Know of a great book that might interest others?

Why not suggest a reading or submit a review? Find out more about Adding or Reviewing Readings.
Bookmark this
Add this page to your development related bookmarks using the button below.

AddThis Social Bookmark Button


Visitors: 4194265
We have 138 guests online
Browse the latest additions below or choose from the list of most viewed Readings & Reviews on the right. Members are invited to add or review readings. To do so, please find out more about Adding or Reviewing Readings.
Child poverty in the developing world
User Rating: / 6
Contributed by The Editors   
Studying Development: Books and Reviews(2003) by David Gordon, Shailen Nandy, Christina Pantazis, Simon Pemberton, and Peter Townsend
This book is the result of a large UNICEF-funded study and constitutes a significant step forward in grasping the extent and severity of poverty experienced by children in developing countries.

The authors undertook what is the first ever comprehensive quantitative estimation of absolute child poverty across developing countries.

Given that much research has highlighted the inadequacy of household measures as a proxy for child poverty, this book and its underlying report offer a much-needed assessment of how widespread and severely poverty affects the lives of children themselves.

Cover description:

none at present

DAVID GORDON is Professor of Social Justice, SHAILEN NANDY is a Research Associate, CHRISTINA PANTAZIS is a Research Fellow, SIMON PEMBERTON is a Post-Doctoral Research Fellow, all at the School for Policy Studies, University of Bristol. PETER TOWNSEND is Professor of International Social Policy at the London School of Economics and Political Science.

This report can currently be found online at:


reproduced with kind permission from 'Children, Youth and Society'

Children, Youth and Environments
Volume 14, Issue 1 (2004)

Child Poverty in the Developing World
Gordon, David et al. (2003).
Bristol, UK: The Policy Press. ISBN# 1 86134 559 3.
This brief report is a summary of a large UNICEF-funded research project undertaken by the London School of Economics and the University of Bristol, which focused on determining the extent and severity of child poverty in the low-income countries. Most of the data used in this study were drawn from existing Demographic and Health Surveys (DHS) for 46 countries (including China, which had its own survey). Based on detailed face-to-face interviews, the data covered a sample of nearly 1.2 million children (0-18 years) from 380,000 households– according to the authors, probably the largest, most accurate survey sample of children ever assembled.

This report serves as a useful complement to the paper by Rainwater and Smeeding in the last issue of this journal, which compared child poverty in the U.S. to that in 14 other rich nations. However, the two projects have important differences in the approaches they take in measuring child poverty. Both projects reject the notion that absolute income is an adequate measure of poverty (the assumption on which the World Bank’s $1 per day poverty line is based.) Rainwater and Smeeding use the widely accepted approach of defining poverty in terms of the median household income within a given country- an approach that views poverty as relative, reflected in the incapacity of those below a certain income level to act and participate as full members of a particular society.

The research described in the report reviewed here takes a different tack. It focuses on absolute rather than relative poverty, but it also rejects the notion that income is an adequate yardstick. It questions the assumption that household poverty is an accurate proxy for the poverty of the children within that household. According to this approach, income is a critical indicator of poverty, but it is by no means the only factor affecting people’s standard of living or their capacity to participate as part of the social mainstream. The cost of food, water and decent shelter, for instance, can vary widely not only from country to country, but also within countries, especially between rural and urban areas. Access to information, education, health care and other basic services, the strength of social networks and the potential for involvement in local political processes all make a difference in the burden of poverty, and these factors are neither wholly dependent on income, nor reliably predicted by income. Nor, according to this project, is it reasonable to assume that the resources of a given household are an adequate predictor of the well-being of the children in that household. This would imply that resources are equally shared within households. In fact, as this report points out, many households make enormous sacrifices to meet children’s needs; in others, few of the benefits of what is earned or produced actually trickle down to reach children.

The research reported here starts instead from a view of poverty as characterized by a range of deprivations which can affect people independent of income, and which can affect children independent of the level of household poverty. Deprivation for children is defined here in terms of the circumstances most likely to have serious consequences for their health, well-being and development. Eight criteria are used in developing a taxonomy of deprivation for children: food, safe drinking water, sanitation facilities, shelter, health, education, access to information and access to health and education services. (A child’s contact with health services and actual exposure to education may not relate solely to the availability of these services– hence, I assume, the added category of access.) The definition of absolute poverty in children, which draws on this taxonomy, is very strict. Children are considered to be living in absolute poverty only if they experience “severe” deprivation in at least two of these areas– and the definitions for severe deprivation go beyond what would generally be considered threatening to children’s well-being. For instance, severe deprivation in the area of food means serious malnutrition– not simply going hungry on occasion; sanitation deprivation means no facilities near the dwelling, not just the absence of indoor plumbing; health deprivation means no immunization at all and contact with only limited non-professional care, not just expensive or inadequate medical care; deprivation in the area of access to health services or schools means facilities a day’s travel away, not just at an inconvenient distance. The extreme quality of these measures reflects an attempt on the part of this research to err on the side of caution in defining the scope and nature of child poverty; no one could argue that deprivation at this level does not reliably represent significant poverty.

