealth Mul-dimensionally status (18%) poor only hh (15%)

Overlap between poverty based on wealth status of hh and muldimensional poverty (45%)

Although the two concepts used yield more or less similar percentages at the national level, there was a clear-cut difference when the results were disaggregated at the urban and rural level. While 17 per cent of children living in urban areas were poor based on the wealth status of their household, 30.1 per cent were multidimensionally poor. In rural areas, on the other hand, there was a higher proportion of poor children based on wealth status (76.2 per cent) than on multidimensional poverty (68.6 per cent).

18 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 5: Overlap between child poverty for children aged 0–17 years based on wealth status of households and multidimensional poverty by area

Child poverty based on Multidimensional Area wealth status of households child poverty (K=2)

Urban 17 30.1

Rural 76.2 68.6

Total 61.4 59.6

The gap between multidimensional deprivation and poverty based on wealth status also varies at the provincial level (Figure 11). The variations observed could be explained by the conceptual difference between the two types of poverty. Asset poverty, used as a proxy for monetary poverty in this case, determines whether a household can, in principle, afford to buy the goods and services necessary to ensure their livelihood if these goods and services are readily available to the household.

The analysis of deprivation, on the other hand, deals with real household access to selected aspects of well-being, for example the quality of drinking water, the type of toilet, school attendance and the nutritional balance. If this access is conditioned by the financial capacity of the household, its effectiveness also depends on the available supply and on the desire of the household to obtain those goods and services. Indeed, many of the needs of children are public or semi-public in nature (Education, Water and Sanitation), and access to basic services can be denied for socio-economic, geographical or religious reasons, through lack of information, etc.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 19 Results

Figure 11a: Child poverty for children aged 0–17 years based on wealth status of households by province

Mashonaland Central Mashonaland West 73.9 65.3

Harare 22.1 Mashonaland East 63.8

Matabeleland North Midlands 85.3 Manicaland 63.5 67.9

Bulawayo 13.9

Masvingo Matabeleland South 73.5 62.2 Total

13.9 - 50.0 50.1 - 70.0 70.1 - 85.3 Missing Value l

Source: ZIMSTAT_MICS_2014

20 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 11b: Child poverty for children aged 0–17 years based on multidimensional poverty (K=2) by province

Mashonaland Central 64.3 Mashonaland West 59.6

Harare 37.1 Mashonaland East 56.3

Matabeleland North Midlands 74.1 60.3 Manicaland 60.5

Bulawayo 26.5

Masvingo 72.1

Total Matabeleland South 64.8 26.5 - 50.0 50.1 - 60.5 60.6 - 74.1 Missing Value l

Source: ZIMSTAT_MICS_2014

Therefore, if access to goods and services is dictated by financial means, monetary poverty would be equal to deprivation. But in reality, there were monetarily poor households that did not necessarily have deprived children, because of the availability of free goods and services (social protection schemes, for example the Basic Education Assistance Module (BEAM), cash transfer programmes, free immunization and food for work), or else it could be that most of the household income was invested in the children. Similarly, there may be monetarily non-poor households with deprived children, simply because there is no supply of a particular good or service in the region, or else religion might play a role in not allowing the child access to those goods and services. While the multidimensional poverty rate was higher than the poverty rate based on the wealth status of households for children living in Bulawayo, Harare and Matabeleland South provinces, the opposite was the case in the remaining provinces.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 21 Results

3.2 Deprivation among children aged 0–23 months

This section concentrates on children aged 0–23 months.

Box 1: Key results for children aged 0–23 months

l Some 70.8% of children aged 0–23 months were multidimensionally poor, that is, they were deprived on at least two dimensions of child well-being.

l Very high deprivation rates were observed for the dimensions Nutrition (54.7 per cent), Health (58.7 per cent) and Sanitation (45.4 per cent).

l The deprivation rates for children were higher in urban than in rural areas.

l Masvingo Province had the highest adjusted deprivation headcount rate (M0), with an index of 0.52, while Bulawayo Province had the lowest (0.066).

l In both urban and rural areas, the Health and Nutrition dimensions contributed most to the deprivation level. In rural areas, the dimension Water contributed more towards the level of deprivation than in urban areas.

3.2.1 Single deprivation analysis

This section presents the deprivation rates for each indicator and dimension that was selected for this analysis (see Figures 12a and 12b). The results give an indication of the policy areas that need particular attention for children in different age groups.

Figure 12a: Percentage distribution of deprivation rates for children aged 0–23 months by indicator and dimension

0102030405060

Nutrion Infant & Young Child Feeding 54.7 Vaccinaon 26.4 Birth assistance 11.6 Health Mother's knowledge on MTCT 30 Cooking fuel 11.5 Stunng 25.5 Physical Wasng 5.22 Development Underweight 12.2

Indicators ofdeprivaon Indicators Water source 21.4 Water Distance to water source 16.9 Sanitaon Handwashing 45.4

Proporon of children deprived as percentage of children aged 0 to 23 months, %

22 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 12b: Percentage distribution of deprivation rates for children aged 0–23 months by dimension

58.7 60 54.7

50 45.4 40 32.3 29.7 30

Percentage 20

0 Nutrion Health Physical Water Sanitaon

Dimensions of deprivaon

For children aged 0–23 months, the dimensions Health, Nutrition and Sanitation had relatively high deprivation rates of 58.7 per cent, 54.7 per cent and 45.4 per cent, respectively. The high deprivation rate for the dimension Nutrition was driven by the indicator infant and young child feeding, measured by exclusive breastfeeding for children below 6 months and meal frequency and diversity, as required by the WHO. The World Food Programme (WFP) asserts that in Zimbabwe dietary diversity is generally poor and consumption of protein is inadequate (WFP, 2015). Handwashing, under the dimension Sanitation, has the second-highest level of deprivation among the indicators, with 45.4 per cent of Zimbabwean children aged 0–23 months not having water and soap available in their household for handwashing. Several studies (e.g. Curtis and Cairncross, 2003) found that handwashing with soap can reduce the risk of diarrhoea by more than 40 per cent. Trials in Zimbabwe itself demonstrated that handwashing with soap is more effective than handwashing with only water in reducing faecal bacteria on hands (Kaltenthaler et al., 1991). Diarrhoea was the fourth-highest cause of death among children under 5, and was the reason for 12 per cent of child hospital admissions.12 Around 32 per cent of children were deprived on the Water dimension.

Deprivation rates for each indicator constituting the Health dimension were around 25 per cent. However, using the union approach, 58.7 per cent of children in this age group were deprived on this dimension. In Zimbabwe, cooking fuel was considered an important indicator under the Health dimension. The results reveal that more than one child in ten below 24 months lived in a household where solid fuels were used for cooking food inside the house. In relation to full immunization, about one child in four was not immunized at all or not immunized on time. With regards to skilled birth attendance, one child in ten aged 0–23 months was born with the assistance of a relative or a friend (see below for provincial disparity on the Health dimension). The indicator with the highest deprivation rate under the Health dimension was MTCT of HIV and AIDS, with 30 per cent of mothers lacking knowledge of all the possible channels of transmission (pregnancy, delivery and breastfeeding).

12 USAID [online]. Article available at http://www.mchip.net/node/2941

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 23 Results

Almost one child in three was deprived on the Physical development dimension. Some 25 per cent and 12 per cent of children aged 0–23 months were stunted and underweight, respectively. These poor physical development indicators will manifest themselves in bigger health problems for the children in the future.

The above results reflect aggregated percentages for the whole country. In an attempt to facilitate identification of the most vulnerable children and to ease policy making, the analysis will profile the children based on their geographical location, individual characteristics, household characteristics and the characteristics of the mother. In the following sections, the profile of deprived children is determined.

3.2.1.1 Deprivation rates based on geographical location of the child

The deprivation levels on all the dimensions was significantly higher for children in rural than in urban areas, as shown in Figure 13. More significant disparities – of around 39 and 17 percentage points between urban and rural areas – were seen on the Water and the Health dimensions, respectively.

Figure 13: Deprivation rates for children aged 0–23 months by area

Physical*

33.2

Water* 20.7 63.5 Health* 43.5 46.8 4.2

37.4 48.7

48.6 57.1 Sanitaon* Nutrion*

Urban Rural

Source: ZIMSTAT_MICS_2014

24 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 14a–e shows the deprivation rates disaggregated at the provincial level. Deprivation rates for Nutrition were above 40 per cent in all provinces. Very high percentages of deprived children were observed in Masvingo (64.3 per cent), Manicaland (62.8 per cent) and Mashonaland Central (62.7 per cent).

The highest deprivation rates were noted for the dimension Health, with deprivation rates ranging from 42.9 per cent in Bulawayo Province to 67.2 per cent in Matabeleland South. The differences for each of the indicators constituting the dimension were analysed. Results for the indicators cooking fuel and skilled birth attendance were notable. In the metropolitan provinces of Harare and Bulawayo, around 1 per cent of children in this age group lived in a household where solid fuel was used for cooking in the household. In contrast, the percentage was around 20 per cent in predominantly rural provinces, such as Masvingo, Matabeleland South and Mashonaland Central. The provincial disparity was also highly significant for the indicator skilled birth attendance, with 15 per cent of children from rural areas having been assisted by a friend or a relative, whereas the deprivation rate was 3.3 per cent in urban areas. Mashonaland Central was the most deprived province regarding skilled birth attendance (19.2 per cent). The deprivation level for the indicator full immunisation13 was also higher in rural areas (28.3 per cent) than urban areas (21.7 per cent). Children from Masvingo and Midlands provinces had the highest deprivation levels (32 per cent) with regards to vaccination, compared to Bulawayo (60.4 per cent) and Matabeleland North (19 per cent) provinces.

The deprivation for Physical development was relatively high for Mashonaland Central, Manicaland and Masvingo provinces, as shown in Figure 14c. With regards to Water, Matabeleland North (48.2 per cent), Mashonaland South (47.3 per cent) and Masvingo (47.3 per cent) provinces yielded the highest deprivation rates. Masvingo Province had the highest proportion of children deprived on the Sanitation dimension (63.8 per cent).

13 A child is deprived if they missed at least one vaccination or did not receive it on time.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 25 Results

Figure 14a: Deprivation rates for children aged 0–23 months for Nutrition by province

Mashonaland Central 62.7 Mashonaland West 56.8

Harare 49.5 Mashonaland East 51.3

Matabeleland North Midlands 44.9 Manicaland 52.5 62.8

Bulawayo 46.0

Masvingo Matabeleland South 64.3 51.3 Total

44.9 - 49.5 49.6 - 52.5 52.6 - 64.3 Missing Value l

Source: ZIMSTAT_MICS_2014

26 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 14b: Deprivation rates for children aged 0–23 months for Health by province

Mashonaland Central 66.6 Mashonaland West 57.0

Harare 49.9

Mashonaland East 62.4 Matabeleland North Midlands 47.5 Manicaland 63.9 63.3

Bulawayo 42.9

Masvingo Matabeleland South 63.3 47.2 Total

42.9 - 49.9 50.0 - 63.3 63.4 - 66.6 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 27 Results

Figure 14c: Deprivation rates for children aged 0–23 months for P Physical development by province

Mashonaland Central 37.1 Mashonaland West 31.8

Harare 20.8

Mashonaland East 25.7 Matabeleland North Midlands 29.3 Manicaland 28.2 36.7

Bulawayo 20.0

Masvingo Matabeleland South 35.4 29.1 Total

20.0 - 25.7 25.8 - 29.3 29.4 - 37.1 Missing Value l

Source: ZIMSTAT_MICS_2014

28 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 14d: Deprivation rates for children aged 0–23 months for Water by province

Mashonaland Central 47.3 Mashonaland West 38.1

Harare 8.6 Mashonaland East 19.8

Matabeleland North Midlands 48.2 Manicaland 41.5 30.0

Bulawayo 5.0

Masvingo 47.3

Total Matabeleland South 39.6 5.0 - 19.8 19.9 - 39.6 39.7 - 48.2 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 29 Results

Figure 14e: Deprivation rates for children aged 0–23 months for Sanitation by province

Mashonaland Central Mashonaland West 35.4 40.9

Harare 39.6

Mashonaland East 38.3 Matabeleland North 54.9 Midlands 37.6

Manicaland 59.1

Bulawayo 29.2

Masvingo 63.8

Total Matabeleland South 29.2 - 37.0 44.4 37.1 - 49.9 50.0 - 63.8 Missing Value l

Source: ZIMSTAT_MICS_2014

3.2.1.2 Deprivation rates based on individual characteristics of the child

Figures 15a and 15b show that the percentage of unregistered (i.e. without a birth certificate and health card) young children aged 0–23 months who were deprived was significantly higher on all dimensions. The highest disparity was seen on the dimension Health, where 97.2 per cent of children without a health card were deprived, compared to 56.8 per cent of children with a health card. It should be noted that only 5 per cent of children in this age group did not have a health card and almost 80 per cent of those children belonged to the Apostolic sect. Some 23 per cent of children did not have a birth certificate, and they appeared to be more deprived across the dimensions analysed.

30 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 15a: Percentage distribution of deprivation rates for children aged 0–23 months with/without a health card

Sanitaon* 56.9 44.8

44.3 Water* 31.7

Physical* 41.7 29.1

Health* 97.2 56.8 Dimensions of deprivaon of Dimensions

64.2 Nutrion* 54.2

010 20 30 40 50 60 70 80 90 100 Percentage of deprived children

Child does not have a health card Child has a health card

Figure 15b: Percentage distribution of deprivation rates for children aged 0–23 months with/without a birth certificate

Sanitaon* 47.1 39.6

Water* 36.2 19.3

Physical* 31.9 22.1

61.4 Health* 49.5 Dimensions of deprivaonof Dimensions 55.7 Nutrion* 51.2

010 20 30 40 50 60 70 Percentage of deprived children

Child does not have a birth cerﬁcate Child has a birth cerﬁcate

Table 6 shows the deprivation rates disaggregated by other characteristics of the child. An important observation is that the sex of the child does not seem to matter when it comes to deprivation rates on any dimension, with the exception of Physical development: there deprivation rates were higher for boys (33.8 per cent) than girls (25.7 per cent). Households with less-educated mothers and less- educated household heads14 have a higher proportion of deprived children. A similar pattern was observed for households with many members (more than six), labour-constrained households and poor households.

