Transnational migration, health and well-being: Nigerian parents in Ireland and the Netherlands

Abstract

The phenomenon of families separated across continents is a result of migratory flows in a globalised world. Transnational families occur because one or both parents migrate internationally requiring children to be raised in transnational child-raising arrangements, with the help of caregivers. This study examines the health and the emotional well-being of Nigerian migrant parents living in Ireland and the Netherlands, using comparative analyses based on a survey of close to 300 migrant parents in each host country. Half of the sample in each country is living in transnational families the other half are not. This paper adds to the existing literature on transnational families by including control groups (migrants who are not separated from their children) and comparing migrant parents from the same origin country who live in different host countries, allowing us to identify the significance of migratory context and legal regimes in shaping the emotional well-being and health of parents.

The results indicate that the factors that drive the health and emotional well-being of migrant parents are not solely related to their separation from their children but rather to other mediating variables such as legal status, socio-economic status, and the normative contexts. While Nigerian child fostering norms ease the influence of separation in both contexts, separate analyses of the Irish and the Netherlands sample show the more pronounced consequences of the mediating factors in the Irish sample, highlighting the differences in the migratory trajectories of Nigerian parents in Ireland and the Netherlands.

Introduction

The phenomenon of families separated across continents is a result of migratory flows in a globalised world. Non-state agencies and Non-Governmental Organisations in migrant sending regions in the Global South estimate that approximately one quarter of all children live with at least one parent living abroad, indicating the scale of this phenomenon (Mazzucato & Schans, 2011).

Transnational families may be defined as families who live apart but retain a sense of collective welfare and identity across national borders (Brycesson & Vuorela, 2002). In this paper the term is used to describe a family where one or both parents migrate internationally requiring children to be raised in transnational child-raising arrangements,¹ with the help of caregivers. A distinct focus of studies on transnational families concentrates on the emotional and educational impacts of transnational separation on children ‘left behind’ in migrant sending communities (see Battistella & Conaco, 1998; Dreby, 2007; Gindling & Poggio, 2010; Graham et al., 2012; Hondagneu-Sotelo & Avila, 1997; Kandel & Kao, 2001; Lahaie, Hayes, Markham-Piper, & Heymann, 2009; Poeze & Mazzucato, 2014; Schmalzbauer, 2005; SOROS Foundation, 2007; Suárez-Orozco & Todorova, 2008; Suárez-Orozco, Todorova, & Louie, 2002). A related but smaller body of literature explores the impacts of family separations on migrant parents, and especially on migrant mothers, in migrant receiving societies (Boccagni, 2012; Coe, 2011; Horton, 2009; Laurie, 2007; Schmalzbauer, 2004; Suárez-Orozco & Bernhard, 2008).

This paper provides a comparative analyses of the emotional well-being and health of Nigerian migrant parents living in Ireland and the Netherlands based on a survey of close to 300 migrant parents in each host country. Half of the sample in each country are parents with at least one child in Nigeria (or ‘transnational families’) and half are migrant parents, who live with their children in the host country and therefore are not living in a transnational arrangement (or ‘non-transnational migrant families’). This paper aims to contribute to the literature in a number of important ways. Firstly, the analysis compares parents in Nigerian transnational families with Nigerian parents who live with their children (i.e. non-transnational migrant families) to allow distinguishing outcomes that might be specific to transnational families from those that may be relevant to migrants generally. Secondly, by examining migrants from the same sending country (Nigeria) in two different European countries (Ireland and the Netherlands), the analysis in this study can offer important insights into how historical, political or cultural factors in different host country contexts may play an important role in the specific effects of living in transnational families. Finally, the analysis focuses on African migrant parents in Europe whereas the literature to date has focused predominantly on Asian and South American migrants in the US or European migrants in Europe. Of interest here are West African traditions and practices of child raising which may play a role in shaping the emotional impact on migrant parents of separation from one’s children. Practices of child fosterage, defined as children living away from their biological parents (Oni, 1995) or as raising another’s child as one’s own but without severing the bonds to biological parents (Renne, 1996), is an established practice in Nigeria. The preference is for a close kinship relationship to exist between a foster carer and child such as a biological grandparent or maternal or paternal aunt (Renne, 1993). Child fosterage is not an indicator of parental inferiority as the child is expected to return to the biological family for days or years; rather fosterage is within normative kinship obligations and is often used to support social mobility of the child and his or her family (Okunola & Ikumola, 2010). There are some suggestions within the literature that traditional practices of fosterage such that the maxim ‘every mother regards the child of her sister as her own child’ are changing; and that in present day Nigeria, biological children may be differentiated from fostered children by the school they attend or the domestic activities they do in their fostered parents’ home (Naidu, 1982, cited in Okunola and Ikumola (2010)). However as remittances from transnational migration are often used to pay left-behind children’s school fees and upkeep (Poeze, Dankyi, & Mazzucato, 2017), this may not be the case in transnational families. Okojie (2009) has suggested that the networks of solidarity used to foster children have in some cases degenerated into financial transactions and so socio-economic status of transnational parents may significantly influence transnational child-raising. This paper aims to examine if being a transnational parent has any effect on the health and emotional well-being of migrant parents. Health was measured using a self-assessed health measure. Emotional well-being was measured using a self-assessed life satisfaction and the General Health Questionnaire (GHQ-12).

Theoretical framework

Although the extent to which being a transnational parent affects migrant parents’ general well-being is unclear, there are substantial small scale transnational family studies which highlight the negative emotional and health consequences of parent-child separation due to international migration. There are also more recent large scale transnational family studies which quantify the well-being consequences of migration induced separation on children (see Mazzucato et al., 2015) and on parents (Haagsman, Mazzucato, & Dito, 2015). This paper aims to add to these quantitative studies of the impact of separation on migrant parents.

