Child:

Original Article
care, health and development
doi:10.1111/j.1365-2214.2011.01276.x

Cross-border marriage and disparities in early

childhood development in a population-based
birth cohort study: the mediation of the
home environment cch_1276 1..9

J. C.-L. Wu,* R. H. Bradley† and T.-L. Chiang*

*College of Public Health, National Taiwan University, Taipei, Taiwan, and
†Family and Human Dynamics Research Institute, Arizona State University, Tempe, AZ, USA

Accepted for publication 14 May 2011

Abstract
Background Taiwan has experienced a large influx of cross-border marriage migrants in recent
years. The majority have been women in their childbearing ages and have come from countries
with lower average standards of living than Taiwan. This trend has changed the ethnic composition
of children who live in Taiwan, and it has generated considerable social concern over the future
health status of Taiwan’s citizens. This study aimed to examine: (1) whether there are disparities in
development between children reared in families characterized by cross-border marriages and
children reared in families with two Taiwanese-born parents; and (2) whether the quality of home
environment explains the group differences in early childhood development.
Methods Data came from the Taiwan Birth Cohort Study. A total of 19 499 participants who
completed 6-month, 18-month and 3-year surveys were included for analysis. Cross-border
marriage status was defined by mother’s original nationality and categorized into three broad
groups: Taiwanese-born, Chinese cross-border and South-East Asian (SEA) cross-border. Early
childhood development was measured at age 3 years, and covered the domains of gross motor, fine
motor, language and socio-emotional competence. Hierarchical linear regressions were used to
Keywords
examine the mediation effects of the home environment.
birth cohort, cross-border
marriage, early childhood Results Children of Chinese and SEA cross-border groups scored lower in fine motor, language
development, home and socio-emotional competence than those of their Taiwanese-born counterpart at age 3 years.
environment, Taiwan
Chinese–Taiwanese group differences in all three developmental domains became insignificant
Correspondence:
after the addition of home environment, while SEA–Taiwanese group differences in fine motor and
Tung-liang Chiang, language development remained, yet were noticeably reduced. The mediation of home
College of Public Health,
environment was further confirmed using the Sobel test.
National Taiwan
University, Room 620, No. Conclusions Home environment plays a central role in reducing the disparities in developmental
17, Xu-Zhou Road, Taipei outcomes among children of different marriage groups. Interventions should be directed towards
10055, Taiwan
E-mail:
enhancing the quality of early home environment for children reared in families of cross-border
[email protected] marriages.

