The impact of mothers eduction on child health bevidence from MICS4 vietnam

UNIVERSITY OF ECONOMICS

INSTITUTE OF SOCIAL STUDIES

HO CHI MINH CITY

THE HAGUE

VIETNAM

THE NETHERLANDS

VIETNAM - NETHERLANDS
PROGRAMME FOR M.A IN DEVELOPMENT ECONOMICS

THE IMPACT OF MOTHER’S EDUCATION ON
CHILD HEALTH:
EVIDENCE FROM MICS4 - VIETNAM
A Thesis Submitted in Partial Fulfilment of the Requirement for the Degree of
MASTER OF ARTS IN DEVELOPMENT ECONOMICS

By
TRAN TRUNG THU

Academic Supervisor:
Dr. TRUONG DANG THUY
HO CHI MINH CITY, December 2014

1

DECLARATION
I would like to declare that this thesis, titled “THE IMPACT OF
MOTHER’S

EDUCATION

ON

CHILDREN HEALTH:

EVIDENCE FROM

MICS4 - VIETNAM”, is original, which is submitted in fulfilment of the
requirements for the degree of Master of Art in Development Economic s to the
Vietnam – The Netherlands Programme. I ensure that this paper has not been
submitted anywhere for the award of any degree.
This thesis was completed with big support from my supervisor. All source s
of data and information have been fully cited in the thesis.

TRAN TRUNG THU
MDE19

2


ACKNOWLEGDEMENT
First of all, I would like to express my greatest gratitude to my respectful
supervisor, Dr. Truong Dang Thuy, for his valuable knowledge in helping me
complete this study. I would like to express my thanks to him, who helped and
supported me through interesting courses, especially his valuable advice, guidance

and inspiration, which motivate me to finish this study.
I also want to express my thanks to all Professors of the MDE Programme
during the past two years (2012 -2014), my friends from Class MDE 19, and
administrative staffs of Economic Development Faculty, University of Economics
HCMC (UEH), who provided me useful documents and materials.
Finally, I would like to express my deeply appreciation to my dear family
during the time I follow this grogram for their spiritual and physical supports. In
particular, I dedicate this thesis to my mother, who has been helping me virtually
during the process of conducting my thesis.
HCMC, December 2014
TRAN TRUNG THU
MDE 19

3


ABSTRACT
This study aims to examine the relationship between mother’s education and
children health in Vietnam. Data are taken from The Vietnam Multiple Indicator
Cluster Survey (MICS) by the General Statistics Office of Vietnam, the United
Nations Children’s Fund (UNICEF) and the United Nations Population Fund
(UNFPA). The main finding of this study in Vietnam is consistent with other studies
in the world about mother’s education and children health. The result of this study
in Vietnam showed that mother’s education is an important factor of the health of
children. Thus, they can provide better environment such as provid ing safe water
for their house and their children. The result also provides new evidence to the
general literature that flushing toilet does not really affect children health in
Vietnam.

4

Table of Contents

DECLARATION ...................................................................................................................... 2
ACKNOWLEGDEMENT....................................................................................................... 3
ABSTRACT .............................................................................................................................. 4
LIST OF TABLES ................................................................................................................... 7
CHAPTER 1: INTRODUCTION........................................................................................... 8
1.1 Problem statement .......................................................................................................... 8
1.2 Research objectives ...................................................................................................... 11
1.3 Research scope and data .............................................................................................. 11
1.4 The structure of this study ........................................................................................... 12
CHAPTER 2: LITERATURE REVIEW............................................................................. 13
2.1 The relationship between mother’s education on child health ............................... 13
2.2 The impacts of all other factors .................................................................................. 17
2.3 Conceptual framework................................................................................................. 20
CHAPTER 3: RESEARCH METHODOLOGY ................................................................ 22
3.1 Model and Data ............................................................................................................ 22
3.1.1.

Model ................................................................................................................. 22

3.1.2.