Evidences of emotional costs for migrant parents

More broadly, the evidence from the small scale transnational family literature from Latin America and East Asia emphasizes the negative emotional experiences of migrant parents, especially its gendered dimension. These are documented as early as the 80s for sub-Saharan African transnational mothers living in France who experienced negative health effects by having a child abroad (Afulani, Torres, Sudhinaraset, & Asunka, 2016). More specifically, the literature from Latin America and South East Asia shows transnational mothers feelings’ of ambivalence about mothering from afar, with mothers experiencing migration as a form of self-sacrifice (see Boccagni, 2012) while exposing them to a higher risk of depression (Suárez-Orozco & Bernhard, 2008). Other studies reported transnational mothers’ experiences of stigma, negative sentiment and growing resistances to their migration, particularly in the context of increased demands on caregivers, especially from female family members who take on caregiving roles (Parrenas, 2010).

A strain on mother-child attachment as noted by Schen (2005) and perceived abandonment by left behind children are reported to be some of the reasons why transnational mothers feel anxious, depressed, desperate, and experience ill health (e.g. Coe, 2011; Laurie, 2007). These are inextricably linked to the social and cultural norms that may impact upon transnational parents who are separated from their children. For example, the literature indicates that one of the most significant challenges transnational parents face is around child-raising arrangement for their younger left behind children (Carling, Menjívar, & Schmalzbauer, 2012). While cultural norms dictate family members’ obligations to support these children, they simultaneously evoke negative attitudes on the absent parent. For example, Hondagneu-Sotelo and Avila (1997) show how initiating separations from communities, families, and (sometimes) spouses results in negative emotions due to feelings of guilt, shame as well as criticism from others. Pustułka (2012) notes similar findings of guilt about being ‘bad mothers’ for Polish transnational migrant mothers in European migrant receiving societies.

Given the focus of these small scale, ethnographic studies on mothers, the emotional cost of transnational parenting is mostly focused on this group, implying that fathers may not have similar experiences. Studies have shown transnational fathers point to the importance of providing for families through remittances and gifts and that this contributes to their sense of themselves as ‘good fathers’ (see Dreby, 2006; Pribilsky, 2004). This highlights how traditional gender roles & norms might shape emotions and parenting practices in transnational families (Carling et al., 2012). In light of this, some studies indicate that while there are similarities in parenting practices of transnational mothers and fathers, coming to terms with distance and separation may be emotionally easier for fathers than it is for mothers (Avila, 2008; Laurie, 2007). Ryan, Sales, Tilki, and Siara (2009) point to migration resulting in a double of caring responsibilities for female migrants as they must care for family members (including in some cases their children) in both emigration and immigration contexts and countries.

While research on transnational parenting has dramatically increased over the last two decades for most of this period transnational parenting was synonymous with transnational motherhood. Souralová and Fialová (2017) argue that explorations of the interaction between gender roles and parenting assumed these were central to the experiences of female migrants but were largely of marginal importance to male migrants. The insertion of fathers into the literature and research on transnational parenting has reproduced gendered stereotypes by focusing on fathers as breadwinners and as remote, distant authoritarian figure (Souralová and Fialová, 2017). However, certain (again small scale qualitative) studies have focused on transnational fathers and explored fathers’ experiences of severe loneliness, depression and abuse of alcohol arising as a result of their separation from children (Schmalzbauer, 2005). Schmalzbauer (2015, p. 214) has developed this perspective arguing that the picture that emerges of the emotional costs of transnational fathering (compared to mothering) is “less clear because fatherhood tends to be constructed around provision and authority, there is no cultural script for how fathers should maintain an emotional connection with children in the context of family separation”.

Potential factors for emotional costs

The small scale transnational family literature paints a picture of guilt ridden, depressed, and anxious migrant parents, particularly in the case of the migrant mothers. However, few of these small scale studies systematically investigated whether these experiences are distinctly associated with only separating from children or are as a result of some other mediating factors. As Carling et al. (2012) argue the care arrangements in transnational families entail the intersection of material and emotional concerns, thus, it is important to disentangle the different factors that contribute to the negative emotional experiences of these parents.

Generally, research has stressed that migrants are more likely to face additional health and emotional difficulties to wellbeing than native-born populations because of the social and economic problems they face due to e.g. discrimination, social isolation, and lack of legal documents as migrants (see Avila, 2008; Bernhard, Landholt, & Goldring, 2005; Jolivet et al., 2012). As such, at the heart of the frustration felt by transnational parents could be their class position in the host country as migrants, in particular if they are undocumented or are in a state of continuous insecure legal status (Carling et al., 2012; Leifsen & Tymczuk, 2012). Bernhard et al. (2005) found that being separated from children was not associated with elevated health risks for transnational parents; however perceived discrimination in the receiving country was associated with decreased emotional well-being for this group. These perceived discriminations could be due to immigrants limited integration in the labor market of the destination country (Teixeira & Dias, 2018).

Other evidence indicates that undocumented migrants and their families have limited access to quality health care (Boccagni, 2015; Devillé et al., 2011) illuminating one of the dimensions of how migrants’ well-being could be affected in the context of international migration. Thus legal status appears to be an important mediating variable in shaping the health of transnational parents as well. Similarly, feelings of sacrifices reported by transnational mothers in some studies can be linked to how they are incorporated differently to the destination country’s labour market, a result due to an undocumented status as found by Abrego (2009) for El Savadorian transnational mothers in the US. Abrego argues that these mothers are exposed to vulnerable working conditions with longer hours of work so that they consistently remit a large proportion of their small income, a sacrifice facilitated by the gendered expectations of El Savadorian motherhood. Recent qualitative study also confirms these consequences of illegality on the working conditions of Latina migrant mothers in the US (Abrego & Schmalzbauer, 2018). The frustration and the anxiety felt by transnational parents could be due to their indefinite separation from their children, making the family reunification project intractable (Schmalzbauer, 2004; Fresnoza-Flot, 2009; González-Ferrer, Baizán, & Beauchemin, 2012). In contrast research on transnational care and migration among Polish migrant communities in Europe suggest that the numbers of Polish children living apart from migrant parents has declined in recent years as families avail of the right to move freely across the EU (see White, 2016).

