Child Development, September/October 2018, Volume 89, Number 5, Pages 1768–1785
A Longitudinal Analysis of Well-Being of Ghanaian Children in
Transnational Families
Victor Cebotari Valentina Mazzucato
Maastricht University and Social Policy Research Institute Maastricht University
Ernest Appiah
University of Ghana
This study is the first to employ panel data to examine well-being outcomes—self-rated health, happiness, life
satisfaction, and school enjoyment—of children in transnational families in an African context. It analyzes data
collected in 2013, 2014, and 2015 from secondary schoolchildren and youth (ages 12–21) in Ghana (N = 741).
Results indicate that children with fathers, mothers, or both parents away and those cared for by a parent, a
family, or a nonfamily member are equally or more likely to have higher levels of well-being as children in
nonmigrant families. Yet, there are certain risk factors—being a female, living in a family affected by divorce
or by a change in caregiver while parents migrate—that may decrease child well-being.
Sub-Saharan Africa is a major exporter of labor (Ghana Statistical Service, Ghana Health Service,
migrants to the world. A significant number of and ICF International, 2014). Whether this is due to
these migrants originate from Ghana. Accurate fig- migration or other reasons is unknown.
ures on the number of Ghanaian migrants abroad Despite the size of this population, relatively lit-
are missing because of the lack of systematic data tle is known about it. Parental migration is often
collection and also because of the illegal status of motivated by a desire to improve the living stan-
migrants overseas. By some estimates, the stock of dards of children and other family members,
Ghanaians who left the country for Western desti- although it is increasingly recognized that such ben-
nations had reached 825,000 by 2010, excluding efits may come with social and emotional costs.
undocumented migrants (World Bank, 2011). By Indeed, studies show that parental migration can
other estimates, a total of 1.5 million Ghanaians have negative consequences on child well-being
reside abroad in Europe, North America, other Afri- measured with outcomes such as education (Cebo-
can countries (predominately in the Economic Com- tari & Mazzucato, 2016; Cortes, 2015; Hu, 2013;
munity of West African States), the Middle East, Kandel & Kao, 2001; Nobles, 2011), physical and
and Asia (Twum-Baah, 2005). Many of these subjective health (Cebotari, Mazzucato, & Siegel,
migrants are parents who, either by choice or due 2017; Dreby, 2010; Wen & Lin, 2012), psychological
to strict migration regulations, leave their children health (Graham & Jordan, 2011; Mazzucato & Ceb-
in Ghana in the care of a family or a nonfamily otari, 2017), and happiness (Jordan & Graham,
member, thus creating transnational families. There 2012). Through detailed analyses, these studies
are no precise statistics on the number of children inform theoretical explanations and highlight the
living in transnational care in Ghana, but nationally challenges that transnational families face in main-
representative data suggest that approximately 37% taining active relationships across borders. At the
of children under the age of 18, excluding orphans,
lived without at least one biological parent in 2014
© 2017 The Authors
Child Development published by Wiley Periodicals, Inc. on behalf of Society
This work was supported by the Netherlands Organization for for Research in Child Development.
Scientific Research, WOTRO Science for Development Division This is an open access article under the terms of the Creative Commons
under grant number W01.65.316. Attribution-NonCommercial License, which permits use, distribution and
Correspondence concerning this article should be addressed to reproduction in any medium, provided the original work is properly cited
Victor Cebotari, Maastricht University/Social Policy Research and is not used for commercial purposes.
Institute. Electronic mail may be sent to victor.cebotari@ 0009-3920/2018/8905-0024
maastrichtuniversity.nl or victor.cebotari@spriglobal.org. DOI: 10.1111/cdev.12879
Well-Being of Children in Transnational Families 1769
same time, these studies show that families con- China. There is a gap in the literature on large-scale
tinue to operate across large distances and over analyses of African cases (although see Carling &
longer periods of time, and are able to be ever more Tønnessen, 2013; Cebotari & Mazzucato, 2016; Ceb-
resilient to the negative effects of migration. It is otari et al., 2017; Gaydosh, 2015; Goldberg, 2013;
now acknowledged that children of migrants live in Madhavan, Schatz, Clark, & Collinson, 2012; Maz-
complex transnational care arrangements in the zucato & Cebotari, 2017). The African geographical
country of origin (Dreby, 2010; Parren~as, 2005). Dif- focus is important due to specifics in family norms
ferent forms of transnational care allow children to governing the function of transnational families.
continue to experience the comfort of a familial The widespread practices of child fostering and
environment at home while simultaneously enjoy- social parenthood, where it is customary for chil-
ing the economic benefits of migration (Antman, dren to live in nonparental households, may result
2012; Donato & Duncan, 2011). Yet how the com- in outcomes that are different from those found in
plexity of transnational family arrangements affects other parts of the world.
child well-being is an under researched area of This study analyzes a sample of Ghanaian
inquiry. schoolchildren and youth between the ages of 12
This study builds on recent research on children and 21 years. The age range reflects the distribution
in transnational care in several sub-Saharan African of pupils in the schools and classrooms of the sur-
countries that used a range of indicators to veyed areas. For convenience, we use the term
investigate psychological, educational, and health “children” for this age group to emphasize the type
well-being outcomes (Carling & Tønnessen, 2013; of relationship they have with the migrant parents.
Cebotari & Mazzucato, 2016; Cebotari et al., 2017; In the following, we discuss the context and the
Mazzucato & Cebotari, 2017; Mazzucato et al., mechanisms under which migration may affect
2015). This study focuses on subjective measures of child well-being.
child well-being and contributes to the literature in
four ways. This is the first study to employ panel Migration and the Well-Being of Children
data to draw inferences with regard to child well-
being in African transnational families. Using panel Since the turn of the century, transnational fam-
information is important because child outcomes ily studies emerged to focus specifically on families
are dynamic events and the existing studies almost living across borders. Within this stream of litera-
exclusively rely on data snapshots to advance theo- ture, the outcomes of children left behind have
retical and empirical arguments. Second, this study become an important area of inquiry. Although
investigates the well-being of children who experi- there is a substantial interest in the well-being of
ence different transnational care arrangements. children in transnational care, theoretical input
Within this setting, we move beyond the classical remains underdeveloped. In general, parental
measure of who the migrant parent is and addition- migration is viewed as a strategy to improve the
ally look at internal and international parental well-being of children and other family members
migration, and who the caregiver of the child is. when there is a state of economic deficiency in the
We distinguish between different transnational fam- country of origin (Stark & Bloom, 1985). However,
ily characteristics also in relation to other conditions questions remain about whether migration is a suc-
that may be affecting child well-being, such as the cessful strategy to improve the well-being of family
stability of the care arrangement, marital status, members who stay behind (Adams & Page, 2005).
and the gender of the child. As a result, we con- When looking specifically at children in transna-
tribute to the literature by bringing the complexity tional care, research indicates that parental migra-
of transnational family life into focus when measur- tion may improve children’s socioeconomic status
ing child well-being. Third, this study adds to (Wen & Lin, 2012), their health, and nutrition (Asis,
small-scale qualitative studies that compose the 2006; Carling & Tønnessen, 2013), as well as their
majority of research on transnational families by educational aspirations, attainment, and perfor-
comparing children in transnational care with those mance (Antman, 2012; Kandel & Kao, 2001). At the
in nonmigrant families, thus enabling conclusions same time, studies articulate concerns about the
of whether the observed outcomes are relevant for emotional costs experienced by children following
a larger population of children. Fourth, the African parental departure, especially when mothers
context is a contribution to the literature, which is migrate (Jordan & Graham, 2012; Parren~as, 2001).
mostly based on evidence gathered in Latin Amer- The effects of migration on child well-being
ica, Southeast Asia, and on internal migration in must, however, be understood in the context
1770 Cebotari, Mazzucato, and Appiah
surrounding the transnational care of children (Gra- children and caregivers back home regarding
ham & Jordan, 2011; Mazzucato & Schans, 2011). reunification and material benefits (Dreby, 2007;
The transnational family literature focuses predomi- Parren~as, 2005). When these expectations are
nantly on specifics of parental migration and finds unmet, children may experience psychological and
that maternal migration often results in poorer out- material difficulties that may affect their well-
comes for children. Specifically, studies conducted being.
