1 Introduction

Caring for children is a task conducted by multiple actors. These can be parents, guardians, relatives, and siblings. In migrant families, in particular, such actors may not be located in the same household, nor the same country. Migrant families that have established themselves with different generations in a host country , may be spread throughout a city or country but still rely on forms of elderly or child care based on extended family systems characteristic of the countries where they come from (Schans 2009). In other configurations, migrant family members, nuclear or extended, may be located in different countries reflecting current migration systems where migrants from the Global South often leave their children or elderly parents in the home country in the care of others. Such families are composed of members in different nation states but they still function as a family unit where provision of care for children and elderly is organized transnationally. In both cases, it is useful to question how the different actors involved in these child care arrangements perceive care and who does the caring in order to understand how such arrangements function. This chapter focuses on a methodology used with transnational families, but the methodology can also be used more generally with migrant families consisting of various generations living in the same country.

Transnational family configurations have remained largely outside of the purview of sociological and demographic family studies. A recent body of literature termed transnational family studies has drawn attention to this important phenomenon showing how such families function and the stresses that living transnationally may entail for the actors involved (Bryceson and Vuorela 2002). Particular relationships such as those between migrant mothers and their ‘left-behind’ children have been favored in such studies. They point to the difficulties for parents and children who are separated by migration to maintain healthy and happy relationships. While caregivers’ roles in the home country have been mentioned as being particularly important in mediating the relationship between migrant parent and child (Dreby 2010; Parreñas 2005) they have not received specific attention, and migrant fathers are relatively a minor focus (Poeze and Mazzucato 2012).

A multiplicity of actors involved in care work for migrants’ children is also a phenomenon amongst migrant families that live together in one country. Migrant families may make use of extended family members who are in the same country, they may bring over a relative expressly to help them with day-to-day care, and finally others can be involved in the care of children such as teachers, social workers, or sports coaches. Finally, even when migrant families live together, their children are often more mobile than non-migrant families, spending extended periods of time in the home country without one or both of their biological parents. During these periods, caregivers in the home country provide the day-to-day care and children and youths can develop affective relations with them as well.

Yet while such networks of care are considered to be important for the functioning of a family (Dankyi 2011; Mazzucato and Schans 2011; Moran-Taylor 2008), seldom do studies focus on the composition of these care networks. They either focus on particular relationships or they assume such networks to exist, without studying their particularities, characteristics and how they function.

Often societies in the Global South, where many migrants to industrialized countries come from, are assumed to function on principals of the extended family, where norms such as child fostering and social parenthood are presumed to make it easy or normal to leave a child in the care of another person in order to be able to migrate (Olwig 2012; Poeze and Mazzucato 2013). Yet how easy is it to find someone to care for one’s child? And how does the resulting transnational child raising arrangement (TCRA) function? Who are the people involved and how do they perform care giving tasks? These were some of the research questions that the TCRA projectFootnote 1 set out to answer by conducting a care network analysis using a multi-sited and multi-actor research design.

In this chapter we describe the methodology used in order to map transnational networks of care and discuss the main insights that such a methodology helped to gain. Further analysis for what may explain the phenomena observed is beyond the scope of this chapter, such as the structural conditions faced by migrant parents overseas (documented status, access to income generating activities), or the pressures felt by caregivers and ambivalence by children stemming from the coexistence of nuclear and extended family norms. Here instead we focus on showing how a network-mapping tool allowed us to see that caregivers have much more restricted networks than is commonly expected, both in public discourse in Ghana, and in literature on African extended family systems. Furthermore, we noted discrepancies between networks as reported by migrant parents, caregivers and children, giving us insights into different notions of care and who is considered a carer. Furthermore, these discrepancies indicated sources of possible tension within TCRAs. We also discuss the importance of applying the care network mapping tool in conjunction with other qualitative methods such as in-depth interviews and show how each serves to augment the information that can be gotten from the other method.