Thus, what complicates the health and emotional wellbeing concerns in transnational caregiving arrangements are the socioeconomic status and the documented status that invariably shapes parents’ propensity to remit (Held, 2017); to reunify (Menjívar, 2006); and to communicate regularly (Leifsen & Tymczuk, 2012; Poeze et al., 2017). In the latter case, Peng and Wong (2013) found that the emotional costs of separation from children may be mitigated by transnational mothers’ use of ICT to maintain meaningful relationships with children. Many studies from different regions indicate the important meaning attached to remittances, in which the remitter feels and is seen by family members back home as fulfilling a fundamental aspect of family obligations (e.g. Hall, Garabiles, & Latkin, 2019; Kelly & Lusis, 2006; Lahaie et al., 2009; Wong, 2009).

These are all important elements that contribute to a trustful and easy relationship in transnational care arrangements (Dankyi, Mazzucato, & Manuh, 2017) and potentially contributing to a better wellbeing for all actors involved. More recent quantitative studies confirm the important role documented status has on transnational parents’ well-being (Dito, Mazzucato, & Schans, 2017; Haagsman et al., 2015).

Children’s and migrant parent relationship, life satisfaction and emotional wellbeing

A recent large scale study on African transnational families in the Netherlands shows the positive correlation between parent-child relationships and higher life satisfaction (Haagsman et al., 2015). This is reflective of the findings from the small scale studies which documented that transnational child care complicates the relationship that transnational parents have with their children in origin countries (Aguilera-Guzmán, de Snyder, Romero, & Medina-Mora, 2004; Coe, 2008; Dreby, 2006).

In order to address the question of whether being separated from children has a negative impact (or not) on a transnational parents’ health and emotional well-being researchers must disentangle effects associated with the characteristics of the transnational parent such as education status, income generating opportunities, lack of documentation a with separate living arrangements as a member of a transnational family.

Across a range of studies, transnational parents report feelings of guilt, loss and loneliness due to their separation from their children (see Boccagni, 2012; Coe, 2011; Hondagneu-Sotelo & Avila, 1997; Laurie, 2007; Parrenas, 2010; Pustułka, 2012; Suárez-Orozco & Bernhard, 2008). In other studies parents’ feelings are located within culturally and contextually specific relations and factors that appear to mitigate these emotional and health-related impacts (see Dankyi et al., 2017; Leifsen & Tymczuk, 2012; Peng & Wong, 2013; Poeze et al., 2017).

Many researchers report that gender plays a key role in shaping the impact of how separation from children influences transnational parents, however these impacts are interpreted in contradictory ways (for example see Avila, 2008; Carling et al., 2012; Dreby, 2006; Laurie, 2007; Pribilsky, 2004; Ryan et al., 2009; Schmalzbauer, 2005, 2015; Souralová & Fialová, 2017). It is important to recognise that these findings are almost all based upon qualitative data produced through small scale research projects. Very few studies are based on large scale survey based quantitative analysis on the effects of separation on transnational parents (for an example see Haagsman et al., 2015).

Many studies do not include control groups (migrants who are not separated from their children) that can assist in determining whether these outcomes are specific to transnational parents. In addition most studies are based on migrants moving to one host context/country (i.e. single flows of migrants between one country of origin and one host country). Some compare the situations of different groups of migrants in the same host contexts (Leifsen & Tymczuk, 2012). None explore the empirical realities of living in transnational families among migrant parents from the same origin country who live in different host countries (which might allow identification of the significance of migratory context and legal regimes in shaping the emotional well-being and health of parents).

This paper aims to address these concerns and gaps in research via an analysis of data produced in a large scale survey with Nigerian migrant parents in Ireland and the Netherlands. This data includes migrant parents living with all family members in the respective destination countries as well as those separated from one or more of their children. The data also includes migrants from one country of origin (Nigeria) living in two host countries (Ireland and the Netherlands). For the purposes of this study we hypothesise that being in a transnational family has a negative impact on the emotional well-being and health of parents from Nigeria in Ireland and the Netherlands that is more pronounced than other social factors – including gender, migrant status, education and socio-economic status. The following section details important comparisons between Nigerian migrants in Ireland and the Netherlands (w.r.t. the history of immigration from Nigeria from the mid-1990s and similarities in the age, gender and household composition of Nigerian migrants in both societies) as well as important differences (w.r.t. the history of immigration generally into both societies and the dominant route ways through which migrants from Nigeria enter Ireland and the Netherlands). An examination of the impact of transnational separation on migrant parents in both contexts will help illuminate the significance of these other social factor (including gender, migrant status, education and socio-economic status).

Context

Both Ireland and the Netherlands had very small numbers of Nigerian migrants up until the mid 1990s when financial collapse and political repression in Nigeria created new forms of emigration patterns, leading to a surge in the number claiming asylum across Europe (Carling, 2006; see also IOM, 2009; and Komolafe, 2008). The majority of Nigerian migrants to Ireland have been prompted by either economic necessity or as a result of political/ethnic/cultural conflict and violence (Komolafe, 2008).

At the turn of the century the numbers of people claiming asylum in the Republic of Ireland increased dramatically from a handful (40) in 1993 to over 11,000 in 2002. Over this time period the majority of asylum applications were made by African nationals and the largest group within this population are Nigerians, so much so that this has been reflected in racialised discourses in Ireland that have conflated ‘asylum-seeker’, ‘African’ and ‘Nigerian’ in debates about immigration, asylum and Irishness (Lentin & McVeigh, 2002). Between 2000 and 2006 Ireland moved from being an insignificant destination point to having the highest number of Nigerian asylum applicants in the EU (Carling, 2006). Asylum applications from Nigerian nationals increased rapidly to about 4000 per year in 2002. These tailed off in subsequent years, approximately the same number of Nigerians claimed asylum over the entire period 2006 to 2011 (IOM, 2009). In 2012 19,780 people who had been born in Nigeria were resident in Ireland (Central Statistic Office, 2012). The asylum process is certainly the most important mode of entry for Nigerian migrants to Ireland, however it is not the only route for Nigerian migrants (Komolafe, 2008).