in Southeast Asia and China found that separation The role of the caregiver in the life of children is
from a mother makes children unhappier (Jordan & conceptually important but does not feature promi-
Graham, 2012), more disadvantaged in health, and nently in the transnational family research (Mazzu-
less engaged in school (Wen & Lin, 2012). Drawing cato & Schans, 2011). Who the caregiver is when
on ethnographic work in the Philippines, Parren~as parents migrate plays a crucial role in the develop-
(2005) found that children are more emotionally ment of children in transnational care. If the parent
insecure following maternal migration than when who migrates is the primary caregiver, then chil-
fathers migrate. A negative relation between mater- dren are more likely to experience difficulties
nal migration and child well-being was found by because of separation (Heymann et al., 2009). When
research conducted in other geographical areas, parents migrate, children must adapt to new care
such as Latin America (Dreby, 2007; Schmalzbauer, arrangements, form new attachments, and accept
2004) and Africa (Goldberg, 2013; Mazzucato et al., new authority figures (Schmalzbauer, 2004). As
2015). However, one must be cautious in making such, the caregiver plays a fundamental role in the
claims about negative consequences on children life of children by providing a good quality parent-
when mothers migrate, as this area of inquiry is ing when one or both parents migrate. In Tanzania,
recent and largely unexplored (Madhavan et al., Gaydosh (2015) finds that many parents decide to
2012; Mazzucato & Cebotari, 2017). migrate only if they are able to find a close family
Transnational family studies, for the most part, member to care for their children. The availability
look at international migration. There is also a of kin to provide care to children is seen as an
large body of research that has been conducted in enabling factor allowing African parents to migrate
China and relates to internal migration, although alone for work (Madhavan et al., 2012). Conversely,
over large distances (Gao et al., 2010; Hu, 2013; when migrant parents have difficulties finding a
Wen & Lin, 2012). A number of recent studies suitable caregiver in the local network of people,
looked at the effects of internal and international then tensions may arise within the transnational
parental migration in relation to well-being out- family (Poeze, Dankyi, & Mazzucato, 2017). The
comes and found that more often than not chil- quality of care provided by the caregiver is also of
dren whose parents migrated abroad do not have relevance and is expected to be higher when the
poorer physical or psychological health than chil- caregiver is one of the parents of the child or a
dren whose parents migrated internally in Ghana close family kin member (Schmalzbauer, 2004).
and Nigeria (Cebotari et al., 2017; Mazzucato The above discussion has focused on the charac-
et al., 2015). Another study found no association teristics of the family members (which parent
between internal or international parental migra- migrate and who cares for the child) and on the
tion and the school performance of Angolan and type of migration (internal or international). There
Nigerian children, but Ghanaian children were is another set of characteristics that pertains to the
found to fare worse in school when parents were context within which migration takes place both at
abroad than when parents were in the country the family and societal levels. These relate to the
(Cebotari & Mazzucato, 2016). Based on this evi- stability of the care arrangement, whether parents
dence, there are reasons to believe that these two are divorced or separated, the gender of the child
forms of migration will impact child well-being (related to gender norms), and more generally to
differently. Insights from the literature suggest that the family normative context. These are discussed
close proximity is necessary between the child and in the following.
the primary caregiver for optimal child develop- Little is known about how stable the care
ment (Suarez-Orozco, Todorova, & Louie, 2002). arrangements are for children in transnational care.
When parents migrate internationally, they usually The role of the caregiver and the frequency with
encounter administrative and financial difficulties, which children change caregivers have been tradi-
which prevents them from physically seeing their tionally the focus of child psychology (Amato &
children on a regular basis. International parental Cheadle, 2005) and demographic scholarship (Gay-
migration also creates higher expectations in dosh, 2015; Goldberg, 2013; Madhavan et al., 2012).
Well-Being of Children in Transnational Families 1771
Research shows that disruptions in care arrange- the investment in girls is sensitive to the household
ments increase the risk of children’s violence, risky income, in that fewer resources from abroad are
health behavior, adolescent childbearing, early used on girls’ education, with negative effects on
school dropout, and greater mental discomfort their school success. In Ghana, qualitative evidence
(Boynton-Jarrett, Hair, & Zuckerman, 2013; Gold- suggests that girls do more domestic work and can
berg, 2013). Not until recently was the stability of move or be moved around more easily, which may
the care arrangement introduced in the transna- alter their quality of life (Whitehead, Hashim, &
tional family literature (Mazzucato et al., 2015). Iversen, 2007). In other contexts, however, the gen-
Studies point to a variety of factors that may dis- der differences were found to advantage girls when
rupt the care arrangements of African children, parents migrate. Evidence from Mexico suggests
including parental migration and the remitting flow that paternal international migration has a net posi-
(Poeze et al., 2017), parental divorce (Madhavan tive effect on girls’ educational attainment, albeit
et al., 2012), child return migration (Bledsoe & Sow, such an advantage ceases to exist when fathers
2011), and attending school or providing help to migrate internally (Antman, 2012).
kin (Goody, 1982). These findings highlight the The normative family context in which transna-
importance of studying the stability of the care tional families operate is also important to take into
arrangement and how it relates with child well- account. Much transnational family research origi-
being in transnational families. nates from Latin American contexts, where nuclear
A recent body of research has focused on the family norms of care are strong and where it is
relation between parental marital status and the common for migrant parents and children to
well-being of children in transnational care. Evi- reunite at the destination or to regularly see each
dence from migration research in Ghana reveals other at the place of origin (Dreby, 2010; Nobles,
that parental absence may strain spousal relations, 2011). By contrast, norms of child fosterage and
especially when women migrate singly or when social parenthood prevail in many African coun-
both parents are away at the same time (Caarls & tries, including Ghana, where many children live in
Mazzucato, 2015). A number of studies compared the care of someone other than the biological par-
child well-being under conditions of parental ents (Goody, 1982; Isiugo-Abanihe, 1985; Radcliffe-
migration in families where parents are together or Brown & Forde, 1950). These norms contribute to a
divorced and found that, more often than not, chil- high proportion of children who live apart from
dren of divorced migrant parents are worse off in their biological parents, irrespective of whether the
terms of nutrition, fever, and diarrhea (Carling & parent is a migrant. Leaving children behind in
Tønnessen, 2013); psychological health (Mazzucato transnational care can be a preferred choice for
& Cebotari, 2017); educational aspirations (Nobles, Ghanaian parents (Dito, Mazzucato, & Schans,
2011); and school performance (Cebotari & Mazzu- 2017). The education landscape in Ghana features
cato, 2016). The negative outcomes associated with high-quality boarding schools and parents at times
migration and divorce are explained by specifics of prefer this option for raising their children (Bledsoe
this transnational context. The migration of parents & Sow, 2011). Finally, Ghanaian caregivers were
in stable unions is motivated by a desire to improve found to build harmonious relationships with chil-
the lives of children, whereas the divorced migrant dren in their foster care and treat them as “their
parents may not always have time and resources to own” (Poeze et al., 2017).
invest in children, particularly if they have formed
new families abroad and have children in those Current Research
unions (Dreby, 2010).
Gender differences between children are also of Drawing on the above-mentioned literature, we
significance when measuring the well-being of chil- conduct an analysis that takes into account the
dren in transnational families. Although consider- complexity of transnational families and the possi-
able heterogeneity exists in gender norms in ble variations occurring in these families over time.
different contexts, the dominant narrative is one in We employ several dimensions pertaining to speci-
which girls are more disadvantaged when parents fics in transnational family forms. Specifically, we
migrate. In China, for example, Gao et al. (2010) look at internal and international migration in con-
found that girls in transnational care are more at nection to parental migration status and who the
risk for unhealthy behaviors such as drinking and caregiver of the child is. In addition, we analyze
smoking than girls in nonmigrant families. In Viet- the diversity in the transnational forms of living in
nam, Behrman and Knowles (1999) concluded that connection to some conditions—the stability of the
1772 Cebotari, Mazzucato, and Appiah
care arrangement, parental divorce, and child’s gen- the analysis. Thus, although detailed protocols were
der—that have been identified in the literature as established to allow for future replication of the
potential risk factors for children’s well-being. We data, the sample is not nationally representative.
do not expect to find uniform impacts on different The survey employed a stratified sampling pro-
well-being outcomes between children in transna- cedure to randomly select eight low- and high-qual-
tional and nontransnational families. In previous ity public and private junior high schools (JHS) and
studies (Cebotari & Mazzucato, 2016; Mazzucato & senior high schools (SHS) in the two locations. Both
Cebotari, 2017), we found that Ghanaian children school types last 3 years and SHS follows upon
with a migrant father and a mother caregiver were JHS. Normally students change schools between
likely to have poorer psychological health com- JHS and SHS, although some schools contain both
pared with children in nonmigrant families. At the levels. The Ministry of Education in Ghana ranks
same time, we found that the psychological well- public and private schools in different quality cate-
being and school grades of Ghanaian children gories based on school enrollment, infrastructure,
whose parents had both migrated were lower than and children’s final exam performance. Schools
those of children in nonmigrant families. We also were approached and asked to take part in the sur-
found that frequent changes in caregivers had a vey, all of which agreed to participate. Within each
negative effect on children’s psychological health school, a class was randomly selected from each of
but not on their educational performance. This the first two grades for the inclusion in the first
study looks at other, more subjective, indicators of round of the survey (2013). We chose to survey the
child well-being. By extending the range of well- first two grades of JHS and SHS to have a sufficient
being dimensions that are being examined, and by sample and to ensure that we would have all stu-
conducting the analysis in a longitudinal fashion, dents for at least 2 years in our panel. In addition,
we intend to have a more detailed insight into how a random but purposive sample of children in
sensitive the results are to the particular measure transnational care was selected in each school to
being analyzed and how this changes over time. ensure a sufficient number of children with migrant
This study aims to answer the following question: parents. In the subsequent panels, the same chil-
To what extent do different forms of transna- dren were surveyed until they graduated from the
tional family life affect the well-being (as measured school. During all rounds of the survey, children
by subjective measures of health, life satisfaction, were reminded about the purpose of the study and
happiness, and school enjoyment) of Ghanaian chil- the voluntary nature of their participation.
dren in transnational care compared with children Overall, the national school attendance rates of
in nonmigrant families? Ghanaian children in urban JHS and SHS are 83%
and 58%, respectively (Ghana Statistical Service,
Ghana Health Service, and ICF International, 2014).
