734232

research-article2017
YASXXX10.1177/0044118X17734232Youth & SocietyCheung

Article
Youth & Society
2020, Vol. 52(1) 99­–118
Gender, Victimization, © The Author(s) 2017
Article reuse guidelines:
and Adolescent sagepub.com/journals-permissions
DOI: 10.1177/0044118X17734232
https://doi.org/10.1177/0044118X17734232
Psychological Health in journals.sagepub.com/home/yas

the Context of Migration:

Evidence From China

Nicole W. T. Cheung1

Abstract
Although gender has become a key dimension in migration research, previous
scholarship has largely focused on adult rather than child migrants. Even
less attention has been paid to the role of gender in migration-related child
health. By comparing rural migrant and urban-born adolescents in the largest
city in south China, this study assessed gender differences in psychological
health and whether such differences were informed by socially induced
victimization stressors and stress-moderation mechanisms related to the
social relationships that link migrants to their host cities (local ties) and
home villages (translocal ties). Victimization was more strongly correlated
with suboptimal psychological health in girls versus boys across the migrant
and urban-born samples. Local and translocal ties directly increased the
psychological health and buffered the deleterious effect of victimization for
migrant girls; yet, the roles of these ties were less notable in migrant boys.
These results indicate gendered susceptibility to social stress in migrant
adolescents.

Keywords
gender, health, mental health, immigration, victimization, stress

1The Chinese University of Hong Kong, Sha Tin, Hong Kong

Corresponding Author:
Nicole W. T. Cheung, Department of Sociology, The Chinese University of Hong Kong, Sha
Tin, New Territories, Hong Kong.
Email: nwtcheung@cuhk.edu.hk
100 Youth & Society 52(1)

There is growing meta-analytic evidence that links the stressful impact of

migration to common mental disorders (depression and anxiety) (Das-
Munshi, Leavey, Stansfeld, & Prince, 2012). This health disadvantage is
especially evident when the healthy migrant effect, which holds that health-
ier individuals tend to migrate, diminishes over time in tandem with difficult
assimilation (S. Lee, O’Neill, Ihara, & Chae, 2013). This disadvantage also
intersects with gender, with migrant women being more vulnerable to health
risks (Malmusi, Borrell, & Benach, 2010). It has been widely recognized
that migration cannot be understood independently of gender, which shapes
the motivations and risks governing the migration movement. Gender also
clarifies the reconfiguration of power relations in household, workplace, and
health care at the destinations, and cultural changes in the origin communi-
ties (Lu & Tao, 2015; Paul, 2015; Pessar & Mahler, 2003). However, most
previous work has focused on adult migrants and has produced limited
knowledge of the gender issue in child migration. This limitation has been
particularly evident in research on migration-related child health (Stevens &
Vollebergh, 2008). Studying child migration in China provides an opportu-
nity to address those research gaps. During the past three decades of eco-
nomic reform, China’s massive rural-to-urban migration, which has often
been transient and circular, has affected the well-being of rural children who
have accompanied their migrant parents to cities (Xu & Xie, 2015). By com-
paring rural-to-urban migrants with urban native adolescents in south China,
this study focuses on gender differences in the psychological health of inter-
nal migrant children, and considers how far these differences are informed
by victimization stressors and stress-moderation mechanisms related to
social support relationships.

Gender and Migration-Related Child Health in

China
By 2015, an estimated 277.47 million adult rural residents in China had
moved from the largely impoverished, agriculture-dominated rural regions of
the western and central inland provinces to cities on the eastern seaboard in
search of better work opportunities (China National Bureau of Statistics,
2016). This figure was 27.3% times higher than the comparable 2010 figure
recorded by the national census. These rural migrants included around 35.8
million adolescents below 18 years of age (Wang & Mesman, 2015). Internal
migration within China is plagued by strong institutional barriers to move-
ment under the long-standing household registration (hukou) policy adopted
in the 1950s. Reinforcing an unbalanced rural–urban divide, the hukou sys-
tem assigns an agricultural (rural) or nonagricultural (urban) residency origin
Cheung 101

to individuals at birth, tying their rights (or lack thereof) to the entitlements
inherent in their hukou status. As a result of that system, migrant children
from rural China share the fate of their migrant parents. They are treated as
second-class citizens who are barred from attaining permanent urban citizen-
ship in their host cities (J. Wu, 2010). They are also denied equal access to
subsidized education, health care, and many other social services in urban
China (Sun, Chen, & Chan, 2016; X. Wu & Zhang, 2015), which in turn hin-
ders their urban adaptation (Yuan, Fang, Liu, Hou, & Lin, 2013).
The vulnerable position of China’s rural migrant children in the urban
migration process has prompted extensive studies of their well-being. To
date, two meta-analyses of this body of literature (Sun et al., 2016; Wang &
Mesman, 2015) have been conducted on their health and social functioning
outcomes. Targeting the health outcomes of 25 selected studies, Sun et al.
(2016) revealed that migrant children in public schools present significantly
more mental health (internalizing and externalizing) problems than their
urban-born peers, whereas migrant children in private, mostly unlicensed
migrant schools have broadly similar outcomes. They also showed that
migrant children are exposed to higher physical health risks (eye problems,
anemia, and oral health problems) due to their lack of vaccination and health
insurance. Wang and Mesman’s (2015) synthesis of 41 selected studies found
evidence for publication bias against studies indicating less favorable aca-
demic and social competence for migrant children. Apart from these meta-
analyses, research documenting and accounting for gender differences in
migrant children’s health has been sparse (Choi, 2016).
It seems important to consider the intersection of gender with child migra-
tion in the context of China. Preference for sons, rooted in traditional patrilin-
eal culture, has been enforced under the one-child policy in rural rather than
in urban China for more than three decades; however, because the new two-
child policy has officially been in effect since early 2016, whether son prefer-
ence will be bolstered across urban and rural China remains to be seen. There
is evidence that urban parents are increasingly willing to invest in urban sin-
gleton girls (such as by providing them with education) under the one-child
policy, which eliminates opportunities for parents to discriminate against
daughters (M. H. Lee, 2012). However, the patrilineal culture in rural China
has been reinforced by economic concerns. In view of agricultural labor
demands putting a premium on sons, the one-child policy has been relaxed in
rural areas, allowing rural couples to have a second child if their first child
was female (J. Li, 2004). Heavy financial reliance on sons is further driven by
the weak institutional support for the rural elderly (Ebenstein & Leung,
2010). Economic constraints in poorly developed rural regions also make it
less rational for patrilineal families to allocate resources to girls (Zhang, Kao,
102 Youth & Society 52(1)