14 Less educated refers to Primary or no education.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 31 Results

Table 6: Deprivation rates for children aged 0–23 months by dimension and child’s characteristics

Physical Profiling variable Nutrition Health develop- Water Sanitation ment

Female 54.1 58.4 25.7* 31.2 45.3 Child’s sex Male 55.3 59.1 33.8* 33.5 45.5

Secondary or higher 51.1* 54.4* 25.9* 25.0* 42.7*

Mother’s Primary education 61.1* 66.8* 36.9* 46.8* 50.6* education No education 72.3* 77.5* 40.8* 49.2* 53.5*

Secondary or higher 53.0* 54.4* 25.8* 24.7* 41.6* Education level of Primary education 57.0* 64.8* 34.6* 43.9* 51.4* household head No education 57.7* 67.0* 40.9* 41.9* 50.3*

More than 6 58.0* 64.8* 32.9* 40.1* 49.9* members

Household 4–6 members 52.5* 56.7* 28.5* 30.3* 44.1* size 1–3 members 56.5* 56.4* 28.4* 27.4* 43.1*

Above median 55.7 60.7* 31.0* 36.8* 47.8* Number of number of children children in household Below or equal to (median=2) median number of 53.3 56.0* 27.9* 26.1* 42.2* children

Labour constrained: Labour- dependency ratio 59.3* 66.4* 33.2* 41.2* 51.0* constrained greater than 2 household Household not 53.3* 56.4* 28.6* 29.7* 43.7* labour constrained

Q5 (Richest) 46.4* 51.8* 20.0* 18.4* 35.2*

Q4 54.1* 53.1* 23.2* 13.8* 42.1* Asset index (proxy for Q3 58.2* 56.1* 30.4* 22.9* 44.6* household wealth) Q2 55.2* 64.3* 33.2* 42.6* 46.1*

Q1 (Poorest) 58.9* 66.7* 39.7* 58.5* 56.8*

32 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.2.2 Multiple Overlapping Deprivation Analysis

3.2.2.1 Deprivation overlap analysis

Deprivation overlap analysis allows for better insight into the nature and severity of children’s deprivations, by highlighting whether sectoral deprivations are individual problems or overlap with other deprivations and need to be addressed in combination.

Figure 16 shows that a very small proportion of children were deprived on only one dimension, while a high proportion of children were deprived on two or more dimensions concurrently. The figure also depicts the percentage of children who would be better off if deprivation on a dimension was eradicated in Zimbabwe. For example, tackling the Nutrition problem would result in 6.1 per cent of children experiencing no deprivation at all, while 17.2 per cent of children would experience only one instead of two deprivations.

Figure 16: Deprivation overlap for children aged 0–23 months for each dimension 80

13.5 13.3

40 12.6 19.0 17.6

11.1 15.2 10.4

Percentage of deprived children deprivedof Percentage 20 18.9 12.3 17.2 9.7 12.9

7.3 7.0 7.3 6.1 4.8 0 1.7 2.0 Nutrion Health Physical Water Sanitaon

Deprived only on the speciﬁed dimension Deprived on one other dimension Deprived on 2 other dimensions Deprived on 3 or more other dimensions

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 33 Results

Deprivation overlap of three dimensions

This section summarizes the results for the proportion of children deprived on any three dimensions at one time. While all combinations of dimensions were analysed, only the most telling results are presented in this report. Overlaps between three dimensions are better represented with the use of Venn diagrams to provide insight into how deprivations relate to each other. A Venn diagram of any combination of three dimensions describes the deprivation levels for: (1) each dimension separately; (2) overlap between two dimensions; (3) overlap between three dimensions; (4) the population not deprived on any of the three dimensions.

The Venn diagram of the deprivation rates on the Nutrition, Health and Sanitation dimensions has the largest proportion of overlap (16.2 per cent) of all possible combinations of three dimensions. Of the 54.3 per cent of children deprived on the dimension Nutrition, only 11.3 per cent were deprived on Nutrition only, while the remaining 43 per cent were also deprived on the Health and/or Sanitation dimension at the same time. It was interesting to see the relatively large overlap of 17.6 per cent between the dimensions of Nutrition and Health. Such an overlap can lead to policy recommendations that target children deprived on both Nutrition and Health at one go. It should be noted that 11.9 per cent of children in this age group were not deprived on any of the dimensions Nutrition, Health or Sanitation.

With regards to the overlap between Nutrition, Health and Physical development, it is demonstrated in Figure 17 that the percentage of children deprived on all three dimensions was 11.9 per cent. Since physical development at this very tender age is likely to have an impact on the whole lifetime of the child, particular emphasis is laid on this dimension. Of the 29.1 per cent of children who were deprived on Physical development, just 5.1 per cent were deprived on Physical development only.

34 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 17: Deprivation overlap for children aged 0–23 months

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 35 Results

Since Nutrition, Health and Sanitation have the highest percentage of overlap, further analysis was carried out by disaggregating the results at the urban and rural level, as shown in Figure 18.

Figure 18: Deprivation overlap for children aged 0–23 months by area

Rural Areas

Urban Areas

36 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

In relation to the overlaps on the three dimensions, there was a notable difference between rural (19.4 per cent) and urban (8.2 per cent) areas. It seems that in rural areas, children in this age group generally tend to be deprived on two or three dimensions simultaneously. The percentage of children deprived on Nutrition only was 5 percentage points lower in rural than in urban areas; on the Sanitation dimension only the figure was 3.7 percentage points lower in rural than in urban areas. The percentage of non-deprived children was also higher in urban areas (17.1 per cent, as compared to 9.8 per cent in rural areas).

Deprivations experienced by children at national and subnational levels

The crux of the MODA methodology, as highlighted in the methodology section, lies in the overlapping of deprivations on different dimensions. Children deprived on several dimensions at a time are more vulnerable. In this section, the number of simultaneous deprivations experienced by each child are analysed.

As Figure 19a shows, 7.4 per cent of children in the age group 0–23 months were not deprived on any of the dimensions under study. A large proportion of children (78 per cent) were deprived on 1–3 dimensions. However, almost a third of all children were deprived on two dimensions simultaneously. There were also children who were deprived on five dimensions (2.9 per cent) of well-being simultaneously. These most deprived children are unlikely to survive or develop into healthy adults.

Figure 19b shows that the majority of children living in urban areas were deprived in one and two dimensions, whereas in rural areas most of the children tended to be deprived in two and three dimensions simultaneously. Although there were urban children deprived on many dimensions at the same time, this percentage was much more for children living in rural areas. The deprivation distribution by province in Table 7 shows that Bulawayo Province had the highest proportion of non-deprived children (18.7 per cent), while Masvingo Province had 6.4 per cent of its children deprived on all five dimensions at the same time.

Figure 19a: Percentage distribution of children aged 0–23 months deprived on each number of deprivations

35 31.6 30 24.6 25 21.8 20

15 11.6 Percentage Percentage 10 7.4 5 2.9 0 012345 Number of simultaneous deprivaons

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 37 Results

Figure 19b: Percentage distribution of children aged 0–23 months deprived on each number of deprivations by area

40 34.7 33.9 30.8 30 28.5

20 16.7 14.1 14.9 15.3

Percentage 10 4.8 2.3 4 0.2 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Urban Rural

Table 7: Percentage distribution of children aged 0–23 months deprived on each number of deprivations by province

Number of simultaneous deprivations experienced Province by the child 0 1 2 3 4 5

Manicaland 4.6* 15.0* 32.8* 27.2* 14.8* 5.7*

Mashonaland Central 4.0* 17.5* 27.8* 31.1* 16.7* 2.8*

Mashonaland East 9.4* 25.7* 34.4* 21.2* 8.3* 1.0*

Mashonaland West 5.4* 20.3* 35.9* 24.7* 11.4* 2.3*

Matabeleland North 7.5* 18.9* 32.5* 27.2* 10.7* 3.3*

Matabeleland South 5.5* 23.6* 27.5* 24.1* 17.1* 2.2*

Midlands 6.8* 20.7* 32.5* 26.4* 10.2* 3.4*

Masvingo 3.0* 13.7* 23.8* 32.3* 20.8* 6.4*

Harare 12.6* 33.3* 31.8* 19.1* 3.2* 0.0*

Bulawayo 18.7* 35.2* 35.3* 10.0* 0.8* 0.0*

In order to profile the most deprived children for efficient targeting, in-depth analysis of the dimension count was again carried out for some of their characteristics (Table 8).

38 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 8: Deprivation distribution for children aged 0–23 months by child’s characteristics Number of simultaneous deprivations experienced by the child Profiling variable 0 1 2 3 4 5

Secondary or 9.7* 25.4* 33.8* 21.5* 8.0* 1.8* higher Mother’s education Primary education 2.9* 15.2* 27.6* 31.1* 18.7* 4.6*

No education 4.8* 8.1* 26.0* 26.8* 19.0* 15.3*

Secondary or higher 9.8* 25.2* 33.2* 21.9* 7.9* 1.9* Education level of household Primary education 4.1* 16.4* 30.0* 28.2* 17.2* 4.1* head No education 2.7* 19.2* 25.1* 30.8* 15.4* 6.8*

Household size 5.4* 16.3* 30.7 27.6* 15.5* 4.5* greater than 6

Household Household 8.4* 23.7* 31 24.3* 10.4* 2.1* size size 4–6

Household 7.2* 24.2* 34.9 21.4* 9.3* 3.0* size 1–3

Above median 7.26 19.0* 29.9* 26.7* 13.8* 3.5* Number of number of children children in household Below or equal to (median=2) median number of 7.6 25.8* 34.1* 21.9* 8.5* 2.1* children

Possession Birth certificate 12.1* 29.5* 32.7 17.9* 6.0* 1.7* of birth certificate No birth certificate 6.0* 19.6* 31.3 26.6* 13.2* 3.3*

Labour constrained: dependency ratio 4.2* 17.1* 29.7 27.1* 16.9* 5.0* Labour- greater than 2 constrained household Household not 8.4* 23.2* 32.2 23.9* 10.0* 2.3* labour constrained

Q5 (Richest) 13.7* 30.4* 32.0* 19.2* 4.2* 0.4*

Q4 10.4* 28.1* 34.7* 19.7* 6.6* 0.6* Asset index Q3 6.4* 24.6* 34.2* 23.1* 9.4* 2.2* (proxy for wealth) Q2 5.2* 17.8* 30.8* 28.0* 14.3* 4.0*

Q1 (Poorest) 2.3* 10.5* 27.4* 31.8* 21.3* 6.6*

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 39 Results

3.2.2.2 Multidimensional deprivation indices

The multidimensional deprivation indices in Table 9 complement the analysis of the distribution and overlap of deprivations, by indicating the overall incidence and intensity of deprivation among children in a specific age group. As previously explained, the deprivation headcount (H) reports the percentage of children who were deprived according to a specified cut-off point (K) of multidimensional deprivations. The average intensity among the deprived (A) indicates the depth or severity of deprivations in children, according to the deprivation cut-off. For example, when the cut-off point was K=2 (2–5 deprivations), 70.8 per cent of children aged 0–23 months were identified as multidimensionally deprived. For all children with at least two deprivations, the average number of deprivations experienced was 2.8 per child, or 56 per cent of all possible deprivations. For this age group (as well as other age groups used in this report), a child was deemed multidimensionally poor if the child was deprived on at least two dimensions.

Table 9: Multidimensional deprivation indices for children aged 0–23 months at the national level

Average Average Deprivation Number of Deprivation intensity across intensity across headcount deprivations headcount the deprived the deprived adjusted for (dimensions) (H) % (A) in % (A) intensity (M0)

1–5 92.6 47.5 2.37 0.44

2–5 70.8 56 2.8 0.396

3–5 39.1 68.9 3.44 0.27

4–5 14.5 84 4.2 0.122

5 2.91 100 5 0.029

Figure 20 shows how the deprivation headcount rate changes at the national level and for rural and urban areas when a different cut-off point is selected. From the graph, it can be seen that in rural areas, the deprivation headcount rate was higher than in urban areas. For example, 47.8 per cent of children in rural areas were deprived on 3–5 dimensions, compared to 17.3 per cent in urban areas.

40 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 20: Multidimensional deprivation headcount for children aged 0–23 months by various cut-off points

100 95.2 90 92.6 85.9 80 78.5 70 70.8 60 51.2 50 47.8 40 Percentage 39.1 30 20 17.3 19.3 14.5 4 10 2.9 0 2.4 0.2 1-5 dimensions 2-5 dimensions 3-5 dimensions 4-5 dimensions 5 dimensions

Urban Rural Naonal

Multidimensional deprivation indices (K=2) for children aged 0–23 months by province

Figure 21 below includes all three multidimensional deprivation indices by province, and shows that Masvingo Province had the highest prevalence of deprivation when a cut-off point of at least two deprivations was used (83.3 per cent). Since the average intensity of deprivation of 62.4 per cent was high in Masvingo Province, the headcount ratio adjusted for the intensity of deprivation resulted in the highest value within the country (0.52). Manicaland and Mashonaland Central provinces also had high proportions of multidimensionally deprived children (80.5 per cent and 78.5 per cent, respectively). The average intensity of deprivation was similar for most provinces for multidimensionally poor children. This implies that the children experience almost the same number of deprivations, irrespective of the province in which they live. Bulawayo Province was relatively better off, with the lowest headcount deprivation rate, average intensity of deprivation and thus adjusted headcount ratio.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 41 Results

Figure 21: Multidimensional deprivation indices (K=2) for children aged 0–23 months by province 100 0.6

80 0.5 0.4 60 0.3 40 0.2 and intensity, % intensity, and 20 Deprivaon headcount Deprivaon 0.1

0 0.0 deprivaon headcountAdjusted

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

In Figure 22, multidimensionally poor children are profiled by some of their characteristics.