Method In our survey, the survival rates of panel children
Data in JHS were 83% in the second round and 96% in
the third round. In comparison, the survival rates
Data were collected in 2013, 2014, and 2015 as of panel children in SHS were 82% and 85% in
part of a panel study on the Effects of Transnational Rounds 2 and 3, respectively. These rates do not
Child Raising Arrangements on Life-chances of include children who graduated from school. Each
Children, Migrant Parents and Caregivers between year, schools were visited a second time to include
Ghana and the Netherlands (TCRA Ghana). The panel children who may have been absent during
survey gathered information on children and youth the first visit. Reasons for other absences were
aged 12–21 in two urban areas with high out- mainly related to school dropout and children being
migration rates: Kumasi and Sunyani. A national transferred to other schools. Of those children who
data sampling strategy was not sustainable in the dropped out or changed schools, 48% were living
context of this study given the geographical spread with both parents, whereas 34% and 18% had one or
of high out migration areas in Ghana. The Ashanti both parents away internally and internationally,
and the Brong Ahafo Regions, in which the two respectively, at the time of data collection.
cities are, respectively, located, are regions with The survey questionnaire was administered in
high rates of international emigration (Twum-Baah, English and filled in by students themselves under
Nabila, & Aryee, 1995) and thus ensured that we the guidance and supervision of the surveying
would be able to sample enough children with team. The same team, composed of five trained
international migrant parents to be able to conduct interviewers, assisted children in completing the
Well-Being of Children in Transnational Families 1773
questionnaires during the three rounds of data col- education of the caregiver, the duration of child–
lection. In total, 985 unique respondents were sam- parent separation, the marital status of parents, and
pled. Of these, 405 took part at two waves and 350 the stability of the care arrangement. The caregiver’s
took part at all three waves. This study includes education is a binary indicator, where one indicates
only children who were present in at least two completed secondary education or more. The infor-
rounds of the survey. To avoid ambiguity in the mation about child’s current caregiver—the mother,
status of parental absence, we omit children who the father, both parents, a family, and a nonfamily
were orphans of both parents at the start of the sur- member—was used in relation to that person’s
vey (N = 14). highest level of education achieved. For children
cared for by both parents, the highest education
Measures level of the two parents was recorded. The use of
caregiver’s education in longitudinal studies is
Self-rated well-being outcomes generally reflect important because it accounts for changes in adult
how people evaluate the quality of their life. We human capital levels when children change resi-
use four measures of well-being to capture the com- dences and live with different caregivers over time.
plexity of the concept: self-rated health, life satisfac- Also, qualitative studies have shown that in con-
tion, happiness, and school enjoyment. In all texts of international migration, the role of the
instances, children were asked to indicate on a scale child’s resident caregiver (whether the other biolog-
from 1 to 5 how they rate their own health, how ical parent or someone else) is important for the
satisfied they are with their life, how happy they child’s well-being (Dreby, 2007; Schmalzbauer,
consider themselves, and if they generally enjoy 2004). These authors argue that a caregiver can help
school. Higher scores on the scale reflect better a child to interpret his/her parent’s migration posi-
well-being outcomes. These are measures that are tively, provide the necessary care and affection to
broadly used in the well-being literature and have make a child feel cared for, and foster communica-
been employed and validated in previous research tion between the child and the migrant parent
on transnational families (Dito et al., 2017; Donato through long-distance phone or Internet calls. The
& Duncan, 2011; Jordan & Graham, 2012). duration of separation records the period of time
The analysis includes two independent variables since children last saw their parents in person. This
pertaining to different forms of transnational family measure allows for three answers: no separation,
life of children in Ghana. The first measurement separation occurred in the past 12 months, and sep-
provides a detailed account of the parental migration aration occurred over 13 months ago or more. The
status: non migrant, father away internationally, marital status of parents is a binary variable with
father away internally, mother away internationally, married/together and divorced/separated as cate-
mother away internally, both parents away interna- gories. The stability of the care arrangement is also a
tionally, and both parents away internally. The sec- binary variable, where one indicates that children
ond variable defines the child’s caregiver: either changed caregivers one or more times during all
both parents, or, for cases where one or both par- parental migration events in a child’s lifetime.
ents migrated, the other parent, an uncle or aunt, a Two socioeconomic indicators that are likely cor-
grandparent, or others. This last category includes related with migration and child well-being are
all relationships (siblings, nonkin adults, nobody, included in the models. Controlling for household
etc.) that could not be assigned separate categories wealth is important because migration is often
due to small sample size. For the two transnational selective along a scale of affluence. Migration of
family life variables, cases where one parent was parents involves resources and those who are able
away internally and the other away internationally to ensure better living conditions for their children
(N = 71), the status of the parent who was abroad may also be more likely to migrate internally or
was counted. The rationale for this choice is based internationally. To address this selectivity, we rely
on evidence from the literature that shows greater on a methodological tool described by McKenzie
difficulties for children when parents migrate over (2005), to create a household asset index. Specifically,
large distances (Cortes, 2015; Suarez-Orozco et al., we employ principal components analysis to gener-
2002). ate an index that comprises information on assets
Individual and family-specific control variables such as the ownership of durable goods (house,
are included in the models. Two individual-level refrigerator, computer, and means of transporta-
controls include child’s gender and age in full years. tion) and the access to private utilities (toilet and
The four family characteristics consist of the bathing facilities). A second measure of children’s
1774 Cebotari, Mazzucato, and Appiah
socioeconomic status summarizes the general living complicated by the issue of migrant selectivity. The
conditions of the child. Children were asked to migration of parents may be positively selected in
assess their living conditions in relation to other many ways, and those who are able to provide
children on a three-category scale: better, the same, more resources for children may also be more likely
or less good. For this analysis, a binary indicator to migrate. A number of controls described earlier
was derived, where one indicates better living con- are included in the models to partially control for
ditions and zero otherwise. The two indicators are migrant selection. Auxiliary analyses (not shown)
used in lieu of a direct measure of household were performed in which we looked at changes in
income because it is generally acknowledged in the the household assets and the subjective living con-
literature that income data are difficult to collect ditions of children whose parents experienced a
accurately and that expenditure and asset data bet- transition from being a nonmigrant to being an
ter explain migration decision making, human capi- internal or an international migrant during the
tal formation, and health investments (Acosta, 2011; panel years. There was no conclusive evidence to
McKenzie, 2005). suggest that more household resources and better
A further two measures account for the total living conditions were available to parents and chil-
number of children living with the child and the dren in the year prior to migration. However, there
number of younger children that are present in the may also be other conditions that influence migra-
household. These indicators comprise biological tion and child well-being that are not captured by
brothers and sisters as well as other children who our data, for example, the prepanel measurements
live under the same roof with the respondent. of wealth, which may induce parents to migrate
Together, these measurements provide an account and affect children’s well-being. We caution readers
of the total number of children and the position of to keep this in mind as they consider the analysis.
the surveyed children within the household. In the In this investigation, we employ a fixed effects
African context where family relations are fluid, modeling strategy, where all time-invariant child
children can derive support from each other but the and family characteristics are captured by the fixed
presence of other children may also entail a reduc- effects method and all time events that may influ-
tion in resources if children are younger and requir- ence both child well-being and different transna-
ing care (Eloundou-Enyegue & Williams, 2006; tional characteristics are accounted for by the wave
Smith, Lalonde, & Johnson, 2004). (year) fixed effects. Compared with other computa-
Based on previous studies on transnational fami- tional techniques available such as the hybrid
lies, a family process variable was included to mea- method (Allison, 2009), this study’s fixed effects
sure the quality of the child–caregiver relationship approach works with the assumption that all mea-
(Cebotari & Mazzucato, 2016; Jordan & Graham, sured and unmeasured time-invariant variables are
2012). Five response categories for the quality of the accounted for as long as they have the same effects
relationship ranged from (1) always open/warm to at all occasions. The fixed effects modeling allows
(5) never open/warm. In the analysis, we created a estimating the effects of change in different forms
binary indicator where the last three categories of transnational family life on the change in child’s
were clustered to indicate a distant relationship and health, life satisfaction, happiness, and school enjoy-
the former two to indicate a close relationship. ment. Given the individual fixed effects modeling,
Interaction terms were fitted to examine the the measures employed in this study are those that
moderating relation between gender, marital status, change over time, except for gender, which is
the stability of care, and different transnational fam- included to produce the interaction terms.