While numbers of Nigerian asylum seekers in Ireland rapidly increased after 1996, in the Netherlands a strikingly small number of Nigerians sought asylum (at this time Nigerians were the fifth largest asylum seeker group in Europe (Carling, 2006). A significant number of Nigerian migrants residing in the Netherlands are undocumented migrants although exact numbers are unknown. In 2012 the total number of Nigerian migrants (including first and second generation) in the Netherlands was 11,196, representing a threefold increase since the mid 1990s. Van Heelsum and Thomas (2006) point out that some of the Nigerian migration to the Netherlands is in order to marry a Nigerian migrant already resident in the Netherlands.

While there are important differences between Nigerian migrants to Ireland and the Netherlands with respect to the routes taken by migrants, there are important characteristics shared by the Nigerian population in both countries. In 2008 the Nigerian community in Ireland was profiled by the Irish Central Statistics Office alongside other ‘non-national’ groups in a “population profile” series. The age profile for the Nigerian population in 2006 differed significantly from other migrant groups (as well as the majority host population) in that one in four Nigerians were aged less than 15 while only 15% were in their twenties. A majority of Nigerians lived in family groups with at least one child who was an Irish citizen. Many of these children were quite young (below 10) and had been born in Ireland. The Nigerian population was concentrated in Irish towns and cities, with only 4% of the Nigerians living in ‘rural areas’ while 40% of the total population of Nigerians lived in Dublin City or suburbs.

In 2011 in Ireland the Nigerian population was female-dominated, with large proportions in their 30s or 40s and very early teens (Central Statistic Office, 2012). In part this was as a result of the specific circumstances of Irish citizenship law (which up until a referendum in 2004 was automatically granted to children born on the island of Ireland). This right to citizenship and the practice in Irish courts to interpret the ‘right to family’ for Irish citizens opened what was argued to be a ‘loophole’ allowing migrant parents remain legally in Ireland with their Irish born citizen children (see White & Gilmartin, 2008 for further details). Iroh (2010) argues that the desire for legal residence and citizenship attracted Nigerian migrants to Ireland in the early years of the twentiethcentury and that this can be understood as a strategy of the ‘feminisation of survival’ enacted through “the strategic deployment and practice of motherhood” (p.19).

Similarly the majority of Nigerian migrants arrived in the Netherlands at a child-bearing age, the Nigerian population in the Netherlands was rather young in comparison to other new immigrant groups a high share of Nigerians were married with children and living in a household with these children (Van Heelsum & Thomas, 2006). According to Van Heelsum & Thomas this was the result of the fact that Nigerians migrated relatively more often to the Netherlands as a marriage partner of an already resident Nigerian migrant. As a result of this at the start of Nigerian migration to the Netherlands men were in the majority, however, over time, there has been a feminization of migration whereby sex ratios have evened out. In 1996, 35% of Nigerian migrants in the Netherlands were female, but by 2011 this increased to 46%. In 2011 the relatively youthful Nigerian population was partly a result of the large group of Nigerian children that were born in the Netherlands (by 2003 40% of the Nigerian population were children born in the Netherlands (Van Heelsum, 2005)). Finally the Nigerian population in the Netherlands is heavily urbanised, most Nigerians reside in Amsterdam, followed by the Hague, Rotterdam and Utrecht.

Currently Ireland has no formal legal framework for family reunification for anyone who is not a recognised Convention refugee (CADIC, 2006). Individuals must apply on a case by case basis to be reunited with family members to the Minster for Justice and Equality. The lack of transparency in the Irish family reunification system coupled with the length of time many Nigerians spend in the asylum process (where they have no access to family reunification) has meant that family reunification rates are very low in Ireland by European standards. The Netherlands has become restrictive on immigration laws in general and family reunification policies in recent years (Bruquetas-Callejo, Garcés-Mascareñas, Penninx, & Scholten, 2007). Laws introduced in 2005 require family members wishing to reunite to learn the Dutch language and Dutch culture before migrating to the Netherlands. Stricter income requirements and qualification on who should be allowed to be reunified have also been introduced. Despite these very real restrictions and limitations, it remains the case that the numbers of Nigerians entering the Netherlands on family reunification permits is much higher in the Netherlands than it is in Ireland. Thus in both countries restrictive family reunification and immigration policies will have contributed to the number of enforced and prolonged family separations. However the relative impact is likely to be greater in Ireland as a result of the lack of any transparent system coupled with the length of time Nigerians are stuck in the asylum system.

Data and methods

This research uses comparative survey data collected on 609 Nigerian migrant parents in Ireland² (309) and the Netherlands (300) between December 2010 and June 2011. In the absence of a baseline survey on the Nigerian migrant population in either context, it was not possible to employ a random sampling strategy. A purposive sampling strategy³ was employed to recruit respondents. Interviews were carried out by trained interviewers in community settings or in participants’ homes. In both contexts, efforts were made to administer the survey in areas where Nigerian migrants have settled. In Ireland surveys were carried out in Dublin city (n = 167), the Greater Dublin Area (n = 77), Cork city and suburbs (n = 45), Galway (n = 11) and the midlands (n = 8). In the Netherlands surveys were administered in the large cities - Amsterdam, The Hague and Rotterdam. In order to allow for as diverse a sample as possible, research teams were asked to use different gateways into the Nigerian community to find respondents from different social backgrounds, e.g. through personal contacts, churches, hometown organisations, cultural projects, city and government departments and also by using snowball sampling.