42 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 22: Multidimensional deprivation headcount (K=2) for children aged 0–23 months by child’s characteristics Percentage of children deprived in 2-5 dimensions 0% 25% 50% 75% 100%

Naonal Naonal 71% Secondary or higher 65% Mother's Primary educaon 82% educaon* No educaon 87% Secondary or higher 65% Educaonal level Primary educaon 80% of hh head* No educaon 78% hhsize >6 78% Household size* hhsize 4-6 68% hhsize 1-3 69% no religion 77% other religion 75% tradional 82% other chrisans 79% Religion* apostolic sect 75% pentecostal 60% protestant 59% roman catholic 61% No child <5 has died in the last 5 years 70% Infant mortality* At least one child <5 has died in the last 5 years 76% Possession of Child has a health card 70% health card* No health card 89% Possession of Child has a birth cerficate 58% birth cerficate* No birth cerficate 74% No. of children in Above median number of children 74% hh (media n =2)* Below or equal to median number of children 67% Labour Labour constraint: dependency rao>2 79% constraint hh* Hld not labour constraint 68% Q5 (Richest) 56% Q4 62% Asset Index (proxy for Wealth)* Q3 69% Q2 77% Q1 (Poorest) 87% rural non-poor 72% Area and poverty urban non-poor 49% status (based on rural poor 80% asset index) of hh* urban poor 59% female headed hh (non-poor) Gender of hh head 55% and poverty status male headed hh(non-poor) 60% (based on asset female headed hh(poor) 79% index) of hh* male headed hh(poor) 78%

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 43 Results

Decomposition of the multidimensional deprivation ratios

The decomposition of the adjusted deprivation headcount allows us to unmask the contribution of each of the dimensions to the overall deprivation rate, without losing the aspects of prevalence and intensity of the general measure. Figure 23 shows that the adjusted deprivation headcount rate of children aged 0–23 months was 0.396 at the national level. It shows a clear difference between children living in urban provinces (0.245) and in rural provinces (0.456), suggesting that the prevalence and depth of deprivation was lower among children in the former category.

Figure 23: Adjusted deprivation headcount rate (K=2) for children aged 0–23 months by area

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5

Urban 0.245

Rural 0.456

Naonal 0.396

At the national level, the main dimensions that contribute to the adjusted headcount deprivation rate are deprivation on the Health, Nutrition and Sanitation dimensions for children (26 per cent, 24.3 per cent and 20.5 per cent, respectively). In rural areas, Water contributed much more to the deprivation level than in urban areas (18 per cent, compared to 3 per cent). The contribution of Nutrition was significantly lower for children in rural than in urban areas (23.1 per cent, compared to 30.3 per cent) as shown in Figure 24.

Figure 24: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 0–23 months by area 100

23.4 19.9 20.5

75 3 18 15.3 14.5

13.8 13.7 50 28.8 26 25.3 Percentage 25

30.3 23.1 24.3

0 nabrU laruR lanoitaN

Nutrion Health Physical Development Water Sanitaon

44 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.3 Deprivation among children aged 24–59 months

This section concentrates on children aged between 24 and 59 months, and is similar to the above section for children aged 0–23 months.

Box 2: Key findings for children aged 24–59 months

l Some 83.4 per cent of children aged 24–59 months were multidimensionally poor: that is, they were deprived on at least two dimensions of child well-being.

l The dimensions with the highest deprivation rates were Sanitation (82.8 per cent), Child development (52 per cent) and Health (48.9 per cent).

l Some 75.6 per cent of children aged 36–59 months do not attend ECD programmes.

l In all, 17.6 per cent of children in rural areas (as compared to 9.5 per cent of children in urban areas) were deprived on the Health, Physical development and Sanitation dimensions simultaneously.

l Masvingo Province had the highest proportion of multidimensionally poor children (90 per cent), while Bulawayo had the lowest (66.6 per cent). The intensity of deprivation among deprived children (A) was also highest in Masvingo Province (60.8 per cent). Matabeleland South and Midlands provinces were the second and third worst-off provinces, with an M0 of 0.401 and 0.392, respectively.

l The Water dimension contributed 17.2 per cent to the level of deprivation in rural areas, while its contribution was 2.6 per cent in urban areas. Conversely, child development contributed relatively more to the level of deprivation in urban (24.8 per cent) than in rural areas (19.9 per cent).

3.3.1 Single deprivation analysis

Within the age group 24–59 months, it is of paramount importance to ensure that a child benefits from adequate health care. The Health dimension comprises the indicators full immunization and use of cooking fuel inside the house. It was observed that nearly half of children (48.9 per cent) were deprived on the Health dimension. About 42 per cent of children had missed at least one of the required vaccinations or did not get it on time; this figure was higher than in the younger age group, 0–23 months (26.4 per cent).

The education sector in Zimbabwe recognizes that early childhood development (ECD) education can contribute significantly to nurturing young children at various levels, e.g. physical, social, emotional, intellectual, cultural and spiritual. In 2004, the Government created a national ECD policy, which requires primary schools

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 45 Results

to offer a minimum of two ECD classes for children from 3 to 5 years of age.15,16 In 2013, of the 5,805 primary schools in the country, 5,670 had ECD classes.

For children in this age group, Child development is particularly important. Studies (Heckman and Carneiro, 2003) have shown that it is between the ages of 0 and 4 that children start developing their mental abilities and mental growth. In fact, Heckman and Carneiro (2003) highlighted the fact that early education is far more effective and promises greater returns than does later remedial training.

Figure 25a–b shows a high deprivation rate for the indicator early childhood education attendance (75.6 per cent) under the dimension Child development.17 In addition to ECD programmes, child–adult interaction and availability of toys or books can boost the level of the child’s imagination, creativity and intellectual capacity. Around 10 per cent of children were deprived in child–adult interaction. This implies that in 10 per cent of cases, one adult in the household did not do any of the following activities with the child: read books, tell stories, sing songs, take the child outside, play with the child or name/count/draw things together. Only 2 per cent of children were deprived for the indicator availability of toys or books. Almost a third of children (32.3 per cent) were deprived on the Physical development dimension, and this was mostly due to stunted children.

The dimension Sanitation had the highest deprivation rate, of 82.6 per cent. Toilet type and sharing of toilet facility were also added as indicators under the dimension Sanitation, alongside the indicator handwashing, since it is usually after the age of 24 months that children start making use of toilet facilities. The health consequences of lacking access to water and sanitation were severe, and particularly important for the growth of the child (Roushdy et al. 2012; Ezeh et al., 2014). The impact of unimproved water and sanitation as a leading cause of childhood diarrhoea has long been recognized and documented in the public health literature. Using a combination of quantitative and qualitative data, Roushdy et al. (2012) found that proper water supply and adequate sanitation have a positive effect on the incidence of childhood diarrhoea among children under 5 years of age in Egypt. Similarly, it was found by Ezeh et al. (2014), using Cox regression analysis, that improved water source and sanitation reduces neonatal, post-neonatal and child mortality significantly in Nigeria.

More than one child in three was deprived on the Water dimension. The distance to the water source is important to the well-being of the child. Around 19.3 per cent of children aged between 24 and 59 months were deprived in the indicator distance to water source. A recent study of 26 countries in sub-Saharan Africa found that children under the age of 5 have increased risk of diarrhoea, mortality and stunting, the further the water source is from the home (Pickering and Davis, 2012). The authors hypothesize that the association between healthier children and a closer water source was due to mothers having more time available for other activities, such as caring for their children, seeking health care services and generating income. A closer water source may also translate into more water collected for hand hygiene, which is known to prevent gastrointestinal and respiratory diseases.

15 Vvob Education for Development Zimbabwe [online]. Available at: http://www.vvob.be/zimbabwe/what-we-do/earlychildhooddevelopment. UNICEF purchased ECD materials for 50 schools, and assisted communities to construct outdoor play equipment in 50 schools; 200 latrines for ECD children were also constructed at 50 schools. 16 UNICEF Zimbabwe. Available at: http://www.unicef.org/zimbabwe/basic_education_1729.html 17 ECD was measured for children aged 36–59 months only.

46 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Fink et al. (2011), using 171 surveys in 70 low- and middle-income countries over the period 1986–2007, demonstrated that access to improved sanitation is associated with lower mortality, a lower risk of child diarrhoea and a lower risk of mild or severe stunting. Access to improved water is associated with a lower risk of diarrhoea and a lower risk of mild or severe stunting, but does not show any association with non-infant child mortality.

Figure 25a: Deprivation rates for children aged 24–59 months by indicators and dimensions 80 75.6

70 67

50 45 42.3 43 40 30 30 22.4 18.6 20 11.9 9.7 10.9 10 Percentage of deprived children deprivedof Percentage 2.2 2 0 ECD Wasng Stunng Toilet Toilet type Vaccinaon Cooking fuel Cooking Shared toilet Underweight Handwashing Water source Water Child-adult interacon Distance to source Distance water Availability of books or toys Availability

Health Child development Physical development Water Sanitaon Indicators of deprivaon

Figure 25b: Deprivation rates for children aged 24–59 months by dimensions 90 82.8 80

60 52 50 48.9

40 34.7 32.3 30

Percentage of depirved children ofdepirved Percentage 20

0 Health Development Physical Water Sanitaon Dimensions of deprivaon

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 47 Results

3.3.1.1 Single deprivation by indicator

The difference between the deprivation rates for girls and boys in this age group was significant only for the indicators early childhood education attendance, stunting and underweight, with a higher proportion of deprived boys. As with the previous age group, the analysis points out that children without a birth certificate were significantly and proportionately more deprived than others. The highest percentage point difference (17 per cent) between children with and without a birth certificate was for the indicator early childhood education attendance. In many cases, ECD centres provide a venue for the identification of children without birth certificates and provide an opportunity to educate parents about the importance of such documentation, as well as early learning and school readiness programmes. ECD also has various other functions: it provides an opportunity for children from deprived backgrounds to receive stimulation and social interaction, which they may lack at home. ECD centres also enable monitoring of children’s health and nutritional status and can make appropriate interventions where needed (Association for Childhood Education International, 2015). Thus, ECD can help improve the well-being of a child in many ways (see Figures 26a–b).

Figure 26a: Deprivation rates for children aged 24–59 months by indicators and gender 78 80 74 70 60 50 42 43 40 33 30 27 Percentage 20 12 10 9 2 2 2 2 0 ECD* Wasng Stunng* Vaccinaon Underweight* Availability of books or toys or books of Availability

Health Child Development Physical Development

Female Male

48 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 26b: Deprivation rates for children aged 24–59 months by indicators and possession of a birth certificate 90 81 80 70 64 60 50 46 40 39 33 Percentage 30 24 20 12 10 9 2 3 2 2 0 ECD* Wasng Stunng* Vaccinaon* Underweight* Availability toys* of books or

Health Child Development Physical Development

Child has a birth cerﬁcate Child does not have a birth cerﬁcate

3.3.1.2 Single deprivation by dimension

There was a higher proportion of deprived children living in rural than in urban areas for all dimensions. A Chi-square test reveals that the differences were significant for all the dimensions under study. Given a high percentage point difference (41 percentage points) between deprivation rates in rural and urban areas, a deeper analysis was carried out to determine in which province the water problem manifests itself most for this particular age group (Figure 27).

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 49 Results

3.3.1.3 Profiling based on geographical location

Figure 27: Deprivation rates for children aged 24–59 months by dimension and area Health*

51.4

Sanitaon*86.7 41.5 Child development* 71.6 53.8 46.7 4.38 25

34.8 45

Water* Physical development*

Urban Rural

Figure 28a demonstrates that the deprivation rates for the Water dimension vary significantly – from only 1.1 per cent in Bulawayo Province to 56 per cent in Masvingo Province. Matabeleland South Province also has a very high proportion of children (53.6 per cent) in this age group who are deprived on the Water dimension. For the dimensions Physical development and Health, it can be seen from Figure 28b–c that the deprivation rates were not very different across the eight provinces.

50 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 28a: Deprivation rates for children aged 24–59 months for Water by province

Mashonaland Central 41.8 Mashonaland West 39.5

Harare 9.7

Mashonaland East 20.2 Matabeleland North Midlands 53.6 Manicaland 41.8 28.8

Bulawayo 1.1

Masvingo Matabeleland South 56.0 38.9

Total

1.1 - 20.2 20.3 - 45.0 45.1 - 56.0 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 51 Results

Figure 28b: Deprivation rates for children aged 24–59 months for Physical development by province

Mashonaland Central Mashonaland West 34.0 30.5

Harare 28.2 Mashonaland East 32.2 Matabeleland North Midlands 29.5 Manicaland 31.4 38.3

Bulawayo 26.5

Masvingo Matabeleland South 31.4 38.8 Total

26.5 - 29.5 29.6 - 35.0 35.1 - 38.8 Missing Value l

Source: ZIMSTAT_MICS_2014

52 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 28c: Deprivation rates for children aged 24–59 months for Health by province

Mashonaland Central 45.0 Mashonaland West 44.0

Harare 40.2

Mashonaland East 55.5 Matabeleland North 45.4 Midlands 49.1

Manicaland 54.6

Bulawayo 43.7

Masvingo Matabeleland South 54.2 50.8 Total

40.2 - 44.0 44.1 - 50.9 51.0 - 55.5 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 53 Results

For children below 5 years of age, it was deemed important to find out the dimensions on which children living in households with infant mortality were deprived, so as to reduce the mortality rate. Figure 29 shows that children living in households with a case of infant mortality in the five years preceding MICS5 were significantly more deprived on the Water, Physical development and Health dimensions (6.4, 8.6 and 13.3 percentage points). Ultimately, actions should be taken very soon to tackle those issues before it is too late.