Variables ............................................................................................................ 23

3.2 Data ................................................................................................................................ 27
3.3 ......................................................................................................................................... 28
Research hypotheses........................................................................................................... 30

CHAPTER 4: EMPIRICAL RESULTS .............................................................................. 32
4.1 Descriptive statistics .................................................................................................... 32

5


............................................................................................................................................... 33
4.2 Bivariate analysis ......................................................................................................... 35
4.3 The regression results .................................................................................................. 39
CHAPTER 5: CONCLUSION ............................................................................................. 43
5.1 Conclusion remarks ...................................................................................................... 43
5.2 Limitations and future research of the study............................................................. 44
REFERENCES ....................................................................................................................... 46
APPENDIX ............................................................................................................................. 50

6


LIST OF TABLES
Table 1: Variables of the study ............................................................................................. 23
Table 2: Mean, Std.dev., minimum and maximum from MICS abour HAZ z-score .... 32
Table 3: Percent of HAZ z-score divided five parts........................................................... 33
Table 4: Percent of mother's education level ...................................................................... 33
Table 5: Percent of wealth index quintiles .......................................................................... 34
Table 6: Percent of households who use safe water and use flushing toilet ................... 34
Table 7: Mother's education and HAZ................................................................................. 35
Table 8: Wealth index quintiles and HAZ........................................................................... 36
Table 9: Variables of use safe water, mother's education and HAZ ................................ 37
Table 10: Variables of use flushing toiler, mother's education and HAZ ....................... 38
Table 11: Ordinary least squares estimates of the effect of mother's education on the

health of children .................................................................................................................... 39

7


CHAPTER 1: INTRODUCTION
1.1 Problem statement
Child malnutrition is not only one of the key issues in the world but also
especially pervasive in almost any low income nations and Vietnam because the
poorest quintile of children usually has the highest ratio of malnutrition.
What is malnutrition? When nutrients in meal are not enough or perhaps are
too much, it can lead to malnutrition. Children who suffer from malnutrition can
have health problems such as a weakened immune system, infections, mental
retardation, brain damage or even HIV/AIDS which lead to a high risk of
developmental delay.
Many children even can be at this damaged risk caused their mother don’t
get proper nourishment. Even after recovering from this issue, they still remain
stunted in their lives. In reality, more than 200 million children in all developing
countries under age 5 years old were not provided with a good enough environment
to reach their developmental potential (Grantham-Mc Gregor et al., 2007).
There have been many efforts in many countries to bring out solutions which
can reduce the number of malnourished children. It is very important for national
security issues in the long time for any nations about human resources and
economics. In particular, this serious problem is occurring in developing countries.
However, it varies from region to region and from country in order to country such
as in South/Southeast Asia, Latin America or perhaps Sub-Saharan Africa.
Therefore, many countries try to estimate which factors affect child health,
particularly health and nutritional inputs, medicines, medical care or perhaps the
quality of household drinking water sources, toilet facilities, and other hygienic
conditions or household assets, parental schooling, community economic and

health-related characteristics (Glewwe, 1999).
From many factors above, there are numerous studies found out the
important role of parental schooling, especially mother’s education. It is believed

8


that increased education of mothers may be an important means for improvement of
child health or perhaps their nutrition.

As a result, they can provide safer

environment from their earnings or beneficial food as well as health habits; thus,
mothers are given chances in order to improve the nutrition of their children.
However, the results of the relationship of mother’s educ ation and child
health were not the same from country to country. Hobcraft et al. (1984)
suggested that because of large socio-economic differentials among countries,
the relationship between mother’s education and child survival were different
amount countries. This result is covered 28 World Fertility Surveys by
Hobcraft et al. (1984). In detail, increased education of mothers was an
important factor in order to improve opportunities of child survival in larg e
area of developing countries. They pointe d out that because of l arge socio economic differentials from country to country in child survival (from age
one and five), increased age of the child were widened. Both mother’s and
father’s levels of education also play an important role in de termining child
survival. The father’s occupation also related to this variable. In detail, the
father’s education variable increased opportunities of child survival. This
result improved even with a small increasing levels of education .
Mensch et al. (1985) also covered from 15 countries and had the same result
as Hobcraft et al. (1984).