ily forms. The interactions helped to establish All models add clusters of variables in a step-
whether there were gender-, care-, and marital-spe- wise fashion to ascertain whether the effects
cific variations according to specifics of migration, remained once the controls were added progres-
as some transnational family studies suggest (Ant- sively. Because of the space limitation, we only pre-
man, 2012; Hu, 2013; Nobles, 2011). sent the results of the full models. A separate
model is run for each transnational characteristic in
Analysis relation to each measure of child well-being. Each
model estimates interactions between gender, mari-
The aim of this study is to estimate the effects of tal status, and changes in caregiver, and the
different forms of transnational family life on child transnational family characteristics. For parsimony,
well-being over time. Modeling transnational family only significant interactions are displayed. In all
formations in relation to child well-being is models, robust standard errors were estimated and
Well-Being of Children in Transnational Families 1775
corrected for clustering of observations at individ- Table 1
ual level. Indicators in the analyses were tested for Means/Percentages (SDs) of Dependent and Independent Variables
collinearity and none was detected. Multilevel mod- Full panel sample
eling was considered but not fitted because, with
only eight schools in the sample, the number of Variables %/M (SD) N/n
units at the second level was too small for a solid
speci cation. Health 4.08 (0.93) 1,835fi
Life satisfaction 3.96 (1.01) 1,835
Happiness 4.01 (1.01) 1,835
School enjoyment 3.92 (0.96) 1,846
Results Parental migration status 100 1,848
Nonmigrant parents 48.92 904
Table 1 provides a descriptive overview of the indi- Father away internationally 15.42 285
cators employed in this study. Data showed that on Father away internally 14.18 262
average, children tended to report positive well- Mother away internationally 2.92 54
being outcomes. Within the sample population, Mother away internally 6.06 112
51% of children had at least one parent away inter- Both parents away internationally 4.46 64
nally or internationally at the time of the survey. Of Both parents away internally 9.04 167
those children living without a parent due to Caregiver 100 1,766
migration, the greatest proportion had a father Both parents, nonmigrant 46.55 822
One parent, other away internationally 13.65 241
away. When a parent migrated, children often One parent, other away internally 17.04 301
stayed in the care of the other parent, followed by Uncle or aunt, parent away 3.85 68
an uncle or aunt and other family members. Female internationally
respondents represented approximately half of the Uncle or aunt, parent away internally 5.32 94
overall sample. The average age of children at Grandparent, parent away internationally 2.66 47
the time of the surveys was 15.61 (SD = 2.01). The Grandparent, parent away internally 1.98 35
number of children separated from their parents for Other, parent away internationally 2.49 44
less than a year due to internal or international par- Other, parent away internally 6.46 114
ental migration was almost equal to those who Child is girl 47.53 884
have been separated from their parents for a longer Child age (years) 15.61 (2.01) 1,860
period of time (24% and 23%, respectively). Data Caregiver’s education 43.80 805
secondary or more
show that 29% of children changed their caregiver Duration of separation: none 52.63 979
one or more times during current and past parental Duration of separation: ≤ 12 months 24.03 447
migration events. Most children had parents who Duration of separation: 13 ≥ months 23.33 434
were married or in a relationship and reported a Parents divorced/separated 33.03 608
good relationship with their caregivers at home. In Child changed caregiver ≥ 1 28.66 533
general, children reported living with up to three Living conditions are better when 48.39 899
other children in the same household, and on aver- compared to other children
age there were up to 1.39 (SD = 1.35) younger chil- Household asset index 3.65 (1.51) 1,860
dren in the residence. Children also declared Total number of children living 2.99 (2.13) 1,860
having, on average, 3.65 (SD = 1.51) assets at home with the child
and approximately half of them reported having Total number of younger children 1.39 (1.35) 1,860
living with the child
better living conditions than other children. Distant relationship with the caregiver 19.07 354
Table 2 shows the transition rates of parental
migration status over panel years. Data revealed Note. Standard deviations in parentheses. The n indicates the
that of all children, the majority experienced no number of observations for categories within each indicator.
change in their parents’ migration status. In the
observed period, 88% of children with nonmigrant
parents remained so; 79% and 81% of fathers and more children had fathers and mothers who
mothers, respectively, remained abroad; 71% and migrated internally (6% and 2%, respectively) than
65% of fathers and mothers, respectively, stayed as internationally (2% and 0.2%, respectively). Of those
internal migrants in Ghana; and 70% of children children who transitioned from having a father
with both parents away internally or internationally or mother migrant to living together with their
remained so. When children transitioned from parent, more did so who had an internal migrant
having a nonmigrant parent to a migrant parent, parent (15% fathers and 23% mothers) than an
1776 Cebotari, Mazzucato, and Appiah
international migrant parent (8% fathers and 9% Table 4 presents the differences in means of
mothers). Transitions from having a single parent well-being outcomes by considering the migration
away to having both parents away were more status of parents and the child’s caregiver. The
common for internal (7% fathers and 9% mothers) bivariate differences showed that the reported
than international migration (3% fathers and 0% levels of health were generally higher among chil-
mothers). dren whose fathers or both parents migrated inter-
Transition rates of parental migration and chil- nationally when compared with children living
dren’s caregiver are shown in Table 3. In the sam- with both parents. Similarly, children of migrants
pled population, 77% and 69% of children with a compared with children of nonmigrants reported
parent away internationally or internally, respec- higher levels of school enjoyment when mothers or
tively, and cared for by the other parent remained both parents migrated internally and when fathers
in these care arrangements from 1 year to another. migrated internationally. The levels of life satisfac-
Of those children who were cared for by an uncle tion and happiness, by contrast, were lower among
or aunt, 81% stayed in the care of these family rela- children of migrants compared with children in
tives over time. Among categories of children that nonmigrant families. Some differences in well-being
experienced the largest percentage change in status were also observed when looking at the caregiver
were those cared for by grandparents when parents of the child. The reported levels of health were
were away internally and those cared for by others higher among children of international migrants
when parents were away internationally; of these compared with children living with both parents,
children, only 44% and 48%, respectively, remained and specifically when children stayed in the care of
in the same care arrangements during the panel a parent, a grandparent, an uncle, or aunt. The
years. When changes in the care arrangement reports of life satisfaction and happiness, by con-
occurred, most transitions were made toward the trast, were predominantly lower among children of
care of one or both parents and toward the care of migrants, irrespective of who the caregiver of the
a close family relative. Most child–parent reunifica- child was compared with children of nonmigrants.
tions occurred from within Ghana rather than from Finally, children of migrants, relative to children in
abroad. These data illustrate that children’s care- nonmigrant families, reported overall higher levels
giver and the migration of parents are dynamic of school enjoyment when parents went away inter-
processes that change over time, giving greater rea- nally and children stayed in the care of a parent, an
son to conduct a longitudinal analysis. uncle or aunt, a grandparent, or others.
Table 2
Transition Rates for Parental Migration Status: 2013–2015
Both Both
Parental Father Mother parents parents
migration Nonmigrant Father away away Mother away away away away
status parents internationally internally internationally internally internationally internally Total N
Nonmigrant 87.83 2.06 6.18 0.19 1.69 0 2.06 100 904
parents
Father away 7.82 79.33 5.59 0 2.23 3.35 1.68 100 285
internationally
Father away 15.33 4 71.33 0.67 1.33 0.67 6.67 100 262
internally
Mother away 9.38 0 3.13 81.25 6.25 0 0 100 54
internationally
Mother away 22.73 0 1.52 1.52 65.15 0 9.09 100 112
internally
Both parents away 12.50 5 2.50 5 0 70 5 100 64
internationally
Both parents away 10.53 5.26 9.47 1.05 4.21 0 69.47 100 167
internally
Notes. The matrix counts transitions from each observation to the next and follows the order of change in time for each individual. The
changes between 2013 and 2015 are reflected in the rows.
