Growing Up in Scotland: father-child relationships and child socio-emotional wellbeing

Chapter 7: Discussion And Recommendations

This Growing Up in Scotland study of father-child relationships considers several important issues relevant for policy makers and practitioners involved with family influences on children's socio-emotional wellbeing, and the role of fathers. The study draws on information from over 2,500 couple families in the first GUS birth cohort, a nationally representative sample. Ten-year olds in these families were asked about several aspects of trust in, and communication with, the resident father or father figure. Children's responses were used to categorise father-child relationships according to poor, good or excellent levels of fathers' supportiveness. The study examines the distribution of poor, good or excellent father-child relationships; risk factors predictive of poor-father-child relationships; and how supportive father-child relationships are linked with other aspects of ten year-olds' socio-emotional wellbeing.

Most ten-year olds in couple families are positive about resident fathers' supportiveness. However, a substantial minority (16%) perceive poor relationships characterised by low supportiveness. Overall, father-child relationships were viewed by children as less supportive than mother-child relationships, which were more uniformly positive. This difference is in keeping with a large national US study of adults' retrospective views of their relationships with their parents in childhood (Mallers, Charles, Neupert & Almeida, 2010). Children's perceptions of their relationship with each parent diverges most at the lower end of the supportiveness scale, where more children perceive low supportiveness from fathers than from mothers. Overall, only 5% of children perceive poor relationships with both parents. An additional 15% perceive poor levels of supportiveness from only one parent, however this is more likely to be with the father (11%) than with the mother (3%). Thus while a small minority of families may benefit from support in strengthening the child's relationship with both parents, father-child relationships may benefit from more targeted measures helping a wider group of families.

Boys report slightly lower supportiveness from fathers than girls. This gender difference is reflected in other aspects of children's wellbeing: boys also perceive lower supportiveness from mothers, have higher levels of behavioural and emotional problems, and report lower wellbeing on other measures used in this study.

Risk factors for poor father-child relationships

The study identifies several family circumstances that are predictive of a poor father-child relationship. These factors are equally important for boys and girls. Two of them, family socio-economic disadvantage and family adversity, appear to have a negative impact on mother-child, as well as father-child relationships.

Our GUS finding for family socio-economic disadvantage (as indicated here by lower parental education) ties in with the negative effect of low family socio-economic status on the quality of both mothers' and fathers' relations with three year-old children in another large birth cohort, the UK Millennium Cohort Study (Malmberg & Flouri, 2011). In addition, research on couple families using the National Child Development Study found lower family socio-economic status predicted lower father involvement when children were aged 7, 11 and 16 years (Flouri & Buchanan, 2003). The effects of socio-economic disadvantage are likely to relate to low parental psychosocial, as well as economic, resources that compromise parenting quality (Belsky, 1984; La Placa & Corlyon, 2016), Focusing on structural societal changes which reduce inequalities is likely to boost such resources in the longer term (Katz, Corlyon, La Placa & Hunter, 2007). Providing additional help for vulnerable parents, including greater access to health visitor support and affordable childcare is high on the Scottish Government agenda (Scottish Government, 2016).

Our study also found that multiple, potentially disturbing, adverse family events predict a poor father-child relationship, even after allowing for their association with family socio- economic disadvantage. An accumulation of adverse events over the pre-school and early school-age years also appears to undermine mother-child relationships. Other research indicates a negative association between adverse life events and children's socio-emotional wellbeing, independent of family socio-economic disadvantage, among young children and older age groups (Flouri & Kallis, 2011; Flouri, Mavroveli & Tzavidis, 2010). However, research using the UK Millennium Cohort Study did not find that the quality of mothers' and fathers' relationships with younger children (aged three) varied according to a score of adverse family events (Malmberg et al., 2011). Our apparently contradictory GUS finding may reflect measurement of events over a longer time period, and their impact on children's feelings, rather than on parenting behaviour - especially as our measures of parent-child relationships (unlike those in the UK Millennium Cohort Study) are child-reported, and involve older children. Adverse family events may compromise children's trust in parents and emotional security, especially with heightened awareness at older ages, and lead to lower perceived parental support. Additional research using the Millennium Cohort Study suggests that supportive parent-child relationships may help to buffer the effects of family adversity on children's socio-emotional outcomes (Flouri, Midouhas, Joshi & Tzavidis, 2015). Further research using future sweeps of data will be required to confirm a similar protective effect in the GUS sample.

