Content uploaded by Wanippol Mahaarcha
Author content
All content in this area was uploaded by Wanippol Mahaarcha on Aug 11, 2014
Content may be subject to copyright.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 1
Impact of Family Structure, Parental Migration,
and Parental Divorce on an Adolescent’s Educational
Enrollment: Evidence from a Longitudinal Study in
Kanchanaburi Province, Thailand
Wanippol Mahaarcha
Sirinan Kittisuksathit
Introduction
Education is the crucial well-being indicator that remains a concern in many
countries because of its attendant demographic, economic, and social consequences for
human and overall national development. Increasing education is expected to contribute
higher population well-being. Adolescents are considered the future of the nation in
which to drive economic development for many decades in Thailand. Education of the
young Thai population has improved corresponding to Thai government
implementation for a long time. However, Thailand is a country that has both a gender
and a socio-economic gap in schooling (Pattaravanich, et al., 2005). Thus, despite legal
and structural changes making an equity and equality of education for Thai people, not
all children are equally likely to make it to secondary school. Although the goals for net
primary school enrollment and lower secondary school enrollment have been
successfully met, about 68 percent of the adolescent population aged 15-17 were
enrolled in upper secondary school and only 60 percent of those aged 18-21 were
enrolled in college or university (Ministry of Education, 2007).
Much of the recent researches have proposed that an adolescent’s education
have been influenced by their family structure, parental migration, and parental divorce.
These are considered the determinants that are highly associated with adolescent
outcomes and well-being (Buchman, 2000; Hao and Xie, 2002; Khun, 2006). Families
in Thailand have changed overtime, for example, family size has declined from an
average of 5.6 in 1960 to 3.6 persons in 2000, while family structure has become
increasingly diverse and complex (Table 1). Considering gender of household head and
family structure, female-headed households are steadily increasing and, surprisingly,
extended families and unrelated individuals are rising progressively during the period
1980-2005 (National Statistical Office, 2006).
2 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Table 1: Percentage distribution of household by gender of household head and
family structure in Thailand, 1980-2005
1980 1990 2000 2003 2004 2005
Gender of household head
Male 83.5 80.6 73.8 73.2 72.1 70.4
Female 16.5 19.4 26.2 26.8 27.9 29.6
Total 100.0 100.0 100.0 100.0 100.0 100.0
Family structure
Nuclear family 70.6 67.6 60.3 54.2 53.2 53.9
Extended family 25.2 26.2 29.6 33.3 34.0 34.5
Unrelated Individuals 4.2 6.2 10.1 12.2 12.8 11.6
Total 100.0 100.0 100.0 100.0 100.0 100.0
Source: National Statistical Office, 2006
These phenomena may impact a family member’s well-being, especially
adolescents in the household. Growing literatures describe correlations between an
adolescent’s well-being and family structure. Popular discussions have emphasized the
distinction between two-parent families and single-parent families. McLanahan and
Sandefur (1994) and other researchers have confirmed that children who live outside
two-parent families tend to have poorer outcomes when compared with those who live
in two-parent families. Nearly all previous studies have studied family structure by
focusing on divorce and most studies are in Western context where divorces are
common (Brown, 2006; Wolfinger, 1998). While Thai context is different, family
structure, migration, and the rising number of divorces seem to shape a family
member’s well-being nowadays (Curran, et al., 2004; Jones and Kittisuksathit, 2003).
Considering these determinants, first, population aging is a consequence of
fertility and mortality declining, which increases the number of extended families in
Thailand. Mortality decline shapes the longevity among Thai people. Increasing life
expectancy leads to grandparents surviving for longer periods and makes extended
families more prevalent. A few studies concern the effect of extended families on
children and an adolescent’s well-being, especially in an Asian context where the
extended families are regarded as the norm. Previous studies found the positive effect of
extended families on an adolescent’s schooling and suggestions that extended families
play a crucial role in mitigating the adverse effects of single-mother families (Deliere
and Kalil, 2002).
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 3
Second, while much work has been done on the impacts of international
migration on receiving countries and on immigrants themselves, the consequences of
migration for those left behind has not received the attention it deserves (Battistella and
Gastardo-Conaco, 1998; Hadi, 1999). Internal and international migrations are an
important phenomenon in Thailand. Migration is also making the living arrangement
among Thai families different. Left behind children in the household are the
consequences of migration. It is generally accepted that the migration of parents usually
benefit children economically. Remittance from migrants may help the household’s
standard of living and increase the chances for children to enroll in school (Curran, et
al., 2004).