© 2011 Blackwell Publishing Ltd 1

2 J.C.-L. Wu et al.

By contrast, marriages to female migrants from other countries

Introduction
were often (about 40%) commercially arranged (Ministry of the
Cross-border marriages are common in a number of countries. Interior 2004).
These marital unions involve the international migration of
Several factors coalesced to account for the mass influx of
one spouse to the country of residence of the other. Because so
female marriage immigrants to East Asia, including Taiwan
many children now reside in households that have parents in
(Tien & Wang 2006; Jones & Shen 2008; Lee 2010). The main
cross-border marriages, concerns about potential negative con-
driving force behind cross-border marriages pertains to the
sequences stemming from the ethnic composition of Taiwan’s
challenges in the domestic marriage market faced by Taiwanese
child population have become prominent in the public dis-
men who are older, of lower social class and live in rural areas.
course. In Taiwan, cross-border marriages frequently involve
This push to seek spouses from less developed countries was
Taiwanese men who have limited social assets or who have
facilitated by Taiwan’s aggressive trading in the South-East
health constraints that marry female migrants from less devel-
Asian (SEA) region, freeing the ban against communication
oped countries. This combination of disadvantages can lead to
with China and a proliferation of marriage brokers (Hsia 2004;
less stimulating and supportive home environments for chil-
Lu 2005; Tseng 2010).
dren, with evidence suggesting that the children may be more
Not surprisingly, many of these female migrants who
vulnerable to developmental challenges. Specifically, they are at
engaged in cross-border marriages were expected to reproduce
greater risk for poor health and school achievement (Bradley &
offspring and take up gendered family roles as mainstream
Corwyn 2002). Furthermore, poor health in childhood and
Taiwanese culture has traditionally stressed (Tien & Wang
poor school achievement often forecast lifelong health prob-
2006; Chen 2008). In the mid-2000s, around one out of every
lems and diminished success with employment (Irwin et al.
eight children was born to mothers of marriage migration
2007). Taking all of this into account suggests that there is an
(Ministry of the Interior 2008). As stated earlier, this sizable
urgent need to document the health and developmental status
group of new Taiwanese children has drawn considerable
of these so-called ‘new Taiwanese children’, with a view towards
social and policy attention out of concern that they may be
addressing social disparities if they are observed. The popula-
prone to developmental problems (Mo & Lai 2004; Tseng
tion level data from the Taiwan Birth Cohort Study (TBCS)
2010).
afford an opportunity to better understand both the living
conditions and the developmental trajectories of these
children.
Previous literature
Research does not yet make clear whether children born to
parents of cross-border marriages in Taiwan are at higher
Cross-border marriages in Taiwan
risks of developmental and health problems. Studies to date
As intra-regional flows of marriage migration became more are rather consistent in demonstrating that newborns of
prevalent since the 1990s, Taiwan has been the receiving country foreign-born mothers have a lower risk of adverse outcomes
with highest rate of cross-border marriages (Lee 2010). At the than those of native-born mothers (Liao et al. 2006; See et al.
peak year of 2003, one out of every three registered marriages in 2007; Chiang & Wu 2011). Social processes that select healthy
Taiwan was in a cross-border form, with more than 98% of brides have been proposed to explain the favourable birth
cross-border marriages involving a foreign-born woman (Min- outcomes observed in children of foreign-born mothers.
istry of the Interior 2010a). Female spouses from China consti- However, there remains a paucity of research on the relation-
tute about two-thirds of marriage migrants in Taiwan, followed ship between cross-border marriage status and early child-
by women from Vietnam, Indonesia and Thailand (Ministry of hood development. Prior research indicates that children of
the Interior 2010b). On average, these female marriage migrants SEA mothers tend to perform less well on measures of com-
were 12 years younger than their Taiwanese spouses (Ministry prehensive mental ability and social behaviour than children
of the Interior 2004). The cross-border marriages tended to be born to native mothers (Chung et al. 2006; Chiang & Lin
geographically clustered in Taiwan’s farming areas (Chi et al. 2009). The same pattern has also emerged with regard to
2010). According to a national survey, most Chinese-born receptive and expressive language development (Chung et al.
women in cross-border marriages met their Taiwanese spouses 2006; Huang et al. 2008). Chen and Chou’s (2008) study of
through informal networks (61.4%) or on their own (27.5%). SEA mothers and their pre-school children showed that

© 2011 Blackwell Publishing Ltd, Child: care, health and development

 Cross-border marriage and early childhood development 3

scores on the Peabody Picture Vocabulary-Revised test

Methods
for all children were below 50th percentile, based on national
norms. Moreover, there were generally low levels of reading
Sources of data and participants
materials and enriching activities in the family.
Further research on home environmental conditions and Our analysis was based on the TBCS. TBCS is the first large-
their role in the development of children from cross-border scale longitudinal study on a whole-year birth cohort in Taiwan,
marriages is important for three major reasons. First, prior and aims to: (1) document health and developmental trajecto-
studies used cross-sectional designs; thus, it is not possible to ries of Taiwanese children; (2) investigate the impact of the
determine the temporal order of exposures and child outcomes. social environment on children’s health and well-being; and (3)
Second, sample sizes have been small, with participants mainly examine the early origins of adult health from a life course
drawn from child care institutions or from a single locale. Third, perspective.
although researchers have made reference to the likelihood that By using two-stage stratified random sampling, a nationally
certain types of social conditions and aspects of the home learn- representative cohort of 24 200 children born in 2005 was ini-
ing environment are likely implicated in these observed group tially drawn from the National Birth Report Database, with a
disparities, the mediating effects of such factors were not explic- sampling rate approximating 12%. Before the study cohort
itly tested. approached school age, four waves of interview surveys were
scheduled at 6 months, 18 months, 3 years and 5.5 years of age
respectively. Each survey was conducted by face-to-face inter-
This present study
view with the mother or primary caregiver given their informed
Bronfenbrenner’s bio-ecological perspective of human develop- consent, to follow up each child’s growth, development and
ment provides a general frame for our research questions. health conditions. In addition, a wide range of information on
According to Bronfenbrenner (1979), people function within a social exposures and physical environment was collected across
set of nested and interactive systems that affect their behaviours developmental stages. To accommodate foreign-born mothers
and development. These include micro-systems such as the potentially with communication problems, translated versions
family and other social contexts which tend to have significant of certain assessment scales and terminology were made avail-
impact on the developing person as well as macro-systems such able during the interview. As rated by the interviewers, overall,
as culture which impact micro-systems and have more distal 85% of these foreign-born mothers comprehended questions
influences on the individual. well.
Bronfenbrenner (1994) further states that proximal processes A subset of 19 499 children who completed 6-month,
connected to these nested systems serve as the primary engines 18-month and 3-year surveys were included in this present
of development. Proximal processes refer to engagements with analysis, representing a response rate of 80.57%. There were
persons, objects or symbols in the immediate environment. more boys (52.50%) than girls (47.50%). A total of 8.32% of the
Generally speaking, these processes must occur regularly over children were preterm while 6.77% were of low birthweight.
an extended period of time in order to trigger the progress of The majority of the mothers had 10–12 years (39.92%) or more
individual development, albeit there are exceptions such as than 12 years (45.87%) of education and were native Taiwanese
traumatic events. Following this framework, we placed our (87.20%).
focus on the role of home environmental processes as respon-
sible for differences in developmental functioning between chil-
Definitions and measures of study variables
dren living in cross-border marriage households, and children
living in families where both mother and father were native-
Marriage group
born Taiwanese, with consideration of key covariates in the
ecological contexts. More specifically, the present study was For this study, marriage group membership was determined by
designed to examine two questions: (1) Do developmental out- mother’s original nationality. Families were initially split into
comes vary between children reared in families characterized by two broad categories: Taiwanese-born (composed of a mother
cross-border marriages and children reared in families with two and a father both born in Taiwan) and cross-border (composed
Taiwanese-born parents? (2) Does the quality of the home of a Taiwanese-born father and a foreign-born mother).
environment explain the group differences in developmental Foreign-born mothers were identified by checking their self-
outcomes? reported nationality of origin to also include those who have