Figure 29: Deprivation rates for children aged 24–59 months by the presence of infant mortality cases in the five years preceding MICS5

63.5 Health* 50.2

54.4 Child development 52.7

Physical development* 40.5 31.9

Water* 40.4 34

Sanitaon 83.7 Dimensions of deprivaon of Dimensions 84.1

0102030405060708090 Percentage of deprived children

At least one child <5yrs has died in the last 5 years No child <5yrs has died in the last 5 years

Table 10 shows the percentages of children deprived on each dimension according to different characteristics of the child.

54 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 10: Deprivation rates for children aged 24–59 months by dimension and child’s characteristics

Develop- Health Physical Water Sanitation Profiling variable ment

Secondary or higher 47.1* 48.1* 28.8* 25.1* 79.3*

Mother’s Primary education 51.5* 57.1* 37.3* 46.9* 87.2* education No education 49.1* 59.5* 33.0* 52.8* 90.8*

Secondary or higher 47.7 49.2* 28.8* 25.5* 79.8* Education level of Primary education 50.4 55.8* 37.0* 45.1* 85.9* household head No education 49.8 54.2* 34.6* 50.8* 90.1*

Household size 49.1 54.0* 35.6* 43.4* 81.5* greater than 6 Household size Household size 4–6 48.1 52.1* 30.8* 32.7* 82.5*

Household size 1–3 51.7 48.3* 32.0* 26.8* 86.6*

Above median 49.1 54.6* 34.3* 39.4* 83.1 Number of number of children children in household Below or equal to (median=2) median number of 48.7 47.9* 29.1* 27.3* 82.5 children

Labour constrained: dependency ratio 51 55.7* 37.1* 43.9* 84.8* Labour- greater than 2 constrained household Household not 48.1 50.5* 30.3* 31.0* 82.0* labour constrained

Q5 (Richest) 46.2* 43.4* 23.0* 19.3* 62.1*

Q4 43.2* 51.1* 26.8* 15.1* 78.7* Asset index (proxy for Q3 49.0* 52.1* 35.3* 25.2* 82.6* wealth) Q2 52.9* 55.2* 34.0* 47.0* 90.0*

Q1 (Poorest) 51.7* 56.5* 39.7* 58.5* 95.8*

3.3.2 Multiple Overlapping Deprivation Analysis

3.3.2.1 Deprivation Overlap Analysis

Deprivation overlap for each dimension

As observed with the younger age group of 0–23 months, a small percentage of children aged between 24 and 59 months was deprived on only one dimension. Most of them tended to be deprived on several dimensions at one time. Ultimately, here again it would be highly recommended to solve the issues facing the children in combination, rather than in isolation. Since more than 80 per cent of children

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 55 Results

were deprived on Sanitation, it would be imperative for the Government to implement policies to alleviate the problem of sanitation for children in this age group. By doing so, 6.4 per cent of children would experience no deprivation at all; 28.5 per cent would have only one instead of two deprivations; 31.7 per cent would have two instead of three deprivations; and so on (Figure 30).

Figure 30: Deprivation overlap for each dimension for children aged 24–59 months 90

80 16.2 70

50 31.7 12.3 13.2 40

30 19

Percentage of deprived children deprivedof Percentage 18.5 13.9 12.5 20 28.5

16.3 17.6 13.6 10 13.6

6.4 3.6 3.2 3.8 4.3 0 0.4 0.2 Heath Child development Physical development Water Sanitaon

Deprived only on the speciﬁed dimension Deprived on one other dimension Deprived on 2 other dimensions Deprived on 3 or more other dimensions

Deprivation overlap of three dimensions

Of interest is the deprivation overlap combination for children in the age group 24–59 months on Health, Physical development and Sanitation, as shown in Figure 31.

56 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 31: Deprivation overlap for children aged 24–59 months

While a large proportion of children were deprived on the Health (49 per cent) and Physical development (30.4 per cent) dimensions, a much lower percentage of children (5.7 per cent and 2 per cent, respectively) were deprived on Health only or on Physical development only. Similar observations were found for the dimension Sanitation. A majority of children deprived on the dimension Physical development were also deprived on Health and Sanitation. This demonstrates a close relationship between the three dimensions. In fact, 15.6 per cent of children were deprived on all three dimensions at the same time. Since all the dimensions under study are of paramount importance for children in this age group, an overlap on two or more dimensions can have an adverse effect on the growth and – most importantly – survival of the child.

The results were disaggregated by urban and rural areas as follows:

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 57 Results

Figure 32: Deprivation overlap for children aged 24–59 months by area

Rural Areas

Urban Areas

In rural areas, the percentage of deprivation overlap was much more than in urban areas (17.6 per cent, compared to 9.5 per cent). A higher proportion of children also tended to be deprived on only one dimension in urban areas. The percentage of non-deprived children was also very different in the two areas, with 5 per cent in rural areas and 13.3 per cent in urban areas (Figure 32).

58 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Deprivations experienced by children at the national and subnational level

The breadth of deprivation for each child was measured as the total number of dimensions on which the child was deprived and how they were distributed over the population. This allowed us to examine overall deprivation severity among children. This section also provides deprivation distributions profiled by the individual and household characteristics of children.

For children aged 24–59 months, it was observed that the distribution of simultaneous deprivations follows a normal distribution. Only 2.7 per cent of the child population in this age group did not experience any deprivations on the dimensions under study. The majority of children (66.9 per cent) were deprived on two or three of the five dimensions. Almost 17 per cent were deprived on four or five dimensions concurrently. Around 2 per cent of children were deprived on all five dimensions simultaneously (Figure 33a).

Disaggregating the results by rural and urban areas shows significant disparities, as presented in Figure 33b.

Figure 33a: Percentage distribution of children aged 24–59 months by number of deprivations at national level 35 33.9 33

15 13.9 14.2 Percentage

5 2.7 2.3 0 012345 Number of simultaneous deprivaons

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 59 Results

Figure 33b: Percentage distribution of children aged 24–59 months by number of deprivations by area 50 46 45 40 37.4 35 29.9 30 25 24.6 20 20 18 Percentage 15 10.2 10 6.6 5 2.7 3.1 1.4 0 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons Urban Rural

The number of simultaneous deprivations experienced by different profiles of children are detailed in Table 11.

Table 11: Deprivation distribution for children aged 24–59 months by child’s characteristics

Number of simultaneous deprivations experienced by each child Profiling variable 0 1 2 3 4 5

Secondary or higher 4.1* 18.2* 37.5* 28.7* 10.2* 1.2*

Mother’s Primary education 1.0* 8.4* 28.7* 39.0* 19.0* 3.8* education No education 0.5* 4.8* 33.7* 35.7* 22.4* 2.9*

Secondary or higher 3.8* 17.2* 38.4* 28.7* 10.2* 1.7* Education level of Primary education 1.5* 10.2* 28.5* 37.6* 19.1* 3.2* household head No education 0.7* 8.0* 28.6* 41.6* 19.1* 2.1*

Household size 2.5* 12.5* 30.1* 33.7 18.4* 2.82 greater than 6 Household size Household size 4–6 2.9* 14.9* 34.2* 33.1 12.5* 2.31

Household size 1–3 2.5* 12.3* 39.6* 31.3 12.8* 1.37

Possession Birth certificate 5.4* 20.6* 38.5* 25.7* 8.7* 1.2* of birth certificate No birth certificate 0.8* 8.6* 32.4* 37.8* 17.5* 3.0*

Above median 2.39 12.5* 31.7* 34.2* 16.2* 3.0* Number of number of children children in household Below or equal to (median=2) median number of 3.26 16.1* 37.6* 31.0* 10.8* 1.2* children

60 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Table 11: Deprivation distribution for children aged 24–59 months by child’s characteristics

Number of simultaneous deprivations Profiling variable experienced by each child 0 1 2 3 4 5

Labour constrained: dependency ratio 1.4* 10.6* 30.3* 35.1* 18.8* 3.8* Labour- greater than 2 constrained household Household not labour 3.3* 15.2* 35.4* 32.2* 12.3* 1.7* constrained

Q5 (Richest) 9.2* 24.4* 39.3* 19.4* 7.2* 0.5*

Q4 2.6* 21.2* 44.0* 25.5* 6.1* 0.7* Asset index Q3 1.5* 15.5* 37.7* 32.3* 10.7* 2.2* (proxy for wealth) Q2 1.0* 7.9* 28.6* 41.3* 18.1* 3.2*

Q1 (Poorest) 0.4* 4.1* 23.9* 42.3* 25.0* 4.3*

3.3.2.2 Multidimensional deprivation indices

Table 12 shows the deprivation indices calculated for children aged 24–59 months. Slightly more than 97 per cent of children in this age group were deprived on at least one dimension, while 2.3 per cent of children were deprived on all five dimensions simultaneously. For the purposes of this study, the cut-off point was two dimensions (K=2).18 This implies that a child was considered multidimensionally poor if he/she was deprived on at least two dimensions. Overall, 83.4 per cent of children in the age group 24–59 months were multidimensionally poor. On average, these deprived children experienced 2.8 out of five deprivations – that is, 56.4 per cent of the total number of deprivations. This gives a deprivation headcount adjusted for intensity (M0) index of 0.47, which is simply the headcount ratio multiplied by the intensity of deprivation.

Table 12: Multidimensional deprivation indices for children aged 24–59 months at national level

1–5 97.3 51.2 2.6 0.498

2–5 83.4 56.4 2.8 0.47

3–5 49.5 67.6 3.4 0.334

4–5 16.5 82.8 4.1 0.136

5 2.3 100 5 0.023

18 Refer to the methodology section of this report for the criteria on the basis of which the value of K was determined.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 61 Results

Figure 34 depicts the analysis of deprivation headcount ratio (H) by urban and rural areas. In all cases, rural areas had a higher proportion of deprived children – for example, 88.4 per cent of children in rural areas were deprived on at least two dimensions, compared to 68.7 per cent in urban areas. Disaggregating the results by province shows that Masvingo Province has the highest proportion of deprived children (90 per cent) with a deprivation headcount adjusted for intensity (M0) of 0.437. Bulawayo Province had the lowest proportion of deprived children (66.6 per cent) and the lowest M0 of 0.131.

Figure 34: Multidimensional deprivation headcount for children aged 24–59 months by various cut-off points

100 98.6 97.3

90 88.4 93.4 83.4 80

70 68.7

60 58.5 50 49.5

Percentage 40

20 22.7 21.1 16.5 10 3.1 2.3 2.7 0 0 1-5 dimensions 2-5 dimensions 3-5 dimensions 4-5 dimensions 5 dimensions

Urban Rural Naonal

Figure 35: Multidimensional deprivation indices (K=2) for children aged 24–59 months by province

0.47 100 0.437 0.5 0.401 0.392 80 0.353 0.364 0.4 0.346 0.332 0.295 60 0.3 40 0.186 0.2 0.131 headcount 20 0.1 and intensity, % intensity, and

0 0.0 deprivaon Adjusted Deprivaon headcount Deprivaon

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

62 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Multidimensionally poor children were profiled by some of their characteristics (Figure 36).

Figure 36: Multidimensional deprivation headcount (K=2) for children aged 24–59 months by profiling characteristics

Percentage of children deprived in 2-5 dimensions 0% 25% 50% 75% 100% 125%

Naonal Naonal 83% female 85% Gender of hh head* male 83% Secondary or higher 78% Mother's educaon* Primary educaon 91% No educaon 95% Secondary or higher 79% Educaonal level Primary educaon of hh head* 88% No educaon 91% hhsize >6 85% Household size* hhsize 4-6 82% hhsize 1-3 85% no religion 89% other religion 84% tradional 92% other chrisans Religion* 85% apostolic sect 88% pentecostal 72% protestant 73% roman catholic 80% No child <5 has died in the last 5 years 85% Infant mortality* At least one child <5 has died in the last 5 years 91% Possession of Health card 85% health card* No health card 96% Possession of Birth cerficate 74% birth cerficate* No birth cerficate 91% No. of children in Above median number of children 85% hh (media n =2)* Below or equal to median number of children 81% Labour Labour constraint: dependency rao>2 88% constraint hh* Hld not labour constraint 82% Q5 (Richest) 66% Q4 76% Asset Index Q3 83% (proxy for Wealth)* Q2 91% Q1 (Poorest) 96% rural non-poor 80% Area and poverty urban non-poor 65% status (based on rural poor asset index) of hh* 91% urban poor 85% Gender of hh head female headed hh (non-poor) 72% and poverty status male headed hh(non-poor) 71% (based on asset female headed hh(poor) 91% index) of hh* male headed hh(poor) 90%

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 63 Results

Decomposition of the multidimensional deprivation ratios

There is a difference of 0.257 points between the adjusted deprivation headcount index (M0) in urban and rural areas, as shown in Figure 37. By decomposing the M0, it was found that the different dimensions used in the study contribute differently to the level of deprivation for children in the age group 24–59 months. While the Water dimension contributes 17.2 per cent to the level of deprivation in rural areas, its contribution was 2.6 per cent in urban areas. Conversely, Child development contributes relatively more to the level of deprivation in urban (24.8 per cent) than in rural areas (19.9 per cent) (see Figure 37).

Figure 37: Adjusted deprivation headcount rate (K=2) for children aged 24–59 months by area 00.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45

Urban 0.142

Rural 0.399

Naonal 0.334

Figure 38: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 24–59 months by area 100

31.4 32.5 37.6 75

2.6 17.2 14.7

50 24.8 20.8 Percentage 19.9

14 25 12.5 12.7

21 0 18.9 19.3 0 nabrU laruR lanoitaN Health Physical development Child development Water Sanitaon

64 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.4 Deprivation among children aged 5–14 years

This section focuses on children aged 5–14 years.