They both pointed out that because of large socio-

economic differentials, the above association was weaker in sub-Saharan Africa
than in Asia or particularly Latin America. Furthermore, Mensch et al. (1985) also
suggested that the relationship between maternal education and child survival was
the same in any regions (include in rural and urban areas).
In Bangladesh, Lindenbaum (1990) suggested that educated women can keep
greater cleanliness which explained differentials in child mortality or incidence of
diarrhoeal episodes. While Cleland (1990) combined the international evidence
between diarrhoean episodes and maternal education. The author believed that
education maybe play an important role in determining health knowledge, which

9


leads to a more innovative attitude for women who can have more opportunities in
order to have some experience of school.
When comparing Bolivia, Egypt and Kenya, Stewart and Sommerfelt
(1991) covered from 25 Demographic and Health Surveys. They showed out that
the woman’s own educational level can effect on prenatal care and mother’s
education in urban was more powerful than in rural in Bolivia and Egypt while
Kenya has weakly significant in this relationship after controlling urban-rural
residence, a possessions

index, father’s

education, age, births, mother’s

education, family-planning use. In Kenya, at least, the urban-rural residence
variable appeared to have strong significant.

Malnutrition of children under five is also the key issue in Vietnam, as
many low-income countries (WHO, 2007). UNICEF reported that 50 percent of
Vietnamese children under 5 were stunted (abnormally low height for age) in
1993. In 2005, this figure has been improved with 25 percent of children
(UNICEF, 2006) because of economic growth and sustained investment in
primary healthcare.
Haughton et al. (1997) also showed large differences across regions and
ethnic groups in malnutrition of children in Vietnam. That means the greater
levels of malnutrition is concentrated on families in the northern regions, rural
households and ethnic families more than the rest. National institute of nutrition
and UNICEF (2011) mentioned that stunting of children under 5 was about 29.3
percent and the average rate reduced at 1.3 percent points per year in 5 years
from 1995 to 2010.
It is reported that around 60% of all under 5 dead who suffer neonatal
mortality; a third of children under 5 are affected by stunting malnutrition,
anaemia. Moreover, the rate of overweight children is increasing year by year; and
children older than 1 year old often die because of drowning and traffic accidents.
There are still many problems regarding children health in Vietnam remain to be
settled due to the limited quality of healthcare services in mountainous areas;

10


neonatal conditions and diseases; stunting malnutrition and anaemia; drowning
and traffic accidents (WHO, Child health in Vietnam - Fact sheet, 2010).
In summary, there have been many factors which can affect on child health in
the above studies in various countries. However, there have only some descriptions in
Vietnamese reports about the role of mother’s education in determining their child
health. Therefore, whether or not this relationship persists the same in Vietnam. In
order to answer this question, some factors mentioned above will be applied to estimate

which can explain in detail.
1.2 Research objectives
This study intents to: Evaluate the relationship between mother’s education
and child health in Vietnam.
1.3 Research scope and data
This study focus on:
- The women’s education levels
- Child health
- Some other variables: such as the height-for-age z-score, the household
income (wealth index quintile), health environment (drinking water and flushing
toilet) in Vietnam from the data of MICS4.

11


1.4 The structure of this study
Except the introduction

and the references chapter, this study is divided

into 4 chapters as follows:
Chapter 2: Literature review which reviews literature and empirical studies
on the relationship between mother’s education and children health. This chapter
discusses the results of these studies and compares the factors which can impact
the relationship such as whether the difference between father’s education and
mother’s education, between adopted children and own birth children, or other
socio-economic dimensions such as ethnicity, geographical location, gender and
assets or a wide variety of social and economic circumstances, behaviors matter
and community factors. Finally, it also provides the conceptual framework.
Chapter 3: Research methodology. This chapter presents the research

methodology, data source and regression technique. Finally, hypotheses are also
mentioned in.
Chapter 4: Empirical results. This chapter will present the statistic descriptions
of the data as well as bivariate analysis. After that, it will discuss and provide evidence
from the regression results on the findings of the association mother’s education and
children health.
Chapter 5: Conclusion. This chapter will summarize the main findings of
the study from the empirical results. It also present the limitations and suggests
the directions for future research.