© 2011 Blackwell Publishing Ltd, Child: care, health and development

4 J.C.-L. Wu et al.

been entitled to Taiwanese citizenship. Among this study’s developmental task.‘Don’t know’ and ‘Not applicable’ responses
foreign-born mothers, close to one-third (n = 803; 32.25%) were coded as ‘never’ to give a conservative estimation and to
were from China while the majority (n = 1664; 66.83%) of reduce missing values.
the rest came from SEA countries that included Vietnam
(51.12%), Indonesia (8.76%), Cambodia (3.09%), the Philip-
Analytical strategies
pines (1.37%), Burma (1.37%), Thailand (1.00%) and Malaysia
(0.16%). Mothers who moved from Japan, South Korea, the We first conducted bivariate analyses to compare marriage
USA and Eastern European countries accounted for less than groups (Taiwanese-born, Chinese cross-border and SEA cross-
1% (n = 23), and were excluded from our further analysis due to border) with respect to family socio-demographic characteris-
the small number. Since female marriage immigrants in Taiwan tics, home environment and child developmental outcomes at
vary in their migration context and patterns, we further age 3 years. We then conducted hierarchical multiple regres-
classified these cross-border families into two groups based on sion analysis to examine whether any observed differences in
mother’s country of birth: Chinese and SEA. the developmental status associated with marriage group
membership might be mediated by the quality of the home
environment.
Home environment When performing the hierarchical regression analyses, we
examined three sequential models for each development
Home environment was measured at age 3 years. A caregiver-
domain. In Model I, 3-year developmental scores were regressed
report measure adapted from the Home Observation for
on dummy variables of marriage group to present estimates of
the Measurement of Environment Inventory – Short Form
crude differences. For Model II, we added child characteristics
(HOME-SF) was used to assess the quality of children’s home
(including sex and low-birthweight status), father’s education
environment (Wu et al. 2010). This measure contains 12 items
as a generic measure of family socio-economic status, and
that cover cognitive stimulation and emotional support pro-
18-month developmental domain scores in an effort to partial
vided to children at home (e.g. ‘Read to or with the child’,
out the variance explained by control variables. In Model III, we
‘Respond to the child’s request or question’). Respondents rated
introduced the variable of home environment. A reduction in
each item on a 5-point Likert scale that addresses the frequency
the magnitudes of the independent variable coefficients from
with which each stated situation happens (1 = Never to 5 =
Model II, and a significant association between home environ-
Always). A total home environment score was obtained by
ment and developmental scores would provide indication of
summing all item scores, with higher scores representing an
mediating effects (Baron & Kenny 1986).
overall more enriching home environment. Construct validity
We further used the Sobel’s (1982) formula to test the media-
was confirmed by comparing mean scores among known
tion of home environment. The test statistic was calculated as
groups such as family socio-economic status and number of
the product of two coefficients on the mediating pathway
children in the household, as well as by showing significant
divided by its standard error, and a value greater than 1.96
associations with child’s developmental functioning.
supports the proposed mediation at 0.05 level of significance.