Half of the sample includes migrant parents who have at least one child in Nigeria (in Ireland 48% of the sample [n = 147], in Netherlands 44% of sample [n = 132]) the other half consists of those parents who have all their children in the respective European country (in Ireland 53% sample [n = 163] in the Netherlands 56% sample [n = 165]). The sample population could be further divided into those respondents who have some their children in Europe and some children in Nigeria (in the Irish sample 29% sample, [n = 88], 12% [n = 32] Dutch sample), or had all their children in Nigeria (18% Irish sample [n = 57] 32% Dutch sample [n = 96]) (see Table 1). As Table 1 shows a greater proportion of Nigerian parents in the Irish sample have children in Nigeria than those in the Dutch sample, in addition a greater proportion of Nigerian parents in the Irish sample have children in both Nigeria and Europe than those in the Dutch sample.

Table 1 Descriptive Statistics

The variables used in this study are constructed in a similar way for the dataset in the two receiving countries. The dependent variables were three continuous variables: self- assessed health status and two subjective well-being outcomes, measured by satisfaction with life and emotional well-being. Self-assessed health status and satisfaction with life were each measured using a scale of 1 (not good) to 5 (very good). A measure of emotional well-being was constructed using 12 items, based the General health questionnaire (GHQ-12), asking respondents to report on their emotional distress (Goldberg, 1978). This is found to be an easy to administer and well-validated instrument, for use with literate African populations (Abubakar & Fischer, 2012). Responses were given on a four-point scale (1 = less than usual, 2 = same as usual 3 = no more than usual, 4 = much more than usual). The first two responses were recorded into one category to capture better emotional well-being (0) and the latter two responses into another category of worse emotional well-being (1). Reliability is checked for each sample (0.93 for the Netherlands’ sample and 0.91 for the Irish sample). A factor analysis yielded only one key dimension; hence the 12 items were summed to construct an emotional well-being outcome, with increase in scores showing lower emotional well-being.

The study aims to analyse these three dependent variables separately as each capture distinct aspects of a person’s well-being. The correlation analysis showed the correlation coefficients among the dependent variables were not convincingly high⁴ enough to collapse them into one well-being outcome. The reliability tests also confirmed this (0.48 for the Irish sample and 0.26 for the Netherlands sample). This is consistent with research in psychology on well-being that argues that distilling well-being into a single scale is less reliable than exploring well-being using multiple scales (Ryff, 1989). Diener (1994) for example, recommends using several interrelated measurements in order to reduce measurement error as well as adequately capture the multidimensional components of well-being.

The main predictor was transnational parenting status. The surveys included questions on the whereabouts of biological children (whether all their children live with them in the destination country, or all children live in the origin country or some of their children live in the origin country). Using this question, the transnational parenting status measure was defined by distinguishing parents who have at least one child in Nigeria (=1) and parents who have all of their children in the Netherlands or Ireland (=0). Parents with some of their children in the origin country and some in the host countries were treated as transnational parents because we assume that the effect of separating from even one child would be consequential for parental well-being. A sensitivity analysis conducted by dropping these cases or treating them as a separate group did not change the results reported in our findings.

Other covariates included sex (1 = male, 0 = female), age, and the highest education level of the migrant, used as a continuous variable, ranging from no schooling (0) to university education (11). The analysis controlled for migrants’ socioeconomic status in the destination country and in the origin country: home-ownership in the respective host countries (1 = owns house in the Netherlands or Ireland, 0 = otherwise), and the number of assets in Nigeria. We use these two indicators to fully capture the wide range of resources that determine migrants’ socioeconomic status. We focus on asset based measures of socioeconomic status to circumvent the measurement problems associated with standard measures like income.

We also included social network and migration characteristics. The social network variable was a continuous variable capturing the number of people in the host country the migrant counts on for help. The migration characteristics included the legal status (1 = documented, 0 = undocumented) and the length of stay (years) in each host country. We also included the quality of parent-child relationship measured as a continuous variable, ranging from 0 (always open or warm) to 5 (hardly ever open or warm) as evidence shows the emotional quality of the mother-child relationship has longitudinal consequences for child wellbeing (Estrada, Arsenio, Hess, & Holloway, 1987) and conversely parental wellbeing is influenced by child characteristics and the quality of relationship with their child (Belsky, 1984). We acknowledge the importance of including variables that capture the normativity of child raising norms in the two destination countries. However, our individual level data limits us from using variables that capture norms. To circumvent this, we analysed the dataset from the two destination countries separately.

Using three separate ordinary linear regressions, we analyse transnational parents’ self-assessed health (H), Satisfaction with life (S) and emotional wellbeing (GHQ). Comparable but separate analyses were performed for the Dutch sample and for the Irish sample to observe any potential contextual driven effects. The analyses were done in a stepwise manner in which we included various sets of control variables in separate regressions in order to examine which of the covariates mediated the effect of being a transnational parent. The first analyses we performed were a baseline model (Model 1) by including only the transnational parent status variable. In model 2, we included sociodemographic variables such as sex, age and education status to examine to what extent the associations between transnational parenting and well-being are due to these characteristics. Model 3 included the two socioeconomic status variables: house ownership in the host country and number of assets in Nigeria to account for whether any associations found between our variables of interest were driven by differences in the socioeconomic status of transnational parents and the control group. In a similar vein, model 4 to model 6 included the social network variable, the migration variables, and the quality of child-parent relationship, respectively. After the estimation of the model which included all the relevant covariates (Model 6), a test of multicollinearity, the variance inflation factor (VIF), for each sample, was performed. It is found that the mean VIF among the covariates in both samples (the Netherlands = 1.37, and Ireland = 1.32) are far below the recommended maximum VIF value of 5 (Rogerson, 2001), indicating that multicollinearity is not a serious problem in our analysis.

Descriptive findings

Table 1 shows that the average self-assessed health status is 4.32 (out of 5) in the Irish sample. 15% of the sample of Nigerian parents in Ireland reported that their health status is ‘unfavorable’. Similar results in terms of self-assessed health status are also found in the Dutch sample. Specifically the Nigerian parents in the Dutch sample reported high levels of health (4.15 out of 5) on average. The average GHQ score of Nigerian parents in Ireland is 2.81 while for those in the Netherlands is 1.79 (lower GHQ scores refer to better health). On average, Nigerian migrants in the Netherlands show better emotional well-being compared to the parents in Ireland. Respondents in both countries reported high levels of life satisfaction (3.87 = in Ireland, 3.95 in the Netherlands).