Despite the high bar that was set in defining poverty for children in this research, the number of children worldwide that fall into this category is shockingly high. By these standards, over a third of the children in the world’s low and middle-income countries qualify as living in absolute poverty. Moreover, over half of all children in these countries are considered to be “severely deprived” (i.e., subject to severe deprivation in at least one of the categories.) The severe deprivations that affect the largest number of children are related to shelter, sanitation, information and water. Over one-third of children live in dwellings with more than five people to a room or with mud floors; 31 percent lack access to any toilet facilities at all; 25 percent lack access at home to newspapers, radio, television and telephones; and over 20 percent are using unsafe water sources, or have to walk more than 15 minutes for water. Severe deprivations in the areas of food, health and education are less common (15 percent, 15 percent and 13 percent, respectively).

Sub-Saharan Africa has the highest rates of absolute child poverty and severe deprivation, followed by South Asia. Within each region, however, there is also considerable variation. There are significant gender discrepancies in the area of education– girls are at least 60 percent more likely than boys to be seriously deprived in this area– but the overall gender differences in other areas tend to be relatively small. Boys are more likely to be severely food deprived, except in South Asia, and also more severely deprived in terms of health in most countries.

The report shows significant differences in all regions and in all areas of deprivation between rural and urban children, and it is here that the inherent weakness in any large one-size-fits-all survey of this kind is most clearly revealed. Statistics have long shown an urban advantage with regard to poverty, but as has been repeatedly demonstrated, disparities in urban areas are enormous, and aggregated statistics mask the depth of poverty in cities (Mitlin and Satterthwaite 2001). It is also important to point out that conditions that might be challenging in rural areas can become unbearably difficult where there are high concentrations of people.

A few examples: the report points out that the single greatest difference in severe deprivation between rural and urban children is in the area of sanitation. In rural areas, 41 percent of children are rated as severely deprived in this area– that is, they lack access to any toilet facilities. Only 9 percent of urban children are rated as similarly deprived. However, the toilet facilities that serve many of these urban children, and that theoretically keep them out of the severely deprived category, are often public or community latrines that are seriously overcrowded and under-maintained. Research from around the world indicates how few young children actually use these toilets; they are pushed out of line by busy adults, they are fearful of harassment, they are put off by the dark and the smells, and when they are small, they are frightened of falling into overly large pit openings. Most of them squat outside instead. A survey conducted by UNICEF’s India office in poor urban areas found that only 1 percent of children under six used latrines, that the feces of an additional 5 percent were thrown into latrines, and that the remainder ended up in drains, streets or yards, increasing the likelihood of contamination and illness (UNICEF 2000). Even in areas where urban children are not rated as “severely deprived” with regard to sanitation, then, the threat to health and the true level of hardship rapidly becomes much greater than it is for similarly deprived children in far less congested rural areas.

The same holds true for shelter. Severely deprived children are defined as those living in homes with more than five people to a room, or with a mud floor. Among rural children, 42 percent experience this level of deprivation; among urban children, only 15 percent. But crowding is a different matter in a congested urban slum where there is limited space outside the home for play and for routine domestic activities. Mud flooring is also more likely to be a serious problem in overcrowded, poorly drained urban slums, where backed up drainage ditches and flooding can be routine events. An assessment of deprivation calls for different standards where the implications of given conditions are so different. Just as income-based poverty lines mask the relative levels of poverty in urban areas, where more of life’s necessities must be purchased, so do standardized assessments of living conditions overlook the relative hardship inflicted by these conditions. The authors of this report acknowledge that the strict standards they bring to bear on identifying poverty probably result in an underestimation of the problem. This is likely to be especially true when it comes to estimating the number of urban children living in absolute poverty or in severe deprivation.