Child developmental status Results

The TBCS developmental scale is a caregiver-report instrument
Descriptive analysis
designed to measure children’s gross motor, fine motor, lan-
guage and communication, and socio-emotional development The cross tabulations of socio-demographic characteristics by
at different ages, basically in sync with the study’s survey sched- marriage group membership are shown in Table 1. On average,
ule. The 6-month (38 items), 18-month (17 items) and 3-year SEA mothers were significantly younger (mean = 24.42 years;
(19 items) scales are all short in length, but have presented good SD = 4.27 years) than Taiwanese (mean = 29.35 years; SD = 4.70
reliability. There is also evidence for construct validity, as well as years) and Chinese mothers (mean = 28.23 years; SD = 4.12
criterion validity with the Bayley Scale of Infant Development – years). Overall, SEA mothers came from most socio-
Second Edition (Lung et al. 2010). Respondents were asked to economically disadvantaged families, followed by Chinese
use a 3-point Likert scale (0 = Never, 1 = Sometimes and 2 = mothers. Specifically, 70% of SEA mothers received less than 10
Always) to indicate how capable the child performing each years of education, compared with 50% of Chinese mothers,

© 2011 Blackwell Publishing Ltd, Child: care, health and development

 Cross-border marriage and early childhood development 5

Table 1. Distribution of family

Taiwanese- Chinese SEA
socio-demographic characteristics by
Variable Total born cross-border cross-border
cross-border marriage group
Mother’s age* (years) (n = 19 476)
<20 1.94 1.57 0.00 6.67
20–24 17.38 13.76 19.68 53.25
25–29 36.62 37.03 46.95 27.52
30–34 31.54 34.03 25.28 9.07
ⱖ35 12.52 13.62 8.09 3.49
Mother’s education (years) (n = 19 442)
ⱕ6 3.87 0.58 8.76 35.22
7–9 10.34 6.35 44.18 34.92
10–12 39.93 41.59 36.55 24.61
ⱖ13 45.85 51.48 10.51 5.25
Father’s education (years) (n = 19 403)
ⱕ6 1.35 0.78 2.52 6.64
7–9 12.11 9.22 24.69 35.63
10–12 49.66 38.58 46.22 47.52
ⱖ13 46.88 51.42 26.57 10.21
Father’s occupation (n = 19 360)
Professionals 13.17 14.68 4.36 2.05
Administrators/managers 26.38 27.79 24.16 13.19
Clerks/service personnel 20.82 21.80 18.06 12.23
Semi-skilled workers 31.82 28.75 39.98 59.10
Non-skilled workers 4.52 4.04 7.22 8.13
Unemployed 3.27 2.94 6.23 5.30
Family monthly income ($NT) (n = 19 422)
ⱕ30 000 11.57 8.58 25.22 35.73
30 000–50 000 30.02 27.26 49.94 48.67
50 000–70 000 26.12 27.82 17.35 12.88
70 000–100 000 21.20 23.86 4.74 1.87
>100 000 11.09 12.48 2.75 0.85
Residential area (n = 19 476)
City 47.07 48.37 45.08 34.74
Township 27.72 27.84 27.15 26.85
Rural township 25.21 23.80 27.77 38.41
Duration in Taiwan* (year; – – 3.53 ⫾ 2.60 3.30 ⫾ 2.30
mean ⫾ SD) (n = 2490)

*Measures are for the year at child’s birth.

SEA, South-East Asian.

and about 15% of Taiwanese mothers. The distribution of never read with/to the child. This compares with less than 30%
spouse’s occupational status was similar for Chinese and reported by Taiwanese mothers. The observed marriage group
Taiwanese mothers, albeit a higher proportion of Taiwanese differences pertaining to indicators of emotional support
mothers had spouses who worked as professionals. Spouses of tended to be small with interesting variations. For example,
SEA women were predominantly semi-skilled workers. Most Taiwanese mothers were least likely to rarely or never caress/kiss
Taiwanese (48.37%) and Chinese (45.08%) mothers lived in the child, but most likely to rarely or never converse pleasantly
cities whereas only 38.41% of SEA mothers lived in cities. with the child. Taken as a whole, children of Taiwanese group
Table 2 displays the percentage of participants in each mar- (mean = 45.61; SD = 6.07) were exposed to significantly more
riage group that responded ‘rarely’ or ‘never’ with respect to supportive and stimulating home environment than their
each of the 12 home environment indicators. Statistically sig- Chinese (mean = 43.40; SD = 5.98) and SEA (mean = 41.67;
nificant differences were observed for most indicators, but mar- SD = 5.72) counterparts.
riage group differences were most pronounced for items dealing Table 3 demonstrates differences in scores for gross motor,
cognitive stimulation in the home. For instance, more than half fine motor, language and socio-emotional competence at ages
of SEA mothers and over 35% of Chinese mothers reported that 18 months and 3 years for children from the three marriage
the child rarely or never had new books, or that they rarely or groups. Univariate linear models indicated that children from