The majority of parents sampled in Ireland are mothers (53%) and the average age of all sampled parents in Ireland is 39 and a very high proportion (94%) are married or in a relationship. A clear majority (59%) of the sample in the Netherlands are fathers. The average age of Nigerian parents in the Netherlands is in the high 30s (39) and a substantial majority (86%) are married or are in a relationship. Nigerian parents in Ireland are highly educated (on average university level education) while this might point to these respondents enjoying quite high socio-economic status in Nigeria these same parents report owning on average low numbers of assets in Nigeria.⁵ On average, Nigerian parents in the Netherlands were educated to vocational school level (i.e. pre university degree). Quite low proportions of the sample of Nigerian parents in Ireland (15%) are homeowners while nearly a quarter (23%) of Nigerian parents in the Netherlands is homeowners and on average they own more than one asset in Nigeria. The sample of parents in Ireland also reported living in Ireland for an average of 6.93 years and a sizeable proportion of Nigerian respondents in Ireland (27%) were undocumented (73% documented). The sample of Nigerian parents have been resident in the Netherlands for longer (8.53 years) and a smaller proportion (16%) were undocumented. Parents in Ireland reported slightly poorer relationships with their children than parents in the Netherlands.

Table 2 showed whether these variables differ by transnational parenting status. Note that the differences in the well-being outcomes between transnational and non-transnational parents are more pronounced in the Dutch sample compared to the Irish sample. Nigerian parents who live with their children in the Netherlands reported higher levels of life satisfaction compared to transnational parents and this difference is statistically significant (p < 0.05). In both Ireland and the Netherlands significantly greater proportions of parents who are in the transnational parent group reported lower levels of life satisfaction than the control group. In the Dutch sample, this difference is more pronounced (the proportion doubles from 19% to 41% between non-transnational and transnational parents (p < 0.01). In a similar vein Nigerian transnational parents in Ireland and in the Netherlands have on average higher GHQ scores (signifying worse emotional well-being) than the control group, and these differences are significant (in the Dutch sample the difference is more pronounced).

Table 2 Descriptive Statistics by Transnational Parenting

Table 2 also illustrates that differences in the socio-demographic characteristics between Nigerian transnational parents and those in the control group are more pronounced in the Dutch sample compared to the Irish sample.

There are more Nigerian transnational fathers (71%) in the Dutch sample whereas almost the same proportion of transnational and non-transnational fathers are found in the Irish sample. Parents in both countries who are in the control group are younger than transnational parents (Ireland [P < 0.05], Netherlands [P < 0.10]), even though non-transnational parents have been living in Ireland and the Netherlands for significantly longer periods of time (P < 0.01). A far higher proportion of non-transnational parents (36%) than transnational parents (8%) own their house in the Netherlands (P < 0.01), and own greater numbers of assets in Nigeria (p < 0.05). Transnational parents in both host countries, on average reported statistically significant (p < 0.01) poorer relationship with their children (2.4 in Ireland and 1.73 in the Netherlands) compared to non-transnational ones.

Results: self assessed health and emotional well-being

Self-assessed health status

Tables 3 and 4 present the results on the ordinary least square regression results of self-assessed health of Nigerian migrants in Ireland and the Netherlands, respectively. The step-wise regression we followed for each sample of migrants revealed interesting differential findings. Model 1 in the case of those in Ireland shows the presence of significant differences between the transnational and non-transnational parents. The significance of this variable disappears when socio-demographic variables are included, with sex of the migrant found to significantly and positively correlate to health status. In particular, being male increases the reported self- assessed health by 19 percentage points compared to being female. Having a better education level and more assets in Nigeria contributes positively to the migrant parents’ health status, as can be observed from the results found from Model 2 to Model 5. However, in Model 6 when documented status is included, the education and ownership of assets back home are not important anymore. This reveals an interesting result, in that, the importance of one’s education status in contributing to better a health outcome is very much linked to being documented, where the documented can utilize their education status to have better access to labor market opportunities in Ireland, which potentially grant these migrants access to improved income.

Table 3 Ordinary least squares regressions predicting self-assessed Health, Ireland

Table 4 Ordinary Least Squares regressions predicting self-assessed Health, The Netherlands

The findings in the Netherlands sample are very different from what we find for the Irish sample. Firstly, the transnational parenting variable is consistently insignificant (from Model 1 to Model 6) in explaining self-assessed health status of Nigerian migrant in the Netherlands included in our sample. A very strong predictor of health status for this group appears to be sex of the migrant and socioeconomic status, measured in this study by house ownership in the Netherlands. It is found that being a migrant father is associated with better health status than being a migrant mother. As expected, better socioeconomic status is an important correlate for a better health status. Other factors we found important for the Irish sample such as social network, and documented status are not associated with the health status of Nigerian migrant parents in the Netherlands.

Satisfaction with life

Tables 5 and 6 present the regression results for satisfaction with life for Nigerian migrant parents in Ireland and the Netherlands, respectively. Results from the regressions in Table 6 indicate that being a transnational parent is significantly (P = 0.05) negatively correlated with life satisfaction for Nigerian migrant parents in the Irish sample. However, this effect becomes less significant as the stepwise regressions proceed, disappearing entirely once documented status (Model 5) is controlled in the analyses. The persistent significance of being a transnational parent on satisfaction with life once socio-demographic variables are introduced is noteworthy. The finding in the Irish data that higher levels of satisfaction are reported by better educated and married respondents is perhaps unsurprising, but again these effects disappear when documented status is controlled. Much like Table 3 the stepwise regression shows that the most significant (p = 0.01) correlation with satisfaction with life in the Irish data is related to the documented status variable (Model 5 & Model 6). We found that having a documented status increases the scale of reported satisfaction by 76 percentage points.