Third, the number of divorces among Thai population is steadily rising from
8.4 divorces in 1960 to 23.4 divorces per 1,000 marriages in 2000 (Ministry of Public
Health, 2000). It is expected that single-parent families are increasing. Previous studies
have examined that parental divorce can have many effects on the well-being of
offsprings. Compared to people from intact families, the children of divorced families
complete fewer years of school (McLanahan and Sandefur, 1994) and do less well
economically as well as occupationally (Amato and Booth, 1991). In addition, parental
divorce often has profound social and psychological consequences for children.
So far, few studies about family structure, parental migration, and parental
divorce have been done. There is no study considering all of these determinants on an
adolescent’s schooling in Thai context. However, with the concept of family changing
in Thailand over the years, it becomes crucial to test these phenomena with an
adolescent’s schooling. It is expected that the results of the present study will propose
the important information to policy makers in order to improve the adolescent welfares.
Research Objectives
This study addresses the relationship between family structure, parental
migration, parental divorce and adolescent’s educational enrollment. The concept of
family structure, migration, and parental divorce are considered for creating the
independent variables that will be taken into account in the following: family structure,
parent’s duration of migration, and parental divorce. To demonstrate how these
4 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
determinants have effected an adolescent’s educational enrollment, the present study
uses longitudinal data from Kanchanaburi Demographic Surveillance System, 2001 to
2004, to examine the relationship of family structure, parental migration, parental
divorce and adolescent’s educational enrollment in Thai context.
Theoretical Considerations
To reach the question on why and how family structure, parental migration,
parental divorce influence an adolescent’s educational enrollment, all of the three main
independent variables are viewed through theoretical considerations as follows.
Family structure and children schooling
The educational benefits of extended families require at least two conditions.
First, there must be a close multigenerational bond between grandparents and
grandchildren so that grandparents provide emotional support for grandchildren.
Second, grandparents are resourceful in terms of their income and education. Coleman
(1988) conceptualized adult-child relationship in a form of social capital that enhances
the production of human capital. The idea that a close adult-child relationship with
educational resources flowing from the adult to the child has educational benefits to the
child and may be based on the concept of social capital within the family.
Compared to nuclear families, extended families consist of greater and more
varied adult-child relationships. Adult-child bonds could be between parents and
children, or between grandparents and grandchildren, and between grandparents and
parents. There is ‘intergenerational closure’ within an extended family (Coleman,
1988), providing a child more monitoring and supervision. One can imagine an
idealized version of an extended family where grandparents provide support and
connections between the grandparents and his/her parents. If wisdom comes with age,
the help provided by grandparents may be even more useful than the help given by
parents.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 5
Parental migration and children schooling
The effects of migration on outcomes in left-behind children have looked at the
economic perspective. Economic arguments stress the future returns of schooling.
Parents consider how to maximize the resources of the entire family and how to
redistribute them among family members. The primary concern is wealth maximization.
In the existing migration literature, remittances received by families have often featured
as an important factor leading to an increase in standard of living. The remittances make
an important financial contribution towards the well-being of left-behind children. It has
been found that remittances are not only used to compensate emigration-related expense
and income lost by left-behind household, but are also used by left-behind household to
meet the daily needs, improve living conditions, and invest in the education of their
children (Jones and Kittisuksathit, 2003). Children whose parents are migrants may be
expected to be better educated than children whose parents are non-migrants due to the
advantage of receiving remittances.
Parental divorce and children schooling
A number of theoretical perspectives suggest why parental divorce may affect
children well-being (McLanahan and Sandefur, 1994; Wu, 1996). The economic
deprivation perspective argues that much of the difference in child outcomes between
single-parent and two-parent families is a result of poverty. McLanahan and Sandefur
(1994) found that family economic resources account for half the differences in child
developmental outcomes between single-mother families and their two-parent
counterparts. The stress perspectives suggest that it is stress caused by the disruption in
family that is most critical to children’s well-being of children (Wu, 1996). Disruptions,
including divorce and remarriage, can create stress for parents and their children. This
stress may lead not only to less effective parenting, but also to changes in the behavior
of a child.
6 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Conceptual Framework
From the theoretical considerations, we can draw the conceptual framework in
relation to the variables of this study. Family structure, parental migration, and parental
divorce are the independent variables affecting an adolescent’s educational enrollment,
while controlling with youth and household characteristics.