In this study a caregiver is defined as the person or persons who is/are responsible for the day-to-day care of a migrants’ child while a carer as anyone providing any type of emotional, social and material care to the child, irrespective of their geographic location. In this study we deal with children and youths between the ages of 11 and 21 and for simplicity we use the terms children and youths interchangeably.

2 Background and Methods

Ghana is a good case in which to study migrant families. There are roughly 1.5 million Ghanaians living abroad (Twum-Baah 2005). The main Western destinations for Ghanaian migrants are the US and Canada, UK, Germany, Italy, and the Netherlands. The majority of migrants to the West have historically been Akan, because they benefited from the cocoa boom of the 1970s; however, recent data show that other ethnicities from central and southern Ghana are also migrating to the West (Nimako 2000; Twum-Baa et al. 1995). This study focuses on Akan and to a lesser extent Ga-Adangbe migrants, reflecting the majority of ethnic groups in the Netherlands, and their families back home. The Akan are a matrilineal ethnic group while the Ga-Adangbe are patrilineal.

It is common for children in Ghana to be raised by someone other than their biological parent, a practice called child fostering, or by only one of their biological parents (Goody 1982; Mensa-Bonsu and Dowuona-Hammond 1996; Nukunya 2003). Ghana’s Demographic and Health Survey indicates that nationally there are 15.6% of children under the age of 15 who are fostered, not including orphans (GSS, GHS, & ICF Macro 2009). Additionally, 24.2% are living with only one of their parents, while the other parent has migrated nationally or internationally. These figures do not indicate exactly what percentage of children lives in a transnational family formation because child fostering does not only occur in the context of international migration and living with one parent is also a consequence of national migration. A recent survey of Ghanaian junior and senior secondary school children in urban and semi-urban centers in southern Ghana, where most international migrants come from, indicated that 15.5% of children had one or both parents who migrated internationally (Mazzucato and Cebotari 2012). Together these figures can be taken to indicate that transnational family life for Ghanaian children is not an exception.

This study focuses particularly on families with at least one migrant parent in the Netherlands. In 2014, there were approximately 22,500 officially registered, first and second-generation Ghanaian migrants in the Netherlands, the second largest group from sub-Saharan Africa (CBS 2015); unofficial estimates attest to a large undocumented population (Mazzucato 2008). In general, migrants from Ghana are both highly skilled and unskilled; the Netherlands tends to attract a majority of unskilled migrants, since the highly skilled prefer Anglophone countries such as the U.S., Canada and the U.K. (Carrington and Detragiache 1998).

The TCRA project uses a mixed method and multi-sited research design, incorporating large scale surveys in Ghana and the Netherlands, interviews with 19 caregivers in Ghana and 41 migrant parents in the Netherlands in-depth, and ethnographic work conducted on 15 ‘matched’ TCRAs, in which migrant parents in the Netherlands were studied at the same time as their children’s caregivers in Ghana. This paper draws on data collected from the matched sample. The selection for the matched sample started in the Netherlands by selecting migrants who participated in the survey, met particular criteria (we searched for diversity in sex, length of stay in the Netherlands, occupational and documented status), and had given consent for follow-up research. Second, we went through different gateways (churches, key informants in the Ghanaian community in Amsterdam, chance encounters, and previous contacts with migrants of one of the researchers). If both parents migrated to the Netherlands, one spouse was randomly selected. The ethnographic work included various methods including observation, in-depth interviews focusing on particular moments in a person’s life, care history and migration history , informal visits, phone calls, and care network mapping. In this chapter we focus on the care network mapping tool .

Care networks were elicited using a name-generator questionnaire conducted among 15 migrant parents, 15 of their children in Ghana and 15 of the children’s primary day-to-day caregivers. Only children between the ages of 11 and 21 were selected and if a migrant had more children in Ghana in that age group, one child was selected based on sex and age to make sure we had a variety of respondents. The same tool was used to collect information from all three types of actors. In this way, multi-actor and multi-sited data were collected.