Other factors, such as the presence of a father figure rather than the biological father, partner supportiveness and family climate, seem to affect father-child relationships more than mother-child relationships.

A child's relationship with a father figure or stepfather has long been identified as problematic, even after taking account of father figures' typically lower socio-economic status (King, 2006; Marsiglio & Hinojosa, 2010). This may partly reflect the lack of societal norms for men who find themselves in the position of being a father figure, perhaps competing with the biological father's claims on the child. Unfortunately, GUS has not collected information on children's relationships with non-resident fathers at the age 10 interview, although this is planned for the next round of data collection when children are in their first year of secondary school. In our study, the gap in perceived supportiveness from fathers and mothers was widest for families with a father figure. This finding tallies with younger (aged 4 to 7) children's views, in studies where children indicated closeness to family members with pictorial representations (Roe, Bridges, Dunn & O'Connor, 2006; Sturgess, Dunn & Davies, 2001). Father figures may become closer to children when they are more "embedded" in family life with clearly-defined responsibilities, perhaps through longer residence, marriage to the child's mother, or having step- and biological children living together in the same household (Marsiglio et al., 2010). Our findings from multivariable models do not clearly show that residing in the household for more than two years or being married to the child's mother improves father figures' relations with children, and we were unable to explore the effects of any children from a previous relationship. However, US research on older age groups found that length of residence, as well as adolescents' positive socio-emotional adjustment and relationship with the mother, all appeared to make it easier to relate to a resident non-biological father figure (King, Amato & Lindstrom, 2015; King, Thorsen & Amato, 2014). Interestingly, this research did not find that the child's relationship with the biological father affected relations with the stepfather.

A less supportive partner relationship and aspects of family climate also appear to affect father-child relations, and to a greater extent than mother-child relations. (Note that although having parents who are unmarried is a "current" predictor of a poor father-child relationship, there is no information collected at the GUS age 10 interview on the current quality of the partner relationship. In the "early predictors" model in section 5.3.4, any effect of parents' marital status became superseded by the quality of their relationship.)

Although we cannot be sure of the direction of causation in our study (it is possible that poor father-child relations contribute to an unsupportive family climate, rather than the reverse), the findings tend to support an ecological-contextual theory of fatherhood (Doherty et al., 1998). This theory suggests that fatherhood is more vulnerable than motherhood to interpersonal and environmental influences. It is based on the wider variation in societal expectations and norms for fathers', as compared to mothers', parenting. The less well- defined role for the father may therefore be shaped to a greater extent by negotiations between parents, and by other contextual factors. These may collectively either support fathers' involvement in parenting or undermine it, leading less committed fathers to withdraw or "opt out". Conflict between parents may spill over to damage the quality of parenting and co-parenting (the extent to which parents work together to parent a child) (Pedro, Ribeiro & Shelton, 2012; Zimet & Jacob, 2001). Several other studies have pointed to the importance of a supportive partner relationship for involved fathering (Cummings, Merrilees & Ward George, 2010; Planalp & Braungart-Rieker, 2016). Less is known about other contextual influences, although a qualitative study of families with children aged 4-8 years (Pike, Coldwell & Dunn, 2006) had similar findings to our own study: father-child relations were negatively affected by a chaotic family atmosphere as well as by poor partner relationships, while mother-child relations were unaffected. Repeated measurement of couple relationship quality and co-parenting in future sweeps of GUS, as well as parent-child relationships, would help clarify the direction of any causal relationship.

Our study failed to find an effect of early father involvement, as measured by how often fathers engaged in play, reading or caregiving, and fathers' use of positive or negative discipline strategies. This might be because these measures did not adequately capture the quality of parent-child interaction or do not bear a strong relationship to later father involvement, which might be more salient for the GUS ten year-olds. Elsewhere, research reviews indicate that father involvement is positive for young children's socio-emotional and cognitive outcomes (Sarkadi et al, 2008; Downer, Campos, McWayne & Gartner, 2008; McWayne, Downer, Campos & Harris, 2013). Nonetheless, there are still gaps in our understanding. Studies included in these reviews adopt variable measures of father involvement, do not always account for maternal involvement, and do not take account of the possible bidirectionality of associations between father involvement and children's socio- emotional wellbeing over time. Recent UK research on young children attempts to plug some of these research gaps, and supports the idea that more frequent father involvement and fathers' more positive attitudes towards their parental role benefit young children's socio-emotional adjustment (Flouri et al., 2016; Kroll et al., 2016; McMunn et al., 2015; Opondo, Redshaw, Savage-McGlynn & Quigley, 2016). However, effects appear limited, and vary with child age both within and across studies.

