Relationships between infant mortality, birth spacing and fertility in Matlab, Bangladesh

Arthur van Soest, Unnati Rani Saha

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Although research on the fertility response to childhood mortality is widespread in demographic literature, very few studies focused on the two-way causal relationships between infant mortality and fertility. Understanding the nature of such relationships is important in order to design effective policies to reduce child mortality and improve family planning. In this study, we use dynamic panel data techniques to analyse the causal effects of infant mortality on birth intervals and fertility, as well as the causal effects of birth intervals on mortality in rural Bangladesh, accounting for unobserved heterogeneity and reverse causality. Simulations based upon the estimated model show whether (and to what extent) mortality and fertility can be reduced by breaking the causal links between short birth intervals and infant mortality. We find a replacement effect of infant mortality on total fertility of about 0.54 children for each infant death in the comparison area with standard health services. Eliminating the replacement effect would lengthen birth intervals and reduce the total number of births, resulting in a fall in mortality by 2.45 children per 1000 live births. These effects are much smaller in the treatment area with extensive health services and information on family planning, where infant mortality is smaller, birth intervals are longer, and total fertility is lower. In both areas, we find evidence of boy preference in family planning.
Original languageEnglish
Article numbere0195940
JournalPLoS ONE
Volume13
Issue number4
DOIs
Publication statusPublished - Apr 2018

Fingerprint

Planning
Family Planning Services
Fertility Agents
Health
Child Mortality
Health Services

Keywords

  • Panel Data

Cite this

@article{4b4cf05947ac40cb95c74a78a7292c0a,
title = "Relationships between infant mortality, birth spacing and fertility in Matlab, Bangladesh",
abstract = "Although research on the fertility response to childhood mortality is widespread in demographic literature, very few studies focused on the two-way causal relationships between infant mortality and fertility. Understanding the nature of such relationships is important in order to design effective policies to reduce child mortality and improve family planning. In this study, we use dynamic panel data techniques to analyse the causal effects of infant mortality on birth intervals and fertility, as well as the causal effects of birth intervals on mortality in rural Bangladesh, accounting for unobserved heterogeneity and reverse causality. Simulations based upon the estimated model show whether (and to what extent) mortality and fertility can be reduced by breaking the causal links between short birth intervals and infant mortality. We find a replacement effect of infant mortality on total fertility of about 0.54 children for each infant death in the comparison area with standard health services. Eliminating the replacement effect would lengthen birth intervals and reduce the total number of births, resulting in a fall in mortality by 2.45 children per 1000 live births. These effects are much smaller in the treatment area with extensive health services and information on family planning, where infant mortality is smaller, birth intervals are longer, and total fertility is lower. In both areas, we find evidence of boy preference in family planning.",
keywords = "Panel Data",
author = "{van Soest}, Arthur and Saha, {Unnati Rani}",
year = "2018",
month = "4",
doi = "10.1371/journal.pone.0195940",
language = "English",
volume = "13",
journal = "PLoS ONE",
issn = "1932-6203",
publisher = "PUBLIC LIBRARY SCIENCE",
number = "4",

}

Relationships between infant mortality, birth spacing and fertility in Matlab, Bangladesh. / van Soest, Arthur; Saha, Unnati Rani.

In: PLoS ONE, Vol. 13, No. 4, e0195940, 04.2018.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Relationships between infant mortality, birth spacing and fertility in Matlab, Bangladesh

AU - van Soest, Arthur

AU - Saha, Unnati Rani

PY - 2018/4

Y1 - 2018/4

N2 - Although research on the fertility response to childhood mortality is widespread in demographic literature, very few studies focused on the two-way causal relationships between infant mortality and fertility. Understanding the nature of such relationships is important in order to design effective policies to reduce child mortality and improve family planning. In this study, we use dynamic panel data techniques to analyse the causal effects of infant mortality on birth intervals and fertility, as well as the causal effects of birth intervals on mortality in rural Bangladesh, accounting for unobserved heterogeneity and reverse causality. Simulations based upon the estimated model show whether (and to what extent) mortality and fertility can be reduced by breaking the causal links between short birth intervals and infant mortality. We find a replacement effect of infant mortality on total fertility of about 0.54 children for each infant death in the comparison area with standard health services. Eliminating the replacement effect would lengthen birth intervals and reduce the total number of births, resulting in a fall in mortality by 2.45 children per 1000 live births. These effects are much smaller in the treatment area with extensive health services and information on family planning, where infant mortality is smaller, birth intervals are longer, and total fertility is lower. In both areas, we find evidence of boy preference in family planning.

AB - Although research on the fertility response to childhood mortality is widespread in demographic literature, very few studies focused on the two-way causal relationships between infant mortality and fertility. Understanding the nature of such relationships is important in order to design effective policies to reduce child mortality and improve family planning. In this study, we use dynamic panel data techniques to analyse the causal effects of infant mortality on birth intervals and fertility, as well as the causal effects of birth intervals on mortality in rural Bangladesh, accounting for unobserved heterogeneity and reverse causality. Simulations based upon the estimated model show whether (and to what extent) mortality and fertility can be reduced by breaking the causal links between short birth intervals and infant mortality. We find a replacement effect of infant mortality on total fertility of about 0.54 children for each infant death in the comparison area with standard health services. Eliminating the replacement effect would lengthen birth intervals and reduce the total number of births, resulting in a fall in mortality by 2.45 children per 1000 live births. These effects are much smaller in the treatment area with extensive health services and information on family planning, where infant mortality is smaller, birth intervals are longer, and total fertility is lower. In both areas, we find evidence of boy preference in family planning.

KW - Panel Data

U2 - 10.1371/journal.pone.0195940

DO - 10.1371/journal.pone.0195940

M3 - Article

VL - 13

JO - PLoS ONE

JF - PLoS ONE

SN - 1932-6203

IS - 4

M1 - e0195940

ER -