A name-generator questionnaire is a tool used in quantitative social network analysis (Burt 1984; Campbell and Lee 1991). In this study, an exchange approach to networks was used in which questions were asked with respect to the exchange of emotional, social and material supportive content between the children and alters (McAllister and Fischer 1978). Material care was asked about through questions such as who provides for school fees; food and cooking; pays for medical bills, etc. Emotional care was asked through questions such as who helps the child when he/she is feeling down; and socially with questions such as who educates the child on how to behave; gives advice to the child; who does the child spend free time with. Positive relations (such as friendships) as well as negative relations (such as people one argues with) and strong and weak ties (Granovetter 1973) relating to kinship, consanguineous and non-kinship relationships were asked. Respondents were asked to mention as many names as they could think of for each question and along with the names, also the location of the person, the relationship with the child (kin/non-kin and for each, specifying what kind of relationship, such as ‘business partner’ or ‘mother’s sister’) and the frequency and means of contact. The tool was tested for cultural relevance of questions and saturation.Footnote 2

The tool was used to elicit the information about care alters, or carers, but also as a means to collect additional information about how care for the child is practiced, such as who makes decisions with regards to particular aspects of the child’s life (e.g. education, health, discipline), who advises on the child’s upbringing, who one argues most regarding care for the child or who the child argues with most. Furthermore, follow-up in-depth interviews were conducted with each of the respondents and the network tool was used as a reference, either to check information, or to ask further about the role of specific people in the child’s care network. In-depth interviews were important in order to contextualize the information obtained from the network tool and to complete any information that may not have been elicited through the network tool. Likewise, the results of the network tool was useful to have during the interviews to jog people’s memories about particular carers and to go more in detail about particulars of these carers. The findings that follow are based on the combination of the two methods: the network tool and the in-depth interviews, focusing specifically on illustrating what a multi-actor and multi-sited data collection methodology helped us to find.

This tool was used as part of a simultaneous matched-sample (SMS) methodology that has been developed in order to study phenomena occurring in multiple sites simultaneously (Mazzucato 2009). This involved working in a team of researchers located in the different sites at the same time so as to be able to capture transactions and exchanges that affect people’s relationships in a transnational setting. Given the often high frequency such as communication via phone calls and the sometimes small nature of such transactions such as small amounts of money sent, an SMS methodology allows capturing these every-day exchanges which methodologies based on recall cannot obtain. Researchers collected information and then shared it with each other in order to inform the questioning that they were conducting with various methodological tools. In this particular case of the network mapping tool , it meant that researchers collected and exchanged the network maps of each of their respondents so that they could observe similarities and differences between such networks. This helped deepen the interview sessions that followed by asking more detailed questions about specific network members so as to understand certain differences in maps. This was done with discretion and attention to not divulge information from one respondent to another respondent in order to protect privacy.

3 Findings

Before presenting the findings, we give some general characteristics of the mapping exercise that was conducted with the members of each of the 15 TCRAs. This is not a representative sample and therefore not made for generalization. Rather it is presented here in order to contextualize the three detailed cases that we present below (Table 14.1).

Table 14.1 Care network characteristics of 15 TCRA cases (TCRA fieldwork 2010–2012 in Ghana and the Netherlands)
Full size table

3.1 Perceptions of Care Within a Transnational Child Raising Arrangement

Mapping networks of care with the different actors involved in a TCRA allowed seeing whether differences exist in the perceptions of each actor as to who provides what kind of care to the migrant’s child. Following up with in-depth interviews with each actor enabled understanding what these discrepancies consisted of and as such allowed us to see what aspects of care were considered important for each actor. In general, we found that caregivers in Ghana perceived to have smaller care networks than migrant parents and their children perceived these networks to be. Caregivers tended to focus on people who provided practical help and material care. Practical help related to people who could relieve them of some of their tasks. For example, people who they could count on to do some of the day-to-day caring or people they could leave the child with if they needed to go somewhere. Material care related to people who helped them to pay for the expenses of the child’s upbringing such as food, clothing, school fees, and health bills.

