Maternal group participation and child nutritional status in Peru

• Published date: 01 May 2018
• DOI: http://doi.org/10.1111/rode.12363
• Date: 01 May 2018
• Author: Marta Favara

REGULAR ARTICLE

Maternal group participation and child nutritional

status in Peru

Marta Favara

University of Oxford, United Kingdom

Also affiliated to IZA (Institute of Labor

Economics), Bonn, Germany

Correspondence

Marta Favara, Department of International

Development, 3 Mansfield Road, Oxford

OX1 3TB, United Kingdom.

Email: marta.favara@qeh.ox.ac.uk.

Abstract

Using data from the Peruvian sample of the Young Lives

study, this paper investigates the association between

maternal group participation and child nutritional status at

the ages of 1 and 5 years. This study finds that the rela-

tionship between child nutrition and maternal group par-

ticipation depends on the level of maternal education. In

fact, maternal group participation is positively associated

with child height-for-age when children were 1 year old,

for those children whose mothers had had no formal edu-

cation. This same association fades at 5 years of age.

Maternal group participation may affect child nutrition

through a number of mechanisms. This paper provides

suggestive evidence on the positive role that this partici-

pation might play in promoting better breastfeeding prac-

tice and a more secure mother–child bond and in easing

mothers’access to additional resources and support.

1

|

INTRODUCTION

Nutritional status during early childhood is a major predictor of health conditions and economic

outcomes in later stages of life (Case, Fertig, & Paxson, 2005; Gertler & Boyce, 2001). This is the

reason why policy interventions during early childhood have been found to be particularly cost-

effective, with higher rates of return than later interventions (Heckman, 1995, 2000).

This paper studies the association between mothers’membership of community organizations in

Peru and their children’s nutritional status in the first years of life. Being a member of such a

group may affect a mother’s decisions about investments in her child health through a number of

mechanisms. These are quite similar to those that the economic literature discusses as being rele-

vant to maternal education (Glewwe, 1999), one of the strongest predictors of child health and

DOI: 10.1111/rode.12363

Rev Dev Econ. 2018;22:459–478. wileyonlinelibrary.com/journal/rode ©2017 John Wiley & Sons Ltd

|

459

nutritional status (Behrman & Deolalikar, 1990; and for a review, see Strauss & Thomas, 1995).

First, group participation might facilitate the mother’s acquisition of basic health knowledge, which

might encourage better decisions, for example, in terms of breastfeeding, vaccination and nutrition,

as well as improving the mother’s ability to prevent, recognize, and treat chil d illnesses. The infor-

mation-sharing might be a valuable resource, particularly for poorly educated mothers who did not

have the chance to learn through schooling. Second, it might ease access to additional resources

and health services. Third, there might be implications for maternal mental health and mother–

child bonding as a result of the interaction with other mothers within the community.

I test the association between maternal group membership and child nutrition at ages 1 and

5 using the Peruvian sample from Young Lives, a longitudinal study of childhood poverty. I

find evidence of a robust association between maternal group participation and child nutrition

at the age of 1. This fades when the children were 5. Interestingly, this paper suggests that

the relationship between child nutrition and maternal group membership depends on the level

of maternal education, being (positive and) stronger for those children whose mothers had no

education.

The most important caveat to this paper is that the results cannot be interpreted causally

because of the self-selection into group membership. However, while potential omitted variable

bias cannot be ruled out, the fact that our estimation controls for an extended set of characteristics

at child, household, and community level, suggests that the size of the omitted variable bias is

unlikely to be large. The inclusion of community-level fixed effects rules out several potential

threats to causal identification, for example, related to the existence (or not) of community organi-

zations where the mother and the child live. Still, the decision to be a member of a community

organization is a voluntary one, and unobservable characteristics at child, mother, and household

level might drive both the mother’s decision to participate and the child’s nutritional status. Fur-

thermore, this paper recognizes the potential for reverse causality to arise when the relationship

between maternal group membership and child nutrition is examined. This issue is overcome by

the use of longitudinal data, which help disentangle the temporal order between maternal group

participation in the child’s early years and the date when the child’s nutritional status was mea-

sured.

Finally, the last section analyzes the potential mechanisms through which the positive relation-

ship found between maternal group participation and child nutrition might work. It provides sug-

gestive evidence on the positive role that maternal group participation might play in easing

mothers’access to additional resources and health services; in equipping poorly educated mothers

with basic health knowledge that might encourage better breastfeeding practice; and in promoting

a more secure mother–child bond.

2

|

CONCEPTUAL FRAMEWORK

The conceptual framework used in this paper fits into the general household product ion model

according to which parents act to maximize a household utility function, subject to budget con-

straints and a child health production function (Behrman & Wolfe, 1987). The household’s deci-

sions about health and nutritional inputs reflect the household environment and characteristics , the

child health endowments and the local community characteristics.

In this section, we briefly discuss the main predictors of malnutrition in young children in

developing countries, as emerging from the economic literature. The intention is not to provide a

comprehensive review of the literature on child health and nutrition but to understand how a

460

|

FAVARA