Box 3: Key findings for children aged 5–14 years

l About one in two children aged 5–14 years was multidimensionally poor, that is, were deprived on at least two dimensions of child well-being.

l Some 7 per cent of children in this age group did not attend school. A higher proportion of boys was deprived onn the Education indicators.

l Multidimensionally poor children in this age group experienced on average 2.5 out of five dimensions, representing 49.3 per cent of the total number of deprivations.

l Matabeleland North Province, with an M0 index of 0.391, was by far the worst-off province.

l In rural areas, the Health, Water and Information dimensions contribute more to the deprivation index than in urban areas.

3.4.1 Single deprivation analysis

As with the previous age group, Sanitation is the dimension with the highest level of deprivation (80 per cent). This was mainly driven by high levels of deprivation for the indicators handwashing (46.2 per cent), toilet type (43.6 per cent) and sharing of toilet facility (61.5 per cent). The second-highest deprivation level was observed for the Water dimension, with more than one in three children aged 5–14 years being deprived. For the Health dimension, 12.4 per cent of children lived in households that used solid fuels (according to the WHO requirements).

With regards to the Education dimension, it was noted that around 8 per cent of children in this age group did not attend school. Zimbabwe performs relatively well in terms of school attendance, compared to other sub Saharan countries (UNESCO, 2010). About 19 per cent of children aged 5–14 years were deprived in the indicator grade-for-age, meaning that they were three or more years late in attending school. For the Information dimension, 11.3 per cent of children aged 5–14 years did not have access to any information or communication devices, such as TV, radio, phone, mobile phone or computer. At this age, exposure to such devices is likely to contribute to general knowledge.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 65 Results

Figure 39a: Deprivation rates for children aged 5–14 years by indicator

010 20 30 40 50 60 70

Health Cooking fuel 12.4

School aendance 7.85 Educaon Grade-for-age 18.6

Water Source 22.4 Water Water Distance 19.3

Handwashing 46.2

Sanitaon Toilet type 43.6 Indicators ofdeprivaon Indicators Shared toilet 61.5

Informaon Availability of informaon devices 11.3

Percentage of deprived children

Figure 39b: Deprivation rates for children aged 5–14 years by dimension 80 80

40 35.7 30 21.8 20

Percentage of deprived children ofdeprived Percentage 12.4 11.3 10

0 Health Educaon Water Sanitaon Informaon Dimensions of deprivaon

The deprivation rates for education were disaggregated by sex for further analysis, as shown in Figure 40 to illustrate gender parity. The analysis shows that the difference between girls and boys for the indicators under the dimension Education was significant. The percentage of deprived boys was more than that of girls for both school attendance and grade-for-age. Nearly 9 per cent of boys were out of school, compared to 7.1 per cent of girls of the same age. Regarding grade-for-age, the difference between boys and girls was even greater (7.2 percentage points).

66 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 40: Deprivation rates for indicators under the dimension Education, children aged 5–14 years 0510 15 20 25

7.1 School aendance* 8.6

Educaon 15.0 Grade-for-age* 22.2 Indicators ofdeprivaon Indicators Percentage Female Male

3.4.1.1 Profiling by geographical characteristics

For all the dimensions under study for children aged 5–14 years, the percentage of deprived children living in rural areas was more than that in urban areas. As was observed for previous age groups, on the Water dimension the difference in the deprivation level between rural and urban areas was huge (40.9 percentage points).

Figure 41: Deprivation rates by urban and rural area, children aged 5–14 years Health*

Informaon* Educaon* 15.4 24.7 14.2 1.7 0.8 11.2 3.7

64.8 44.6

84.3 Sanitaon* Water*

Urban Rural

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 67 Results

Disaggregating the dimensions by province (Figure 42a–b) shows that Bulawayo and Harare provinces had the lowest deprivation rates (10.5 per cent and 11.5 per cent, respectively) on the Education dimension. With respect to the Information dimension, only 1 per cent of children in Harare and Bulawayo provinces were deprived, while the remaining (predominantly rural) provinces had deprivation rates of 10–13 per cent, with the exception of Matabeleland North (20 per cent). The deprivations for all the dimensions used were analysed by various profiling variables (Table 13).

Figure 42a: Deprivation rates for children aged 5–14 years for Education by province

Mashonaland Central 45.0 Mashonaland West 44.0

Harare 40.2

Mashonaland East 55.5 Matabeleland North 45.4 Midlands 49.1

Manicaland 54.6

Bulawayo 43.7

Masvingo Matabeleland South 54.2 50.8 Total

40.2 - 44.0 44.1 - 50.9 51.0 - 55.5 Missing Value l

Source: ZIMSTAT_MICS_2014

68 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 42b: Deprivation rates for children aged 5–14 years for Information by province

Mashonaland Central 45.0 Mashonaland West 44.0

Harare 40.2

Mashonaland East 55.5 Matabeleland North 45.4 Midlands 49.1

Manicaland 54.6

Bulawayo 43.7

Masvingo Matabeleland South 54.2 50.8 Total

40.2 - 44.0 44.1 - 50.9 51.0 - 55.5 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 69 Results

Table 13: Deprivation rates for children aged 5–14 years by dimensions and child’s characteristics

Profiling variable Health Education Water Sanitation Information

Living with both 11.5 19.1* 35.3 81.2* 8.1* parents

Both parents dead 13.4 34.2* 38.8 80.3* 16.9*

Parent/s alive but does/do not live with 13 23.5* 36.3 76.4* 12.8* child

Living with father 11.8 24.5* 37.3 78.8* 12.7* Caregiver only, mother alive

Living with father 12.7 38.9* 39.7 75.2* 11.7* only, mother dead

Living with mother 13.7 17.7* 33.6 82.8* 12.0* only, father alive

Living with mother 11.8 26.4* 37.4 82.7* 16.8* only, father dead Secondary or higher 10.3* 12.6* 24.3* 74.3* 4.3*

Mother’s Primary education 14.5* 28.6* 43.6* 83.9* 15.0* education No education 12.9* 34.0* 54.5* 89.8* 27.5*

Secondary or higher 9.9* 14.0* 25.5* 74.7* 4.7* Education level of Primary education 14.9* 27.9* 43.4* 83.7* 14.7* household head No education 13.3* 32.8* 51.8* 89.8* 28.8* More than 6 14.4* 26.3* 42.4* 79.5 8.0* members Household size 4–6 members 11.5* 19.7* 33.2* 80 11.8* 1–3 members 11.3* 19.6* 29.7* 81.7 17.9* Labour constrained: dependency ratio 14.7* 25.3* 41.6* 82.4* 16.9* Labour- greater than 2 constrained household Household not 10.8* 19.3* 31.6* 78.4* 7.4* labour constrained

Above median 13.0* 23.0* 38.4* 81.0* 11.2 Number of number of children children in household Below or equal to (median=2) median number of 10.8* 18.6* 28.7* 77.4* 11.5 children

Q5 (Richest) 10.6* 12.5* 21.5* 59.9* 0.2*

Q4 6.3* 14.6* 17.2* 73.3* 0.4* Asset Index Q3 14.3* 21.1* 30.9* 78.4* 2.8* (proxy for Wealth) Q2 15.5* 24.4* 43.3* 88.1* 14.6*

Q1 (Poorest) 14.1* 32.7* 57.3* 95.4* 31.3*

70 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.4.2 Multiple Overlapping Deprivation Analysis

3.4.2.1 Deprivation overlap analysis

Deprivation overlap for each dimension

Unlike for the two younger age groups, it was observed that a very small percentage of children aged 5–14 years were deprived on three or more dimensions. Most were deprived on one or two dimensions. Deprivation on Sanitation only was high (31.8 per cent). By addressing the problem of Sanitation, 31.8 per cent of the children in this age group would experience no deprivation, while 29.5 per cent would experience only one deprivation. As for the other dimensions of Health, Education, Water and Information, it would be more advisable to deal with them in combination, given the small proportions of children deprived on only one dimension (Figure 43).

Figure 43: Deprivation overlap for children aged 5–14 years for each dimension 80 4.0

14.8

29.5

3.7

Percentage of deprived children deprivedof Percentage 12.1

20 3.5

31.8 8.4 2 17.3 3 5 8.3 5 4 2.6 3 0 1 1.6 0 Health Educaon Water Sanitaon Informaon

Deprived only in the speciﬁed dimension Deprived in one other dimension Deprived in 2 other dimensions Deprived in 3 or more other dimensions

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 71 Results

Deprivation overlap of three dimensions

Some 51 per cent of children aged 5–14 years were deprived on at least one of the dimensions shown in Figure 44. Around a third of children deprived on the dimension Information were also deprived on Education. Access to information can promote education among children in two ways. First, it increases awareness about the importance of education and second; secondly, it can be used as a school aid tool. Nearly 22 per cent of children were deprived on the Water dimension only. About 2 per cent of children in this age group were deprived on all three dimensions.

Figure 44: Deprivation overlap for children aged 5–14 years

For the dimensions Water, Education and Information, great disparities were noted between urban and rural areas. Notably, 85.5 per cent of children in this age group in urban areas were non-deprived, compared to 38.9 per cent in rural areas. While 24.8 per cent, 44.6 per cent and 14.1 per cent of children in rural areas were deprived on Education, Water and Information, respectively, the percentages were much lower in urban areas, with 11.1 per cent, 3.7 per cent and 0.8 per cent. The shape of the overlap sections of the Venn diagrams shows that overlapping of dimensions prevails more in rural than in urban areas (Figure 45).

72 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 45: Deprivation overlap for children aged 5-14 years by area

Rural Areas

ural areas

Urban Areas rban areas

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 73 Results

Deprivations experienced by children at the national level

Compared with the younger children, the distribution of the deprivation was skewed towards the left,19 implying that there were more children deprived on a smaller number of dimensions simultaneously. About 12 per cent of children were not deprived on any of the dimensions, 37.5 per cent were deprived on one dimension and 31.1 per cent on two dimensions. Around 4 per cent were deprived on four or five dimensions simultaneously. Significant and large differences were noted for the deprivation distribution between urban and rural areas (Figure 46a–b).

Figure 46a: Percentage distribution of number of deprivations experienced by children at national level 40 37.5 35 31.1 30

20 15

Percentage 15 12.3 10

5 3.6 0.4 0 012345 Number of simultaneous deprivaons

Figure 46b: Percentage distribution of number of deprivations experienced by children by area

60 57.6

40 37 30.9 32 30 18.8 Percentage 20 10 10 7.1 4.5 1.3 0.2 0 0.5 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Urban Rural

19 Skewness skewed left Skewness define is asymmetry in a statistical distribution, in which the curve appears distorted or either to the or to the right. can be quantified to the extent to which a distribution differs from a normal distribution.

74 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

The deprivation distribution for different profiles of children was analysed (Table 14).

Table 14: Deprivation distribution of children aged 5–14 years by child’s characteristics

Number of simultaneous deprivations Profiling variable experienced by each child 0 1 2 3 4 5 Living with both 13.3* 39.7* 29.7* 13.5* 3.2* 0.6 parents Both parents dead 8.6* 31.5* 34.5* 19.2* 6.0* 0.3 Parent/s alive but does/do not live with 13.3* 34.0* 33.8* 15.3* 3.3* 0.2 child Living with father 11.3* 37.0* 30.8* 17.4* 3.5* 0.0 Caregiver only, mother alive Living with father 10.9* 32.8* 29.1* 22.3* 4.5* 0.5 only, mother dead Living with mother 11.0* 41.0* 30.1* 13.9* 3.7* 0.4 only, father alive Living with mother 9.9* 35.3* 30.3* 19.2* 4.9* 0.4 only, father dead Secondary or higher 18.5* 47.4* 25.0* 7.9* 1.2* 0.0* Mother's Primary education 7.6* 31.3* 35.7* 19.6* 5.1* 0.8* education No education 4.0* 19.7* 39.0* 28.6* 8.2* 0.5* Education Secondary or higher 17.8* 46.7* 25.9* 8.2* 1.3* 0.1* level of Primary education 8.1* 31.2* 35.2* 19.6* 5.2* 0.7* household head No education 3.7* 21.3* 38.4* 28.2* 8.0* 0.4* More than 6 11.2 32.5* 35.7* 16.4* 3.57 0.6* members Household size 4–6 members 13 39.9* 29.0* 14.5* 3.38 0.3* 1–3 members 12.1 40.2* 28.6* 13.9* 4.71 0.4* Above median 10.9* 36.2* 32.9* 15.9* 3.73 0.429 Number of number of children children in household Below or equal to (median=2) median number of 15.9* 41.2* 26.6* 12.8* 3.23 0.311 children Labour constrained: Labour- dependency ratio 8.7* 32.4* 34.8* 18.5* 4.9* 0.7* constrained greater than 2 household Household not labour 14.9* 41.2* 28.5* 12.6* 2.7* 0.2* constrained Q5 (Richest) 29.0* 43.7* 21.4* 5.5* 0.4* 0.0* Q4 19.3* 55.3* 20.0* 5.1* 0.3* 0.0* Asset index (proxy for Q3 11.2* 44.4* 31.8* 11.2* 1.5* 0.0* Wealth) Q2 4.4* 33.2* 39.0* 19.2* 4.1* 0.1* Q1 (Poorest) 1.5* 18.9* 39.6* 28.9* 9.6* 1.5*

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 75 Results

3.4.2.2 Multidimensional deprivation indices

Table 15 shows that 87.7 per cent of children in the age group 5–14 years experienced at least one deprivation. Given that K=2, 50.2 per cent of children were considered multidimensionally poor. Those multidimensionally poor children experienced, on average, 2.5 out of five dimensions, representing 49.3 per cent of the total number of deprivations.

Table 15: Multidimensional deprivation indices at national level for children aged 5–14 years

1–5 87.7 36.8 1.9 0.323

2–5 50.2 49.3 2.5 0.247

3–5 19 64.6 3.2 0.123

4–5 4.0 82 4.1 0.033

5 0.4 100 5 0.004

The multidimensional deprivation headcount ratio (H) was, in all cases, higher in rural areas than in urban areas and at the national level, as shown in Figure 47.