In some cases, there were some stark contrasts with the care networks as reported by migrant parents, especially migrant fathers, who tended to give more names than the caregivers. At first, we thought that this might indicate that fathers were not very attuned to the practical realities of care. Yet upon further questioning we found that for these fathers, the additional names tended to be friends or people in the extended family whom they asked to check on the child, just to see how they are doing, in order to get information about their child also from another source, and not just the caregiver . Some of these additional people also gave advice to the child, such as which school to go to, how to behave in school, to focus on school work. And indeed, some were asked to give help to the child with regards to school work, especially when caregivers were illiterate or semi-literate women such as grandmothers or aunts who may not be able to help children with their school work. These forms of care were all ways in which fathers maintained some control over the child’s upbringing, practicing their fathering from afar. While both migrant mothers and fathers are concerned with their children’s education and both send remittances to pay for school fees, our interviews revealed that fathers also considered disciplining and advising as their primary care tasks especially when children are older (Poeze and Mazzucato 2012). This reflects also practices in Ghana where disciplining of older children is seen as a task for a father or male figure in the extended family (Twum-Danso 2009).

Figure 14.1 shows the care network as given by Kofi (migrant father), Agyeiwaa (caregiver and migrant’s mother-in-law) and Osei (migrant’s son) who was 19 years old at the time of fieldwork. Each person in the network has a color according to by whom they were mentioned. Four people were mentioned by the caregiver. The father mentioned nine people. In particular, the five people mentioned only by the father are friends, father’s extended kin and a teacher.

Fig. 14.1
figure 1

Solomon’s care network as reported by multiple actors (TCRA fieldwork 2010–2012 in Ghana and the Netherlands)

Full size image

We therefore investigated further this discrepancy between Kofi’s perception and Agyeiwaa’s perception of the care network.

Kofi is well educated and earns a living in the Netherlands. Although physically separated from his two children, he considers himself and his wife in London as providing the emotional, social and material care that his children in Ghana need. He accomplishes this from a distance as well as through frequent trips to Ghana. Both he and his wife send regular remittances from which school fees, school supplies and health expenses are paid. The children send their report cards by mail. Kofi decides which schools his children are to attend and makes sure that Osei attends extra-curricular classes. He calls his children twice a day and with his mother-in-law once a week. Kofi mentions a large network of friends and family who assist him in caring for the children, none of whom are mentioned by Agyeiwaa, the caregiver . Kofi’s father provides some material care and advises the children. A friend of Kofi’s who is a teacher visits the children on Sundays and makes sure that their studies are going well. Whenever Osei tells his father that he has a problem with a particular subject in school, Kofi calls his teacher friend to ask him to speak with Osei and help him solve his problem. It is this friend who attends the parent-teacher meetings at Osei’s school. Kofi also mentions three friends and an uncle who sometimes talk with Osei and give him advice on whatever issue necessary.

Agyeiwaa, on the other hand, mentioned the housemaid with whom she would leave Osei when he was younger when she needed to go to her shop where she sold wax cloth at the market. She also mentioned her daughter who helped her care for Osei. She recounts that Osei used to be ill quite often as a young child and she would always be the one to care for him, not allowing anyone else close to him, as she felt that he caught infections easily.

This example shows how Agyeiwaa considers care as the daily tasks that need to be done to care for Osei and in this respect only mentions people who help her with such tasks while Kofi’s main concern is his son’s education: that he performs well in school and behaves appropriately. He aims to have an active role in the parenting of his child, even from afar, and therefore has many people, friends and extended family to help him in this, both in terms of tasks they perform (helping with homework, advising) as well as in their checking on Osei in order to get information on Osei from various sources.