12


CHAPTER 2: LITERATURE REVIEW
This chapter will review the literature on the relationship between mother’s
education and children health. The empirical studies on this association will be
discussed as well as compared in detail such as the difference between father’s
education and mother’s education, between adopted children and own birth
children, or perhaps among other socio-economic dimensions, and then finally,
between at national and individual level.
2.1 The relationship between mother’s education on child health
The mother’s education has been the focus of economist. There also have
some papers also showed out the same result that women’s schooling appears to
have an important positive impact on children’s health as well as nutrition in
developing countries.
The conclusion of Grossman M., (1972) suggested that an increasing in an
individual’s stock of knowledge may improve an ability to process information
regarding, fertility options and healthy pregnancy behaviors. In detail, the more
educated could have opportunities to choose a larger optimal stock of health
through an improvement in wage rates. This conclusion is generally about what a

person’s stock of knowledge could impact on by estimating the shadow cost or the
opportunity cost of the time. However, education is seemly not enough to consider
child health. While the conclusion leads to some other questions related to father’s
education or mother’s education, household income, options for houses and
behaviors for child health.
Mother’s income is applied to consider in study of Willis J.R. (1973). This
author covered from his individual data on the number of children born in America
when focusing on the demand for child quality and the supply of child services. The
author has pointed out that the association of education and health will make an
increasing in women’ permanent income which leads to their optimal choices for

13


taking care of their children. That means the demand for child quality increased
which leads to the supply of child services because of an improvement of mothers’
income when mothers’ income is associated with their education.
In other studies, there have also been some evidence illustrating the effect
of mother to their children. Case, A. (2000) had some important information
which we applied for our paper. The important application of this analysis is to
project the impact of mothers’ education on their children that when mother has
better education as well as job, their earnings is used for influencing health
through nutrition, and helped to reduce mother’s stress while taking care their
children. There are good reasons to expect policies from government to improve
education for mothers in order to improve children’s health, education, and
productivity in all their life.
Why is children’s health important? To explore this question, after Currie
and Stabile (2003) covered from panel data on Canadian children, they studied
that health shocks of children impact on test scores and future health. That means
their potential development will be damaged by health shocks. Thus, it is

important to identify which factors can affect health of the next generation of any
countries.
After Currie & Moretti (2003) covered from Vital Statistics Natality data
for 1970 to 1999, they found out that higher maternal education play an important
role in improve infant health (measured by birth weight, gestational age). They
believed that more educated mother can be able to afford more health care and
have healthier behaviors as well as better earnings or even raise family income by
marrying a highly educated man.
However, Berhman & Wolfe (1987) found out the different result that
children’s health outcomes were not improved by mother’s schooling, however it
may inprove their nutrient. That means this result was different with the standard
estimates when they covered their cross-section data from Nicaragua. They
suggested that mother’s schooling generally develops children health. However,

14


they also brought out an evidence of an indirect impact of mother’s schooling on
children health through nutrition – the length of breastfeeding.
The health gap between own birth child and adopted child are also concerned
in some studies, however Y.Chen & Li (2009) showed that child health is effected
by the mother’s education through the nurturing effect while the nature effect is not
really necessary. Because after their series of sensitivity tests and other papers, the
authors could not find out an evidence in order to ensure that the different between
adoptees and own birth children.
Sacerdote (2000) also used three long-term panel data sets (adopted children,
their adoptive parents, and their biological parents) from the British National Child
Development

Survey,

the

Colorado

Adoption Project, and the

National

Longitudinal Survey of Youth (NLSY79) in order to examine the above problem.
This author found out that the association of adoptive parents’ education, income
and children developmental potential. In detail, adoptive parents’ education and
income appeared to have a large impact on college attendance, marital status, and
earnings of children’ future.
Following Plug and Vijverberg (2003, 2004) and Plug (2005), they also did
not show out evidence that their estimations was affected by the adoptees and own
birth children variables. However, adopted children’s health is believed that they are
better because of the nurturing effect. That means the mother’s education played an
important role in determining the health of adopted children. Thus, they can find
information to know how to take care carefully health of children. Even after
controlling these variables: income, the number of siblings, health environment, and
other socio-economic, the impact of education of women on adoptee sample or the
own birth sample is similar. Therefore, there is no difference of the effect of the
mother’s education on the adoptee sample and the own birth sample. However, it is
doubted that parents can choose in among abandoned children which help them
have children in better situation.