& Hannum, 2007). It turns out that son preference impedes female child
health care and intensifies health inequalities in rural China (J. Li, 2004).
Patriarchal belief may be stronger among Chinese rural children, with the
result that an enduring gender bias against rural daughters, along with the
marginalization of rural migrant status, may produce more feelings of inferi-
ority among migrant girls versus migrant boys. This view is echoed by the
broader immigration literature, which speaks to gender-based health inequal-
ities within both internal and transnational immigrant societies of origin char-
acterized by a strong patriarchal system (Malmusi et al., 2010). Through
processes of patriarchal domination and disempowerment, females are more
likely than males to experience relative deprivation and disadvantages in
access to health care opportunities within immigrant groups (Borrell et al.,
2008).

Gender and Migrant Victimization

The victimization of juveniles ranges from minor aggression (e.g., verbal/
relational aggression) to violent aggression (e.g., being robbed by force or
assaulted) (Hamby, Finkelhor, & Turner, 2013). Links between victimization
and the psychiatric impairment (e.g., posttraumatic stress and depressive dis-
orders) of adolescents have been well established (Ford, Elhai, Connor, &
Frueh, 2010). Understanding victimization in migration is particularly worth
considering because immigrant youths are more likely to be victimized by
peers and community violence than their nonimmigrant counterparts (Jaycox
et al., 2002; Sulkowski, Bauman, Wright, Nixon, & Davis, 2014).
Victimization rates also vary across immigrant groups as a function of inter-
generational segmented assimilation (Peguero, 2009), racial–ethnic diversity
(Juvonen, Nishina, & Graham, 2006), and routine lifestyle (Peguero, 2013).
This is a promising line of inquiry, but few migration scholars have exam-
ined migrant victimization by gender, even though gender has been identified
as a major correlate in the victimization field in general (Hong et al., 2014).
Limited evidence (Koo, Peguero, & Shekarkhar, 2012; Le & Wallen, 2009;
Peguero, Popp, & Koo, 2015) suggests that Latino and Asian immigrant boys
confront more school victimization than their female counterparts, findings
that parallel research on general populations showing that young males gen-
erally report increased victimization (Hamby et al., 2013). From the gender
socialization (Hamby et al., 2013) and lifestyle perspectives (Koo et al.,
2012), boys are more reinforced for aggression than girls, due to gender ste-
reotypes, and are prone to unsupervised extracurricular or delinquent activi-
ties, thereby putting them at an elevated risk of victimization. Notwithstanding
the available evidence of differential exposure to migrant victimization by
Cheung 103

gender, the effects of those gender disparities on the well-being of adolescent

migrant populations remain obscure.
As far as China’s internal migration milieu is concerned, little research has
been done on the prevalence of victimization among rural migrants. Findings
on interpersonal discrimination against rural migrants in urban China, arising
from institutional discrimination within the hukou system (Chen, 2013), may
imply a greater risk of victimization. Rural migrants are often blamed for
poor public safety in cities, and this perception has spurred urban residents’
fear of migrant crime (Cheng & Smyth, 2015) and hostility toward migrants
(Nielsen & Smyth, 2008; Yuan, Fang et al., 2013). Nevertheless, existing
studies have not considered whether rural migrants respond differentially to
discrimination based on gender.
The present study connects the social stress, conservation of resources,
and gender identity perspectives to elucidate the effect of the migration–
gender–victimization intersection on health. The social stress paradigm in
the fields of deviance (Ganem, 2010) and health (Thoits, 2011) unanimously
argues that strain-producing situations including victimization trigger nega-
tive affects ranging from anger to anxiety and depression. As further speci-
fied by the conservation of resources theory (Hobfoll, 2001), certain
populations such as teenagers and migrants are more susceptible to the
adverse mental health consequences of resource loss emanating from vic-
timization because these people are at a disadvantage in terms of their
resource reserves (Hill, Kaplan, French, & Johnson, 2010). Migrant suscep-
tibility to victimization may also be a function of gender identity. Abada,
Hou, and Ram (2008) argued that when adolescents develop gender role
identifications, boys become more prone to developing instrumental charac-
teristics that can protect against depressive symptoms. At the same time,
girls become less confident in their problem-solving capabilities; yet, the
opposite is true for boys. The gender role identifications and concomitant
differences in coping strategies may explain why young females are less
able to withstand the problem of victimization (Abada et al., 2008). However,
the Chinese context may shape the gender identity development in ways not
considered by scholars in a Western context. Gender bias in the patriarchal
culture of developing societies such as rural China, as discussed earlier, may
intensify female identification with problem-solving incompetence. We
therefore argue that young female migrants from rural China suffer the dou-
ble disadvantage of being marginalized migrants and disempowered females
and could be even more sensitive to the health-undermining effect of victim-
ization than male migrants.
We should be aware that victimization does not necessarily confer more
health deficits on migrant girls relative to their male peers if the level of stress
104 Youth & Society 52(1)

from victimization is moderated. The social stress and coping perspective

(Thoits, 2011) maintains that social support ties offer a salient psychosocial
resource that can lessen the deleterious psychological effects of stressful con-
ditions. In terms of victimization, the psychosocial functions of social ties are
that they enable teenaged victims to reappraise the stress of victimization in
light of the knowledge that help and support are available; they are encour-
aged to seek constructive, problem-focused coping rather than avoidance-
focused alternatives to prevent repeated victimization (Hammack, Richards,
Luo, Edlynn, & Roy, 2004; Hill et al., 2010). There may be gender disparities
in the stress-moderation effect of social support. Given their stronger inter-
personal caring orientation (Gore, Aseltine, & Colten, 1993), females tend to
report larger social networks than males (Fuhrer & Stansfeld, 2002) and
mobilize others for support in stressful circumstances, including victimiza-
tion, more than males do (Robbers, 2004; Taylor et al., 2000). These observa-
tions in the West replicate in the context of China, as suggested by Bao, Haas,
and Pi (2007), who evinced the impact of gendered utilization of social sup-
port in reaction to strain on delinquency among urban and rural teenagers. In
the migration setting within China, it is conceivable that social ties are of
greater importance to migrant girls versus migrant boys, and that social ties
may matter more for migrant girls in attenuating the negative effect of stress,
such as victimization, on mental well-being.