Figure 1
Conceptual framework for the impact of family structure, parental migration,
and parental divorce on an adolescent’s educational enrollment
Three hypotheses can be drawn from theoretical considerations as above. (1)
adolescents in extended families have better chance of schooling than those in nuclear
families, (2) adolescents whose parents migrate have better chance of schooling than
those with non-migrant parents, and (3) adolescents with no parental divorce have better
chance of schooling than those whose parents have divorced.
Family structure
Parental migration
Parental divorce
Youth characteristics
- Sex
- Age
- Living in Thai speaking household
Household characteristics
- Standard of living
- Household assets
- Received remittances
- Household size
- Stratum Adolescent’s
educational
enrollment
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 7
Data and Methods
Data
The data come from the Kanchanaburi Demographic Surveillance System
(KDSS), longitudinal study during the year 2000 to 2004, conducted by the Institute for
Population and Social Research, Mahidol University, which was supported by the
Wellcome Trust, United Kingdom. The KDSS is a field research and training centre that
is dedicated to monitoring population change and evaluating intervention-based
research. The main research activity is the creation of a database on the demographic,
health, social, and economic composition of the population in 100 villages and
communities. The project began by annually collecting all the households in the
sampled villages in the year 2000.
The KDSS project consists of five strata, categorized by ecological features
and economic activities; urban and semi-urban, rice field, plantation, uplands, and
mixed economy. There are twenty villages or communities in each stratum.
The KDSS is not specifically designed to study the impact of family structure,
parental migration, and parental divorce on an adolescent’s educational enrollment;
however, the longitudinal nature of data has made this study possible. The measure of
family structure, parental migration, and parental divorce, data from round 2 to round 5
(KDSS 2001-2004) were used to predict the impact of family structure, parental
migration, and parental divorce on an adolescent’s school enrollment.
Sample
The study sample contains adolescents aged 13-18 years living in round 2
census (2001). The study subjects were limited to children aged 13-18 because this
group are expected to finish the primary and secondary levels of education. Samples of
this study were limited to those who are single and living with parent(s) because this
study explores the impact of family structure, parental migration, and parental divorce.
The sample was also restricted to those who were enrolled in school at the time of the
second census (2001) to demonstrate the impact of family structure on the adolescent’s
8 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
educational enrollment at the time of the last census (2004). The sample was from
separate household in order to avoid the statistical bias for those who have co-resident
adolescents.
Measurement of Variables
Dependent Variable
The dependent variable in this study is educational enrollment, which refers to
whether adolescents enrolled in upper secondary school or college, which was higher
than the compulsory level of education in Thailand, at the time of the last census
(2004). It was measured as a dichotomous variable.
Independent Variables
The independent variables include family structure, parental migration, and
parental divorce, which are explained in detail as below:
(1) Family Structure
There are 3 types of family structure; nuclear families, extended families with
grandparents, and extended families with other adults. Nuclear family refers to family
that has parent(s) and children. Extended family with grandparent presence refers to
parent(s), children, and grandparents in the household. Extended family with other adult
presence refers to parent(s), children, and other adults aged more than 22 in the
household. This information was derived from the household roster table in the
household questionnaire.
(2) Parental migration
Parental migration is the length of move-out during the year 2001 to 2004.
This study categorize parental migration as father and mother’s duration of migration to
be the set of dummy variables as follow: (1) not move between 2001 to 2004, (2) move-
out for one year, (3) move-out for two years, (4) move-out for three years, and (5)
move-out for four years.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 9
(3) Parental divorce
Parental divorce is categorized into dummy variable, which refers to whether
their parents divorce during the year 2001 to 2004.
Other Variables
Other variables cover the variables at individual and household level. All other
independent variables were assessed in the second census (2001). Individual
characteristics include sex, age, and whether adolescents lived in a household that uses
other languages speaking at home.
Household characteristics comprise standard of living, household asset, living
in household that received remittance, household size, and residential area. As for the
household standard of living score, this study used a combination of five variables;
material of roof, material of housing, electricity available, pipe water available, and type
of toilet. This ordinal scale variable is a composite score from 1-5. Household assets
have been measured from the ownership of household assets to create a composite score
from 1-5 as well. Living in the household that received remittance is a variable which
depend on whether the household has receive some remittance from out-migrants
during the time of 2001 to 2004. Household size was therefore included in the model,
which is a continuous variable. Place of residence is measured on whether adolescents
lived in the urban/semi-urban, rice field, plantation, upland, and mixed economy strata.