© 2011 Blackwell Publishing Ltd, Child: care, health and development

6 J.C.-L. Wu et al.

Table 2. Responses to the home environment items by marriage group: percentage responding ‘rarely’ or ‘never’
Variable Taiwanese-born Chinese cross-border SEA cross-border Group differences*
Child has new toys 23.37 28.52 28.25 T < C, SEA
Child has new books 29.80 38.48 52.82 T < C < SEA
Child taken to grocery 12.75 18.18 18.03 T < C, SEA
Child taken on an outing such as to a park 16.60 24.53 30.47 T < C < SEA
Mother reads to or with the child 26.32 36.74 56.67 T < C < SEA
Mother plays with the child 18.39 32.00 36.54 T < C < SEA
Mother caresses or kisses the child 3.07 4.98 4.51 T < C, SEA
Mother attends to the child while doing household work 1.82 1.74 3.06 T, C < SEA
Mother talks to the child while doing household work 6.70 6.85 7.03 T < C < SEA
Mother responses to the child’s request or question 1.49 2.86 3.55 T < C, SEA
Mother spontaneously praises the child 2.02 2.99 3.73 T, C, SEA
Mother converses pleasantly with the child 1.66 1.25 1.14 T, C, SEA
Total home environmental score (Mean ⫾ SD) 45.61 ⫾ 6.07 43.40 ⫾ 5.98 41.67 ⫾ 5.72 T > C > SEA†

*Statistically significant group differences in the percentage responding ‘rarely’ or ‘never’ of each item.
†Statistically significant group differences in the mean of total home environment scores.
SEA, South-East Asian; T, Taiwanese; C, Chinese.

Table 3. Developmental domain scores (mean

Variable Taiwanese-born Chinese cross-border SEA cross-border
and standard deviation) at age 18 months and
18 months 3 years by marriage group
Gross motor 6.81 (1.44) 6.85 (1.45) 7.01 (1.44)***
Fine motor 6.74 (1.46) 6.71 (1.44) 6.55 (1.61)***
Language 6.02 (2.18) 5.97 (2.07) 5.55 (2.27)***
Socio-emotional 8.08 (1.79) 8.45 (1.80)*** 8.25 (1.88)***
3 years
Gross motor 10.68 (1.76) 10.61 (1.80) 10.59 (1.92)
Fine motor 6.01 (1.73) 5.72 (1.80)*** 5.55 (1.87)***
Language 7.86 (0.72) 7.78 (0.90)*** 7.60 (1.13)***
Socio-emotional 8.75 (1.48) 8.67 (1.61) 8.54 (1.75)***

Significant differences from Taiwanese-born group: *P < 0.05; **P < 0.01; ***P < 0.001.
SEA, South-East Asian.

the SEA cross-border group performed less well in gross motor, multiple regression analyses were only performed on fine
fine motor and language development than those of the motor, language and socio-emotional scores. Results from these
Taiwanese-born group at age 18 months. For the same develop- analyses are presented in Table 4.
mental domains, no significant differences were found between
children of Chinese cross-border and Taiwanese-born groups.
Fine motor development
However, children from the Taiwanese-born group scored
lowest in socio-emotional competence at age 18 months. At age In Model II where previous developmental status, child charac-
3 years, there was no group difference in gross motor develop- teristics and father’s education were controlled, differences in
ment. Nonetheless, children from the Taiwanese group scored performance in motor skills between both cross-border mar-
highest in the other three areas examined (fine motor, language riage groups and the Taiwanese group remained significant.
and socio-emotional), whereas children born to SEA mothers When home environment scores were added in Model III,
scored lowest. Chinese cross-border group was no longer different from
Taiwanese-born group. However, while the coefficient for SEA
cross-border group reduced substantially, it remained signifi-
Multiple regression analyses
cant (P < 0.05). Home environment scores were significantly
Because bivariate analyses revealed no marriage group differ- associated with fine motor development (P < 0.001) and con-
ences in children’s gross motor performance at age 3 years, tributed 3.3% to the total variance explained. We employed