Current Study
We propose three hypotheses based on the foregoing discussion:

Hypothesis 1: Migrant boys may experience more victimization than

migrant girls.
Hypothesis 2: Victimization may be more strongly related to lower psy-
chological health for migrant girls than for migrant boys.
Hypothesis 3: The link between victimization and psychological health
may be moderated by social ties, and as such, the buffering effect may dif-
fer by gender for migrant adolescents.

In analyzing the buffering effect, we adopt a richer perspective on social ties

by distinguishing the local ties that link migrants to their receiving cities from
the translocal ties that link them to their origin villages. The salience of dif-
ferentiating translocal and local ties lies in the burgeoning evidence of
translocal networks on the health benefits (Donato & Duncan, 2011), favor-
able status evaluation (Jin, Wen, Fan, & Wang, 2012), and identity formation
Cheung 105

(Halilovich, 2012) for migrants both in transnational and internal migration

settings. We hypothesize a positive interaction between victimization and
local/translocal ties, whereby the unfavorable psychological health effect of
victimization is reduced by the presence of social ties among young migrants
and especially migrant girls.
Although we focus on rural-to-urban migrant adolescents, we also include
a sample of urban native adolescents for comparison. This enables us to
determine the possible unique features of victimization experiences and men-
tal health in both genders of migrant adolescents.

Method
Sample
The data were used from “Stuck in the City: Migration and Delinquency
Among Migrant Adolescents in Guangzhou,” a study conducted between
June 2010 and November 2011. Guangzhou, the largest city in Southern
China, has been the hub of rural–urban migration influx in coastal China (L.
Li & Li, 2010). We used a stratified cluster sampling of the middle schools in
Guangzhou based on the 10 administrative urban districts and the school
funding types (public school or private school for migrant children) to recruit
rural migrant and urban native participants. Overall, 22 public schools (with
between one and three public schools in each of the 10 districts) and 10 infor-
mal schools for migrant children (from nine of the 10 districts, as one district
has no schools for migrants) joined the study. Migrant children who do not
have household registration in their receiving cities are forced to enter infor-
mal schools due to discriminatory barriers to entering public schools (X. Wu
& Zhang, 2015).
Each participating school randomly assigned one eighth-grade class for
data collection. Research assistants administered anonymous, self-report
questionnaires to the participants in classrooms. Our analytical sample com-
prised 482 rural-to-urban migrant students (males = 276, females = 206) and
838 Guangzhou (urban) native students (males = 435, females = 403). The
Survey and Behavioral Research Ethics Committee of the author’s university
approved the study procedures. Informed consent was obtained from the stu-
dents and their parents through the coordination of the participating schools.

Variables
Dependent variable of psychological health. We adapted the Kessler Psycholgo-
ical Disress Scale (K6) (Kessler et al., 2002), which diagnoses nonspecific
106 Youth & Society 52(1)

mental illness in the general population, to measure psychological health. The

respondents were asked how often they felt nervous, hopeless, anxious,
depressed, or worthless within the previous month (often = 1, never = 4).
Higher scores represented better psychological health (five-item scale, α = .87).

Victimization. This predictor gauged how often the respondents were victimized
during the previous 12 months (never = 0, often = 3). The scale covered minor
and severe victimization: having something stolen by someone, being robbed
by someone, being deceived for money or things by someone, being threatened
by someone with force to get money or things, being physically assaulted by
someone, being teased or insulted by schoolmates, being hit by schoolmates,
and being teased or insulted by teachers (eight-item scale, α = .78).

Local ties with host community. Local ties were indexed by two kinship mea-
sures (parent–child ties and local ties with relatives) and three nonkinship
measures (teacher–child ties, ties with local friends, and neighborhood cohe-
sion) in the receiving city (Guangzhou). Parent–child ties were gauged by
eight questions: “How often do your parents show interest in your school
life?” (never = 1, often = 4); “How often do your parents engage in leisure
activities with you?” (never = 1, often = 4); “How well do your parents take
care of your daily needs?” (very poorly = 1, very well = 4); “How often do
you talk to your parents when you are unhappy?” (never = 1, often = 4);
“How often do your parents give you support and encouragement?” (never =
1, often = 4); “How much do your parents understand you?” (not at all = 1,
very much = 4); “How satisfied do you feel with your communication with
your parents?” (very dissatisfied = 1, very satisfied = 4); and “How is your
relationship with your parents?” (very poor = 1, very good = 4; eight-item
scale, α = .87). Local ties with relatives were measured by one item, which
inquired about the number of relatives in Guangzhou with whom the respon-
dents’ families had regular contact (none = 1, many = 4).
For the nonkinship measures, the respondents’ relationships with their
teachers in Guangzhou were captured by five questions: “How many students
are the teachers close to?” (none = 1, most = 4); “How often do you talk to
your teachers?” (never = 1, often = 4); “How much do you think the teachers
care about the students?” (not at all = 1, very much = 4); “How fairly do the
teachers treat the students?” (unfairly = 1, fairly = 4); and “How does your
class master teacher rate?” (very poor = 1, very good = 4; five-item scale, α =
.84). To estimate ties with local friends, the respondents were asked the fol-
lowing question: “How often do you seek help from friends in Guangzhou?”
(never = 1, often = 4). Neighborhood cohesion was constructed based on four
statements, rated on a five-point scale (strongly disagree = 1, strongly agree
Cheung 107

= 5): “Most people in this neighborhood know each other”; “People in this
neighborhood are willing to help each other”; “People in this neighborhood
get along well with each other”; and “People in this neighborhood are trust-
worthy” (four-item scale, α = .89).