We also did not find that father-child relations are negatively affected by reduced potential availability. Working hours may not in themselves, for most men, jeopardise fathering responsibilities. Indeed, many UK fathers acknowledge full time employment as essential to their typical role as the main family breadwinner (Hatten, Vinter, Williams & Mori Social Research Institute, 2002; Kadar-Satat & Koslowski, 2015; O'Brien & Shemilt, 2003). It is more likely to be other aspects of working life that "spill over" and affect parenting quality, although these could be exacerbated by long hours. Fathers who suffer from work overload, nonstandard work schedules or other work place stressors such as lack of control show less sensitive, engaged parenting of infants (Goodman et al., 2008; Goodman et al., 2011). Father's work-family conflict was linked to less warm, and more irritable and inconsistent parenting in Australian families with young children (Cooklin et al., 2016), although high levels of work-family conflict were reported most often by fathers working long hours as the sole breadwinner. Commuting time might also increase fathers' stress levels and have a negative impact on parenting quality. A German study found that fathers who travelled more than 40 km to work every day were more likely to have a child with poor socio-emotional adjustment, even after taking account of both parents' working hours (Li and Pollman- Schult, 2016). Further research is needed to see how these factors affect GUS fathers. We did not find that GUS fathers' work-life balance when children were very young helped to predict their relationship with the ten year-old child. Nonetheless, it seems possible that some negative effects we found for fathers' occupational class and remote home location when children were young might respectively relate to greater spill-over of work into family life for small employer or self-employed fathers, and to fathers' long work-related travel times for some families living in remote parts of Scotland. Even though a recent study suggests small employers/self-employed occupations are least likely to suffer work-related stress, depression and anxiety (Health and Safety Executive, 2015), both the father's occupational group and remote home location could signal a less family friendly workplace environment. While more research is required to uncover the reasons for our findings, the need to boost family-friendly flexible working practices among small employers is highlighted in a recent Scottish Government policy document (Scottish Government, 2016).

Associations with other aspects of children's socio-emotional wellbeing

Our study finds that father supportiveness is positively associated with the child's perceptions of being supported by the mother, and with other aspects of children's socio- emotional wellbeing at the same age. Although fathers' supportiveness is generally at lower levels than mothers', relations with both parents appear equally important for ten year-olds' school adjustment, relations with peers and overall wellbeing. Findings suggest that fathers make an independent contribution to children's wellbeing, even after taking account of mothers' supportiveness and other family circumstances. This is in line with the "Important

Father" hypothesis, which suggests that fathers do make an important, even if not unique, contribution to children's development (Pleck, 2010).

Two limitations to this conclusion need to be borne in mind. The first is that some associations between father-child relationships and other aspects of wellbeing may be inflated because the child was reporting on both the predictor and its outcome ("shared method variance"). Where there was a different informant (the child's parent) for two outcome measures (behavioural and emotional difficulties, poor school adjustment), the finding of an association with father-child relationship quality appears stronger. We plan to triangulate these results using teacher-reported measures of children's behavioural and emotional difficulties, gathered shortly after the age 10 home interviews. A second limitation is that parental supportiveness and other aspects of wellbeing were all measured at the same time point, so we cannot be sure of the direction of any effect. For example, it is possible that the child who has difficulty adjusting to school or who experiences peer victimisation will tend to feel less supported at home, because it may be difficult for parents to help with matters outwith their immediate sphere of influence. Fathers may also find it less easy to develop a supportive relationship with a "difficult" child.

Longitudinal studies collecting information on fathering and child adjustment at several time points can help to answer the question of how father-child relationships and child socio-emotional adjustment are inter-related. Research on father involvement and child adjustment using three waves of the UK Millennium Cohort Study (ages 3, 5 and 7) found that children's behavioural problems appeared to reduce fathers' participation in play and caregiving activities more often than the reverse (lower father participation increasing children's problems (Flouri et al., 2016). Longitudinal studies on fathers' supportiveness, rather than participation in activities, are scarce among families with young children. A recent study of anxiety trajectories over a wide age range (Parrigon & Kerns, 2016) found low early attachment to fathers predicted that young children would maintain anxiety levels until adolescence. This did not look at how anxiety may have affected father attachment at young ages, although studies of adolescents have found bidirectional associations between father-child attachment and anxiety or depressive symptoms over time (Branje, Hale, Frijns & Meeus, 2010; van Eijck, Branje, Hale & Meeus, 2012).