Table 5 Ordinary Least Squares regressions predicting satisfaction with life, Ireland

Table 6 Ordinary Least Squares regressions predicting satisfaction with life, The Netherlands

In the Netherlands (Table 6) being a transnational parent has a significant (P = 0.05) negative association with satisfaction with life. This result confirms our hypothesis and findings from qualitative transnational family literature that highlight the emotional difficulties experienced by transnational parents. Results found on other controls are interesting, in that the socio-demographic and education variables along with the socioeconomic status variable stand out in their importance in explaining satisfaction with life for Nigerian migrant parents in The Netherlands, rather than documented status. What this shows is that, more demographic and socioeconomic variables such as marital status, and education status and socio-economic status which we measured through house ownership in the Netherlands explain migrant parents’ satisfaction with life.

Emotional well- being

Table 7 presents the ordinary least square regression results on emotional well-being. In model 1, transnational parenting is positively and strongly associated with the GHQ scores, indicating negative emotional well-being. This result has not changed when the socioeconomic status and the education variables are included in model 2, although variables such as marital status and education status are found significant and increase the GHQ scores, hence better emotional well-being. However when we control for house ownership in Ireland, the transnational parenting is not significantly correlated with emotional well-being anymore. Owning a house in Ireland, which we use as an indicator of socioeconomic status, is found strongly significant and reduces the GHQ scores, thus indicating how emotional well-being improves with better socioeconomic status. The migrant’s background and social network are not found to matter but migration characteristics captured by documented status reduces the GHQ scores, showing better emotional well-being among the documented compared to the undocumented migrants.

Table 7 Ordinary Least Squares regressions predicting emotional well-being, Ireland

Table 8 presents results for the Nigerian migrant parents in the Netherlands. Model 1 shows that transnational parenting significantly (p = 0.01) increases the GHQ scores, thus negatively correlating with the emotional well-being of Nigerian migrant parents. This result remains with the inclusion of the socio-demographic and education variables (Model 2); socioeconomic variables (Model 3); and background and social network variables (Model 4). However, the variable becomes insignificant in Model 5, where we control for the documented status of the respondent. We find that having documented status strongly and significantly (p = 0.01) reduces the GHQ scores, showing its positive effect on one’s emotional well-being. In Model 6 where the length of stay in the Netherlands is controlled, the result obtained in model 5 remains. This finding is interesting as it indicates how transnational parenting is highly correlated with one’s documented status. It appears that much of the negative association found with being a transnational parent on emotional well-being is mediated by one’s documented status.

Table 8 Ordinary Least Squares regressions predicting emotional well-being, The Netherlands

The results on the other control variables are of interest. We only focus on those that are found significant for reasons of brevity. Other variables found to be associated with a migrant’s emotional well-being are marital status, levels of education, social network, and where the migrant grew-up in Nigeria. Our analyses shows that being married or in a relationship, increase in education level, and expanded social network are all associated with improved emotional well-being, while urban upbringing negatively affects it.

Conclusion

Before reflecting on the conclusions that may be drawn from this research some limitations of our study are worth mentioning. The data used in this study are not representative due to the absence of baseline surveys on Nigerian migrants in both Ireland and the Netherlands. Furthermore, the data used are cross-section data, which measure health and emotional well-being and the status of being a transnational parent at a point in time. Future research could expand on the results obtained in this study by using representative longitudinal data, which allows examining whether changes in transnational parent status is associated with changes in health and emotional well-being over time, by controlling unobservable time invariant traits of the migrants such as motivations and preferences towards migration as well as transnational parenting.

The analyses presented in this paper – drawn from surveys of migrant parent populations (that include both parents who are separated from their children and those who are not separated from their children as well as roughly equal numbers of mothers and fathers) - is sufficiently distinct from the analytical approaches in the ‘transnational parenting’ literature (which is largely based upon small single country samples of migrant mothers all of whom are separated from their children). This allows this paper make some important contributions to the debate about the effects of transnational separation on migrant parents. At least three important interrelated insights into the literature on transnational parenting can be drawn – the significance (or not) of separation; the significance (or not) of gender; and the importance of context in the host country/society.

Effects of separation

Separation from children is seen (understandably) as of prime importance in explaining the incidence of depression and emotional ill health for migrant parents in many studies on transnational parenting (see Hondagneu-Sotelo & Avila, 1997; Leifsen & Tymczuk, 2012; Parrenas, 2010; Schen, 2005). This is not surprising given the samples of parents involved in these studies are typically all separated from their children and the methodologies employed in these studies is usually based around narrative interviews.

However our findings reveal that being separated from children has no significant effect on self-assessed health, self-assessed life satisfaction and emotional well-being (GHQ score) of Nigerian migrants in Ireland. Rather in Ireland self-assesed health, satisfaction and emotional well-being are strongly associated with the migrant parent being documented. Thus migrant status (being documented or undocumented) in Ireland is more important than separation in terms of the impact of these factors on migrant parents emotional well-being and health. In the Netherlands the regression tables show a slightly more mixed picture, being separated from their children impacts negatively on migrant’s life satisfaction but has no significant association with the other outcome variables. In the Netherlands being documented along with other demographic characteristics (being in a relationship; owning one’s home, levels of education and urban upbringing) show significant (but not as strong) associations with emotional well-being (as measured by the GHQs).

Thus in both Ireland and the Netherlands the step-wise regression analysis allows the influence of mediating factors on self-reported heath, life satisfaction and levels of well-being of migrant parents to be measured alongside separation from children. The results indicate that the factors that drive the health, and emotional well-being of migrant parents are not solely related to their separation from their children. While the transnational literature focuses upon and emphasises separation as explaining the health and emotional well-being of migrant parents, the findings presented here show a more mixed picture. It is not really separation that drives negative wellbeing and health but rather other mediating variables (legal status, socio-economic status and demographic factors etc.). Migrant parents who are not separated from their children are just as likely as those who are separated from their children to suffer poor health and levels of emotional well-being as a result of these factors. However, transnational parents are the most represented in the undocumented and low socioeconomic status categories, hence more effects of from these two variables are experienced by these groups of parents.