3.2 Understanding How Transnational Child Care Functions

Discrepancies between how different actors reported on a child’s care network not only give insight into people’s perceptions of care but also on how transnational child raising arrangements function. While discrepancies in care networks are not enough to understand the functioning of TCRAs, they can provide some key insights to gain a better understanding on their functioning. We illustrate this with the case of Joyce (migrant mother), Cecilia (migrant’s maternal aunt) and Ama (migrant’s daughter) who was 17 years old at the time of fieldwork. For each network we grouped alters according to the type of care they were mentioned as providing: emotional, social and material care.

In Table 14.2 we see that there are hardly any discrepancies between the care network as reported by the migrant mother and the caregiver . Yet when separating alters by the kind of care they provide there was a surprising omission in the alters mentioned by Joyce as providing material care: the caregiver’s name is missing whereas it is present in the caregiver’s network.

Table 14.2 Ama’s care network as reported by multiple actors, by type of care (TCRA fieldwork 2010–2012 in Ghana and the Netherlands)
Full size table

When investigating this through in-depth interviews on both sides we found the following:

Joyce left her 5 children with Cecilia, her mother’s sister, when she left to go to the Netherlands. Cecilia reluctantly accepted the children and only on the condition that Joyce would bring them to the Netherlands within 6 months of her departure. Now, three years later, Cecilia is at the limits of her material and non-material resources: she cannot work because she needs to care for Joyce’s children. Joyce who is documented in the Netherlands but does not have a steady job sends remittances irregularly and too little, according to Cecilia, to cover the costs of living for the children. Cecilia needs to contribute her own resources to make ends meet, which she does with great difficulty given her already meager resources and limited ability to earn an income. Cecilia does not inform Joyce about the expenses because she does not keep track of them as they concern small every-day expenditures and at the same time, she is under the same social pressure as many caregivers: that of not seeming to only be doing it for the money.

Cecilia is caught in a bind: she needs to care for five children without having the resources to do so and she does not receive enough remittances from the migrant to help her in this task. Joyce is frustrated with Cecilia’s constant reminders take her children to the Netherlands, feeling that Cecilia does not understand the hardship she is going through in the Netherlands. This lack of comprehension between the two women is augmented by the fact that Cecilia does not inform the migrant of the material resources that she puts into the care of the children whereby Joyce does not seem to fully realize the difficulties she is under and the fact that her remittances are not enough. This small, yet very significant difference in the two women’s care networks helped to identify a discrepancy in how they perceive who is providing the material care for Ama and her four siblings and allowed us to investigate this more deeply, getting at some of the aspects explaining why this TCRA seems to be fraught with incomprehension and tensions on both sides.

3.3 Giving Children and Youth a Voice

Much research on children relies on accounts of adults regarding their health, educational performance and emotional well-being, such as teachers, caregivers and parents (Mazzucato 2013; Suarez-Orozco et al. 2002). Yet children are active agents, shaping and interpreting their own experiences. They too have ideas about their care and who provides them with such care. By conducting the network mapping tool from a multi-actor perspective, we found aspects of children’s perceptions of care that had not been revealed by questioning adults. Children tended to mention other children, in the same household or from school, from whom they drew especially emotional care, i.e. whom they spent time with and who was important to them.

A clear example is Abigail’s care network (Fig. 14.2). Even though the children she mentions are living in the same household as she and are well known to her caregiver and migrant parents, she is the only one who mentions them. Abigail mentions her cousins from whom she draws emotional support, i.e. spends her free time with. For example, she says about her step-cousin who is 2 years her elder, “Oliver is my most favorite but he is the one I fight with most”. She has developed a close relationship with her youngest cousin who also lives with her. When Abigail’s parents came for a visit from the Netherlands and Abigail went to live with them, she brought her young cousin with her.