Figure 47: Multidimensional deprivation headcount for children aged 5–14 years by various cut-off points and area

100 92.9 90 87.7 80

70 69.1 60 60.9 50 50.2 40 Percentage 30 23.9 20 19 5 0.5 10 11.4 4 0.4 1.4 0.2 0 0 1-5 dimensions 2-5 dimensions 3-5 dimensions 4-5 dimensions 5 dimensions

Urban Rural Naonal

76 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

The deprivation headcount ratio also varies significantly across the provinces, with the lowest deprivation headcount rate observed in Bulawayo Province (8.4 per cent) and the highest in Matabeleland North Province (69.5 per cent). The average intensity of deprivation across deprived children was not very different, implying that deprived children have on average the same number of deprivations in all provinces. To grasp the overall performance of the provinces, it was best to analyse the M0 index – Matabeleland North Province, with an index of 0.391, was by far the worst-off province.

Figure 48: Multidimensional deprivation indices (K=2) for children aged 5–14 years by province 80 0.5

0.391 0.382 0.4 60 0.343 0.332 0.332 0.32 0.331 0.314 0.247 0.3 40 0.188 0.2 0.144 and intensity, % intensity,and 20

Deprivaon headcount Deprivaon 0.1

0 0.0 deprivaonAdjusted headcount

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

The percentage of multidimensionally poor children according to some of their characteristics is shown in Figure 49.

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 77 Results

Figure 49: Multidimensional deprivation headcount (K=2) for children aged 5–14 years by child’s characteristics Percentage of children deprived in 2-5 dimensions 0% 25% 50% 75% 100%

Naonal Naonal 50% female 48% Child's gender* male 52% living with both parents 47% both parents dead 60% parent/s alive but does/do not live with child 53% Caretaker* living with father only, mother alive 52% living with father only, mother dead 56% living with mother only, father alive 48% living with mother only, father dead 55% female 52% Gender of hh head* male 49% Secondary or higher 34% Mother's educaon* Primary educaon 61% No educaon 76% Secondary or higher 36% Educaonal level of hh head* Primary educaon 61% No educaon 75% hhsize >6 56% Household size* hhsize 4-6 47% hhsize 1-3 48% no religion 62% other religion 52% tradional 67% other chrisans 55% Religion* apostolic sect 56% pentecostal 33% protestant 37% roman catholic 40% No. of children in Above median number of children 53% hh (media n =2)* Below or equal to median number of children 43% Labour Labour constraint: dependency rao>2 59% constraint hh* Hld not labour constraint 44% Q5 (Richest) 27% Q4 25% Asset Index Q3 45% (proxy for Wealth)* Q2 63% Q1 (Poorest) 80% rural non-poor 44% Area and poverty urban non-poor 9% status (based on asset index) of hh* rural poor 66% urban poor 26% Gender of hh head female headed hh (non-poor) 28% and poverty status male headed hh(non-poor) 26% (based on asset female headed hh(poor) 64% index) of hh* male headed hh(poor) 65%

78 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Decomposition of the multidimensional deprivation ratios

The comparison of M0 shows that the adjusted deprivation rate was much higher in rural (0.302) than in urban (0.049) areas. Decomposing the M0 index shows that the Sanitation and Education dimensions contributed most to deprivation levels in both urban and rural areas. In rural areas, the Health, Water and Information dimensions contributed more to the deprivation index than in urban areas.

Figure 50: Adjusted deprivation headcount rate (K=2) for children aged 5–14 years by area

00.05 0.1 0.15 0.2 0.25 0.3 0.35

Urban 0.049

Rural 0.302

Naonal 0.247

Figure 51: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 5–14 years by area

3.3 9.2 9

45.8 100 38.7 39

75 12.3 27.4 26.8

Percentage 50

33.8 25 15.5 16.3

9.1 9 0 4.9 nabrU laruR lanoitaN

Health Educaon Water Sanitaon Informaon

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 79 Results

3.5 Deprivation among children aged 15–17 years

This section focuses on children aged 15–17 years.

Box 4: Key findings for children aged 15–17 years

l Some 56 per cent of children aged 15–17 years were multidimensionally poor, that is, they were deprived on at least two dimensions of child well- being.

l The highest deprivation rates were observed for the dimensions Sanitation (75.3 per cent), Education (43 per cent) and Water (31.9 per cent).

l Deprivation rates were significantly lower in urban than in rural areas.

l The combination of the dimensions Education, Water and Sanitation yielded the highest percentage of deprivation overlap (14.4 per cent).

l Some 89.1 per cent of children aged 15–17 years were deprived on at least one dimension, while 0.8 per cent were deprived on all five dimensions simultaneously.

l Bulawayo Province had the lowest deprivation index (M0) of 0.084, while Matabeleland North Province was the worst-performing province, with an M0 of 0.386.

l Decomposing the M0 revealed that the Water (22.9 per cent), Health (8.2 per cent) and Information (6.9 per cent) dimensions contributed relatively more to the deprivation level in rural areas, while the main contributors in urban areas were Education (43.3 per cent) and Sanitation (25.2 per cent).

3.5.1 Single deprivation analysis

Again, Sanitation was the dimension with by far the highest deprivation level (75.3 per cent) for children aged 15–17 years. There was a much higher and significant proportion of children deprived on the dimension Education (40 per cent), compared to children aged 5–14 years (21.8 per cent). This result was driven by high deprivation in the indicators school attendance (38.6 per cent) and primary school attainment (13.5 per cent). While 7 per cent of children aged 5–14 years did not attend school, this was true of 38.6 per cent of children in the age group 15–17 years. This may be due to the fact that some children aged 15–17 years may already have completed lower secondary education (Form 4). See Figure 52a–b.

80 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 52a: Deprivation rates for children aged 15–17 years by indicator

0 102030405060

Health Cooking fuel 12.4

School Aendance 38.6 Educaon Primary School Aainment 13.5

Water Source 19.6 Water Water Distance 17.4

Handwashing 45.4

Indicators ofdeprivaon Indicators Sanitaon Toilet type 36.7

Shared toilet 54.7

Informaon Availability of informaon devices 9.3

Percentage of deprived children

Figure 52b: Deprivation rates for children aged 15–17 years by dimension 80 75.3 70

50 43 40 31.9 30

Percentage of deprived children Percentage deprived of 12.4 10 9.3

0 Health Educaon Water Sanitaon Informaon Dimensions of deprivaon

The results for the Education indicators are disaggregated by sex of the child in Figure 53. The figure shows that at the 5 per cent level of significance, there was a significantly higher proportion of boys than girls deprived in both school attendance and primary school attainment (although the proportions differ by only a few percentage points). Boys seemed more likely than girls to drop out of school after the age of 15 (39.5 per cent and 37.4 per cent, respectively).

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 81 Results

Figure 53: Percentage distribution of Education indicators for children aged 15–17 years by sex

14.4 Primary school aainement* 12.4 Educaon 37.4 School aendance* 39.5 Indicators of deprivaon Indicators

0510 15 20 25 30 35 40 Female Male Percentage of deprived children

Almost one child in three aged 15–17 years experienced deprivation in terms of access to safe drinking water and 12.4 per cent lived in a household where solid fuels were used for cooking. In relation to the Information dimension, 9.3 per cent of children had no access to any devices such as radio, TV, computer or mobile phones.

3.5.1.1 Profiling by geographical characteristics

Figure 54 shows that deprivation rates were significantly different for urban and rural areas. For some of the dimensions, the difference was huge. For example, for the Information dimension, in rural areas the deprivation rate was 12.1 per cent, while in urban areas it was only 0.7 per cent. A big disparity was also observed for the Water dimension, as for the three other age groups presented earlier (0–23 months, 24–59 months and 5–14 years).

82 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 54: Deprivation rates for children aged 15–17 years by urban and rural areas Health*

Informaon* Educaon* 15.8 46.7 31.4 12.1 1.8 0.7

3.5

41.1 57.7

81 Sanitaon* Water*

Urban Rural

At the provincial level, it can be observed from Figure 55a that the deprivation rates for the dimension Education varied from 34 per cent in Harare Province to 57 per cent in Matabeleland South Province. The lowest deprivation rate for the dimension Information was 1 per cent in both Harare and Bulawayo provinces, while the highest was 17.6 per cent in Matabeleland North Province (Figure 55b).

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 83 Results

Figure 55a: Deprivation rates for children aged 15–17 years for Education by province

Mashonaland Central Mashonaland West 51.8 43.9

Harare 34.1

Mashonaland East 36.0 Matabeleland North Midlands 50.7 Manicaland 42.1 39.3

Bulawayo 34.9

Masvingo 42.7

Total Matabeleland South 57.0 34.1 - 36.0 36.1 - 49.9 50.0 - 57.0 Missing Value l

Source: ZIMSTAT_MICS_2014

84 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 55b: Deprivation rates for children aged 15–17 years for Information by province

Mashonaland Central Mashonaland West 12.1 11.4

Harare 1.0

Mashonaland East 9.3 Matabeleland North Midlands 17.6 Manicaland 8.4 9.1

Bulawayo 1.0

Masvingo 10.9

Matabeleland South Total 8.6 1.0 - 8.9 9.0 - 10.9 11.0 - 17.6 Missing Value l

Source: ZIMSTAT_MICS_2014

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 85 Results

Table 16 shows the deprivation rates on the different dimensions by various characteristics of the child.

Table 16: Deprivation rates for children aged 15–17 years by dimension and child’s characteristics Profiling variable Health Education Water Sanitation Information Sex of Female 13.2 40.9* 31.1 78.1* 11.9* household head Male 11.7 44.5* 32.5 73.1* 7.2* Secondary or 9.75 23.8* 22.1* 66.2* 5.0* higher Mother’s Primary education 13 46.2* 42.0* 85.0* 10.4* education No education 12.1 58.2* 50.0* 92.0* 23.6* Secondary or 8.7* 35.5* 23.8* 68.2* 4.7* Education higher level of Primary 16.2* 48.5* 38.5* 80.6* 11.1* household education head No education 12.9* 55.8* 43.0* 86.4* 23.6* More than 6 14.1 45.7* 39.4 74.6 5.6* members Household size 4–6 members 11.2 40.4* 28.2* 74.6 9.9* 1–3 members 12.1 44.8* 27.1* 78.7 15.3* Above median Number of number of 12.4 43.3 34.9* 76.1 8.4* children in children household Below or equal to (median=2) median number 12.4 42.3 25.4* 73.4 11.3* of children Labour constrained: 13.9 44.8 37.1* 77.6* 13.8* dependency ratio Labour- greater than 2 constrained household Household not labour 11.2 41.6 28.1* 73.6* 6.0* constrained Girl has been/ 11.6 98.8* 43.8* 82.3* 11.1 Early is married marriage Girl has never (girls only) 11.3 34.4* 28.5* 71.9* 8.29 been married Girl has been/ 10.5 98.0* 47.5* 81.9* 10.8 Early is pregnant pregnancy Girl has never (girls only) 11.4 38.0* 29.0* 72.5* 8.47 been pregnant Q5 (Richest) 10.2* 34.7* 20.2* 54.9* 0.0* Q4 6.5* 39.1* 17.2* 67.8* 0.3* Asset index (proxy for Q3 16.9* 40.8* 27.8* 76.9* 2.0* wealth) Q2 16.1* 43.1* 42.6* 87.3* 16.0* Q1 (Poorest) 12.8* 58.3* 53.2* 94.1* 29.4*

86 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.5.2 Multiple Overlapping Deprivation Analysis

3.5.2.1 Deprivation overlap analysis

Deprivation overlap for each dimension

No children in this age group were deprived on the dimension Information only. On the other hand, 23 per cent were deprived on the dimension Sanitation only. Given the small percentage of children who were deprived on only one dimension, with the exception of Sanitation, as shown in Figure 56, it is prudent to deal with problems faced by children aged 15–17 years in combination.

Figure 56: Percentage distribution of children aged 15–17 years by deprivation overlaps for each dimension 80

60 18

40 4 30 Percentage 15 4

20 15

18 2 23 4 12 3 5 4 6 2 2 2 0 0 Health Educaon Water Sanitaon Informaon

Deprived only on the speciﬁed dimension Deprived on one other dimension Deprived on 2 other dimensions Deprived on 3 or more other dimensions

Deprivation overlap of three dimensions

After analysing all possible combinations of three of the five dimensions, we found that the combination of Education, Water and Sanitation yielded the highest percentage of deprivation overlap – 14.4 per cent. Some 20 per cent of children were simultaneously deprived on the Education and Sanitation dimensions; slightly over 13 per cent were simultaneously deprived on Water and Sanitation; about 7 per cent of children aged 15–17 years were deprived on Education only. The high percentage of overlap for this combination of dimensions shows that for children aged 15–17 years, it would be efficient to tackle the problems of Education, Water and Sanitation at the same time (Figure 57).

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 87 Results

Figure 57: Deprivation overlaps for children aged 15–17 years

Figure 58 shows disparities between urban and rural areas for all three dimensions. While in rural areas, the deprivations on dimensions tend to overlap more for children, the opposite occurs in urban areas, where a much higher proportion of children tend to be deprived on only one dimension (18.7 per cent and 1 per cent, respectively). The percentage of non-deprived children was also much higher in urban areas (29.8 per cent) than in rural areas (8.8 per cent).

88 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 58: Deprivation overlaps for children aged 15–17 years by area

Rural Areas

Urban Areas

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 89 Results

Deprivations experienced by children aged 15–17 years

The distribution of the number of simultaneous deprivations experienced by children aged 15–17 years is presented in Figures 59a–b. Some 12 per cent of children experienced no deprivation at all; a large proportion of children were deprived on one, two or three dimensions at the same time (31.9 per cent, 32.8 per cent and 18.9 per cent, respectively); and around 4 per cent of children were deprived on four or five dimensions simultaneously.