Analysis
Multivariate analysis is used to assess the complex impact of independent
variables on dependent variables with the set of other independent variables. In this
study, since the dependent variable-educational enrollment is a dichotomous variable,
binary logistic regression is employed. It is used for making prediction of enrolling in
school probability. In order to obtain adequate descriptions and useful predictions, there
are number of independent variables and other variables included in the regression
model.
10 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Results
Table 2 provides summary statistics of the variables included in the analysis.
About 33 percent of the samples were enrolled in school in the last census (2004). There
were more females than males and the mean age of the samples was 15 years. More
than 90 percent of the samples are living in a Thai speaking household. About 14
percent of adolescents were living in a household that received remittance from out-
migrants. The average household size in this study was about 5 people in the household.
The largest proportion of adolescents was living in the urban and semi-urban stratum
(25 percent), followed by the adolescents living in the rice field (22 percent), mixed
economy (22 percent), upland (17 percent), and plantation strata (15 percent).
Most of adolescents were living in nuclear families, which include only
parent(s) and children in the household without other adult presence. About 14 percent
of the adolescents were living in extended families with grandparents and 9 percent of
adolescents were living in extended families with other adults. Eighty percent of
samples’ parents did not move out from the household. About 3 percent of parents were
moving out 4 years. However, we cannot trace the migration before the years of
analysis, which was before the year 2001. Only 3 percent of adolescents had parental
divorce during the year 2001 to 2004.
Table 2: Percentage of adolescents by characteristics, KDSS 2001-2004
Variables Percentage S.D.
Enrolled school in 2004 32.87
Male adolescent 45.03
Mean age of adolescents 15.18 (1.59)
Living in Thai speaking household 92.47
Standard of living score 2.87 (1.25)
Household assets score 3.00 (1.42)
Living in a household that received remittance 14.14
Average household size 4.94 (1.81)
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 11
Table 2: (Continued)
Variables Percentage S.D.
Stratum
Urban and semi-urban 24.66
Rice field 22.06
Plantation 14.72
Upland 17.13
Mixed economy 21.43
Family structure
Nuclear family 77.17
Extended family with grandparents 13.66
Extended family with other adults 9.17
Father’s duration of migration
Not move 82.0
Father move-out for 1 year 7.67
Father move-out for 2 years 4.01
Father move-out for 3 years 2.85
Father move-out for 4 years 3.47
Mother’s duration of migration
Not move 82.31
Mother move-out for 1 year 7.43
Mother move-out for 2 years 4.01
Mother move-out for 3 years 2.70
Mother move-out for 4 years 3.55
Having parental divorce 3.01
N 2,072
Note: Standard deviation for continuous variables
Table 3 presents the relationship between independent variables, other
variables, and dependent variable by using chi-square test. Higher proportions of female
were enrolled in school more than that of male. The proportion of enrolling in school
was increased by age of adolescents. Adolescents living in Thai speaking households
had higher proportions of enrolling in school (33.7 percent) than those living in non-
Thai speaking household (22.4 percent).
12 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Standard of living and household assets increased the proportion of adolescent
who were enrolled in school. Surprisingly, there were higher proportion of adolescent
who were living in household that have not received remittance (34.1 percent) than
those who were not (25.6 percent) in terms of educational enrollment. Higher
proportions of samples that were living in urban and semi-urban stratum (42.1 percent)
were enrolled in school when compared with those from other strata.
Adolescent who lived in extended families with grandparents have higher
proportion of enrolling school, nearly half of them, than those who were not. Higher
proportion of the sample who have non-migrate parent were enrolled in school when
compared with migrate parent. Adolescents with no parental divorce during the year
2001 to 2004 have higher proportion of enrolling school than those who has parental
divorce.