Fig. 14.2
figure 2

Abigail’s care network from a multi-actor perspective (TCRA fieldwork 2010–2012 in Ghana and the Netherlands)

Full size image

4 Some Gaps in Care Network Mapping

There are some gaps in the mapping tool , not particular to care networks, but more generally to network mapping that we encountered during our study, making it all the more important to complement network mapping with in-depth interviews. This pertained to the presence of important incidental care, which turned out to be fundamental in certain TCRAs for ensuring the livelihoods of children and their caregivers. We term incidental care, that care which comes from an unexpected person, such as an acquaintance or even someone whom the child nor the caregiver know directly. Furthermore, in incidental care, the care provided is neither continuous nor recurring but usually a once-off help, often in the form of material care that is received at a moment in time when such care makes a real difference in assuring the livelihoods of the child and caregiver. Such care is important to take into account and to understand how it works, why the incidental carer feels compelled to provide care at that moment, in order to understand how care networks function. Systematic network data collection tools such as the one used, tend to favor stable, frequent or continuous care relationships thus easily overlooking incidental carers.

Naki’s network is presented here to show the importance of incidental care. This information was revealed during in-depth interviewing. In particular, this TCRA experienced three important moments of care provided by incidental helpers. Here we present one of the instances:

“I remember a particular day when my boys and I had to eat gari for supper (author: starchy food made from cassava. It is usually eaten together with protein foods). I had no money on me and all I had was gari and palm oil. I soaked the gari until it was very soft and then I mixed it with the palm oil and we ate together. I was even thankful for that meal because I didn’t know where our next meal for the next day will come from. Then something miraculous happened the next day. I received a visitor in the morning, he happened to be the driver of the then National Chairman of my church. The driver said that the chairman had sent for us – I and the children. You can imagine the shock on my face. I had no idea why he sent for us but I knew we might be fed there at the least. When we got to the house, he and his wife welcomed us warmly and there was a buffet table set for us. There was so much food that I could not eat. I was both happy and sad at the same time. After eating he asked the kids to go out and play and he told me that they had met my husband in Holland a week earlier and he told them about us. He even brought pictures that they took with him for us (that’s the latest pictures we have of him). He also told me how kind my husband had been to him when he came to Ghana after the war in Liberia. He said my husband was the first person to meet him at the gate of the headquarters. He gave him a big hug and told him how happy they are that he is safe in Ghana. He also said that he believed that God spared his life for a purpose and that God will make him a big person in the future. His words were so kind and they became friends from there. The Chairman added that if he had known about my husband’s intention to travel he would never have encouraged him to go. They promised him to look out for us when they got back and that was why they invited us over. We had a nice time and later in the day, the driver was asked to drop us off at home. We were sent back with a car full of foodstuff and money. He did that for us from time to time until his tenure as chairman was over and he relocated. That source was cut from then”. (Naki, Accra, January 2012)

Interestingly, neither the caregiver nor the migrant husband in the Netherlands mentioned the Chairman in their network mapping, yet according to the caregiver this help came at a time when she did not have anything but starch to feed the children. We know from our multi-sited fieldwork that the husband is undocumented in the Netherlands and faces various hardships. During the first 3 years of his migration to the Netherlands, this often led him not to remit any money for long periods of time which was when Naki received this incidental and essential help.

Although we cannot say anything about the prevalence of incidental helpers, two of the fifteen TCRAs were characterized by having received essential help from incidental helpers. The case of the incidental helpers draws attention to the fact that it is inadequate to use the network mapping tool alone to map out care networks of migrants’ children. Rather such a tool needs to be used together with other instruments such as in-depth interviews. The network mapping tool, however, proved to be a very useful exercise since it paved the way for lengthier and more meaningful discussions around the names and circumstances that were generated by the tool and later also for asking about names that came up during in-depth interviews as to why these names were not mentioned in the tool.

5 Conclusion

Networks are considered important in migration studies, yet often they are assumed to exist and little work specifically focuses on mapping out networks: what relationships they are made of, what kinds of support are received and how networks evolve over time. One of the aims of the TCRA project is to specifically study networks and their role in the provision of care in a transnational context, rather than assume their existence. To this aim, we developed a care network mapping tool which was applied to each of the main respondents in transnational child raising arrangements: Ghanaian migrants in the Netherlands, and migrants’ children and their caregivers in Ghana. By mapping networks in a multi-sited and multi-actor framework some important insights were gained into perceptions of care and how networks function. Furthermore, coupling the network mapping tool to in-depth interviews allowed more detailed questioning during the interviews that provided important additional information on how care is provided and how TCRAs function. Here we presented four particular insights that can be gained from the combination of care network mapping and in-depth interviews.