Well-Being of Children in Transnational Families 1777
Table 3
Transition Rates for Parental Migration and Child’s Caregiver: 2013–2015
Uncle or
Uncle or aunt,
Both One parent, One parent, aunt, parent parent Grandparent, Grandparent, Other, Other,
parents, other away other away away away parent away parent away parent away parent away
The caregiver nonmigrant internationally internally internationally internally internationally internally internationally internally Total N
Both parents, nonmigrant 87.01 1.86 7.01 0.62 1.44 0 0.41 0 1.65 100 822
One parent, other away 10.46 77.12 7.19 1.96 0 0 0 2.61 0.65 100 241
internationally
One parent, other away 17.98 3.93 68.54 0.56 2.25 0 2.81 0.56 3.37 100 301
internally
Uncle or aunt, parent away 2.70 0 0 81.08 5.41 2.70 0 5.41 2.70 100 68
internationally
Uncle or aunt, parent away 5.66 3.77 3.77 0 81.13 0 3.77 0 1.89 100 94
internally
Grandparent, parent away 6.90 3.45 3.45 0 3.45 72.41 6.90 3.45 0 100 47
internationally
Grandparent, parent away 11.11 5.56 11.13 0 11.09 5.56 44.44 0 11.11 100 35
internally
Other, parent away 6.90 10.34 0 17.24 0 3.45 0 48.28 13.79 100 44
internationally
Other, parent away 12.50 1.56 9.38 0 4.69 1.56 0 1.56 68.75 100 114
internally
Notes. The matrix counts transitions from each observation to the next and follows the order of change in time for each individual. The changes between 2013 and 2015 are
reflected in the rows.
1778 Cebotari, Mazzucato, and Appiah
Table 4
Means and Standard Deviations of Child Well-Being by Different Transnational Family Characteristics
Health Life satisfaction Happiness School enjoyment
Transnational family characteristics M (SD) N/n M (SD) N/n M (SD) N/n M (SD) N/n
Parental migration status F(6, 1,816) = 3.26** F(6, 1,816) = 4.42*** F(6, 1,816) = 3.27*** F(6, 1,827) = 2.20*
Both parents resident, nonmigrant 4.10 (0.90) 890 4.06 (0.95) 890 4.10 (0.96) 890 3.91 (0.96) 895
Father away internationally 4.16 (0.86) 283 3.93 (0.96) 283 4.02 (0.97) 283 3.96 (0.92) 284
Father away internally 4.10 (0.93) 258 3.80 (1.05) 258 3.83 (1.08) 258 3.91 (0.96) 259
Mother away internationally 3.72 (1.18) 54 3.64 (1.16) 54 3.81 (1.02) 54 3.61 (0.85) 54
Mother away internally 3.88 (0.94) 110 3.86 (0.94) 110 3.91 (0.91) 110 4.09 (1.01) 111
Both parents away internationally 4.26 (0.76) 64 3.96 (1.01) 64 3.97 (1.06) 64 3.76 (1.00) 64
Both parents away internally 4 (1.11) 164 3.82 (1.11) 164 3.96 (1.14) 164 4.01 (0.92) 167
Caregiver F(8, 1,733) = 1.61 F(8, 1,733) = 4.44*** F(8, 1,733) = 3.02*** F(8, 1,744) = 0.93
Both parents, nonmigrant 4.11 (0.87) 809 4.08 (0.940 809 4.12 (0.94) 809 3.93 (0.95) 814
One parent, other away 4.15 (0.88) 239 3.94 (0.95) 239 4.04 (0.94) 239 3.93 (0.93) 241
internationally
One parent, other away internally 4.08 (0.91) 296 3.86 (1.01) 296 3.87 (1.03) 296 3.95 (0.98) 298
Uncle or aunt, parent away 4.17 (0.80) 68 3.91 (0.98) 68 4.01 (1.05) 68 3.76 (0.97) 68
internationally
Uncle or aunt, parent away internally 3.97 (1.04) 91 3.93 (0.92) 91 3.90 (1.04) 91 3.96 (0.94) 94
Grandparent, parent away 4.19 (0.85) 47 3.95 (1.08) 47 3.93 (1.11) 47 3.95 (0.97) 47
internationally
Grandparent, parent away internally 3.91 (1.19) 35 3.65 (1.25) 35 4.08 (1.12) 35 4.14 (0.87) 35
Other, parent away internationally 3.79 (1.21) 44 3.57 (1.14) 44 3.72 (0.99) 44 3.69 (0.83) 43
Other, parent away internally 3.93 (1.08) 113 3.66 (1.18) 113 3.86 (1.18) 113 4.01 (0.94) 113
Note. Analysis of variance was used for all comparisons. Means presented with standard deviations in parentheses. Higher scores indi-
cate higher levels of health, life satisfaction, happiness, and school enjoyment. *p < .05. **p < .01. ***p < .001.
The bivariate results revealed some interesting when mothers migrated internally as compared
patterns on how children in different transnational with children living with both parents (b = 1.54,
care arrangements assessed their well-being out- Model 1; b = 2.08, Model 3; and b = 0.55, Model 4,
comes. However, to answer the research question, respectively). Results also showed that international
we had to additionally employ multivariate models maternal migration was associated with higher
to test the combined effect of the independent vari- levels of child happiness (b = 2.43, Model 3). Nota-
ables on well-being outcomes. Tables 5 and 6 dis- bly, the estimated coefficients for paternal migration
play the full fixed effects models for each and the internal migration of both parents were not
transnational characteristic in relation to well-being statistically significant in any of the models.
outcomes. Because the aim of this study is to inves- In addition, significant interaction effects were
tigate the relation between different transnational found between transnational family characteristics
characteristics and child well-being, we limit the and gender, divorce, and the stability of care,
interpretation and discussion of results to these respectively. Specifically, the interaction terms
associations. revealed that girls were less likely to report better
The relations were not uniform when considering health when mothers migrated internally (b of the
the transnational characteristics across different interaction term = 
0.85, Model 1) and when both
dimensions of child well-being. When the parental parents migrated internationally (b of the interac-
migration status was considered (Table 5), children tion term = 
0.85, Model 1). Furthermore, being a
with both parents away internationally were more girl reduced the overall positive impact of interna-
likely to have higher levels of health, life satisfac- tional maternal migration on child happiness (b of
tion, and happiness as compared with children in the interaction term = 
1.45, Model 3). The higher
nonmigrant families (b = 2.17, Model 1; b = 2.02, coefficient of the interaction term between mother
Model 2; b = 2.84, Model 3, respectively). Further- abroad families and being a girl, in relation to hap-
more, results showed that children had higher piness, may indicate greater difficulties for girls in
levels of health, happiness, and school enjoyment this particular family type.
Well-Being of Children in Transnational Families 1779
Table 5
Parental Migration Status and Child Well-Being—Fully Adjusted Models
Health Life satisfaction Happiness School enjoyment
model 1 model 2 model 3 model 4
b (SE) b (SE) b (SE) b (SE)
Parental migration status
Both parents resident, nonmigrant — — — —
Father away internationally 0.68 (0.43) 
0.06 (0.46) 0.79 (0.61) 0.14 (0.19)
Father away internally 0.40 (0.40) 
0.22 (0.44) 0.08 (0.57) 0.24 (0.16)
Mother away internationally 0.77 (0.76) 0.71 (0.59) 2.43 (0.85)** 0.04 (0.21)
Mother away internally 1.54 (0.58)** 1.06 (0.74) 2.08 (0.85)* 0.55 (0.23)*
Both parents away internationally 2.17 (0.69)** 2.02 (0.80)* 2.84 (1.04)** 0.43 (0.31)
Both parents away internally 0.25 (0.47) 
0.04 (0.46) 0.09 (0.91) 0.10 (0.20)
Child is girl 0.27 (0.33) 0.19 (0.52) 0.56 (0.43) 0.47 (0.28)
Child age (years) 
0.01 (0.06) 0.05 (0.05) 0.09 (0.05) 0.11 (0.05)*
Caregiver’s education secondary or more 
0.09 (0.08) 0.04 (0.09) 0.07 (0.08) 
0.03 (0.09)
Duration of separation: none — — — —
Duration of separation: ≤ 12 months 
0.30 (0.13)* 
0.17 (0.17) 
0.23 (0.16) 
0.27 (0.13)*
Duration of separation: 13 ≥ months 
0.36 (0.13)** 
0.06 (0.17) 
0.23 (0.15) 
0.26 (0.13)*
Parents divorced/separated 0.05 (0.10) 0.17 (0.15) 0.24 (0.14) 
0.10 (0.10)
Child changed caregiver ≥ 1 0.07 (0.11) 0.11 (0.13) 0.13 (0.12) 
0.02 (0.09)
Living conditions are better when compared to other children 0.01 (0.06) 0.19 (0.06)** 0.22 (0.06)*** 0.12 (0.07)
Household asset index 0.07 (0.03)* 0.05 (0.03) 0.03 (0.03) 0.01 (0.03)
Total number of children living with the child 0.01 (0.02) 0.01 (0.02) 
0.01 (0.03) 0.00 (0.02)
Total number of younger children living with the child 0.04 (0.03) 0.05 (0.03) 0.02 (0.03) 0.02 (0.04)
Distant relationship with the caregiver 
0.06 (0.08) 
0.26 (0.09)** 
0.26 (0.09)** 
0.25 (0.08)**
Interactions
Mother Away Internally 9 Female 
0.85 (0.39)*
Mother Away Internationally 9 Female 
1.45 (0.40)***
Both Parents Away Internationally 9 Female 
0.85 (0.41)*
Mother Away Internally 9 Divorce 
0.76 (0.31)*
Both Parents Away Internationally 9 Divorce 
1.42 (0.64)* 
1.49 (0.39)***
Both Parents Away Internationally 9 Changed Caregiver ≥ 1 
0.67 (0.25)** 
0.84 (0.27)** 
0.93 (0.35)**
Wave-fixed effects Yes Yes Yes Yes
Child-fixed effects Yes Yes Yes Yes
Unique number of children 741 741 741 741
Total number of observations 1,777 1,777 1,777 1,789
R2 0.04 0.06 0.08 0.07
Note. Standard errors in parentheses (adjusted to account for clustering within individuals). *p < .05. **p < .01. ***p < .001.