© 2011 Blackwell Publishing Ltd, Child: care, health and development

 Cross-border marriage and early childhood development 7

Table 4. Hierarchical regression analyses for predicting developmental outcomes at age 3 years
Fine motor (b) Language (b) Socio-emotional (b)
I II III I II III I II III
Cross-border marriage (ref. Taiwanese-born)
Chinese cross-border -0.30 -0.18** -0.11 0.09** -0.06* -0.04 -0.08 -0.11* -0.04
SEA cross-border -0.47*** -0.23*** -0.09* -0.26*** -0.18*** -0.14*** -0.21*** -0.15*** -0.03
Child characteristics
Boys -0.31*** -0.32*** 0.03** 0.02* -0.11*** -0.13***
Low birthweight -0.17*** -0.17*** -0.15*** -0.15*** -0.17*** -0.17***
Father’s education (ref. <10 years)
10–12 years 0.20*** 0.13*** 0.07*** 0.05** 0.04 -0.02
ⱖ13 years 0.47*** 0.27*** 0.13*** 0.08*** 0.23*** 0.06
18-month developmental 0.34*** 0.30*** 0.09*** 0.08*** 0.23*** 0.21***
scores
Home environment score 0.05*** 0.02*** 0.05***
0.006 0.111 0.144 0.009 0.081 0.092 0.001 0.087 0.117

*P < 0.05; **P < 0.01; ***P < 0.001.

SEA, South-East Asian.

Sobel’s test, and confirmed that the home environment is a significant mediator for both SEA (t = -20.00, P < 0.001) and
significant mediator of differences in fine motor development Chinese groups (t = -9.63, P < 0.001).
for both Chinese (t = -9.70, P < 0.001) and SEA (t = -20.81,
P < 0.001) groups when compared with Taiwanese-born group.
Discussion
Our study is the first to use survey data from a large-scale
Language development population-based birth cohort to explore the association
between cross-border marriage and early childhood develop-
For language development, adjusted group differences in Model
ment for the Taiwan’s context. We first demonstrated that chil-
II for both Chinese and SEA cross-border groups were signifi-
dren born to parents of cross-border marriages constituted a
cant. When home environment scores were introduced into
potentially vulnerable group as a consequence of family socio-
the equation, difference between children with Chinese and
economic disadvantages. Consistent with descriptive profiles
Taiwanese mothers disappeared while differences between chil-
from previous studies (e.g. Huang et al. 2008; Chiang & Lin
dren with SEA and Taiwanese mothers remained significant
2009), children born to migrant mothers from SEA countries
(P < 0.001), with the coefficient slightly reduced. Also shown in
performed less well in fine motor, language and socio-economic
Model III, home environment was significantly associated with
development at age 3 years than children born to native Taiwan-
language ability (P < 0.001), and accounted for an additional
ese mothers. In addition, our results suggest that maternal
1.1% of the variance. Again, Sobel’s test confirmed the mediated
familiarity with the language and mores of the country of des-
path for both SEA (t = -11.32, P < 0.001) and Chinese groups
tination may also make a difference in the likelihood children
(t = -3.06, P = 0.002).
from cross-border marriages will do well. Specifically, children
born to Chinese mothers generally performed better than those
born to SEA mothers.
Socio-emotional development
Importantly, our findings support the mediating role played
The same analytical approach was conducted for socio- by home environment in relations between marriage group
emotional development (see Table 4). In Model II, adjusted status and early child development. When indicators of the
group differences were significant, but no longer existed after quality of the home environment were included in our analy-
the addition of home environment scores in Model III. Home tical models, differences between children from Chinese
environment was a significant predictor of socio-emotional cross-border families and children from Taiwanese families dis-
development (P < 0.001) and independently explained 3.0% appeared. As well, differences were reduced or eliminated for
model R2. Sobel’s test revealed that home environment was a children from SEA cross-border families. Although its mediat-