15

A next question related to adopt child is whether more educated mothers
select their adopt child through health. Generally, it is believed that adopted
children did not have their information on the birth parents or their health included
in the case of China (Y.Chen & Li (2009)). Because of one-child policy, most
families in China have been trying to make sure their next generation who is boy.
Chinese parents only abandoned disabled or ill boys, while abandoned girls are
generally normal. Therefore, most of adopted children are girls (around 90 percent
of abandoned infants and around 80 percent of the adoptees). Because
abandonment in China is illegal, thus these children are abandoned in very early
months, in detail, most of parents have to abandon children in the first six months,
therefore adoptive parents cannot select children. Finally, there is no evidence to
ensure that parents will choose children whose health is better. Thus, after finding
out about adopt children, education of parents is considered: mother’s education or
father’s education?
The next question is that the association of father’s education or perhaps
mother’s education and child health. Whether or not the difference of the child
health is when comparing father’s education and mother’s education. To explore
this question, Behrman (1997) believed that increasing education of women is an
important means for children as well as has a greater beneficial educational than
increasing men’s schooling. This author also showed out that even when mothers
have the same abilities, those with higher levels of schooling maybe provide a
better environment for their children in order to receive potential development
(such as greater academic and then labor-market performances).
Following Berhman & Rosenzweig (2002) concluded that an increase in
women’ schooling would not impact on the schooling of children. That means it
is difficult to ensure women help their children to study more, because it is still a
decision of children for their future. However, increasing mother’s education in
the similar environment can improve children outcomes, such as their health.

16

In Vietnam, generally fathers have more opportunities to attain qualified
education than mothers do, thus, father’s education can be an important variable.
Y.Chen & Li (2009) covered their data in China and found that fathers have more
education than mothers generally. The tradition of China and Vietnam is similar.
Thus, in China, the authors also believed that father’s education may play an
important role. However, after testing, Y.Chen & LI (2009) stated that the mother’s
education is still more important than, though the difference is small.
2.2 The impacts of all other factors
There have many papers argued that health status can be explained by other
socio-economic dimensions. Following Gwatkin et al (2007), they applied 120
indicators as in four categories: health status (child nutritional level); use of basic
health services (antenatal care, treatment of common childhood illnesses); healthrelated behaviors (smoking and alcohol use); and other health-status determinants
(education). It is also mentioned some other indicators such as income, ethnicity,
geographical location, gender and assets or perhaps a wide variety of social as well
as economic circumstances, behaviors matter and community factors (the
psychological state of the primary care-giver, weaning and other feeding practices,
the social norms and behaviors that govern sexual transmission of diseases, and the
natural occurrence of trace minerals and vitamins available in soils and foods, water
and sanitation, vaccination coverage).
From above variables, O’Donnell et al (2008) suggested that income may
play an important role in determining children health. They covered data from the
1993 and 1998 Vietnam Living Standards Surveys in order to estimate the
association of the changes of the distribution of child nutritional status (through the
distribution of child height) and the changes in the level and distribution of income.
Their result showed that one-half of the 15-pong fall in the ratio of children
malnutrition (stunted) can be explained by changes in the distributions of income.
Therefore, income appears to have strong positive impact on children health. As a

17


result, household income help to expand opportunities to purchase not only health
care but also sanitation, food and other determinants of nutrition.
Case et al. (2002) pointed out other evidence from income that children
health is related to household income. They found out that a part of this
relationship can be explored by the effect and the arrival of chronic conditions.
Indeed in this research, they believed that children health from lower-income
households are worse than those from higher-income households. Therefore,
higher-income families can have opportunities to provide a better environment
(water and toilet) in order to take care their children who will have better health
with effective child care arrangements, greater academic and performance in
labor-market.
In addition, it is also believed that the intergeneration transfer of socioeconomic status begins in very early life or even in the womb. Thus, many
researches tried to find out which factors appear to have strong positive to
children’s health, as a part of the intergeneration transfer. Following Case et al
(2002) showed out that the impact of parents’ income on children’s health c an
explain a part of this issue by transferred their income to nutrition of their children.
While Y.Chen & Li (2009) provided information that the mother’s education has a
nurturing effect on children’s health. Therefore, children health can be considered to
be determined by the households’ income and mother’s education.
However, Pradhan, et al. (2003) argued that the determinants of health status
need to be looked at national and individual level. Because at national or individual
level, child health can be impacted on at many levels. At national level, it can be
considered as health care systems of nations or growth rate. And at individual level,
it can be considered as sanitation of households of household income. Which are
these indicator important?
At the national level, national incomes can be considered to be determined
by health status (health expenditures, social service infrastructure, education, and