Translocal ties with home community. This migrant-specific measure was

indexed by two variables relating to ties with relatives and friends in the
sending villages. The migrants were asked to report how often they had been
in contact with relatives and friends in their home community while living in
Guangzhou (never = 1, often = 4).

Covariates. We controlled for gender, age, school type (public vs. private for
migrant children), and standard of living, which we measured by asking the
respondents to state how many of 14 amenities (e.g., self-contained kitchen,
air-conditioner, computer, motorcycle) were available in their living quarters.
We also considered three migrant-specific covariates: years of residence in
Guangzhou, fluency in the Guangzhou dialect (cannot understand or speak it
at all = 1, understand and speak it well = 4), and perceived receptivity of
Guangzhou natives to nonnatives (not accepted at all = 1, well accepted = 5).

Statistical Analysis
All analyses were performed in Stata 14.2. Missing observations (constituting
0.3%-5.5% of cases) were imputed by using the multiple imputation method
(Amelia program; King, Honaker, Joseph, & Scheve, 2001). Ordinary least
squares regressions were used to model the relative contributions of victimiza-
tion and social ties to the continuously measured psychological health out-
come for different genders in the migrant and urban-born samples. To assess
whether social ties condition the victimization–psychological health relation-
ship in different genders, the interactions between the victimization and social
ties measures were individually added to the models. Examination of the vari-
ance inflation factors for the interaction and its component parts revealed that
none exceeded 1.52 and, therefore, the multicollinearity problem was modest.
A z statistic (Paternoster, Mazerolle, & Piquero, 1998) was computed to fur-
ther determine whether the regression estimates of victimization, social ties,
and interaction effects differ significantly across genders in the migrant and
urbanite samples. To accommodate our data nested within schools and mini-
mize the misestimation of standard errors, the regressions were based on
robust standard errors clustered by school (Williams, 2000). We also included
school fixed-effects estimators in all the regressions to adjust for the school’s
unobserved effects on psychological well-being.
108 Youth & Society 52(1)

Results
Descriptive and Bivariate Statistics
Table 1 depicts the sample statistics of the variables stratified by gender and
migrant status. As anticipated, migrant girls fared significantly more poorly
in terms of psychological health than migrant boys, and this gender pattern
also existed in the urbanites. Both migrant and urban native boys encountered
significantly more victimization than their female counterparts.
Looking at local ties with the host community forged by migrants, there
were significantly more local ties with relatives, teachers, and friends for
migrant girls compared with their male counterparts. However, migrant
boys perceived significantly stronger neighborhood cohesion. No signifi-
cant gender difference in parent–child ties was found in the migrant sample.
In terms of translocal ties with relatives and friends in the rural home among
the migrants, the gender difference was not statistically discernible. Among
the urban natives, girls reported having significantly more ties with local
friends, whereas boys reported having stronger neighborhood cohesion.
Other forms of ties did not differ significantly between genders among the
urbanites.

Main Effects of Victimization and Social Ties

Table 2 describes how victimization, local ties, and translocal ties relate to
the mental health of different genders in the fixed school-effects regression
models. The association between victimization and suboptimal psychological
health was stronger in migrant girls (b = –1.341, p < .01) than in migrant boys
(b = –0.920, p < .01). A similar pattern was observed in the urbanites, with
victimization being more predictive of the poor psychological health of
urban-born girls (b = –1.772, p < .001) than that of urban-born boys (b =
–1.233, p < .001). The z statistics confirmed that the victimization effect var-
ied significantly across genders in both samples.
Closer teacher–child ties in the host city (b = 0.221, p < .05) and closer
translocal ties with friends in the rural home (b = 0.714, p < .05) were signifi-
cantly related to better psychological health for migrant girls but not for migrant
boys, and those gender differences were significant in terms of z statistics.
Although the psychological health of migrant boys and girls was significantly
predicted by parent–child ties in the host city (b = 0.075, p < .10, and b = 0.076,
p < .05, respectively), the effect of such ties was insignificant across genders.
Local ties with relatives, friends, and neighbors in the host city, and translocal
ties with relatives in the rural home, were not predictive of within-gender and
between-gender variations in psychological well-being among the migrants.
 Table 1. Descriptive and Bivariate Statistics for Gender-Specific Rural-to-Urban Migrant Adolescents (n = 482) and Urban Native
Adolescents (n = 838).
Migrant sample Urban native sample

Boys Girls Boys Girls

(n = 276) (n = 206) (n = 435) (n = 403)

Range M/% (SD) M/% (SD) Test statistics M/% (SD) M/% (SD) Test statistics

Psychological health 5-20 13.96 (3.92) 13.04 (3.87) t = 3.53** 13.59 (4.12) 12.87 (3.84) t = 2.06*
Victimization 0-24 3.13 (3.57) 1.92 (2.62) t = 4.55*** 2.88 (3.55) 1.78 (2.05) t = 4.05***
Local ties with host community
Parent–child ties 8-32 21.20 (4.88) 20.99 (5.27) t = 0.55 22.59 (4.78) 23.02 (4.77) t = 1.18
Local ties with relatives 1-4 2.92 (0.88) 3.05 (0.80) t = 2.44* 2.44 (0.92) 2.38 (0.86) t = 1.08
Teacher–child ties 5-20 13.56 (3.31) 14.16 (3.29) t = 2.93** 13.87 (3.28) 14.09 (2.88) t = 0.93
Ties with local friends 1-4 2.59 (0.88) 2.77 (0.80) t = 1.99† 2.93 (0.76) 3.08 (0.66) t = 3.47**
Neighborhood cohesion 4-20 15.41 (3.89) 14.81 (4.36) t = 1.74† 16.12 (3.83) 15.71 (3.74) t = 2.05†
Translocal ties with home community
Contact with home relatives 1-4 3.06 (0.85) 3.01 (0.93) t = 0.63 — — —
Contact with home friends 1-4 2.44 (0.92) 2.33 (0.95) t = 1.52 — — —
Sociodemographics
Age (years) 12-18 14.77 (0.80) 14.63 (0.80) t = 2.18* 14.43 (0.60) 14.40 (0.62) t = 0.76
School type
  Public school 24.3% 28.6% χ2 = 1.09 97.9% 98.5% χ2 = 1.01
   Private school for migrant children 75.7% 71.4% 2.1% 1.5%
Household amenities (number) 0-14 8.43 (2.98) 9.12 (2.70) t = 3.49** 9.72 (2.44) 9.84 (2.26) t = 0.92
Migration variables
Years of residence in Guangzhou 1-18 8.60 (4.24) 8.76 (4.67) t = 0.32 — — —
Fluency in Guangzhou dialect 1-4 3.24 (0.85) 3.25 (0.79) t = 0.10 — — —
Perceived receptivity of Guangzhou natives to nonnatives 1-5 3.80 (0.94) 3.76 (0.94) t = 0.27 — — —

