Growing Up in Poverty

Findings from Young Lives

M. Bourdillon, J. Boyden, M. Bourdillon, J. Boyden

• Publisher: Palgrave Macmillan
• Year: 2014
• Print ISBN: 9781137404022
• eBook ISBN: 9781137404039
• Topic: Politics & International Relations
• Subtopic: Development Economics
• Index: Politics & International Relations

This book presents the latest evidence from Young Lives, a unique international study of children and poverty. It shows how the persistence of inequality amid general economic growth is leaving some extremely poor children behind, despite the promises of the Millennium Development Goals.

Part I

Poverty and the Lives of Growing Children

1

How Does Where Children Live Affect How They Develop? Evidence from Communities in Ethiopia and Vietnam

Paul Dornan and María José Ogando Portela

Introduction: why place matters for children’s development

The importance of the environment around the family for children’s development has been long recognized (for example, Bronfenbrenner 1986: 724).

As well as conceptually important, this makes intuitive sense: few would argue that differences in the life chances of Swedish or Afghan children can be sensibly explained by household circumstances alone. Yet our understanding of the impact of child poverty in developing countries is often reliant on household-based explanations, with comparatively little known about how where children live affects how they develop.

Understanding how children’s chances are affected by where they grow up is important for policy: knowing more about the links between area characteristics and child development can inform policy priorities and principles for allocation or targeting. If research used to understand the intergenerational transmission of poverty underplays area characteristics as an explanation, these may focus attention on individualized, household-level explanations, whilst underplaying the wider economic opportunity and social environment of the community. In either case, there is a need for some caution in explaining ultimate causes – ‘[t]here is no area problem, merely national structural problems that find their place locally’ (Glennerster et al. 1999, paraphrasing Peter Townsend).

In analysing the causes and consequences of area deprivation, Diez Roux suggests a combination of ‘compositional’ and ‘contextual’ explanations (2001: 1784). Compositional explanations suggest areas are measured as poor because there is geographic clustering of households that are themselves deprived (for example, by ethnicity, livelihood, or education level). In this case, place is a proxy for household disadvantage. Contextual explanations highlight something additional about the nature of the community – infrastructure, social networks, or economic opportunities – which affect the circumstances of the individuals in a particular area.

For policy, these compositional and contextual distinctions overlap. Both the countries considered in this chapter use area-based policy. For example, Ethiopia’s Productive Safety Net Programme (PSNP), which provides cash or food primarily in return for work, has been implemented in food-insecure rural areas. (It has an area-based element, alongside a household-targeting mechanism within these areas – see World Bank no date: 18–23.) The Vietnamese Government has used the area-based Programme 135 (the ‘Programme for socio-economic development in communes faced with extreme difficulties’) to target poorer communities and to improve the local infrastructure (see MOLISA and UNDP 2004: 25–7; UNDP 2011: 88). Since these programmes cover populations which are typically very poor, they are relatively efficient at targeting poor households though they exclude any poorer people ­living in less poor areas. But alongside this, both policies also aim at area development, implying a contextual understanding of why people in specific areas face deprivation. (Programme 135 improves infrastructure like schools, while the PSNP uses public works to, for example, improve irrigation for better land quality.) Area-based policy therefore could address both contextual and compositional aspects of area deprivation; however, an understanding of what is driving poorer outcomes may be useful for the design of initiatives.

Having established that there is an important aspect of poverty that relates to where children live, in this chapter, we test the extent to which this features in child poverty research. We then demonstrate dramatically changing community context in Ethiopia and Vietnam between 2002 and 2009. The chapter then explores variations in indicators of child well-being in middle childhood according to where children live (indicating the extent to which area targeting would reach children experiencing low well-being, higher stunting rates, and other disadvantages) and provides some exploratory analysis of the role of neighbourhood factors in children’s development (seeking to unpick what the channels for these impacts may be).

How analysis of child poverty features the impact of where children live

There is considerable agreement that there is a spatial element to poverty. Studies demonstrate dramatic differences in the extent of deprivations affecting children between urban and rural areas or by region.1 Gordon et al., for example, demonstrate (for developing countries overall) that nearly half of the children in rural areas (48 per cent), compared with one in ten (12 per cent) in urban communities, experienced two or more severe deprivations (2003: 24). More detailed child poverty analysis has emerged through UNICEF’s Global Study on Child Poverty and Disparities. A number of the studies produced within the Global Study initiative appear in Minujin and Nandy (2012). For example, the number of children without access to hygienic sanitation varied from three-quarters (74.6 per cent) in the mountainous North West region of Vietnam to one in ten (13.4 per cent) in the Red River Delta area around Hanoi (Roelen and Gassmann 2012: 314).

Much of the analytic work in this area has used panel studies to consider the effects of neighbourhood. Reviewing literature2 suggests this is dominated by developed world data, primarily from the USA (Brooks-Gunn et al. 1997a, 1997b and Ellen and Turner 1997 provide reviews). Related evidence from developing countries tends to be more focused on economic change associated with communities rather than on impacts on children (see for instance Jalan and Ravallion 1997). Ellen and Turner identify relevant channels of influence: quality of local services, socialization by adults, peer influences, social networks, exposure to crime and violence, physical distance from job opportunities, and isolation (1997: 836–42). The extent to which poorer communities may experience lower service quality despite higher need has been described elsewhere as the ‘inverse care law’ (see Tudor Hart 1971).

US research reviews indicate persistent evidence of neighbourhood effects and that these effects are smaller than those associated with household characteristics (Brooks-Gunn et al. 1997a: 267). It is difficult to distinguish neighbourhood effects causally from household-level factors (Ellen and Turner 1997: 843). The use of experiments is a way of overcoming the challenge of ascribing causality – the Moving to Opportunity study started in five US cities in 1994, randomly assigning 4,606 low-income households to a control group (no change) and two treatment groups (given incentives either simply to move or to move to relatively affluent areas). The experiment therefore tested the impact of moving from an initial (poorer) location on later outcomes. Fifteen years later the treatment groups were more likely than the control groups to live in areas with lower poverty rates; to have social ties to more affluent households; and to report better physical and psychological health than the control groups. However there was little difference in some outcomes – the treatment and control groups reported similar incomes, employment levels, and reading and mathematics results (Sanbonmatsu et al. 2011: xvi–xxxi). US experimental and observational data therefore broadly concur: neighbourhood effects exist, matter for children, and are hard to unpick; but they are of a smaller magnitude than those directly associated with household characteristics. These findings are broadly similar to those presented here for data from Vietnam and Ethiopia.

Cross-country analysis by the Organisation for Economic Co-operation and Development (OECD) of the Programme for International Student Assessment (PISA) dataset (again mostly OECD countries) estimated learning impacts associated with a child’s household and the average socio-economic background of pupils in the school. The OECD study suggests both household circumstances and school peer group had an impact on literacy and that factors associated with the peer group were of greater magnitude than the individual background factors (OECD 2001: 199). Though clear in its inference (school social mix matters for poorer children’s performance), this study was not able to unpick the causal chain analytically (lacking before/after data). Instead, the authors suggest a couple of factors, which might explain these contextual effects, including greater support from parents, peer effects from children working together or competing, and a better school environment (fewer discipline problems and more motivated teachers) (ibid.: 198).

Having established evidence ...

Table of contents

• Cover page
• Half Title page
• Series page
• Title page
• Copyright page
• Contents
• List of Tables
• List of Figures
• Acknowledgements
• Notes on Contributors
• Introduction: Child Poverty and the Centrality of Schooling
• Part I Poverty and the Lives of Growing Children
• Part II How Does Schooling Help Poor People?
• Index