Interaction terms also showed negative associa- terms = 
1.42, Model 2). Significant interaction
tions between higher levels of happiness and paren- effects for the stability of care are also worth not-
tal divorce when mothers migrated internally or ing. Results showed that one or more changes in
when both parents migrated internationally (b of caregivers reduced the overall positive effect of
interaction terms = 
0.76 and 
1.49, Model 3, international parental migration on children’s
respectively). The relative effect size of the interac- health, life satisfaction, and happiness (b = 
0.67,
tion “both parents away internationally and Model 1; b = 
0.84, Model 2; b = 
0.93, Model 3,
divorce” was larger than for the comparable inter- respectively).
action “mother away internally and divorce” attest- Table 6 summarizes the results of models featur-
ing for a stronger negative effect of the former on ing the caregiver of the child in relation to well-
children’s reports of happiness. Similarly, children being outcomes. Net of other factors, there were no
were less likely to report higher levels of life satis- statistically significant effects between specifics of
faction when they lived in divorced families in the caregiver arrangement and different well-being
which both parents were abroad (b of interaction outcomes of children in transnational care as
1780 Cebotari, Mazzucato, and Appiah
Table 6
Caregiver and Child Well-Being—Fully Adjusted Models
Health Life satisfaction Happiness School enjoyment
model 5 model 6 model 7 model 8
b (SE) b (SE) b (SE) b (SE)
Caregiver
Both parents, nonmigrant — — — —
One parent, other away internationally 0.12 (0.16) 
0.23 (0.69) 0.68 (0.44) 
0.04 (0.62)
One parent, other away internally 0.26 (0.15) 
0.97 (0.65) 
0.14 (0.53) 
0.08 (0.46)
Uncle or aunt, parent away internationally 
0.01 (0.28) 2.92 (2.15) 2.68 (1.64) 
0.15 (1.57)
Uncle or aunt, parent away internally 0.27 (0.23) 
0.25 (1.01) 0.18 (0.88) 
0.62 (1.02)
Grandparent, parent away internationally 0.12 (0.28) 2.17 (1.40) 0.59 (1.53) 0.63 (1.12)
Grandparent, parent away internally 
0.06 (0.27) 
1.17 (2.38) 
1.41 (2.16) 0.54 (0.84)
Other, parent away internationally 0.40 (0.22) 0.77 (1.46) 1.00 (1.07) 0.31 (0.98)
Other, parent away internally 0.12 (0.25) 
0.54 (0.96) 0.34 (1.22) 1.58 (0.99)
Child is girl 0.14 (0.29) 0.08 (0.51) 0.64 (0.40) 0.42 (0.32)
Child age (years) 
0.02 (0.06) 0.04 (0.05) 0.07 (0.05) 0.11* (0.05)
Caregiver’s education secondary or more 
0.11 (0.08) 0.05 (0.09) 0.12 (0.09) 
0.05 (0.09)
Duration of separation: none — — — —
Duration of separation: ≤ 12 months 
0.20 (0.13) 
0.01 (0.17) 
0.16 (0.15) 
0.24 (0.13)
Duration of separation: 13 ≥ months 
0.26 (0.13) 0.14 (0.17) 
0.16 (0.15) 
0.21 (0.14)
Parents divorced/separated 0.04 (0.11) 0.27 (0.18) 0.24 (0.17) 0.01 (0.16)
Child changed caregiver ≥ 1 
0.05 (0.09) 0.09 (0.13) 0.02 (0.10) 
0.05 (0.12)
Living conditions are better when 0.03 (0.06) 0.18 (0.07)** 0.20 (0.07)** 0.10 (0.07)*
compared with other children
Household asset index 0.06 (0.03)* 0.05 (0.03) 0.02 (0.03) 0.02 (0.03)
Total number of children 0.02 (0.02) 0.01 (0.02) 
0.01 (0.03) 0.02 (0.02)
living with the child
Total number of younger 0.04 (0.03) 0.05 (0.03) 0.03 (0.03) 0.01 (0.04)
children living with the child
Distant relationship with the caregiver 
0.21 (0.09)* 
0.25 (0.09)** 
0.23 (0.09)* 
0.27 (0.08)***
Interactions
One Parent Caregiver, the 0.70 (0.32)*
Other Parent Away Internally 9 Female
Other Caregiver, Parent Away 0.77 (0.39)*
Internationally 9 Female
Uncle or Aunt Caregiver, Parent 
1.02 (0.54)* 
1.20 (0.59)*
Away Internationally 9 Divorce
Uncle or Aunt Caregiver, Parent 
0.71 (0.30)*
Away Internationally 9 Changed Caregiver ≥ 1
Grandparent Caregiver, Parent 
0.82 (0.40)*
Away Internally 9 Changed Caregiver ≥ 1
Wave-fixed effects Yes Yes Yes Yes
Child-fixed effects Yes Yes Yes Yes
Unique number of children 711 711 711 711
Total number of observations 1,702 1,702 1,702 1,714
R2 0.04 0.08 0.08 0.08
Note. Standard errors in parentheses (adjusted to account for clustering within individuals). *p < .05. **p < .01. ***p < .001.
compared with children in nonmigrant families. interaction term = 0.70, Model 6). Similarly, girls
However, interaction terms revealed some interest- were more likely to report higher levels of school
ing facts. Specifically, we found that girls were enjoyment when parents migrated internationally
more likely to report higher levels of life satisfaction and children were cared for by others (b of the
when they were cared for by a parent while the interaction term = 0.77, Model 8). At the same time,
other parent was away internally (b of the children in divorced families who were being cared
Well-Being of Children in Transnational Families 1781
for by an uncle or aunt when parents were abroad an important finding because it nuances the results
were generally less likely to report higher levels of from previous studies that associate parental migra-
life satisfaction (b of interaction term = 
1.02, tion with poorer well-being outcomes of children in
Model 6) or happiness (b of interaction transnational care. First, in contrast to small-scale
term = 
1.20, Model 7). Significant interaction studies on which most knowledge about transna-
effects of the stability of the care arrangement were tional family life stems, our study contains a com-
also found; children who changed their caregiver parison group of children in families without a
one or more times were less likely to report higher migration experience, allowing us to discern if neg-
levels of life satisfaction when they were cared for ative well-being outcomes are indeed particular to
by an uncle or aunt while parents were away inter- transnational family life or may be associated with
nationally (b of the interaction term = 
0.71, Model other characteristics of the population at large. Sec-
6) and of school enjoyment when they stayed in the ond, being one of the only longitudinal studies that
care of a grandparent and parents were away inter- looks specifically at child well-being in transna-
nally (b of the interaction term = 
0.82, Model 8). tional families, this study captures variations in
Taken together, the regression results seem to sug- well-being over time. We know from qualitative
gest that children were less likely to report higher studies (Dreby, 2010; Schmalzbauer, 2004) that chil-
levels of well-being when they lived in transna- dren can adapt to new caregivers as time goes by
tional settings that were characterized by divorce or and are better able to cope with parental absence as
instability of care. Furthermore, there were gender they grow older. Although demographic longitudi-
differences in children’s reports of their well-being nal studies in the area of child well-being in sub-
dependent on who the caregiver of the child was Saharan Africa exist (see Gaydosh, 2015; Goldberg,
and parents’ migration status. 2013; Madhavan et al., 2012), these are mainly
focused on the South African context that is charac-
terized by internal migration. This study is the first
Discussion to collect such data for international migration.