In Nigeria, child fosterage is a recognised family strategy in the context of international migration and fosterage “tends to ease some of the constraints on migration of parents. Parents who migrate (internally or internationally) can leave children behind with relatives, either temporarily or permanently” (Isiugo-Abanihe, 1985, p. 55). These practices of sharing the burdens of child rearing when migrating is continued when Nigerians (and other west Africans) migrate to Europe or America (Bernard & Gupta, 2008; Philpott, 2001; Selwyn & Nandy, 2012). Transnational, migration contributes to maintain fosterage as a practice as parents often leave children behind rather than bringing them along to presumed short term stays, or uncertain living, schooling, working conditions in a foreign place (Isiugo-Abanihe, 1985), including uncertain legal conditions. Of relevance here is that in Ireland, children’s caregiver in Nigeria was the other biological parent (mother, 37%; father 16%), a maternal relative (31%), paternal relative (8%) or an adult sibling (3%) while non-kinship arrangements accounted for only 3% of caregiving arrangements; in the Netherlands, three quarters of caregivers were the other biological parent (63% mothers, 13% fathers) and maternal kin 15% and non kin accounted for only 1% of caregivers (White, 2013). In spite of the hardships, there were opportunities for Nigerian migrants in Ireland and the Netherlands; one migrant mother in our sample stated that life in Ireland is better than in Nigeria because in Ireland at least you know you will eat every day. The use of technologies such as social media supported Nigerian migrant parents and children to engage in imaginative mobility and and to imagine the life of the other which supported the quality of everyday relating (Veale & Anders, 2014).

Gender effects

As noted above many of the studies exploring transnational parenting focus on the impacts on mothers arguing that these are manifold and almost always negative (see Bernhard et al., 2005; Hondagneu-Sotelo & Avila, 1997; Parrenas, 2010; Schen, 2005 and Suárez-Orozco & Bernhard, 2008). Studies that do include gender as a variable (i.e. that include both mothers and fathers) have pointed to very different experiences for mothers and fathers, that the impact of distance and separation while difficult, may be easier for transnational fathers than it is for transnational mothers (see Avila, 2008; Laurie, 2007; Schmalzbauer, 2005). Research on transnational fathers have emphasised the significance of providing for left behind children as a way of being a ‘good father’ (Pribilsky, 2004). The regression analysis presented in this paper complicates this picture of very different gendered experiences of separation by transnational parents. In both the Irish data and data from the Netherlands gender did not seem to matter as having any impact on the health and well-being of transnational parents. In the Netherlands gender did have a (weak) significant negative effect on the satisfaction with life scores (for mothers) but this effect disappeared once other co-variates (marital status, levels of education) were taken into account. In much of the literature on transnational parenting these co-variates are not usually taken into account (at least not in any systematic way), the analysis presented here suggests that these co-variables (including socio-economic status, legal status and other demographic characteristics including marital status) play a more important role than gender of the parent.

The importance of context in the host country/society

As discussed above much of the published literature on transnational families has provided evidence that separation due to migration may impact negatively on parental emotional well-being and health, however these are mostly based upon qualitative data produced through small scale case studies with single groups of migrants. By including sample populations of migrant parents from the same origin country who live in different host countries, the analysis identifies the significance of the migratory context and legal regimes in shaping the emotional well-being and health of migrant parents.

Considering the practices of child fosterage in Nigeria, it is not surprising that parent-child separation is not readily associated with negative well-being outcomes for the parents in our data. As mentioned above, Nigerian child fosterage norms help families to take advantage of social and economic alliances to provide care for children when one or both parents cannot manage to bring them up (Goody, 1978; Oni, 1995), for reasons that include migration (Isiugo-Abanihe, 1985). The existence of such child fosterage norms may ease negative emotions such as guilt and anxiety reported by transnational migrant parents in other contexts. Additionally, it could also inform a transnational child raising arrangement which assigns migrant parents’ transnational obligations to remitting and to engage in frequent communication with the caregiver in the home country. In Ghana, Poeze et al. (2017) found that migrant parents and caregivers carefully manage their communications to maintain a mutually trusting relationship and feelings of co-presence through long distance communication and this pattern of communication was also observed among transnational families in Nigeria (Veale & Anders, 2013).

The strongly significant impact of legal status on the health and emotional well-being of migrant parents (particularly in Ireland) points to the significance of the social, legal and migratory context of host societies. The findings in this paper reflect the significance of the migratory routes used by Nigerians into the EU generally and to Ireland in particular. In the Irish sample more than 80% of the sample had been in the asylum system and while many had transitioned out of this system a large proportion (26%) were undocumented (which may account for the stronger association between legal status and emotional well-being and health in the Irish sample). In Ireland undocumented migrants are denied access to even essential health care and services (Migrant Rights Centre of Ireland, 2007). In the Netherlands healthcare for undocumented migrants is provided for by what Romero-Ortuño (2004) calls a softened utilitarian system (whereby undocumented migrants receive medical care in part as a result of the professional code of practice of doctors as well as a public policy to protect the health of the host population). This may explain why in Ireland legal status played such an important role in shaping the health and emotional well-being of parents.

In the Dutch sample close to 82% of the sample did not seek asylum in the Netherlands and a much lower proportion of the sample (16%) were undocumented. Higher proportions of Nigerians in the Netherlands (23%) were owner occupiers (as compared to 16% of Nigerians in Ireland) and it may be that owning property in the host society indicates a greater level of stability and feeling of permanent settlement for Nigerian parents. This feeling of permanence might in turn translate into a reduced sense of the importance of legal status on the well-being of parents in the Netherlands.

In both contexts documented status has a strong effect on parents’ capacity to provide materially for their children and families (whether they are separated from each other or not). This is supported by the significance of access to socio-economic resources (owning ones house, having higher levels of education, growing up in a city) for the emotional-well-being of respondents (especially in the Netherlands). These findings suggest that being able to be a responsible and caring parent by providing for the needs of your children is as important as, or more important, for these migrant parents than their physical presence.