Note. To compare gender-specific responses, t and chi-square tests are used for the continuous variables and the categorical variable of school type, respectively. Both
tests include the survey design effect to account for school clustering.

109
†p < .10. *p < .05. **p < .01. ***p < .001.
 Table 2. School Fixed-Effects Regression Estimates (Unstandardized) for Main Effects of Victimization and Social Ties on

110
Psychological Health of Different Genders in Rural-to-Urban Migrant and Urban Native Adolescents.
Migrant sample Urban native sample

z test of
z test of difference difference of
Boys (n = 276) Girls (n = 206) of coefficients Boys (n = 435) Girls (n = 403) coefficients

across across
Independent variables b (SE) b (SE) genders b (SE) b (SE) genders

Victimization −0.920 (0.192)** −1.341 (0.210)** 1.480† −1.233 (0.187)*** −1.772 (0.296)*** 1.539†
Local ties with host community
Parent–child ties 0.075 (0.047)† 0.076 (0.033)* 0.017 0.101 (0.048)* 0.144 (0.040)** 0.688
Local ties with relatives −0.233 (0.241) −0.144 (0.391) 0.194 −0.030 (0.173) −0.215 (0.125) 0.867
Teacher–child ties 0.080 (0.068) 0.221 (0.082)* 1.324† 0.049 (0.055) 0.177 (0.066)* 1.490†
Ties with local friends −0.107 (0.234) −0.448 (0.323) 0.855 0.343 (0.191)† 0.365 (0.253) 0.069
Neighborhood cohesion −0.114 (0.311) 0.194 (0.297) 0.186 −0.042 (0.211) −0.097 (0.225) 0.178
Translocal ties with home community
Contact with home relatives 0.064 (0.354) 0.175 (0.367) 0.218 — — —
Contact with home friends 0.292 (0.193) 0.714 (0.133)* 1.800 * — — —
Sociodemographics
Age −0.228 (0.277) 0.151 (0.284) — 0.051 (0.212) −0.011 (0.195) —
School type (1 = public; 2 = private) −5.760(1.397)*** −5.118 (0.922)*** — −0.990 (0.541)† −0.028 (0.573) —
Household amenities −0.289 (0.292) −0.164 (0.383) — 0.027 (0.239) 0.674 (0.177)** —
Migration variables
Years of residence in Guangzhou −0.018 (0.349) −0.196 (0.261) — — — —
Fluency in Guangzhou dialect −0.092 (0.316) 0.170 (0.392) — — — —
Perceived receptivity of Guangzhou natives 0.525 (0.344) 0.173 (0.213) — — — —
to nonnatives
Model R2 .237 .357 .279 .304
F statistics 3.546*** 4.978*** 12.082*** 16.208***

Note. Robust standard errors clustered by school are in parentheses.

†p< .10. *p < .05. **p < .01. ***p < .001.
Cheung 111

Among the urbanites, ties with teachers significantly predicted the girls’
psychological health (b = 0.177, p < .01) but did not predict that of their male
peers, with the z statistics indicating a significant gender disparity. Although
parent–child ties were significantly associated with psychological health in
urban-born boys and girls (b = 0.101, p < .05, and b = 0.144, p < .01, respec-
tively), the effect across genders was not discernible as informed by the z
scores. Ties with local friends predicted the urban-born boys’ psychological
health only (b = 0.343, p < .10), but the predictability of such ties across gen-
ders was nonsignificant. Ties with local relatives and neighborhood cohesion
predicted neither the within-gender nor the between-gender variations in psy-
chological wellness of the urbanites.

Moderation of Victimization
Table 3 summarizes the findings of our interaction analysis, which estimated
whether the relationship between victimization and psychological health was
moderated by the social ties of different genders while adjusting for the main
effects of the predictors. Within the migrant sample, none of the interaction
terms was a significant predictor of the boys’ psychological health. However, we
detected three significant interaction effects on the migrant girls’ psychological
health in the expected positive direction: between victimization and teacher–
child ties in the host community (b = 0.092, p < .05), between victimization and
contact with relatives in the home community (b = 0.663, p < .05), and between
victimization and contact with friends in the home community (b = 0.748, p <
.01). The z statistics further indicated that the interaction effects of victimization
with those three forms of social ties differed significantly across genders, thereby
illustrating the greater influence of social ties moderation for migrant girls.
For the urban native sample, victimization interacted only with parent–
child ties for the boys in the hypothesized direction (b = 0.036, p < .10).
Victimization also had positive interactions with teacher–child ties and
neighborhood cohesion for urbanite girls (b = 0.129, p < .10, and b = .436, p
< .10, respectively). Similar to the migrant sample, the z statistics showed
that the interaction effect mattered significantly more for urbanite girls.