Finally, most of the negative associations found
This analysis is the first, to our knowledge, to longi- between transnational families and child well-being
tudinally investigate the effects of living in transna- pertain to studies in Latin America, China, and
tional families on the well-being of children in an Southeast Asia. The African context may differ
African context. This allows us to more accurately because of the prevalent norms around family life
ascertain the effects that have been previously characterized by social parenthood and child foster-
tested in studies based on cross-sectional data. Four ing, making it common for children to change resi-
results, in particular, warrant discussion because dency and be raised by people other than their
they add a different perspective to current knowl- biological parents, irrespective of parental migration
edge. First, living transnationally does not necessar- status (Radcliffe-Brown & Forde, 1950). This may
ily cause poorer well-being, contrary to what is contribute to the fact that Ghanaian children do not
often found in the literature. Second, the forms that necessarily feel underprivileged if they do not live
transnational families take are an important differ- with their parents (Poeze et al., 2017) as has been
entiating factor as to whether children experience a found in other contexts (Cortes, 2015; Parren~as,
decline in well-being as a consequence of parental 2005). Finally, an important distinguishing factor is
migration. Third, multiple indicators of well-being that few studies on migration and child well-being
are necessary in order to more accurately reflect the ask children to evaluate their well-being and
multifaceted nature of how children experience liv- instead rely on the assessments of adults such as
ing in a transnational family. Finally, the gender of parents, other caregivers, or teachers. As Jordan
the child, marital discord, and instability of care are and Graham (2012) note, children experience their
particular risk factors that heighten the likelihood transnational realities differently from adults, and it
of poorer outcomes for children within transna- is, therefore, important to take children’s reports
tional families. These findings are discussed in into account.
detail in the following. The results also highlight the importance of con-
Results indicate that living transnationally in and sidering multiple indicators of well-being, as differ-
of itself do not cause poorer well-being. In fact, ent associations were found in relation to specific
under some transnational family compositions, chil- well-being measures, attesting to the multiplicity of
dren may have better well-being outcomes than dimensions that comprise a child’s experience in
their counterparts living with both parents. This is transnational families. We also found that although
1782 Cebotari, Mazzucato, and Appiah
on the whole transnational family life does not lead caregiver by circumstantial characteristics such as
to worse well-being outcomes, particular family the caregiver’s level of education and the certainty
configurations do put children in more vulnerable that children will receive a proper disciplinary
positions. The stability of the care arrangement, supervision (Poeze et al., 2017).
something that only recently has been taken into We also find an important story about gender as
account in migration research (Mazzucato et al., a risk factor for children in transnational care. Girls’
2015), was found to be an important determining health is negatively affected depending on which
factor in the analysis. Care arrangements in which parent migrates: when mothers migrate internally
children move residences and are in the care of dif- or both parents migrate internationally, the health
ferent people are more likely to negatively reflect advantages diminish for girls as compared with
on children’s well-being. Changing caregivers may boys. The same is true for their happiness when
be traumatic for children because they need to mothers migrate internationally. In terms of care-
adapt to new environments, form new attachments, givers, girls’ life satisfaction and school enjoyment
and accept new authority figures (Schmalzbauer, were found to reach higher levels than that of boys
2004; Smith et al., 2004). Findings of this study clo- when they live with one parent and the other par-
sely align with evidence from demographic longitu- ent is an internal migrant and when they are in the
dinal research in Africa that reflects on the care of a nonparental caregiver and the parent is
importance of the stability of care for children’s life away internationally. Yet, girls who change care-
paths (Goldberg, 2013; Madhavan et al., 2012). givers fare worse than boys in terms of life satisfac-
Who the caregiver is, whether a parent, an aunt tion and school enjoyment when they are cared for
or uncle, or someone else, does not make a differ- by uncles or aunts and their parents are away inter-
ence for children’s well-being. Yet who the migrant nationally and when cared for by a grandparent
parent is and whether he or she has migrated inter- and their parents are away internally. The negative
nally or internationally does make a difference. Bet- gender differences may reflect the unequal distribu-
ter outcomes are found for the health of children tion of workload and resources in the Ghanaian
when both parents have migrated internationally, households, in that girls do more chores and benefit
greater happiness for children when mothers have less than boys from the remitted capital (Whitehead
migrated either internally or internationally, and et al., 2007). Although the reasons need to be fur-
more school enjoyment and better health for chil- ther investigated, it is important to note that under
dren whose mothers have migrated internally com- certain circumstances girls may suffer or gain more
pared with children living with both parents. These than boys.
are important distinctions especially with regard to Finally, divorce is also an important risk factor.
mothers’ migration as previous studies have Children who were cared for by an aunt or uncle
emphasized the suffering that children experience while their parent migrated internationally experi-
when mothers migrate (Cortes, 2015; Jordan & Gra- enced a decline in life satisfaction and happiness
ham, 2012; Wen & Lin, 2012). Yet, in our study, we when parents were divorced or separated. Likewise,
find that transnational family life when mothers divorce or separation leads to lower levels of happi-
migrate actually leads to greater health, happiness, ness in children who live in families where the
and school enjoyment for children in Ghana. Migra- mother migrated internally and to lower levels of
tion studies have shown that women tend to remit happiness and life satisfaction when both parents
more regularly and their remittances are preva- migrated internationally. These findings are consis-
lently used for the upkeep of the family (Pfeiffer & tent with recent research on divorce, migration, and
Taylor, 2008). Furthermore in our study, the care- child well-being in Ghana (Cebotari et al., 2017;
givers of children whose mothers migrated either Mazzucato & Cebotari, 2017) and other contexts
internationally or internally were noticeably more (Carling & Tønnessen, 2013; Nobles, 2011). The evi-
likely to have higher levels of education (results not dence presented here leads to an important asser-
shown). This may correspond with the fact that the tion: parent–child separation may take various
Akan ethnic group, to which most of our respon- forms through migration, divorce or both, among
dents belong, is matrilineal, in that mothers are which the separation that involves both migration
more involved in the decision making related to the and divorce may reflect on children’s outcomes
care of their children. Indeed, a recent study that more negatively. Future research that looks into the
looks into the transnational child-care practices, dynamics of parental migration and marital discord
mostly among the Akan in Ghana, concluded that in relation to child well-being will do well to fur-
migrant parents are guided in their choice of a ther explore this line of inquiry.
Well-Being of Children in Transnational Families 1783
The limitations of this study should be noted. behind. Journal of Population Economics, 25, 1187–1214.
Although we employ panel data and control for the https://doi.org/10.1007%2Fs00148-012-0423-y
socioeconomic status of the child’s household, the Asis, M. B. (2006). Living with migration: Experiences of
issue of migrant selectivity persists. There is no children left-behind in the Philippines. Asian Population
empirical literature that is able to completely con- Studies, 2, 45–67. https://doi.org/10.1080/
trol for migrant selectivity over time, but one needs 17441730600700556Behrman, J. R., & Knowles, J. C. (1999). Household
to be alert to this condition. Another limitation is income and child schooling in Vietnam. World Bank
that we conducted the study over a period of Economic Review, 13, 211–256. https://doi.org/10.1093/
3 years and therefore only measure short- to med- wber/13.2.211
ium-term effects on child well-being. Subsequently, Bledsoe, C., & Sow, P. (2011). Back to Africa: Second
we are not able to assess how the circumstances of chances for the children of West African immigrants.
migration affect children in their transitions into Journal of Marriage and Family, 73, 747–762. https://doi.
adulthood, once they leave school for example. org/10.1111/j.1741-3737.2011.00843.x
Finally, this study includes children who attended Boynton-Jarrett, R., Hair, E., & Zuckerman, B. (2013). Tur-
schools in two urban areas with high out migration bulent times: Effects of turbulence and violence expo-
rates in Ghana. Therefore, the results may not be sure in adolescence on high school completion, health
generalizable to children who dropped out of risk behavior, and mental health in young adulthood.Social Science & Medicine, 95, 77–86. https://doi.org/10.
school or reside in areas with lower migration rates 1016/j.socscimed.2012.09.007
in the country. Caarls, K., & Mazzucato, V. (2015). Does international
These limitations notwithstanding, this study migration lead to divorce? Ghanaian couples in Ghana
adds a longitudinal perspective to the current and abroad. Population, 70, 127–151. https://doi.org/
knowledge on the effects of living in transnational 10.3917/popu.1501.0135
care on child well-being. The fact that children Carling, J., & Tønnessen, M. (2013). Fathers’ whereabouts
across different transnational family forms were and children’s welfare in Malawi. Development Southern
equally or more likely to have better levels of Africa, 30, 724–741. https://doi.org/10.1080/0376835x.
well-being compared with children in nonmigrant 2013.859068
families is encouraging and challenges some of the Cebotari, V., & Mazzucato, V. (2016). Educational perfor-
existing theoretical constructs about child well- mance of children of migrant parents in Ghana, Nigeriaand Angola. Journal of Ethnic and Migration Studies, 42,
being in the transnational family literature. The 834–856. https://doi.org/10.1080/1369183x.2015.
results of this study help avoid discourses that 1125777
stigmatize children of migrant parents and Cebotari, V., Mazzucato, V., & Siegel, M. (2017). Child
strengthen the argument that migration in and of development and migrant transnationalism: The health
itself is not detrimental for children in transna- of children who stay behind in Ghana and Nigeria.
tional care. Journal of Development Studies, 53, 444–459. https://doi.
org/10.1080/00220388.2016.1187723
Cortes, P. (2015). The feminization of international migra-
References tion and its effects on the children left behind: Evidence
from the Philippines. World Development, 65, 62–78.
Acosta, P. (2011). School attendance, child labour, and https://doi.org/10.1016/j.worlddev.2013.10.021
remittances from international migration in El Salvador. Dito, B., Mazzucato, V., & Schans, D. (2017). The effects
Journal of Development Studies, 47, 913–936. https://doi. of transnational parenting on the subjective health and
org/10.1080/00220388.2011.563298 well-being of Ghanaian migrants in The Netherlands.