Table 3: Percentage distribution of adolescents by independent variables and
dependent variable, KDSS 2001-2004
Independent variables
Not
enrolled
in 2004
Enrolled
in 2004
Total
(%)
Total
(n) χ
2
Sex of adolescents
Male 68.6 31.4 100.0 933 1.646
Female 65.9 34.1 100.0 1,139
Age of adolescents
13 37.6 62.4 100.0 388 396.357
14 46.6 63.4 100.0 395 p<0.001
15 72.2 27.8 100.0 431
16 84.6 15.4 100.0 357
17 88.1 11.9 100.0 302
18 91.5 8.5 100.0 199
Living in Thai speaking household
Living in non-Thai speaking household 77.6 22.4 100.0 156 8.319
Living in Thai speaking household 66.3 33.7 100.0 1,916
Standard of living score
Lowest 72.5 27.5 100.0 429 16.906
Lower 71.6 28.4 100.0 408 p<0.01
Medium 61.1 38.9 100.0 321
Higher 64.8 35.2 100.0 840
Highest 64.9 35.1 100.0 74
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 13
Table 3: (Continued)
Independent variables
Not
enrolled
in 2004
Enrolled
in 2004
Total
(%)
Total
(n) χ
2
Household assets score
Lowest 76.1 23.9 100.0 414 34.790
Lower 71.6 28.4 100.0 415 p<0.001
Medium 66.1 33.6 100.0 414
Higher 62.3 37.7 100.0 414
Highest 59.3 40.7 100.0 415
Received remittances
Not received 65.9 34.1 100.0 1,779 87.211
Received 74.4 25.6 100.0 293
Household size
Less than 5 persons 67.7 32.3 100.0 986 14.408
5 persons and over 66.6 33.4 100.0 1,086
Stratum
Urban and semi-urban 57.9 42.1 100.0 511 46.368
Rice field 67.4 32.6 100.0 457 p<0.001
Plantation 66.6 33.4 100.0 305
Upland 80.0 20.0 100.0 355
Mixed economy 67.6 32.4 100.0 444
Family structure
Nuclear family 69.1 30.9 100.0 1,599 12.351
Extended family with grandparents 54.1 45.9 100.0 283 p<0.001
Extended family with other adults 70.0 30.0 100.0 190
Father’s duration of migration
Not move 63.3 36.7 100.0 1,699 25.706
Father move-out for 1 year 75.5 24.5 100.0 159 p<0.001
Father move-out for 2 years 76 24.0 100.0 83
Father move-out for 3 years 79.7 20.3 100.0 59
Father move-out for 4 years 80.7 19.3 100.0 72
Mother’s duration of migration
Not move 61.6 38.4 100.0 1,705 75.622
Mother move-out for 1 year 77.3 22.7 100.0 154 p<0.001
Mother move-out for 2 years 76.1 23.9 100.0 83
Mother move-out for 3 years 86.7 13.3 100.0 56
Mother move-out for 4 years 90.4 9.6 100.0 74
Parental divorce
Not have 65.3 34.7 100.0 2,008 27.795
Have 83.3 16.7 100.0 64 p<0.001
N 67.1 32.9 100.0 2,072
14 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Table 4 displays two regression models, all of which take educational
enrollment as the dependent variable. The first model includes the other variables, while
the independent variables are taken into account in the second model.
We begin our analysis by focusing on the relationship between educational
enrollment and demographic factors, including individual and household characteristics,
in Model 1. The results showed that males were less likely to enroll in school than
female. The age of adolescents decreased the chance of enrolling in school. Household
asset score increased the probability of enrolling in school. Adolescents who were
residing in the ‘medium’, ‘higher’, and ‘highest’ quintile of household asset score
tended to have more chance to enroll in school than those who were residing in the
lowest quintile. Comparing with adolescents in urban and semi-urban, those dwelling in
other stratum were less likely to enroll in school in the last census (2004).
In Model 2, we add the key independent variables in the analysis. For family
structure, adolescents who are living in extended families with grandparents were more
likely to enroll in school than those living in nuclear families. The results suggested that
the duration of migration of the fathers and the mothers mattered differently.
Adolescents whose father moved out from the household for a short time, only 1 year,
had lower probabilities of enrolling in school than those who had non-migrant father.
Adolescents whose mothers moved out from the households were less likely to enroll in
school than those who had non-migrant mothers. Adolescents whose parents divorced
during the year 2001 to 2004 have lower probabilities of educational enrollment than
those who have no parental divorce. When the independent variables were added into
the second model, the significant relationship between other variables and dependent
variable were not changed.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 15
Table 4: Odds ratios of logistic regression analysis of the impact of family
structure, parental migration, and parental divorce on adolescent’s
educational enrollment, KDSS 2001-2004
Variables Model 1 Model 2
Male 0.796 (0.086)** 0.790 (0.088)**
Age 0.479 (0.019)*** 0.471 (0.020)***
Living in Thai speaking household 1.289 (0.293) 0.818 (0.203)
Standard of living score (Lowest: ref.)