Discussion
Central to our study is the question of how victimization relates, along gender
lines, to the psychological health outcome of Chinese rural-to-urban migrant
vis-à-vis urban-born adolescents. In brief, our findings complement the wider
migration and victimization literature by underlining the importance of the
gender dimension.
112
Table 3. Summary of School Fixed-Effects Regression Estimates (Unstandardized) for Victimization/Social Ties Interactions on
Psychological Health of Different Genders in Rural-to-Urban Migrant and Urban Native Adolescents.
Migrant boys (n = 276) Migrant girls (n = 206) z test of difference
of coefficients
Independent variables b (SE) Model R2 F statistics b (SE) Model R2 F statistics across genders

Victimization/Local Ties interaction

× Parent–Child Ties 0.013 (0.038) .237 3.326*** 0.038 (0.035) .361 4.694*** 0.484
× Local Ties With Relatives −0.171 (0.265) .239 3.442*** −0.019 (0.415) .357 4.630*** 0.309
× Teacher–Child Ties 0.023 (0.042) .237 3.333*** 0.092 (0.030)* .372 4.952*** 1.337†
× Ties With Local Friends 0.326 (0.203) .250 3.764*** −0.371 (0.300) .363 4.710*** 0.124
× Neighborhood Cohesion −0.211 (0.245) .239 3.338*** 0.379 (0.304) .362 4.716*** 0.430
Victimization/Translocal Ties interaction
× Contact With Home Relatives 0.226 (0.199) .241 3.516*** 0.663 (0.255)* .379 4.275*** 1.351†
× Contact With Home Friends 0.091 (0.293) .237 3.291*** 0.748 (0.199)** .385 4.115*** 1.855 *

Urban native boys (n = 435) Urban native girls (n = 403) z test of difference
of coefficients
b (SE) Model R2 F statistics b (SE) Model R2 F statistics across genders

Victimization/Local Ties interaction

× Parent–Child Ties 0.036 (0.019)† .283 11.054*** 0.044 (0.039) .306 14.704*** 0.184
× Local Ties With Relatives 0.185 (0.130) .282 11.109*** 0.157 (0.327) .304 14.611*** 0.080
× Teacher–Child Ties 0.007 (0.041) .279 10.848*** 0.129 (0.077)† .310 15.012*** 1.399†
× Ties With Local Friends 0.116 (0.171) .280 11.021*** −0.421 (0.372) .307 14.711*** 0.745
× Neighborhood Cohesion −0.063 (0.150) .279 10.875*** 0.436 (0.241)† .310 15.026*** 1.314†

Note. Interaction terms are entered individually to the models while controlling for the main effects of all the victimization, local ties, translocal ties, and
sociodemographic variables. Models for migrant boys and girls additionally adjust for the migration variables. Robust standard errors clustered by school are in
parentheses.
†p < .10. *p < .05. **p < .01. ***p < .001.
Cheung 113

This study demonstrates that vulnerability to victimization stress within

migrant populations is gendered. The consistent effect of victimization
uncovered by our analysis supports our conjecture that victimization is more
closely related to poor psychological health for migrant girls than for migrant
boys. Although we found, according to our hypothesis, that migrant boys
were exposed to more victimization than migrant girls, migrant girls appeared
to be more vulnerable to the influence of victimization, which was more det-
rimental to their mental well-being than to that of migrant boys. This could
have been one of the main factors accounting for the gender differences in
health: that, migrant girls suffered worse effects to their psychological well-
being than migrant boys. Such gendered vulnerability similarly existed in
urbanites, yet migrant girls were lower in vulnerability than urbanite girls
possibly due to counteractive migrant resilience (Cheung, 2014).
This study also substantiates the role of gender in the mitigation of migrant
victimization stress. Females tended to benefit more than males from the
direct effect of certain social ties on their mental health status, and this fea-
ture was more pronounced for migrants than for urban natives. An even more
salient result is that the buffering effect of social ties on the victimization–
health link was far stronger in migrant girls versus migrant and urban-born
boys, which conforms to our posited hypothesis. Specifically, victimized
migrant girls were more likely than their male counterparts to turn to social
ties with their teachers in the host city and translocal ties with relatives and
friends in the rural home, which in turn reduced the adverse mental health
sequelae of victimization for migrant girls. This buffering effect was also
stronger than that in urban-born girls. Altogether, our results point to the par-
ticular importance of supportive social ties for female migrants under stress,
which might embody the patriarchal influence of China discussed at the out-
set of the article. More broadly, they add to the knowledge of the gendered
stress-buffering mechanisms of the social stress model.
Our results can inform policy making on the provision of health care to
rural migrant children in China. Studies of gender inequality in migrant
health have cast considerable light on the interventions of adults’ sexual risk
behaviors and maternal health (Yang & Xia, 2006; Yuan, Qian, & Thomsen,
2013). Nevertheless, the gender issue has been overlooked in the health care
of the expanding population of migrant children (Choi, 2016). Based on our
findings, effective measures to promote mental health in teenaged migrants
and, thus, gender equity in child health should recognize and manage particu-
larly the migrant girls’ vulnerability to victimization stress and their avail-
ability of social support for stress coping.
These findings should be considered in light of their methodological limita-
tions. First, given the cross-sectional nature of the study, we were able to capture
114 Youth & Society 52(1)

the contemporaneous effects of victimization and their moderation by social

ties, but the long-term causal effects on the gendered variations in the health of
young migrants remain unknown. Second, we did not distinguish the forms of
victimization (e.g., relational/verbal vs. overt) between genders to evaluate their
effects on health status because the gender samples were small, and the highly
and negatively skewed victimization data might not have sufficient variation to
detect any statistical significance. Improved data that are longitudinal and
involve larger gender samples would enable future research to draw more defin-
itive conclusions about the gender–victimization–health nexus. Third, we lacked
the relevant measures for gender beliefs, and so, the influence of patriarchal
culture in the context of China is not clear yet. However, cultural changes in
rural origins with the increasingly migration-driven diffusion of gender-egalitar-
ian values from urban destinations are possible (Lu & Tao, 2015) and may abate
the marginalization and stress vulnerability facing the female migrants. Further
investigation should look into beliefs in patriarchal versus egalitarian values,
which may elaborate the mechanisms underlying the pathway from gender and
victimization to health outcomes within migrant populations.
Despite these constraints, this study, which focuses on internal migrating
adolescent populations in China, has implications beyond the study setting.
Investigating gendered vulnerability to sources of social stress such as vic-
timization and gendered perceptions of coping resources would reveal useful
approaches to clarifying the relationship between gender and child health in
future migration studies.