Figure 59a: Percentage distribution of children aged 15–17 years by number of deprivations at national level

35 32.8 31.9 30

20 18.9

Percentage 12.1 10

5 4

0.4 0 012345 Number of simultaneous deprivaons

Figure 59b: Percentage distribution of children aged 15–17 years by number of deprivations by area 50 48.6 45 40 36.8 35 30 29.3 26.5 25 24.5 20.2 20 Percentage 15 10 6.5 5.2 5 1.6 0.2 0.1 0.5 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Urban Rural

90 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

There was a clear disparity between the number of deprivations experienced by children in urban and rural areas. While the majority of children in urban areas had no or only one deprivation, the majority in rural areas experienced two or more deprivations simultaneously. The deprivation distribution by other characteristics of the child is shown in Table 17.

Table 17: Deprivation distribution for children aged 15–17 years by child’s characteristics

Profiling variable 0 1 2 3 4 5

Secondary or higher 18.4* 38.5* 29.2* 12.0* 1.8* 0.2 Education level of Primary education 6.6* 28.6* 34.9* 23.8* 5.7* 0.5 household head No education 5.7* 14.9* 40.5* 31.0* 7.3* 0.7

More than 6 11.1 28.3* 35.5* 20.5 4.41 0.2 members Household size 4–6 members 13.2 34.9* 30.5* 17.6 3.31 0.5

1–3 members 10.9 30.6* 33.9* 19.4 4.97 0.3

Above median 10.9* 31.6 33.8* 19.3 4.18 0.2* Number of number of children children in household Below or equal to (median=2) median number of 14.8* 32.6 30.5* 18 3.47 0.7* children

Labour constrained: dependency ratio 8.8* 29.1* 34.2 22.6* 4.67 0.7* Labour- greater than 2 constrained household Household not 14.6* 34.0* 31.7 16.2* 3.44 0.2* labour constrained

Girl has been/ 0.2* 11.2* 39.6* 39.9* 7.8* 1.2* Early is married marriage Girl has never been (girls only) 14.8* 36.8* 31.3* 13.4* 3.6* 0.1* married

Girl has been/ 0.9* 9.5* 40.6* 39.4* 8.4* 1.3* Early is pregnant pregnancy (girls only) Girl has never been 14.0* 35.6* 31.7* 14.9* 3.8* 0.1* pregnant

Q5 (Richest) 25.7* 38.2* 27.1* 8.3* 0.6* 0.0*

Q4 17.8* 44.2* 28.4* 8.7* 1.0* 0.0* Asset index Q3 9.8* 36.5* 36.9* 13.2* 3.4* 0.1* (proxy for wealth) Q2 3.3* 25.2* 38.8* 28.6* 4.0* 0.1*

Q1 (Poorest) 0.8* 15.2* 34.1* 37.2* 11.2* 1.7*

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 91 Results

3.5.2.2 Multidimensional deprivation indices

Table 18 shows that 87.9 per cent of children aged 15–17 years were deprived on at least one dimension, while 0.4 per cent of them were deprived on all five dimensions at the same time. Given that a child is considered multidimensionally poor if he/she is deprived on at least two dimensions of well-being, 56 per cent of children aged 15–17 years were multidimensionally poor; they experienced on average 2.5 out of five deprivations simultaneously, implying 50 per cent of the total number of dimensions. The deprivation indices for the various thresholds are shown in Table 17 at the national level, while Figure 60 depicts the ratio at the urban and rural level as well. A much higher prevalence of deprivation was noted for children in rural areas for all the values of K used.

Table 18: Multidimensional deprivation indices for children aged 15–17 years at national level

1–5 87.9 39.1 2.0 0.344

2–5 56 50 2.5 0.280

3–5 23.2 64 3.2 0.149

4–5 4.33 81.7 4.1 0.035

5 0.371 100 5 0.004

Figure 60: Multidimensional deprivation headcount for children aged 15–17 years by various cut-off points 100 93.5 90 87.9 80

70 70.7 66.9 60 56 50

Percentage 40

30 30.1 23.2 20 22.1 5.6 0.3 0.1 0.4 10 4.3 0 1.9 0.5 1-5 dimensions 2-5 dimensions 3-5 dimensions 4-5 dimensions 5 dimensions

Urban Rural Naonal

92 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Although the deprivation headcount varies significantly across the provinces, there is less variance for the average intensity of deprivation across multidimensionally poor children, apart from in Harare and Bulawayo provinces. This implies that multidimensionally poor children in those provinces have on average the same level of deprivation. Bulawayo Province again had the lowest M0 of 0.084, while Matabeleland North Province had the highest M0 of 0.386.

Figure 61: Multidimensional deprivation indices (K=2) for children aged 15–17 years by province

80 0.5

0.4 60 0.386 0.353 0.346 0.3 0.304 40 0.282 0.285 0.288 0.280 0.261 0.2 headcount 20 and intensity, % and intensity, 0.118 0.1 0.084 deprivaon Adjusted Deprivaon headcountDeprivaon 0 0.0

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 93 Results

Multidimensionally poor children aged 15–17 years are profiled in Figure 62.

Figure 62: Multidimensional deprivation headcount (K=2) for children aged 15–17 years by profiling characteristics Percentage of children deprived in 2-5 dimensions 0% 25% 50% 75% 100%

Naonal Naonal 56% female 54% Child's gender* male 58% Secondary or higher 43% Educaonal level Primary educaon 65% of hh head* No educaon 80% hhsize >6 61% Household size* hhsize 4-6 52% hhsize 1-3 59% no religion 67% other religion 72% tradional 77%

Religion* other chrisans 61% apostolic sect 60% pentecostal 42% protestant 43% roman catholic 50% No. of children in Above median number of children 58% hh (media n =2)* Below or equal to median number of children 53% Labour Labour constraint: dependency rao>2 62% constraint hh* Hld not labour constraint 52% Early marriage Girl has been/is married 89% (girls only)* Girl has never been married 48% Early pregnancy Girl has been/is pregnant 90% (girls only)* Girl has never been pregnant 51% Q5 (Richest) 36% Q4 38% Asset Index Q3 54% (proxy for wealth)* Q2 72% Q1 (Poorest) 84% rural non-poor 54% Area and poverty urban non-poor 19% status (based on asset index) of hh* rural poor 72% urban poor 40% female headed hh (non-poor) 41% Gender of hh head and poverty status male headed hh(non-poor) 34% (based on asset female headed hh(poor) 67% index) of hh* male headed hh(poor) 74%

94 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Decomposition of the multidimensional deprivation ratios

The adjusted deprivation headcount (M0) of 0.34 in rural areas (compared to 0.093 in urban areas) shows the higher prevalence, depth and intensity of deprivation among rural children aged 15–17 years. Decomposing the M0 reveals that Water (22.9 per cent), Health (8.2 per cent) and Information (6.9 per cent) contributed relatively more to deprivation levels in rural areas, while the main contributors in urban areas were Education (43.3 per cent) and Sanitation (25.2 per cent). See Figure 63a–b.

Figure 63a: Adjusted deprivation headcount rate for children aged 15–17 years by area

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4

Urban 0.093

Rural 0.34

Naonal 0.28

Figure 63b: Adjusted deprivation headcount rate and its decomposition (K=2) for children aged 15–17 years by area

100 1.4 6.9 6.5

75 46.5 36.7 37.5

50 6.1 22.9 21.5 Percentage

25 43.3 25.2 26.7

8.2 7.8 0 2.6 nabrU laruR lanoitaN

Health Educaon Water Sanitaon Informaon

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 95 Results

3.6 Analysis of multidimensional poverty according to the wealth status of children

For this analysis, wealth quintiles were used to categorize children in each age group. Since there were no data on income or consumption in the MICS datasets, an arbitrary value of wealth index was assigned to each household using the multiple correspondence analysis (MCA) technique, based on a selection of assets and some household characteristics. Households were then classified in increasing order of wealth status, thus determining the quintiles used throughout this section.20 According to the Poverty Income Consumption and Expenditure Survey (PICES) 2011/12, about 63 per cent of households were deemed poor (ZimStat, 2012); this implies that households in the first three wealth quintiles were poor in absolute terms. These were considered such in this analysis. Poverty based on the wealth index is here referred to as monetary poverty.

The analysis is divided into three parts: single deprivation analysis, multiple deprivation analysis and overlap between deprivation and poverty.

3.6.1 Single deprivation analysis

The single deprivation analysis by wealth quintile for each age group is shown in Figures 64–67. It can be observed that in all cases the difference between the five quintiles was significant on all dimensions. Across the age groups, big disparities of between 33 and 41 percentage points were found between the poorest and the richest quintiles on the Water dimension. On the Sanitation dimension, the disparities ranged from 22 to 39 percentage points.

For the dimensions Nutrition and Health, although the Chi-square test revealed that the difference is significant, the variance in the deprivation rates is relatively small, compared to the Water and Sanitation dimensions. For children aged 5 years and above, access to Information seems to be driven by the wealth status of the household. Despite the fact that in most cases, the percentage of deprived children is higher in poorer households, high deprivation rates still exist among the richest households. For example, 62 per cent of children aged 24–59 months in the wealthiest quintile were deprived on Sanitation. This confirms that wealth is not always a good indicator of well-being when it comes to children.

20 The method used to determine the wealth quintile is documented in the Annex.

96 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 64: Deprivation rates for children aged 0–23 months by wealth quintile 80

70 67 64 59 60 58 59 57 54 55 56 52 53 50 46 45 46 43 42 40 40 35 33 30 30 23 23 20 20 18 14

Percentage of deprived children deprived of Percentage 10

0 Nutrion* Health* Physical Development* Water* Sanitaon* Dimensions of deprivaon Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Figure 65: Deprivation rates for children aged 24–59 months by wealth quintile

100 96 90 90 83 80 79

70 62 60 59 55 57 53 52 51 52 50 49 46 47 43 43 40 40 35 34 30 27 25 23 19

Percentage of deprived children deprivedof Percentage 20 15 10

0 Health* Child Development* Physical Development* Water* Sanitaon* Dimensions of deprivaon Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 97 Results

Figure 66: Deprivation rates for children aged 5–14 years by wealth quintile 100 95 90 88

80 78 73 70

60 60 57

50 43 40 33 31 31 30 24 21 22

Percentage of deprived children deprivedof Percentage 20 17 14 16 15 15 14 13 11 10 6 3 0 0 0 Health* Educaon* Water* Sanitaon* Informaon* Dimensions of deprivaon Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Figure 67: Deprivation rates for children aged 15–17 years by wealth quintile

100 94

90 87

80 77

70 68

60 58 55 53 50 43 43 41 40 39 35 29 30 28

Percentage of deprived children deprived of Percentage 20 20 17 16 17 16 13 10 10 6 2 0 0 0 Health* Educaon* Water* Sanitaon* Informaon* Dimensions of deprivaon Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

98 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

3.6.2 Multiple deprivation analysis

3.6.2.1 Number of simultaneous deprivations experienced by children by wealth quintile

The number of simultaneous deprivations experienced by children in every age group by wealth quintile is shown in Figures 68–71. The majority of children in the richest quintile tended to be deprived on few dimensions, while those in the poorest quintile were deprived on many dimensions at the same time. Nonetheless, it is worth noting that even children from the wealthiest quintile experienced several deprivations simultaneously. For example, 19 per cent of children aged 0-23 months in the richest quintile were deprived on three dimensions simultaneously.

Figure 68: Number of simultaneous deprivations for children aged 0–23 months by wealth index 35 35 34 32 32 30 31 30 28 27 28 25 25 23 21 20 19 20 18

15 14 14

10 11 10 9 6 7 7

Percentage of deprived children deprived of Percentage 5 5 4 4 2 2 0 1 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Figure 69: Number of simultaneous deprivations for children aged 24–59 months by wealth index 50

45 44 41 42 39 40 38 35 32 30 29 26 25 25 24 24 21 19 20 18 16 15 11 10 9 8 7

Percentage of deprived children ofdeprived Percentage 6 4 4 5 3 3 2 2 1 0 0 1 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 99 Results

Figure 70: Number of simultaneous deprivations for children aged 5–14 years by wealth index

60 55 50 44 44 40 39 40

33 32 30 29 29

21 20 19 19 20 19

11 11 10 10 4 6 5 4 Percentage of deprived children deprivedof Percentage 1 1 2 0 0 0 0 0 0 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Figure 71: Number of simultaneous deprivations for children aged 15–17 years by wealth index

50 45 44

40 38 39 37 37 37 35 34

30 28 29 27 26 25 25

20 18 15 15 13 11 10 10 8 9

Percentage of deprived children children ofPercentage deprived 5 3 3 4 1 1 1 2 0 0 0 0 0 0* 1* 2* 3* 4* 5* Number of simultaneous deprivaons

Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

3.6.2.2 Multidimensional deprivation ratios by wealth quintile

To compute the percentage of deprived children for the multiple deprivation analysis, it was assumed that K=2. This implies that a child was considered to be deprived if he/she was deprived on at least two dimensions. In this section, the multidimensional deprivation ratios for each age group are disaggregated by the wealth quintile. The higher the wealth quintile, the lower the multidimensional deprivation headcount rate for all age groups. In general, the average intensity of deprivation was almost the same for multidimensionally poor children across the different wealth quintiles. This implies that on average deprived children experienced the same number of deprivations, irrespective of their wealth quintile (Figures 72–75).