A first insight concerns the fact that care is conducted by multiple actors. This is all the more so in a transnational setting where parents migrate to ensure a better future for their children and leave their children in the care of others, back home, to do the day-to-day care work. We therefore translated this to our methodology by conducting care network mapping with the different actors who provide care. This resulted in insights into the different perceptions that multiple carers have about what care is and who provides what care. In our sample, migrant fathers tended to report larger care networks because they included people whom they asked to check up on their children or who provided help with schooling and advice to their children. Such help was either not recognized or not deemed important enough by caregivers who did not report such help but rather focused on those people who could help them with the day-to-day tasks of providing care. This reflects the often time-constrained schedules of caregivers who, due to the care they need to provide, find it difficult to continue with other income-generating activities that they may have been engaged in before taking on the care of a migrant’s child. They therefore attach high importance to people who can relieve them of some of their care tasks. This resulted in the insight that caregiver networks seem to be more constrained than is often implied in literature on child fostering and extended family systems in migrants’ home countries. Although not mentioned explicitly, literature on migration from societies where extended family systems prevail, conjure the image of a society where it is relatively easy to leave one’s child with a relative in order to be able to migrate. Our findings rather, reflect literature on changing family life and child care in West African societies, which emphasizes the growing importance of the nuclear family ideal and the declining support of the extended family system for child care, especially in urban centers (Ardayfio-Schandorf and Amissah 1996; Oppong 1974; Van Dijk 2002). Such findings have implications for social policies addressing the interconnection between family, care and migration at both ends of the transnational migration space.

A second insight relates to the discrepancies between different actors’ network maps within the same TCRA, which helped to better understand how particular TCRAs function. Differences in the perceptions of carers’ specific role in care networks highlighted points of miscommunication and tensions within a TCRA. By investigating such discrepancies further through in-depth interviews it was possible to understand some of the sources of miscommunication and tensions.

A third aspect of conducting a care network mapping tool through a multi-actor perspective, is that it allowed us to give voice to children and youth’s perceptions. The network mapping tool was easily understood by young respondents between the ages of 11 and 21. In fact it also provided a means through which to start interviewing young actors who at times can be difficult to interview due to shyness or, as is the case in Ghana, in cultures where children are not expected to express opinions in conversation with adults. Conducting network mapping with children and young adults revealed their perceptions of care as including other youths who they looked up to, who gave them advice and who they could spend leisure time with. Never were these youths reported in the networks of adults. While peers are generally found to be important for children’s well-being, these relationships have not been subject to specific investigation in a transnational context where children are living far from their parents, making the emotional and social support they receive from their peers especially important for their well-being. These findings are potentially of relevance for the literature on transnational care and intergenerational relationships and influences, pointing out at the importance of peer influences.

Finally, the care network mapping tool was useful when used in conjunction with in-depth interviews as it helped obtain more specific information, for example, about different carers in a network and also by being able to focus on discrepancies between respondents’ networks maps within the same TCRA. At the same time care network mapping, alone, was not enough to capture an important form of help: incidental care. This type of care was particularly fundamental in ensuring the livelihoods of children and caregivers in certain TCRAs. It is also an important type of care that results in societies where social support is given through generalized reciprocity (Sahlins 1972) where support given by one person can result in other people receiving help in some future moment, in reciprocation for this initial support action. Such help is often overlooked by systematic network data collection as this form of data collection favors continuous and frequent relationships whereas incidental care, by its very nature, is ad hoc and infrequent. As such it is important to conduct care network mapping in conjunction with another qualitative data collection tool such as in-depth interviews.