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.

Funding
The author(s) disclosed receipt of the following financial support for the research,
authorship, and/or publication of this article: Data of this article were funded by the
South China Programme, Hong Kong Institute of Asia-Pacific Studies, The Chinese
University of Hong Kong.

References
Abada, T., Hou, F., & Ram, B. (2008). The effects of harassment and victimization on
self-rated health and mental health among Canadian adolescents. Social Science
& Medicine, 67, 557-567.
Bao, W. N., Haas, A., & Pi, Y. (2007). Life strain, coping, and delinquency in the
People’s Republic of China: An empirical test of general strain theory from
Cheung 115

a matching perspective in social support. International Journal of Offender

Therapy and Comparative Criminology, 51, 9-24.
Borrell, C., Muntaner, C., Sola, J., Artazcoz, L., Puigpino, R., Benach, J., & Noh,
S. (2008). Immigration and self-reported health status by social class and gen-
der: The importance of material deprivation, work organisation and household
labour. Journal of Epidemiology and Community Health, 62, e7. doi:10.1136/
jech.2006.055269
Chen, J. (2013). Perceived discrimination and subjective well-being among rural-to-
urban migrants in China. Journal of Sociology & Social Welfare, 40, 131-156.
Cheng, Z., & Smyth, R. (2015). Crime victimization, neighborhood safety and happi-
ness in China. Economic Modelling, 51, 424-435.
Cheung, N. W. T. (2014). Social stress, locality of social ties and mental well-being:
The case of rural migrant adolescents in urban China. Health & Place, 27,
142-154.
China National Bureau of Statistics. (2016). Statistical communiqué of the People’s
Republic of China on the 2015 national economic and social develop-
ment. Retrieved from http://www.stats.gov.cn/english/PressRelease/201602/
t20160229_1324019.html
Choi, S. Y. P. (2016). Children and migration in China. Asian Population Studies,
12, 131-134.
Das-Munshi, J., Leavey, G., Stansfeld, S. A., & Prince, M. J. (2012). Migration, social
mobility and common mental disorders: Critical review of the literature and
meta-analysis. Ethnicity & Health, 17, 17-53.
Donato, K. M., & Duncan, E. M. (2011). Migration, social networks, and child health
in Mexican families. Journal of Marriage and Family, 73, 713-728.
Ebenstein, A., & Leung, S. (2010). Son preference and access to social insurance:
Evidence from China’s rural pension program. Population and Development
Review, 36, 47-70.
Ford, J. D., Elhai, J. D., Connor, D. F., & Frueh, B. C. (2010). Poly-victimization
and risk of posttraumatic, depressive, and substance use disorders and involve-
ment in delinquency in a national sample of adolescents. Journal of Adolescent
Health, 46, 545-552.
Fuhrer, R., & Stansfeld, S. A. (2002). How gender affects patterns of social relations
and their impact on health: A comparison of one or multiple sources of support
from “close friends.” Social Science & Medicine, 54, 811-825.
Ganem, N. M. (2010). The role of negative emotion in general strain theory. Journal
of Contemporary Criminal Justice, 26, 17-185.
Gore, S., Aseltine, R. H., & Colten, M. E. (1993). Gender, social-relationship involve-
ment, and depression. Journal of Research on Adolescence, 3, 101-125.
Halilovich, H. (2012). Trans-local communities in the age of transnationalism:
Bosnians in diaspora. International Migration, 50, 162-178.
Hamby, S., Finkelhor, D., & Turner, H. (2013). Perpetrator and victim gender patterns
for 21 forms of youth victimization in the national survey of children’s exposure
to violence. Violence and Victims, 28, 915-939.
116 Youth & Society 52(1)

Hammack, P. L., Richards, M. H., Luo, Z., Edlynn, E. S., & Roy, K. (2004). Social
support factors as moderators of community violence exposure among inner-city
African American young adolescents. Journal of Clinical Child & Adolescent
Psychology, 33, 450-462.
Hill, T. D., Kaplan, L. M., French, M. T., & Johnson, R. J. (2010). Victimization in
early life and mental health in adulthood: An examination of the mediating and
moderating influences of psychosocial resources. Journal of Health and Social
Behavior, 51, 48-63.
Hobfoll, S. (2001). The influence of culture, community, and the nested-self in the
stress process: Advancing conservation of resources theory. Applied Psychology:
An International Review, 50, 337-421.
Hong, J. S., Peguero, A. A., Choi, S., Lanesskog, D., Espelage, D. L., & Lee, N. Y.
(2014). Social ecology of bullying and peer victimization of Latino and Asian
youth in the United States: A review of the literature. Journal of School Violence,
13, 315-338.
Jaycox, L. H., Stein, B. D., Kataoka, S. H., Wong, M., Fink, A., Escudero, P., &
Zaragoza, C. (2002). Violence exposure, posttraumatic stress disorder, and
depressive symptoms among recent immigrant schoolchildren. Journal of
American Academy of Child & Adolescent Psychiatry, 41, 1104-1110.
Jin, L., Wen, M., Fan, J. X., & Wang, G. (2012). Trans-local ties, local ties and
psychological well-being among rural-to-urban migrants in Shanghai. Social
Science & Medicine, 75, 288-296.
Juvonen, J., Nishina, A., & Graham, S. (2006). Ethnic diversity and perceptions of
safety in urban middle schools. Psychological Science, 17, 393-400.
Kessler, R. C., Andrews, G., Colpe, L. J., Hiripi, E., Mroczek, D. K., Normand, S. L.
T., Walters, E. E., & Zaslavsky, A. M. (2002). Short screening scales to moni-
tor population prevalences and trends in non-specific psychological distress.
Psychological Medicine, 32, 959-976.
King, G., Honaker, J., Joseph, A., & Scheve, K. (2001). Analyzing incomplete politi-
cal science data: An alternative algorithm for multiple imputation. American
Political Science Review, 95, 49-69.
Koo, D. J., Peguero, A. A., & Shekarkhar, Z. (2012). Gender, immigration, and school
victimization. Victims & Offenders, 7, 77-96.
Le, T. N., & Wallen, J. (2009). Risk of non-familial violent physical and emotional
victimization in four Asian ethnic groups. Journal of Immigrant and Minority
Health, 11, 174-187.
Lee, M. H. (2012). The one-child policy and gender equality in education in China:
Evidence from household data. Journal of Family and Economic Issues, 33,
41-52.
Lee, S., O’Neill, A. H., Ihara, E. S., & Chae, D. H. (2013). Change in self-reported
health status among immigrants in the United States: Associations with measures
of acculturation. PLoS ONE, 8(10), e76494. doi:10.1371/journal.pone.0076494
Li, J. (2004). Gender inequality, family planning, and maternal and child care in a
rural Chinese county. Social Science & Medicine, 59, 695-708.
Cheung 117