environmental infrastructure). O’Donnell, et al. (2008), also suggested that

18


through improvements in commune infrastructure and reduction in negative public
health externalities – the nutrition quality in community could be improved.
However, it is difficult to estimate the association of health care systems and
health of each child in each family.
Skoufias (1998) covered from cross-sectional household data from the 1994
Integrated Household Survey of Romania. This author estimated the associatio n of
socioeconomic, demographic or perhaps environmental factors and growth
attainment of pre-school (0-5 year old). And finally they concluded that nation
income does not affect child health in urban areas when estimating child health
during the economic transition in Romania. Thus, child health cannot be explained
by nation income based on regions. Therefore, it is required to find out other
monetary variables to determine the relationship.
After using from household survey data from 12 countries and data on
malnutrition rates in a cross-section of countries from the 1970s, Haddahs el al.
(2003) also found that it is required to have 6% growth in incomes per capita for
20 years to help reducing the ratio of stunted children aged 5 years old, however
in fact, the ratio of stunting in this age group was decreased 15 percent points
while growth ratioin five years was close to the forecast. They concluded that
growth ratio can explain how malnutrition reduces, however the malnutrition
variable cannot explained by only this growth income. That means that the target
and programmes for nutrition cannot rely on economics growth alone.
At the individual level, household income can be considered to play an
important variable to explain child health or malnutrition, because when families
have greater incomes at the household level, they can expand more opportunities
in order to invest more in food consumption, clean water as well as good hygiene.

As a result, those can help parents to afford better child care arrangements
(Haddahs et al 2003).
On the other hand, O’Donnell et al. (2008) pointed out that household
income helped to explain 15% of the fall in the ratio of children stunted. Moreover,

19


adding safe drinking water and sanitation variables, helped to explain 35% of the
fall in this. Thus, malnutrition is declined by other factors or household income
rather than GDP. Therefore, child health can be explained by community
infrastructure variables, such as water and sanitation, drug availability or household
level covariates, such as education of various members, composition.
Sahn and Alderman (1997) also used data from Maputo, Mozambique in
order to estimate the impact of houshold resources or health inputs. They found out
two conclusions. First, mother’s education is a good variable to explain nutritional
status of children two years of age and younger. Second, an increase in household
incomes affect child health of children two years of age and older.
In general, the results of the studies examining the association of mother’s
education and children health are not the same level among countries, however,
their tendency of this relationship is closed. This study is to identify the relationship
between mother’s education and children health in Vietnam from MICS4.
2.3 Conceptual framework
The below factors are mentioned in the above parts. It is easily to realize that
three points of view to identify the association of mother’s education and children
health.
Firstly, Berhman (1997), Sahn & Alderman (1997), Berhman & Rosenzweig
(2002) and Gross M. (1972) suggested that mother’s education impacts on children
health directly, such as an increasing in healthy pregnancy behaviors or a good
variable to explain nutritional status of children two years of age and younger.

Secondly, more educated mother can have better earnings which can impact
on child health. Indeed in Willis J.R (1973) suggested that mother’s income is
improved which help an increasing in demand of child quality. While Currie and
Stabile (2003), O’Donnell et al (2008) pointed out mother’s income can afford
more health care and have healthier behaviors.

20


Thirdly, authors agreed that mother’s education have strong positive
significant with their income. And their income can be transferred to nutrition. It is
showed in studies of Case, A. (2000), Case et al (2002), Y.Chen & Li (2009) and
(Haddahs et al 2003). While Berhman & Wolfe (1987) pointed out in detail about
the length of breastfeeding.
While Sahn and Alderman (1997) found out that an increase in household
incomes affect child health of children two years of age and older, O’Donnell et al
(2008) showed in detail that household income helps to expand opportunities to
purchase not only health care but also sanitation, food and other determinants of
nutrition. Case et al. (2002) and Haddahs et al (2003) also pointed out higherincome families can have opportunities to provide a better environment (water
and toilet) in order to take care their children who will have better health with
effective child care arrangements, greater academic and performance in labormarket.
Finally, this conceptual framework helps to provide an overview of all
variables which used in this research.

