Dynamics of Child Malnutrition in Rural and Small‐Town Ethiopia

• Author: Ilana Seff,Dean Jolliffe,Sarah Baird
• Date: 01 June 2018
• DOI: http://doi.org/10.1111/saje.12190
• Published date: 01 June 2018

DYNAMICS OF CHILD MALNUTRITION IN RURAL

AND SMALL-TOWN ETHIOPIA

ILANA SEFF*, SARAH BAIRD

†

AND DEAN JOLLIFFE

‡

Abstract

Improving the status of malnourished children, and preventing children from becoming

malnourished, lies at the heart of several of the Sustainable Development Goals. While many

cross-sectional studies examine correlates of stunting, they largely cannot identify drivers of

change in stunting. We use two waves of panel data from Ethiopia and incrementally larger sets

of fixed effects to control for time-variant observable characteristics and time-invariant unobserv-

able characteristics. After controlling for these potential confounders, our analysis reveals that

many factors that are associated with stunting in the cross-section do not impact stunting dynam-

ics. We also estimate individual fixed-effects regressions, separately, according to baseline stunting

status. We find evidence suggesting that while improved societal conditions drive many children

out of a stunted state, certain exogenous factors may lead previously healthy children to become

stunted. Overall, policymakers and practitioners would be wise to consult research utilising both

cross-sectional and panel data analyses in order to more effectively target already stunted children

as well as vulnerable children who may be at risk of becoming stunted.

JEL Classification: I15, I31, C23

Keywords: Ethiopia, child malnutrition, stunting, panel data analysis

1. INTRODUCTION

Under-5 malnutrition affects more than 26% of children worldwide and accounts for

anywhere between one-third and one-half of child mortality cases globally (Black et al.,

2008). Undernutrition makes children more vulnerable to common infections, increases

the frequency, duration and severity of these infections, and raises overall susceptibility

to death.

1

Thus, improving child nutrition lies at the heart of achieving many of the

Sustainable Development Goals (SDGs) (IFPRI, 2014).

* Corresponding author: Research Associate Department of Population and Family Health,

Columbia University Mailman School of Public Health, 722 West 168th Street, New York,

NY 10032, USA. E-mail: ijs2120@cumc.columbia.edu

†

Department of Global Health, Milken Institute School of Public Health, George Washing-

ton University.

‡

World Bank, Development Data Group and the Institute for the Study of Labor (IZA), Bonn.

The authors are grateful to the UK Department for International Development Ethiopia for

funding assistance. The authors would also like to thank Tassew Woldehanna, Assefa

Admassie, Solomon Shiferaw, Alemayehu Seyoum Taffesse, Demirew Getachew and Tadele

Ferede for their feedback on earlier versions of this paper.

1

The relationship between illness and stunting has been extensively studied. For example,

Kimani-Murage et al. (2011) analysed HIV and stunting status among 671 children in rural South

Africa and found that HIV-positive children were much more likely to be stunted. Stephenson

et al. (2000) detail the causal relationship between parasitic helminth infections and stunting in

high burden countries.

V

C2018 The World Bank. South African Journal Economics V

C2018 Economic Society of South Africa. doi: 10.1111/saje.12190

127

South African Journal of Economics Vol. 86:2 June 2018

South African Journal

of Economics

Stunting, defined as having a height-for-age z-score (HAZ) less than 22, is a

commonly used measure of chronic malnutrition in children 6-59 months. It is widely

considered the most comprehensive and objective measure of children’s health inequal-

ities, as it conveys information about a child’s physical development, health status, envi-

ronment and socioeconomic conditions (Beaton et al., 1990; Pradhan et al., 2003;

Wam ani et al., 2007). It is also associated with a number of long-term health outcomes

such as deficiencies in nutrition (chronically inadequate levels of protein and energy and/

or micronutrient deficiencies), frequent infections and inappropriate feeding practices

over a sustained period (Muller and Krawinkel, 2005).

The effects of childhood stunting are long-term and potentially irreversible. Stunting

can cause delayed motor function and diminished cognitive ability, thus children with

low HAZ in their early years may exhibit lower school engagement, poor academic per-

formance and diminished adult earnings later in life (Daniels and Adair, 2004; Alderman

et al., 2006a; Hoddinott et al., 2008; Maluccio et al., 2009; Crookston et al., 2011).

Given stunting’s impact on earning potential later in life, both through mortality and

lowered cognitive ability, Hoddinott et al. (2013) estimate significant and substantial

benefit-cost ratios for interventions targeting stunting in high burden countries.

The literature identifies a number of risk factors correlated with stunting and HAZ in

developing countries.

2

One recent study of children under-5 in Ghana found mother’s age

and household size to be associated with stunting status (Darteh et al., 2014). Parent’s

schooling, particularly mother’s education and mother’s height were correlated with lower

stunting prevalence in Nigeria (Senbanjo et al., 2011; Adekanmbi et al., 2013). Evidence

also suggests that income and wealth are associated with better stunting outcomes as addi-

tional income enables households to allocate more funds to food, clean water and

improved sanitation (Senbanjo et al., 2011). Using cross-sectional data from Ethiopia, the

context for this analysis, the literature shows evidence of a correlation between water source

and stunting in the Tigray region (Alemayehu et al., 2015); a relationship between food

diversity, but not quantity, and stunting in the Amhara region (Motbainor et al., 2015);

and an association between dairy cow ownership and lower stunting prevalence in the

highland regions of Amhara, Oromiya, SNNP and Tigray (Hoddinott et al., 2015).

These cross-sectional studies help us understand the risk factors for stunting at a given

point in time, but suffer two related shortcomings. First, the cross-sectional estimates may

be biased due to unobservable, confounding factors. Second, these studies largely do not

provide insight into the dynamics of stunting, namely what characteristics determine

whether a child transitions from stunted to not stunted or vice-versa. Woodruff et al.

(2017) and Headey (2014) provide two noteworthy exceptions, using repeated cross-

sections from Demographic Health Surveys (DHS) to identify variables that are associated

with height-for-age in the cross-section and exhibittrendsovertimeconsistentwithchanges

in height-for-age. The authors of these two papers note the limitations of using cross-

sectional data and recommend future studies explore their findings using panel cohorts.

While a few recent studies have utilised longitudinal data to explore the dynamic deter-

minants of stunting over time (see Dearden, et al., 2017 for an example), the majority of

2

Very little research has looked at correlates of stunting in developed countries due to low preva-

lence, but the few existing studies suggest that the majority of stunting cases in these environments

are a result of social, and not nutritional, factors (Wright and Garcia, 2012).

128 South African Journal of Economics Vol. 86:2 June 2018

V

C2018 The World Bank. South African Journal Economics V

C2018 Economic Society of South Africa.