Li, L., & Li, S. M. (2010). The impact of variations in urban registration within cit-
ies. In M. K. Whyte (Ed.), One country, two societies: Rural-urban inequality in
contemporary China (pp. 188-215). Cambridge, MA: Harvard University Press.
Lu, Y., & Tao, R. (2015). Female migration, cultural context, and son preference in
rural China. Population Research and Policy Review, 34, 665-686.
Malmusi, D., Borrell, C., & Benach, J. (2010). Migration-related health inequalities:
Showing the complex interactions between gender, social class and place of ori-
gin. Social Science & Medicine, 71, 1610-1619.
Nielsen, I., & Smyth, R. (2008). Who wants safer cities? Perceptions of public safety
and attitudes to migrants among China’s urban population. International Review
of Law and Economics, 28, 46-55.
Paternoster, R., Mazerolle, P., & Piquero, A. (1998). Using the correct statistical test
for the equality of regression coefficients. Criminology, 36, 859-866.
Paul, A. M. (2015). Negotiating migration, performing gender. Social Forces, 94,
271-293.
Peguero, A. A. (2009). Victimizing the children of immigrants: Latino and Asian
American student victimization. Youth & Society, 41, 186-208.
Peguero, A. A. (2013). An adolescent victimization immigrant paradox? School-
based routines, lifestyles, and victimization across immigration generations.
Journal of Youth and Adolescence, 42, 1759-1733.
Peguero, A. A., Popp, A. M., & Koo, D. J. (2015). Race, ethnicity, and school-based
adolescent victimization. Crime & Delinquency, 20, 1-27.
Pessar, P. R., & Mahler, S. J. (2003). Transnational migration: Bringing gender in.
International Migration Review, 37, 812-846.
Robbers, M. L. P. (2004). Revisiting the moderating effect of social support on strain:
A gendered test. Sociological Inquiry, 74, 546-569.
Stevens, G. W. J. M., & Vollebergh, W. A. M. (2008). Mental health in migrant chil-
dren. Journal of Child Psychology and Psychiatry, 49, 276-294.
Sulkowski, M. L., Bauman, S., Wright, S., Nixon, C., & Davis, S. (2014). Peer vic-
timization in youth from immigrant and non-immigrant US families. Social
Psychology International, 35, 649-669.
Sun, X., Chen, M., & Chan, K. L. (2016). A meta-analysis of the impacts of internal
migration on child health outcomes in China. BMC Public Health, 16, 66-76.
Taylor, S. E., Cousino Klein, L., Lewis, B. P., Gruenewald, T. L., Gurung, R. A. R., &
Updegraff, J. A. (2000). Biobehavioral responses to stress in females: Tend-and-
befriend, not fight-or-flight. Psychological Review, 107, 411-429.
Thoits, P. A. (2011). Mechanisms linking social ties and support to physical and men-
tal health. Journal of Health and Social Behavior, 52, 145-161.
Wang, L., & Mesman, J. (2015). Child development in the face of rural-to-urban
migration in China: A meta-analytic review. Perspectives on Psychological
Science, 19, 813-831.
Williams, R. L. (2000). A note on robust variance estimation for cluster-correlated
data. Biometrics, 56, 645-646.
118 Youth & Society 52(1)

Wu, J. (2010). Rural migrant workers and China’s differential citizenship. In M. K.

Whyte (Ed.), One country, two societies: Rural-urban inequality in contempo-
rary China (pp. 55-81). Cambridge, MA: Harvard University Press.
Wu, X., & Zhang, Z. (2015). Population migration and children’s school enrollments
in China, 1990-2005. Social Science Research, 53, 177-190.
Xu, H., & Xie, Y. (2015). The causal effects of rural-to-urban migration on children’s
wellbeing in China. European Sociological Review, 31, 502-519.
Yang, X., & Xia, G. (2006). Gender, migration, risky sex, and HIV infection in China.
Studies in Family Planning, 37, 241-250.
Yuan, B., Qian, X., & Thomsen, S. (2013). Disadvantaged populations in mater-
nal health in China who and why? Global Health Action, 6(1), Article 19542.
doi:10.3402/gha.v6i0.19542
Yuan, X., Fang, X., Liu, Y., Hou, S., & Lin, X. (2013). Development of urban adap-
tation and social identity of migrant children in China: A longitudinal study.
International Journal of Intercultural Relations, 37, 354-365.
Zhang, Y., Kao, G., & Hannum, D. (2007). Do mothers in rural China practice gender
equality in educational aspirations for their children? Comparative Education
Review, 51, 131-157.

Author Biography
Nicole W.T. Cheung, PhD, is an associate professor of the Department of Sociology,
The Chinese University of Hong Kong. Her research interests include sociology of
deviance, juvenile delinquency, drug abuse, gambling, and migration and adolescent
health. Her work appears in Addictive Behaviors, Social Science & Medicine, Health
& Place, Journal of Gambling Studies, and Journal of Youth & Adolecence.