Lower 1.087 (0.194) 1.149 (0.213)
Medium 1.209 (0.234) 1.231 (0.246)
Higher 1.057 (0.177) 1.138 (0.199)
Highest 0.674 (0.229) 0.752 (0.269)
Household assets score (Lowest: ref.)
Lower 1.169 (0.223) 1.176 (0.232)
Medium 1.595 (0.305)** 1.521 (0.303)**
Higher 1.921 (0.377)*** 1.853 (0.376)***
Highest 1.943 (0.417)*** 1.757 (0.394)***
Received remittances 1.000 (0.000) 1.000 (0.000)
Household size 0.981 (0.030) 0.975 (0.035)
Stratum (Urban and semi-urban: ref.)
Rice field 0.591 (0.098)*** 0.475 (0.083)***
Plantation 0.624 (0.115)*** 0.540 (0.104)***
Upland 0.322 (0.066)*** 0.258 (0.056)***
Mixed economy 0.643 (0.104)*** 0.549 (0.093)***
Family structure (Nuclear family: ref.)
Extended family with grandparents 1.731 (0.298)***
Extended family with other adults 0.932 (0.192)
Father’s duration of migration (Not move:
ref.)
Father move-out for 1 year 0.772 (0.195)*
Father move-out for 2 years 0.861 (0.309)
Father move-out for 3 years 2.233 (1.036)
Father move-out for 4 years 1.283 (0.207)
Mother’s duration of migration (Not move:
ref.)
Mother move-out for 1 year 0.395 (0.101)***
Mother move-out for 2 years 0.357 (0.143)***
Mother move-out for 3 years 0.073 (0.058)***
Mother move-out for 4 years 0.044 (0.089)***
Having parental divorce 0.387 (0.099)***
Log-likelihood -1265.394 -993.365
N 2,072
Note: *p<0.05; **p<0.01; ***p<0.001 and standard errors in parentheses
16 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Discussion
Using the longitudinal data of KDSS, the present study provides three
interesting findings from the logistic regression models to examine the relationships
between family structure, parental migration, parental divorce and adolescent’s
educational enrollment. The first interesting finding is that extended families with
grandparents have a positive effect on an adolescent’s educational enrollment. The
second interesting finding showed that the duration of migration has negatively affected
the schooling of adolescents, especially mother’s migration. The third interesting
finding is that parental divorce has the negative effect on an adolescent’s education
enrollment.
For the first interesting finding, showing the positive effect of grandparents in
the household, supports literature which argues that resources in the hands of
grandparents will improve adolescent’s welfare outcomes. Grandparents can be an
important source for extra income and child care, especially in single-parent families
where economic deprivation is considered high (Deliere and Kalil, 2002). Using the
KDSS data, the effects of extended household suppress negative impacts of the long-
term absence of the mother on the children’s school enrollment (Jampaklay, 2006).
Grandparents in Thai context have an important role in terms of economic advantage
and socialization. We do not believe that grandparents are substitutes for fathers and
mothers and that this effect holds true across all level of socioeconomic status. Instead,
we suspect that the presence of grandparents interacts with socioeconomic status and is
particularly beneficial among low-income families.
For the second interesting finding about the negative effect of duration of
migration, there are differences in the effect of the length of migration between fathers
and mothers. Short-term migration of the father reduces the likelihood of an
adolescent’s education; while long-term migration of the mother reduces the chance of
enrolling school. It can be said that long-term migration of the father may imply
successful migration, but long-term migration of mother cannot make the benefits for
adolescent in terms of education (Jampaklay, 2006). The present study provides support
for previous findings that it is hard to find a substitute for a mother’s absence.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 17
The third interesting finding proposes that adolescents who have parental
divorce during the year 2001 to 2004 are less likely to enroll in school than those who
have not. This finding is consistent with other research that has shown that adolescents
who have parental divorce often experience poorer educational enrollment and
outcomes (Painter and Levine, 2000). Parental divorce typically involves a shift in
household membership and a reorganization of family roles that disrupt family routine,
resulting in inconsistent parenting. Poor parenting can contribute to emotional
insecurity among children and lower quality parent-child relationships. In addition, the
disadvantage of economic hardship among disrupted families can have negative effects
on an adolescent’s educational enrollment. However, several studies suggest that the
impact of divorce on adolescents may begin long before their parent separate, probably
as a result of conflict, family dysfunction, and economic difficulties before divorce
(Cherlin, et al., 1991; Elliot and Richards, 1991).