21


CHAPTER 3: RESEARCH METHODOLOGY
First, this chapter will present the model and data source of this study then
the conceptual framework. Finally, hypotheses will be mentioned and discussed the

impact of each factor on the dependent variable.
3.1 Model and Data
3.1.1. Model
I am interested in these variables: mother’s education, water and sanitation
conditions and household income which can explain child health through HAZ zscore. In detail, HAZ z-score is affected by mother’s education, safe drinking water,
flushing toilet and the wealth index quintiles.
This is the standard model, currently, there has no study which provide
another model for this problem (The βo, β1, β2, β3, β4, β5, β6, β7 and β8 are the
corresponding vectors of coefficients, ε is the residual.). Thus, I began by
decomposing the standard function into components related with the above
variables. This study uses model below to examine the effect of the mother’s
education on child health (Y.Chen & Li, 2009):
HAZi = βo + β1 melevel1i + β2 melevel2i + β3 wiq 1i + β4 wiq 2i + β5 wiq 3i +
+ β6 wiq4i + β7 dtoilet i + β8 dwater i + εi

22


3.1.2. Variables
Table 1: Variables of the study
Variable
HAZ

Definition

Expected
sign

Height-for-age z-score (WHO) of child i.
melevel1i = 1 if level of mother i’s education is upper

Melevel1i secondary.
= 0 otherwise.
Melevel2 i

Wiqi

Wiqi

Wiqi

melevel1i = 1 if mother i studies teritarty level.
= 0 otherwise.
wiq 1i = 1 if the wealth index quintiles of householdi is second.

= 0 otherwise.
wiq 2i = 1 if the wealth index quintiles of householdi is middle.

= 0 otherwise.
wiq 3i = 1 if the wealth index quintiles of householdi is fourth.

= 0 otherwise.

(+)

(+)

(+)

wiq 4i = 1 if the wealth index quintiles of householdi is richest.

Wiqi

Dtoileti

Dwateri

(+)

= 0 otherwise.
dtoilet i = 1 if household i uses a flushing toilet.
= 0 otherwise.
dwater i = 1 if household i uses safe water.
= 0 otherwise

23

(+)

(+)


The above variables in Table 1 are explained in detail:
HAZ
Pradhan, et al. (2003) also argued that there have been some difficulties
problems when trying to compare incomes or perhaps socio-economic variables to
health status by time or even place. Especially, constructing income as well as
expenditure measures is too complex in poor countries. They also agreed that
standardized height should be used. Because it avoids not only the measurement
pitfalls of other health indicator (such as morbidity, mortality or even life
expectancy) but also the measurement problems associated to using monetary

variables (such as income or perhaps expenditure). Therefore, the height-for-age zscore (HAZ) is mentioned because of its widely used indicator of health and in
order to prevent other measures of health unsuitable for the particular quantitative
exercise (O'Donnell, Nicolas, & Doorslaer, 2008).
There also have other anthropometric indicators, such as the weight-forage z-score, the weight-for-height z-score, as well as the body mass index. They
can be used for estimating the regression in this research. However they are not
appropriate because more weight or perhaps too thin does not mean that they can
have better health. Thus, height is better to explain model (Pradhan, E.Sahn, &
D.Younger, 2003).
However, in the paper of Y.Chen & Li (2009), the regression found out that
using weight-for-age and weight-for-height (BMI) also gave the similar results. In
addition, they also mentioned that even after controlling income, health
environment, and other socioeconomic variables, their result still showed that the
determinant of the HAZ is the mother’s education.
Height is used to measure of short run and long run health status. O’Donnell,
Nicolas & Doorslaer (2008) covered from the 1993 and 1998 Vietnam Living
standards Survey and used height-for-age z-scores as a measure of long-term
nutritional status because they believed that child nutritional status could be

24


explained by the complete distribution of HAZ. Thus, this paper only use HAZ
variable to measure child health (more details in Appendix).
The height-for-age z-score is defined:

: the observed height of child i in group k (child sex and the birth month);
: the median of the height in group j;
: the standard deviation of the height in group j.
It is formed as the following categories
-1 < HAZ < 0

Normal

-2 < HAZ < -1

Marginally stunted

-3 < HAZ < -2

Moderately stunted

HAZ < -3

Severely Stunted

MELEVEL
Variable melevel is the mother’s education. In previous parts, the study show
out that the mother’s education can affect their children health.
SAFE DRINK WATER
A dummy dwater is used for testing whether a household uses safe drink
water. Water is called “safe drinking water” when it is met WHO guidelines
national standards on drinking water quality and took from sources such as
household connection, public standpipe, borehole, protected dug well, protected
spring and rainwater. (WHO)

25