Adams, R. H., & Page, J. (2005). Do international migra- Population, Space and Place, 23(3). Advance online publi-
tion and remittances reduce poverty in developing cation. https://doi.org/10.1002/psp.2006
countries? World Development, 33, 1645–1669. https://d Donato, K. M., & Duncan, E. M. (2011). Migration, social
oi.org/10.1016/j.worlddev.2005.05.004 networks, and child health in Mexican families. Journal
Allison, P. D. (2009). Fixed effects regression models. Thou- of Marriage and Family, 73, 713–728. https://doi.org/10.
sand Oaks, CA: Sage. 1111/j.1741-3737.2011.00841.x
Amato, P. R., & Cheadle, J. (2005). The long reach of Dreby, J. (2007). Children and power in Mexican transna-
divorce: Divorce and child well-being across three gen- tional families. Journal of Marriage and Family, 69, 1050–
erations. Journal of Marriage and Family, 67, 191–206. 1064. https://doi.org/10.1111/j.1741-3737.2007.00430.x
https://doi.org/10.1111/j.0022-2445.2005.00014.x Dreby, J. (2010). Divided by borders: Mexican migrants and
Antman, F. M. (2012). Gender, educational attainment their children. Berkeley, CA: University of California
and the impact of parental migration on children left Press.
1784 Cebotari, Mazzucato, and Appiah
Eloundou-Enyegue, P. M., & Williams, L. B. (2006). Fam- Mazzucato, V., Cebotari, V., Veale, A., White, A., Grassi,
ily size and schooling in sub-Saharan African settings: M., & Vivet, J. (2015). International parental migration
A reexamination. Demography, 43, 25–52. https://doi. and the psychological well-being of children in Ghana,
org/10.1353/dem.2006.0002 Nigeria, and Angola. Social Science & Medicine, 132,
Gao, Y., Li, L. P., Kim, J. H., Congdon, N., Lau, J., & Grif- 215–224. https://doi.org/10.1016/j.socscimed.2014.10.
fithss, S. (2010). The impact of parental migration on 058
health status and health behaviours among left behind Mazzucato, V., & Schans, D. (2011). Transnational fami-
adolescent school children in China. BMC Public Health, lies and the well-being of children: Conceptual and
10, 1–10. https://doi.org/10.1186/1471-2458-10-56 methodological challenges. Journal of Marriage and Fam-
Gaydosh, L. (2015). Childhood risk of parental absence in ily, 73, 704–712. https://doi.org/10.1111/j.1741-3737.
Tanzania. Demography, 52, 1121–1146. https://doi.org/ 2011.00840.x
10.1007/s13524-015-0411-4 McKenzie, D. J. (2005). Measuring inequality with asset
Ghana Statistical Service, Ghana Health Service, and ICF indicators. Journal of Population Economics, 18, 229–260.
International. (2014). Ghana Demographic and Health Sur- https://doi.org/10.1007/s00148-005-0224-7
vey 2014 [Dataset]. Demographic and Health Surveys Nobles, J. (2011). Parenting from abroad: Migration, non-
Program and ICF International [Distributor]. Retrieved resident father involvement, and children’s education
from http://www.dhsprogram.com/ in Mexico. Journal of Marriage and Family, 73, 729–746.
Goldberg, R. E. (2013). Family instability and pathways https://doi.org/10.1111/j.1741-3737.2011.00842.x
to adulthood in Cape Town, South Africa. Population Parren~as, R. S. (2001). Mothering from a distance: Emo-
and Development Review, 39, 231–256. https://doi.org/ tions, gender, and inter-generational relationships in
10.1111/j.1728-4457.2013.00590.x Filipino transnational families. Feminist Studies, 27, 361–
Goody, E. N. (1982). Parenthood and social reproduction: 390. https://doi.org/10.2307/3178765
Fostering and occupational roles in West Africa. Cam- Parren~as, R. S. (2005). Children of global migration: Transna-
bridge, UK: Cambridge University Press. tional families and gendered woes. Stanford, CA: Stanford
Graham, E., & Jordan, L. P. (2011). Migrant parents and University Press.
the psychological well-being of left-behind children in Pfeiffer, L., & Taylor, E. (2008). Gender and the impacts
Southeast Asia. Journal of Marriage and Family, 73, 763– of international migration: Evidence from rural Mexico.
787. https://doi.org/10.1111/j.1741-3737.2011.00844.x In A. R. Morrison, M. Schiff, & M. Sjo€blom (Eds.), The
Heymann, J., Flores-Macias, F., Hayes, J. A., Kennedy, international migration of women (pp. 99–125). Washing-
M., Lahaie, C., & Earle, A. (2009). The impact of migra- ton, DC: Palgrave Macmillan.
tion on the well-being of transnational families: New Poeze, M., Dankyi, E. K., & Mazzucato, V. (2017). Navi-
data from sending communities in Mexico. Community, gating transnational childcare relationships: Migrant
Work & Family, 12, 91–103. https://doi.org/10.1080/ parents and their children’s caregivers in the origin
13668800802155704 country. Global Networks, 17, 111–129. https://doi.org/
Hu, F. (2013). Does migration benefit the schooling of 10.1111/glob.12135
children left behind? Evidence from rural northwest Radcliffe-Brown, A. R., & Forde, C. D. (Eds.). (1950). Afri-
China. Demographic Research, 29, 33–70. https://doi. can systems of kinship and marriage. London, UK: Oxford
org/10.4054/demres.2013.29.2 University Press.
Isiugo-Abanihe, U. (1985). Child fosterage in West Schmalzbauer, L. (2004). Searching for wages and moth-
Africa. Population and Development Review, 11, 53–73. ering from afar: The case of Honduran transnational
https://doi.org/10.2307/1973378 families. Journal of Marriage and Family, 66, 1317–1331.
Jordan, L. P., & Graham, E. (2012). Resilience and well- https://doi.org/10.1111/j.0022-2445.2004.00095.x
being among children of migrant parents in South-East Smith, A., Lalonde, R. N., & Johnson, S. (2004). Serial
Asia. Child Development, 83, 1672–1688. https://doi. migration and its implications for the parent–child rela-
org/10.1111/j.1467-8624.2012.01810.x tionship: A retrospective analysis of the experiences of
Kandel, W., & Kao, G. (2001). The impact of temporary the children of Caribbean immigrants. Cultural Diversity
labor migration on Mexican children’s educational aspi- and Ethnic Minority Psychology, 10, 107–122. https://d
rations and performance. International Migration Review, oi.org/10.1037/1099-9809.10.2.107
35, 1205–1231. https://doi.org/10.1111/j.1747-7379. Stark, O., & Bloom, D. E. (1985). The new economics of
2001.tb00058.x labour migration. American Economic Review, 75, 174–
Madhavan, S., Schatz, E., Clark, S., & Collinson, M. 178.
(2012). Child mobility, maternal status, and household Suarez-Orozco, C., Todorova, I. L., & Louie, J. (2002).
composition in rural South Africa. Demography, 49, 699– Making up for lost time: The experience of separation
718. https://doi.org/10.1007/s13524-011-0087-3 and reunification among immigrant families. Family
Mazzucato, V., & Cebotari, V. (2017). Psychological well- Process, 41, 625–643. https://doi.org/0.1111/j.1545-
being of Ghanaian children in transnational families. 5300.2002.00625.x
Population Space and Place, 23(3). Advance online publi- Twum-Baah, K. A. (2005). Volume and characteristics of
cation. https://doi.org/10.1002/psp.2004 international Ghanaian migration. In T. Manuh (Ed.),
Well-Being of Children in Transnational Families 1785
At home in the world? International migration and develop- Whitehead, A., Hashim, I. M., & Iversen, V. (2007). Child
ment in contemporary Ghana and West Africa (pp. 55–77). migration, child agency and intergenerational relations in
Accra, Ghana: Sub-Saharan Africa Press. Africa and South Asia. Brighton, UK: Development
Twum-Baah, K. A., Nabila, J. S., & Aryee, A. F. (1995). Research Centre on Migration, Globalisation and Pov-
Migration research study in Ghana. Accra, Ghana: Ghana erty. Retrieved from http://r4d.dfid.gov.uk/PDF/Out
Statistical Service. Retrieved from http://www.popline. puts/MigrationGlobPov/WP-T24.pdf
org/node/309536 World Bank. (2011). Migration and remittances factbook
Wen, M., & Lin, D. (2012). Child development in rural 2011. Washington, DC: Author. Retrieved from http://
China: Children left behind by their migrant parents siteresources.worldbank.org/INTLAC/Resources/Factb
and children of nonmigrant families. Child Development, ook2011-Ebook.pdf
83, 120–136. https://doi.org/10.1111/j.1467-8624.2011.
01698.x