Some longitudinal studies have been carried out pertaining to long-term effects
of parental divorce on offspring adjustment (Hetherington and Kelly, 2002; Sun, 2001).
It is generally found that the effects of divorce are most pronounced shortly after a
divorce, while the long-term post divorce effects are rather inconsistent (Hetherington
and Stanley-Hagan, 1999; Sun and Li, 2002). This study uses the parental divorce
within 4 years of the census (KDSS 2001-2004), considered the short-term period the
negative effects on an adolescent’s educational enrollment. Long-term post divorce
effect should be examined in Thai context.
This study found that females are more likely to enroll in school than male,
which is consistent with the finding of Pattaravanich (2005) that gender gap favoring
boys has closed at the national level and that girls now move up to upper secondary
school more often than boys. Surprisingly, there is no difference in educational
enrollment between adolescents living in Thai speaking households and those who live
in non-Thai speaking households.
Household assets score has significantly predicted educational enrollment in
this study. Household assets can increase the likelihood of schooling in Thailand where
inequalities of upper secondary school attendance still exist among different
socioeconomic background of families (Pattaravanich, 2005). This study does not
18 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
support the idea that remittances can contribute to household well-being and an
adolescent’s education as was found from other findings (Jampaklay, 2006; Jones and
Kittisuksathit, 2003). There is no significant relationship between remittances and
adolescent’s educational enrollment. It might be that the household economic resources
tend to have stronger influence on adolescent’s schooling than remittances. Household
wealth often predicts the ability to migrate and the likelihood of migrant success
(Taylor and Wyatt, 1996; Vanwey, 2003). Migrants may not be from the poorest
household that lacking of ability to move.
Although the data from KDSS were not designed to examine directly about the
relationship between family structure, parental migration, parental divorce and an
adolescent’s educational enrollment, its nature of longitudinal data can help us to
capture the migration effects. The finding that extended families lead to a higher chance
of an adolescent’s schooling and not only on the migration status of parents but also on
the duration of the mother’s migration reduce the chances of an adolescent’s schooling,
requires careful policy consideration.
Recommendations
According to the findings, the intervention related to adolescent developments
should consider more about grandparents. As the results that existence of grandparents
is important for an adolescent’s well-being, grandparents should be trained through the
Training of Trainer (TOT) program in order to improve their skills of caring for their
grandchildren.
To extend the results that a mother’s presence is necessary for an adolescent’s
educational enrollment, the income generating project in the household and community
ought to be promoted such as the project of developing village and community capacity
(SML project), which was run by the Thai government. The SML project should be
focusing on the women by generating the available work for women in the community
level, which can reduce the likelihood of migration among women.
Due to the lesser probability of educational enrollment of adolescents in
divorce homes, the office of the Non-formal and Informal Education, Ministry of
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 19
Education should encourage the Life Skill Program to them with the intention of
increasing the capacity for their careers in the future and also their self-esteem.
Acknowledgements
The data for the present study draw upon the Kanchanaburi Demographic
Surveillance System (KDSS), which was funded by the Wellcome Trust, UK. and
collected by the Insitute for Population and Social Research, Mahidol University. We
would like to thank the Kanchanaburi project, IPSR, for allowing us to use their data for
analysis.
References
Amato, P. R., and Booth, A. 1991. Consequences of Parental Divorce and Marital
Unhappiness for Adult Well-being. Social Forces. 69: 905-914.
Battistella, G., and Gastardo-Conaco, M. G. C. 1998. The Impact of Labour Migration
on the Children Left Behind: A Study of Elementary School Children in the
Philippines. Sojourn. 13(2): 220-241.
Brown, S. L. 2006. Family Structure Transitions and Adolescent Well-being.
Demography. 43(3): 447-461.
Buchman, C. 2000. Family Structure, Parental Perceptions, and Child Labor in Kenya:
What Factors Determine Who is Enrolled in School? Social Forces. 78(4):
1349-1379.
Cherlin, A.J., Furstenberg Jr., F.F., Chase-Lansdale, P.L., Kiernan, K.E., Robins, P.K.,
Morrison, D.R., and Teitler, J.O. 1991. Longitudinal Studies of Effects of
Divorce on Children in Great Britain and the United States. Science. 252:
1386-1389.