© 2011 Blackwell Publishing Ltd, Child: care, health and development

8 J.C.-L. Wu et al.

ing effects were only partial for SEA group in fine motor and Although our study only involves cross-border marriages in
language development, the corresponding coefficients were Taiwan, the conditions that favour such marriages exist elsewhere
reduced by 60.87% and 22.22% respectively. This sheds light on as well. Consequently, it will be useful to examine the impacts of
the vital role of home environment in mitigating the develop- quality home environment on children’s development in other
mental disadvantages among children born in families of societies where cross-border marriages are prevalent.
cross-border marriages. Interestingly, caregivers from different
cultures may not display actions in the same manners or pat-
terns to manage the home environment for their young chil- Key messages
dren; it still holds true that more stimulating and supporting
home environment can predict better developmental outcomes • Taiwan has been the receiving country with highest rate of
(Bradley 2011). cross-border marriages under recent growth trends in
Why a complete mediation of home environment was intra-regional flows of marriage migration.
observed for Chinese cross-border but not SEA cross-border • Children reared in families of cross-border marriages tend
group remains unclear. Given that Chinese and Taiwanese are to be of lower socio-economic position and perform less
generally homogenous with respect to race, and share similar well in fine motor, language and socio-emotional develop-
cultural roots as opposed to SEA, we suggest that further ment than children whose parents are native Taiwanese.
research may examine biological (e.g. genes, prenatal history) • Home environment considerably mediates the effects of
and cultural (e.g. parental beliefs about roles and child devel- marriage group on early childhood development regard-
opment) factors that could help explain the findings. For less of children’s socio-economic position.
example, according to Reifsnider and colleagues’ (2005) eco- • Early interventions geared towards enhancing the home
logical framework, ethnic disparities in child development can environment for families of cross-border marriage is a key
also result from differential access to facilitating resources policy strategy.
largely affected by mother’s ability of mastering host country
language, assimilation level and other ethnicity-related factors.
A detailed discussion of these issues is beyond the scope of our Acknowledgements
study, but worthy of future research attention.
This study was based on longitudinal data from the Taiwan
Another finding to note from the hierarchical regression
Birth Cohort Study, funded by the Bureau of Health Promotion,
analyses is that the introduction of home environment consis-
Department of Health, Taiwan. The authors gratefully acknowl-
tently attenuated the coefficients of father’s education, with
edge the time and continuing support from the participating
home environment itself remaining a significant predictor
families to this project.
across all developmental domains. This implies that home
environment not only has a direct effect on developmental
outcomes but also functions as a mediator of family socio- References
economic conditions. This is also congruent with previous lit-
erature linking family socio-economic to child outcomes Baron, R. M. & Kenny, D. A. (1986) The moderator-mediator
variable distinction in social psychological research: conceptual,
through the home environment (Linver et al. 2002; Raviv et al.
strategic, and statistical consideration. Journal of Personality and
2004; Lugo-Gil & Tamis-LeMonda 2008).
Social Psychology, 51, 1173–1182.
This study underscores the importance of home environment Bradley, R. H. (2011) The HOME inventory. In: A Developmental
in reducing the disparities in developmental outcomes among Environment Measurement Handbook (eds L. C. Mayes & M.
children from different marriage groups. Hence, enhancing the Lewis). Cambridge University Press, New York, NY, USA
quality of home environment for children reared in families of (in press).
cross-border marriages should be placed at the core of policies. Bradley, R. H. & Corwyn, R. F. (2002) Socioeconomic status and
We recommend that policies be designed to directly influence the child development. Annual Review of Psychology, 53, 371–399.
Bronfenbrenner, U. (1979) The Ecology of Human Development:
home environment or via indirect measures such as more
Experiments by Nature and Design. Harvard University Press,
upstream socio-economic conditions. Future research, based on Cambridge, MA, USA.
this representative birth cohort, is needed to document the Bronfenbrenner, U. (1994) Nature-nuture reconceptualised in
trajectory of gaps in child development by marriage group development perspective: a bioecological model. Psychological
and investigate the role of the wider social ecological context. Review, 101, 568–586.