100 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 72: Multidimensional indices (K=2) for children aged 0–23 months by wealth quintile

100 0.6

80 0.537 0.5 0.445 0.4 60 0.373 0.3 0.315 40 0.281

0.2 headcount intensity, % intensity, 20 0.1 Adjusted deprivaonAdjusted 0 0.0 Deprivaon headcount and headcount Deprivaon Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

Figure 73: Multidimensional indices (K=2) for children aged 24–59 months by wealth quintile

100 0.7 0.6 80 0.592 0.538 0.5 60 0.453 0.4 0.384 40 0.336 0.3 headcount intensity,% 0.2 20

0.1 deprivaon Adjusted

Deprivaon headcount and headcount Deprivaon 0 0.0 Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

Figure 74: Multidimensional indices (K=2) for children aged 5–14 years by wealth quintile

100 0.5

80 0.424 0.4

60 0.3 0.305 40 0.2

0.206 headcount and intensity, % intensity,and

20 0.122 0.1 deprivaon Adjusted

Deprivaon headcount Deprivaon 0.113

0 0.0 Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Deprivaon headcount (H) %* Average intensity across the deprived (A) in %* Deprivaon headcount adjusted for intensity (M0)*

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 101 Results

Figure 75: Multidimensional indices (K=2) for children aged 15–17 years by wealth quintile

100 0.5

0.465 80 0.4

0.36 60 0.3

0.256 40 0.2 headcount

and intensity, % intensity,and 0.163 0.173 Adjusted deprivaonAdjusted Deprivaon headcount Deprivaon 20 0.1

0 0.0 Q5 (Richest) Q4 Q3 Q2 Q1 (Poorest)

Deprivaon headcount (H) %* Average intensity across the deprived (A) in % Deprivaon headcount adjusted for intensity (M0)*

3.6.2.3 Decomposition of the deprivation headcount adjusted for intensity (M0) by poverty status

The deprivation headcount adjusted for intensity (M0) shown in the above section was decomposed to identify the contribution of each dimension to the deprivation level for each age group according to poverty status based on the wealth index. In most cases, it was found that the dimensions Sanitation, Physical development, Child development and Health contributed almost the same to the deprivation level, irrespective of the poverty status of the child. However, while the Water and Information dimensions contributed relatively more to the deprivation rate for children living in poor households, Education contributed more in non-poor households. It is important to note that each of the contributions was relative. Thus, actual deprivation in Education might not be more severe for children living in non-poor households than in poor households, but its relative importance was higher (Figures 76–80).

Figure 76: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 0–23 months based on poverty status 100

90 20.5 20.2 21.4 80 9.5 70 15.3 17.8 60 12.9 13.8 14.2 50 40 28.2 Percentage 26 25 30 20 28 10 24.3 22.8 0 lanoitaN rooP hh roop-noN hh

Nutrion Health Physical development Water Sanitaon

102 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 77: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 24–59 months based on poverty status 100 90 32.5 31.9 80 34.1 70

60 14.7 16.7 9.4 50 23.1 40 20.8 19.9 Percentage 30 12.7 12.8 12.5 20

10 19.3 18.7 20.9 0 lanoitaN rooP hh roop-noN hh

Health Physical development Child development Water Sanitaon

Figure 78: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 5–14 years based on poverty status 100 0.4 9 10.4 90

80 42 70 39 38.4 60 50 27.6 40 Percentage 26.8 26.6 30 20 18.6 16.3 15.8 10 9 8.4 11.4 0 lanoitaN rooP hh roop-noN hh

Health Educaon Water Sanitaon Informaon

Figure 79: Decomposition of the adjusted deprivation headcount rate (K=2) for children aged 15–17 years based on poverty status 100 0.1 6.5 8.7 90 80 39.5 37.5 36.9 70 60 50 20.2 21.5 22 40 Percentage 30 31.9 20 26.7 24.9 10 7.8 7.6 8.3 0 lanoitaN rooP hh roop-noN hh

Health Educaon Water Sanitaon Informaon

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 103 Results

3.6.3 Overlap analysis between monetary poverty and multidimensional deprivation

In this section, the overlap of monetary poverty and multidimensional deprivation was analysed for each age group. This type of analysis provides an indication of the ability of monetary poverty to measure the overall well-being of children, compared to the multidimensional deprivation used in the MODA methodology.

For the purposes of this section, children of all age groups were considered to be poor if they belonged to a household in the first three quintiles of economic well- being. On the other hand, they were considered to be deprived if they experienced two or more deprivations (K=2).

Figure 80 shows that 47.9 per cent of children aged 0–23 months were both poor and deprived. It is noteworthy that 22.8 per cent of children were deprived but non-poor, while 13.2 per cent were poor but non-deprived. This illustrates that with children, using monetary poverty to measure deprivation can be misleading. Although the overlap between poverty and deprivation was high (47.9 per cent), it is not a complete overlap between the two concepts of poverty measurement. Some 16 per cent of children in this age group were neither poor nor deprived. See Figures 81–83 for age groups 24–59 months, 5–14 years and 15–17 years.

Figure 80: Overlap between poverty and deprivation for children aged 0–23 months Non-poor and non-deprived (16.0%)

Poor only Deprived only (13.2%) (18%)

Overlap between poverty and deprivaon (47.9%)

104 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Results

Figure 81: Overlap between poverty and deprivation for children aged 24–59 months Non-poor and non-deprived (10.4%)

Poor only Deprived only (6.2%) (25.7%)

Overlap between poverty and deprivaon (57.7%)

Figure 82: Overlap between poverty and deprivation for children aged 5–14 years Non-poor and non-deprived (27.2%)

Deprived only Poor only (9.7%) (22.7%)

Overlap between poverty and deprivaon (40.4%)

Figure 83: Deprivation overlap between poverty and deprivation for children aged 15–17 years Non-poor and non-deprived (15.7%)

Deprived only Poor only (15.7%) (17.2%)

Overlap between poverty and deprivaon (40.2%)

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 105 References

Alkire, S., and Foster, J. (2011). Counting and multidimensional poverty measurements. Journal of Public Economics, 95, pp. 476–487. Alkire, S., and Santos, M.E. (2010). Acute multidimensional poverty: A new index for developing countries. OPHI Working Paper No. 38, University of Oxford. Association for Childhood Education International (August 2015). Education considerations in Zimbabwe. Available at: https://acei.org/zimbabwe Curtis, V., and Cairncross, S. (2003). Effect of washing hands with soap on diarrhoea risk in the community: a systematic review. The Lancet Infectious Diseases, 3(5), pp. 275–281. Ezeh, O.K., Agho, K.E., Dibley, M.J., Hall, J., and Page, A.N. (2014). Determinants of neonatal mortality in Nigeria: evidence from the 2008 demographic and health survey. BMC Public Health, 14(1), p. 521. Fink, G., Günther, I., and Hill, K. (2011). The effect of water and sanitation on child health: evidence from the demographic and health surveys 1986–2007. International Journal of Epidemiology, 40(5), pp. 1196–1204. Gordon, D., Nandy, S., Pantazis, C., Pemberton, S., and Townsend, P. (2003). The Distribution of Child Poverty in the Developing World. University of Bristol. Heckman, J., and Carneiro, P. (2003). Human capital policy. NBER Working Paper No. 9495. Kaltenthaler, E., Waterman, R., and Cross, P. (1991). Faecal indicator bacteria on the hand and the effectiveness of handwashing in Zimbabwe. Journal of Tropical Medicine and Hygiene, 94(5), pp. 358–363. de Neubourg, C., Chai, J., de Milliano, M., Plavgo, I., and Wei, Z. (2012). Step-by-step guidelines to the Multiple Overlapping Deprivation Analysis (MODA). UNICEF Office of Research, Florence. Pickering, A.J., and Davis, J. (2012). Freshwater availability and water fetching distance affect child health in sub-Saharan Africa. Environmental Science & Technology, 46(4), pp. 2391–2397. Roushdy, R., Sieverding, M., and Radwan, H. (2012). The Impact of Water Supply and Sanitation on Child Health: Evidence from Egypt. Population Council. UNESCO (2010). Education for All Global Monitoring Report 2010. Regional fact sheet – Sub-Saharan Africa. Available at: http://en.unesco.org/gem-report/sites/gem-report/files/gmr2010-fs-ssa.pdf UNICEF (2007). Global Study on Child Poverty and Disparities 2007–2008: Guide. Division of Policy and Planning, New York. UNICEF Zimbabwe. Available at: http://www.unicef.org/zimbabwe/basic_education_1729.html USAID. Available at http://www.mchip.net/node/2941 VVOB Education for Development Zimbabwe. Available at: http://www.vvob.be/zimbabwe/what-we-do/earlychildhooddevelopment WFP (2015). What are the current issues in Zimbabwe? Available at: https://www.wfp.org/countries/zimbabwe Zimbabwe Situation (2013). Water-borne diseases resurface. Available at: http://www.zimbabwesituation.com/news/zimsit_water-borne-diseases-resurface/ ZimStat (2012). Poverty and Poverty Datum Line Analysis in Zimbabwe 2011/12. Available at: http://www.zimstat.co.zw/dmdocuments/Finance/Poverty2011.pdf

106 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Annexes

ANNEX 1: Selected dimensions, indicators and thresholds for each age group

Age groups 0–23 24–59 5–14 15–17 Dimension Indicator Threshold months months years years 0–5 months: not exclusively Infant and young breastfed child feeding Nutrition (incl. exclusive 6–23 months: not meeting WHO X breastfeeding for standards for meal frequency and 0–5 months) food diversity

0–4 years old: child did not receive Full immunization all of the following vaccination on X X time: BCG, DPT, Polio, PCV, MMR

0–23 months: mother does not have knowledge of HIV/AIDS Mother’s transmission via pregnancy, knowledge delivery and breastfeeding or was X of MTCT not offered an HIV test during antenatal care or did not accept the Health test or did not receive the results

0–23 months: unskilled birth Skilled birth attendant/s (relative, friend, parent, X attendance other, no one) assisted with the child’s birth

0–17 years: household uses solid cooking fuel (wood, straw/shrubs/ Cooking fuel X X X X grass, other, etc.) and cooking is done inside the house

Early childhood X 36–59 months: child does not education (36–59 attend any ECD programme attendance months)

24–59 months: mother/father/any household member does not do any of the following activities with the Child–adult Child child: read books, tell stories, sing X X interaction development songs, take the child outside, play with the child or name/count/draw things together

24–59 months: child has none of Availability of the following: books, bought toys, X toys or books home-made toys or household items used as toys

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 107 Annexes

Age groups 0–23 24–59 5–14 15–17 Dimension Indicator Threshold months months years years 0–59 months: child’s height for Stunting age is less than -2 SD from X X international median (WHO 2006) 0–59 months: child’s weight for Physical Wasting height is less than -2 SD from X X development international median (WHO 2006) 0–59 months: child’s weight for Underweight age is less than -2 SD from X X international median (WHO 2006) School 5–17 years: child does not attend X X attendance school (ECD, primary or secondary) Primary school 15–17 years: child did not Education X attainment complete primary school 5–14 years: child is more than Grade-for-age X 2 years late for school 0–17 years: household’s main source of drinking water is unimproved (WHO) (unimproved sources: unprotected well, unprotected spring, tanker truck, small cart with tank/drum, surface Water source X X X X water (river, lake, dam, etc.), Water bottled water if non-drinking water is unimproved, other) and is not or is inappropriately treated (inappropriate: let it stand, strain through cloth, other) 0–17 years: time needed to go, Distance to get water, and come back is more X X X X water source than 30 minutes (WHO) 0–17 years: water and soap is not Handwashing available for handwashing in the X X X X household 2–17 years: household uses unimproved toilet (WHO) (unimproved sources: flush/ Sanitation Toilet type X X X pour flush to elsewhere, pit latrine without slab/open bit, bucket, no facilities, bush or field) Sharing of 2–17 years: toilet facilities are X X X toilet facility shared by two or more households Availability of 5–17 years: household has not information/ reported having any of the following: Information X X communication TV, radio, phone, mobile phone, devices computer

108 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016 Annexes

ANNEX 2: Indicator ‘Wish-list’

Age group 15–17 15–17 Dimension Indicator 0–23 24–59 5–14 years years months months years (girls) (boys) School drop-out – specifying the Education X X X reason for dropping out X (9–14 Child pregnancy Health years) Maternal mortality X X Birth certificate/formal ID X X X Child labour (include domestic Protection X X X chores) Sexual abuse X X X Quantity of water used by the Water X X X X X child per day Food frequency, food diversity, Nutrition X X X calories Handwashing arrangements (rather than handwashing place); Sanitation X X X X X handwashing before a meal, after going to the toilet, etc. Access to number of hours of light per day/night (energy poverty is an X X X important concept in Zimbabwe) Type of fuel used for lighting and X X X X X Housing heating No. of square metres per person X X X X X in a household Availability of an improved oven X X X X X

Desirable but not available indicators in the MICS5 to be advocated for in the next MICS, PICES or any other survey

Wealth index

Given that the MICS datasets do not have enough information on household income or consumption, a wealth index was created to determine the poverty status of a household. An arbitrary value for the wealth index was assigned to each household, based on a selection of assets and some household characteristics listed in Annex 3 which reflect wealth status. The assets/ characteristics selected for the analysis differ by the geographical location of the household. Using multiple correspondence analysis (MCA), an adjusted index applicable at the national level was constructed using linear

July 2016 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis 109 Annexes

regression models. Households were then classified in increasing order of wealth status; this determined the quintiles used throughout this chapter. Those quintiles were used as profiling variables in the N-MODA analysis for Zimbabwe.

According to the PICES 2011/12, 62.6 per cent of Zimbabwean households were deemed poor (ZimStat, 2012); this implies that households constituting the first three quintiles of the wealth index were poor in absolute terms. In this analysis, a child was considered poor if his/her household was in the first three quintiles based on the wealth index.

The following annex shows the assets and characteristics that were chosen to build the wealth index in Zimbabwe.

ANNEX 3: Assets used to build the wealth index

Assets in urban households Assets in rural households Floor material Floor material

Wall material Wall material

Roof material Roof material

Radio Radio

Television Television

Mobile phone Mobile phone

DVD/VCD DVD/VCD

Bicycle Bicycle

Motorcycle/scooter Motorcycle/scooter

Battery/generator Battery/generator

Computer Solar panel

Satellite dish Animal-drawn cart

Deep freeze Wheel barrow

Wristwatch Livestock, herds or farm animals

Refrigerator

Car/truck

110 Child Poverty in Zimbabwe: A Multiple Overlapping Deprivation Analysis July 2016