Coleman, J. 1988. Social Capital in the Creation of Human Capital. American Journal
of Sociology. 95: 95-120.
20 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009
Curran, S., Chang, Y., Cadge, W., and Varangrat, A. 2004. Boys and Girls' Changing
Educational Opportunities in Thailand: the Effects of Siblings, Migration, and
Village Remoteness. Research in Sociology and Education. 14: 59-102.
Deliere, T., and Kalil, A. 2002. Good Things Come in Threes: Single-Parent
Multigenerational Family Structure and Adolescent Adjustment. Demography.
39(2): 393-413.
Elliot, J. and Richards, M. 1991. Children and Divorce: Educational Performance, and
Behavior, Before and After Parental Separation. International Journal of Law
and the Family. 5: 258-276.
Hadi, A. 1999. Overseas Migration and the Well-being of Those Left Behind in Rural
Communities of Bangladesh. Asia-Pacific Population Journal. 14(1): 43-58.
Hanson, T. L., McLanahan, S., and Thomson, E. 1998. Windows on Divorce: Before
and After. Social Science Research. 27: 329-349.
Hao, L., and Xie, G. 2002. The Complexity and Endogeneity of Family Structure in
Explaining Children's Misbehavior. Social Science Research. 31: 1-28
Hetherington, E. M., and Kelly, J. B. 2002. For Better or for Worse. New York: W.W.
Norton.
Hetherington, E. M., and Stanley-Hagan, M. M. 1999. The Adjustment of Children with
Divorced Parents: A Risk and Resiliency Perspective. Journal of Child
Psychological Psychiatry. 40: 129-140.
Jampaklay, A. 2006. Parental Absence and Children's School Enrollment: Evidence
from a Longitudinal Study in Kanchanaburi, Thailand. Asian Population Studies.
2(1): 93-110.
Jones, H., and Kittisuksathit, S. 2003. International Labor Migration and Quality of
Life: Findings from Rural Thailand. International Journal of Population
Geography. 9(6): 517-530.
Khun, R. 2006. The Effects of Fathers' and Siblings' Migration on Children's Pace of
Schooling in Rural Bangladesh. Asian Population Studies. 2: 69-92.
JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009 21
McLanahan, S., and Sandefur, G. 1994. Growing Up with a Single-Parent: What Hurts,
What Helps. Cambridge, MA: Harvard University Press.
Ministry of Education. 2007. Educational Statistics of Thailand. [Online]. Available
from: http://www.moc.moe.go.th/filedata/t1_2_50.htm [Accessed 2008 July 17].
Ministry of Public Health. 2000. Thailand Health Profile 1999-2000. Nonthaburi:
Ministry of Public Health.
National Statistical Office. 2006. Core Social Indicators of Thailand 2006. Bangkok:
Statistical Forecast Units.
Painter, G. and Levine, D.I. 2000. Family Structure and Youths' Outcomes: Which
Correlations Are Causal? Journal of Human Resources. 35: 524-249.
Pattaravanich, U., Williams, L. B., Lyson, T. A., and Archavanitkul, K. 2005. Inequality
and Educational Investment in Thai Children. Rural Sociology. 70(4): 561-583.
Sun, Y. 2001. Family Environment and Adolescents' Well-being Before and After
Parents' Marital Disruption: A Longitudinal Analysis. Journal of Marriage and
the Family. 63: 697-713.
Sun, Y., and Li, Y. 2002. Childrens' Well-being During Parents' Marital Disruption
Process: A Pooled Time-series Analysis. Journal of Marriage and the Family.
64: 472-488.
Taylor, J. E., and Wyatt, T. J. 1996. The Shadow Value of Migrant Remittances,
Income and Inequality in a Household-farm Economy. Journal of Development
Studies. 32(6): 899-912.
Vanwey, L. K. 2003. Land Ownership as a Determinant of Temporary Migration in
Nang Rong, Thailand. European Journal of Population. 19: 121-145.
Wolfinger, N. H. 1998. The Effects of Parental Divorce on Adult Tobacco and Alcohol
Consumption. Journal of Health and Social Behavior. 39: 254-269.
Wu, L. L. 1996. Effects of Family Structure and Income on Risks of Premarital Birth.
American Sociological Review. 61: 386-406.
22 JOURNAL OF POPULATION AND SOCIAL STUDIES Volume 18 Number 1 July 2009