Parental absence, remittances and educational investment in children left behind: Evidence from Nepal

• Author: Ryuichi Tanaka,Nirmal Kumar Raut
• Date: 01 November 2018
• DOI: http://doi.org/10.1111/rode.12410
• Published date: 01 November 2018

REGULAR ARTICLE

Parental absence, remittances and educational

investment in children left behind: Evidence from

Nepal

Nirmal Kumar Raut

1

|

Ryuichi Tanaka

2

1

Tribhuvan University and GRIPS

2

University of Tokyo

Correspondence

Nirmal Kumar Raut, Central Department

of Economics, Tribhuvan University,

GPO: 3821, Kathmandu, Nepal.

Emails: phd12101@grips.ac.jp /

nimsraut@gmail.com

Funding information

Japan Society for the Promotion of

Science, Grant/Award Number: 24330077

Abstract

This paper utilizes the third wave of the Nepal Living

Standards Survey to assess the causal impact of parental

absence and remittances on the educational investment in

children left behind. Unlike previous studies, we sepa-

rately identify parental absence and remittances, with

careful consideration of self-selection into migration and

endogeneity of remittances. Using a two-step estimation

strategy, we show that parental absence has a substantive

disruptive effect on education of children left behind,

while remittances have a positive effect. In addition, we

show that non-parental absence has no effect on educa-

tion. Further, we provide supporting evidence about the

channels to disruption. Finally, we show some hetero-

geneity results by the child’s gender and education status

of the mother. The results by child’s gender confirm that

remittances relax the liquidity constraints of households

by investing in education of the girls. Further, the results

by education status of the mother provide evidence that

an educated mother can play a role in neutralizing or mit-

igating the negative effects of parental absence.

1

|

INTRODUCTION

Parental migration is a widely observed phenomenon in South Asian countries. One of the concerns

of parental migration is its effect on children left behind. In particular, the effect of parental absence

on the education of children is a crucial issue. While the remittances received from a migrated parent

help meet the short-term consumption and long-term physical and human capital investment needs of

households (see, for example, Rapoport & Docquier, 2006; Yang, 2008; Adams & Cuecuecha, 2010),

DOI: 10.1111/rode.12410

1642

|

©2018 John Wiley & Sons Ltd wileyonlinelibrary.com/journal/rode Rev Dev Econ. 2018;22:1642–1666.

parental absence does result in family disruption. This disruptive effect may be in the form of less of

the parental input required to maintain children’s health and educational needs. It may also require

that children left behind help to bridge the household labor gap resulting from the absence of adult

members from the household. In extreme cases, children may respond to migration by reducing their

study hours and increasing hours of outside paid work (Antman, 2011).

To examine the effect of parental absence due to migration, it is important to isolate the direct

impact of parental absence from the effect of remittances. Otherwise, when the total effect of

migration is positive, for example, the adverse effects of parental absence may be concealed, which

could be mitigated by other policy measures.

The results are mixed for the strands of literature on the evaluation of the overall impact of migra-

tion on the education of children left behind (see, for example, McKenzie, 2005; Mansuri, 2006;

McKenzie & Rapoport, 2011). Likewise, the overall impact of remittances without the separate iden-

tification of the effect of parental absence on children’s education has been examined in many studies,

and the results are again mixed (see, for example, Edwards & Ureta, 2003; Yang, 2008; Acosta,

2011; Alcaraz, Chiquiar, & Salcedo, 2012; Kroeger & Anderson, 2014) More related to this study,

Amuedo-Dorantes, Georges, and Pozo (2010) and Amuedo-Dorantes and Pozo (2010) show that

remittances received by a migrant household weakly explain improvements in the educational attain-

ment of children left behind compared with those received by a non-migrant household.

This study attempts to determine whether parental migration disrupts the education of children,

with the explicit simultaneous consideration of both parental migration and remittances as the vari-

ables of interest. We investigate the effects of parental absence and remittances on both the inten-

sive and the extensive margins of educational investment: school enrollment and education

expenditure. We exploit the latest wave of the Nepal Living Standards Survey (NLSS), which

unlike previous rounds affords rich information on both absentees and remittances and thus facili-

tates the identification of parental absence with more accuracy.

To identify the causal impacts of parental migration status and remittance amount, it is impera-

tive to correct biases due to the self-selection of parental migration status and endogeneity of

remittances. In our regression analysis, we address these issues explicitly by applying a two-step

estimation method with instruments for migration status and remittances. Our estimation results

reveal the substantial disruptive effects of parental absence on educational investment, while the

absence of family members other than parents (‘non-parental absence’) has no significant effect.

Remittances have a positive effect on all the education variables of children left behind. Besides,

heterogeneous analysis indicates that an educated mother can play a role in neutralizing the nega-

tive effects of parental absence. Likewise, the analysis further confirms that remittances relax

household liquidity constraints by investing on girls’education.

This study is closely related to Bansak and Chezum (2009) and Acharya and Leon-Gonzalez (2014)

in that they evaluate the effect of the absence of household members and remittances on children’sedu-

cation separately by using data from Nepal. Our study has at least two advantages over these studies,

however. First, we use detailed information on the migration status of household members that was

unavailable in previous household surveys in Nepal. This enables us to examine the effect of parental

and non-parental absence separately. Second, and more importantly, our study assesses the causal

effects of migration and remittances through the careful treatment of sample selection and endogeneity

of remittances by exploiting a detailed set of instruments for parental absence and remittances.

The structure of the remainder of this paper is as follows. Section 2 describes the historical

development of the migration and education system in Nepal. Section 3 discusses the estimation

strategy of the study. Section 4 presents the data and descriptive statistics. Section 5 reports the

RAUT AND TANAKA

|

1643