© 2011 Blackwell Publishing Ltd, Child: care, health and development

 Cross-border marriage and early childhood development 9

Chen, Y. H. (2008) The significance of cross-border marriage in a low Lung, F. W., Chiang, T. L., Lin, S. J., Lee, M. C. & Shu, B. C. (2010)
fertility society: evidence from Taiwan. Journal of Comparative Child developmental screening instrument from six to thirty-six
Family Studies, 39, 331–352. months in Taiwan Birth Cohort Study. Early Human Development,
Chen, Y. H. & Chou, J. T. (2008) The language development and 86, 17–21.
home literacy environment of foreign spouse’s children at Ministry of the Interior (2004) Report of 2003 survey on living
preschool. Chia-Nan Annual Bulletin, 34, 543–554. conditions of foreign and Chinese spouses in Taiwan (Chinese).
Chi, Y. L., Jhou, M. S. & Hsieh, Y. S. (2010) Spatial analysis of Ministry of the Interior, Executive Yuan, Taipei, Taiwan. Available
foreign-brides in Taiwan (Chinese). Journal of Population Studies, at: http://www.ris.gov.tw/ch4/0930617-1.html (last accessed 20
38, 67–113. December 2010).
Chiang, T. Y. & Lin, C. Y. (2009) The research of child social Ministry of the Interior (2008) White Paper on Population Policy in
behaviour development in Southeast Asia new immigrants family Taiwan (Chinese). Ministry of the Interior, Executive Yuan, Taipei,
(Chinese). Journal of Early Childhood Education, 20, 97–124. Taiwan.
Chiang, T. L. & Wu, J. C. L. (2011) Cross-border marriage. In: The Ministry of the Interior (2010a) Statistics on registered marriages with
Health Profile of Taiwanese Children in the 21st Century (Chinese) foreigners (Chinese). Available at: http://www.immigration.gov.tw/
(eds T. L. Chiang & S. J. Lin). Bureau of Health Promotion, public/Attachment/01227953494.xls (last accessed 5 January
Ministry of Health, Taipei, Taiwan (in press). 2011).
Chung, F. C., Wang, K. C. & Chen, Y. L. (2006) A comparison study Ministry of the Interior (2010b) Statistics on foreign and Chinese
of language, mental abilities development, and learning behaviour spouses (Chinese). Available at: http://www.immigration.gov.tw/
for children of foreign bride families in Pingtung (Chinese). public/Attachment/112413442941.xls (last accessed 5 January
Bulletin of Educational Psychology, 37, 411–429. 2011).
Hsia, H. C. (2004) Internationalization of capital and the trade in Mo, L. L. & Lai, P. L. (2004) A preliminary discussion of low fertility
Asian women – the case of ‘foreign brides’ in Taiwan. In: Women and children of foreign spouses in Taiwan (Chinese). Community
and Globalization (eds D. Aguilar & A. Lacsamana), pp. 181–229. Development Quarterly, 105, 55–65.
Humanity Press, New York, NY, USA. Raviv, T., Kessenich, M. & Morrison, F. J. (2004) A meditational
Huang, K. Y., Chung, L. C., Liao, Y. C. & Chen, P. C. (2008) Speech model of the association between socioeconomic status and
intelligibility and productivity in preschool children of foreign three-year-old language abilities: the role of parenting factors.
spouse. Journal of the Speech-Language-Hearing Association of Early Childhood Research Quarterly, 19, 528–547.
Taiwan, 21, 55–91. Reifsnider, E., Gallagher, M. & Forgione, B. (2005) Using ecological
Irwin, L. G., Siddiqi, A. & Hertzman, C. (2007) Early Child models in research on health disparities. Journal of Professional
Development: A Powerful Equalizer. Human Early Learning Nursing, 21, 216–222.
Partnership, Vancouver, BC, Canada. The Knowledge Networks See, L. C., Shen, Y. M. & Lo, Y. J. (2007) The differential effect of
final report for the World Health Organization’s Commission on foreign-born status on preterm birth: a Taiwan population-based
the Social Determinants of Health. birth registry study. BJOG: An International Journal of Obstetrics
Jones, G. & Shen, H. H. (2008) International marriage in East and and Gynaecology, 114, 1023–1027.
Southeast Asia: trends and research emphasis. Citizenship Studies, Sobel, M. E. (1982) Asymptotic intervals for indirect effects in
12, 9–25. structural equations models. In: Sociological Methodology
Lee, H. K. (2010) Family migration issues in the North East Asia. (ed. S. Leinhart), pp. 290–312. Jossey-Bass, San Francisco, CA,
Background paper for the World Migration Report 2010. USA.
International Organization for Migration, Geneva, Switzerland. Tien, C. Y. & Wang, H. Z. (2006) Masculinity and cross-border
Liao, J. J., Lin, C. L., Chan, W. Y., Sung, F. C. & Liao, C. C. (2006) marriages: why do Taiwanese men seek Vietnamese women
Comparison of low birth weight infants born to alien and native to marry? Taiwan Journal of Southeast Asian Studies, 3, 1–36.
brides. Taiwan Journal of Family Medicine, 16, 226–236. Tseng, Y. F. (2010) Marriage migration to East Asia: current issues
Linver, M. R., Brooks-Gunn, J. & Kohen, D. E. (2002) Family and proposition in making comparisons. In: Asian Cross-Border
processes as pathways from income to young children’s Marriage Migration: Demographic Patterns and Social Issues
development. Developmental Psychology, 38, 719–734. (eds W. S. Yang & M. C. W. Lu), pp. 31–45. IIAS, Amsterdam
Lu, M. C. W. (2005) Commercially arranged marriage migration: case University Press, Amsterdam, the Netherlands.
studies of cross-border marriages in Taiwan. Indian Journal of Wu, J. C. L., Chiang, T.-L. & Bradley, R. H. (2010) Adaptation
Gender Studies, 12, 275–303. and validation of the HOME-SF as a caregiver-report home
Lugo-Gil, J. & Tamis-LeMonda, C. S. (2008) Family resources and environment measure for use in the Taiwan Birth Cohort Study
parenting quality: links to children’s cognitive development across (TBCS). Early Child Development and Care, Published online
the first 3 years. Child Development, 79, 1065–1085. 8 September 2010, DOI: 10.1080/03004430.2010.504881.

© 2011 Blackwell Publishing Ltd, Child: care, health and development