Absolute poverty, as calculated here, affects almost a third of the world’s children; by contrast, less than 22 percent of the world’s population fell under the $1 a day poverty line in 2000. A more in-depth comparison than this between child poverty and general poverty, however, would have been an interesting addition to this report– in particular a comparison of the extent of child poverty as calculated by the means used in this research, and as calculated by measures that use the household as unit of analysis. Are children, on the whole, better or worse off than the rest of their households? How does this vary from region to region? An analysis of this issue would be a very interesting addition to our understanding of child poverty.


Mitlin, Diana and David Satterthwaite (2001). “Urban Poverty: Some Thoughts about Its Scale, Nature and about Responses to It.” In Shahid Yusuf, Simon Evenett and Weiping Wu (editors), Facts of Globalization: International and Local Dimensions of Development. Washington, D.C.: World Bank, 193-220.

UNICEF (2000). Multiple Indicator Survey. Delhi: UNICEF.

International Institute of Environment and Development

Sheridan Bartlett is a senior research associate in the Human Settlements Program at the International Institute of Environment and Development (IIED) in London, England. She is currently involved in issues pertaining to urban children, and is the managing editor of IIED’s journal, Environments and Urbanization. Dr. Bartlett is also a research associate at the Children’s Environments Research Group in New York, and works as a consultant to Save the Children’s International Alliance, conducting research and developing programs for young children in Nepal and Bangladesh. Recent publications include a UNICEF Innocenti Digest on urban children with David Satterthwaite, a review of children’s rights and the physical environment for Save the Children, Sweden, and articles in various journals on topics related to children’s environmental health. Dr. Bartlett is on the Editorial Board of Children, Youth and Environments.


Children, Youth and Environments.
Vol 14, No.1 (2004)

Response to Review of
Child Poverty in the Developing World
Dave Gordon, Shailen Nandy, Christina Pantazis,
Simon Pemberton, and Peter Townsend

Citation: Gordon, Dave et al. (2004). “Response to Review of Child Poverty in the Developing World.” Children, Youth and Environments 14(1). Retrieved from

We thank both the reviewer and CYE for reviewing Child Poverty in the Developing World. The report summarized a project conducted for UNICEF by the Townsend Centre for International Poverty Research at the University of Bristol and the London School of Economics. A more detailed publication is in preparation, which discusses many of the salient points raised by the reviewer, and also includes data from a number of other countries and regions.

The review highlights a number of very important issues regarding the extent of children's absolute poverty in urban areas. We are in complete agreement with the reviewer on the inadequacies of currently available survey data to identify the enormous disparities in some urban areas. We welcome the recent proposal that UNICEF should include an urban “slum” identifier in its third round of Multiple Indicator Cluster surveys (MICs3). This should help to produce more detailed measures of urban disparities in child poverty, however getting an international agreement on the definition of a “slum” may prove to be problematic.

Our research aimed to produce the first scientific estimates of child poverty in the developing world based on the definition agreed upon by the governments of 117 countries at the 1995 World Summit on Social Development and within the framework of International Human Rights conventions. Absolute poverty was defined as

a condition characterized by severe deprivation of basic human needs, including food, safe drinking water, sanitation facilities, health, shelter, education and information. It depends not only on income but also on access to social services.

We have operationalized this definition specifically for children using definitions of deprivation that are much more severe than have been used previously. In this way we have erred on the side of caution and our estimates of child poverty should only be considered a minimum- the “true” extent of absolute poverty in both urban and rural areas is certainly much higher. For example, the use of crowding as a measure of shelter deprivation originates from the work of Charles Booth on London data from the 1891 Census. In 1895 Booth argued that

A man and his wife and one child, or a widow with two children may occupy only one room; or a family of six or seven may have only two rooms; and yet not be ‘very poor’ in the sense of suffering ‘chronic want.’ But when four or more persons live in one room or eight or more in two rooms, there must be great discomfort, and want of sufficient food, clothing, and firing must be a frequent incident. I have therefore drawn the line at this point.

In our research, we have defined severe housing quantity deprivation as more than five people per room thus we have erred on the side of caution by using a much harsher definition of crowding than that used by Charles Booth to estimate the number of “very poor” people living in the worst slum conditions of nineteenth century London. If evidence were available which demonstrated that crowding was “worse” for children in a congested urban slum than in a rural area then it would be possible to use different crowding thresholds in urban slums, non-slum urban areas and rural areas (once these area types can be identified in survey data). Similarly, different thresholds for sanitation deprivation could be used for young children in urban slum areas.

We agree with the reviewer on the current limitations of our work and hope to remedy some of these in a much larger study for which we have received funding from the UK Department for International Development.

click to purchase via

< Prev   Next >