Our findings do not point to different roles for fathers and mothers, or to different effects of fathers' and mothers' supportiveness for sons' and daughters' wellbeing. Research on older children has, however, suggested differential effects according to parental and child gender, although results appear to vary across outcomes. For example, one study points to a greater protective role for fathers than mothers in adolescent depression (Desjardins & Leadbeater, 2011), and others show stronger associations between father-child relationship quality and reduced anxiety or depression for boys (Branje et al., 2010; van Eijck et al., 2012). However, a meta-analysis of associations between adolescent attachment and delinquency found stronger protective effects of mother than father attachment, as well as stronger effects when parent and child were of the same sex (Hoeve et al., 2012).

Future plans

GUS plans to collect further information on parent-child relationships and socio-emotional wellbeing when children are in their first year at secondary school (sweep 9). This should help us answer questions on how fathers and mothers may support children's socio- emotional development in more detail. It will be important to revisit the question of whether effects of parenting differ according to whether the parent and child are of the same or different sexes, since gender differences may become more pronounced as children enter the teenage years.

Also planned is qualitative research with GUS fathers in order to explore how parents' own experiences of being parented influence how they go on to parent their own children. This process is referred to as 'intergenerational transmission of parenting'; research evidence suggesting there is significant, though modest, continuity in parenting across generations (Belsky, Conger & Capaldi, 2009). This study should provide better understanding of the complexities which lie behind different aspects of father involvement in relation to particular family circumstances.

Children's relationships with non-resident fathers are a notable omission from this study: other research has pointed to the quality of the child's relationship with the non-resident father as a significant influence on socio-emotional adjustment (Adamsons & Johnson, 2013). However, sweep 9 of GUS will incorporate information on this subject. It is also planned to collect data on fathers' perspectives, to supplement children's views, and to consider other dimensions of parenting such as parental control and encouragement of autonomy.

Concluding remarks

Overall, the results from this report highlight the importance of father-child relationships in couple families, showing that these relationships in middle childhood are closely bound up with several other aspects of children's socio-emotional wellbeing. While the majority of children perceive high levels of supportiveness from resident fathers, a significant minority perceiving low levels of supportiveness also have lower overall wellbeing, regardless of other family circumstances. This lower wellbeing does not simply reflect children's negative perceptions of family life, but extends outside the family to include lower enjoyment of school, and poorer relations with teachers and peers. Future work will seek to strengthen this finding using teacher-reported measures of child wellbeing collected at age 10 (but not available for this study), and by examining father-child relationships at age 10 in relation to children's wellbeing measured in future sweeps of GUS.

The report findings are based on information from over 2,500 couple families in a nationally representative cohort of children, and contribute to the limited research base on father- child relationships in middle childhood. They rely on children's perceptions of fathers' supportiveness. Furthermore, they use cross-sectional observational data, rather than being the results of an experiment or trial, and cannot demonstrate that improving father-child relationships will necessarily increase children's socio-emotional wellbeing. Nonetheless, the study lends support to the idea that for some families, improving the quality of fathering may be a suitable target for future interventions directed at improving children's socio-emotional wellbeing.

Various factors appear to compromise the development of supportive father-child relations among couple families, including a high level of family socio-economic disadvantage, adverse family events, an unsupportive partner relationship, a more disruptive or less cohesive family climate, and the presence of a non biological father figure rather than the biological father. These findings may help to identify families most at risk, as well as suggest intervention targets. Especially among families with high levels of socio-economic disadvantage or family adversity, children's relations with both parents might benefit from greater family access to professional parenting support. Yet previous GUS research suggests that disadvantaged groups are likely to perceive many barriers to professional help, including stigma and fear of interference (Mabelis & Marryat, 2011). More generally, there are multiple barriers to engaging fathers (La Placa & Corlyon, 2014; Lundahl, Tollefson, Risser & Lovejoy, 2008; Panter-Brick et al., 2014; Ramchandani & Iles, 2014). These include fathers perceiving that available services are not relevant to them, and services not, indeed, being 'father-friendly' in terms of delivery, for example in their opening hours or simply because fathers are not targeted or even included in recruitment practices. To be effective, professional services need to devise better ways of engaging and retaining disadvantaged families, particularly fathers, in parenting support programmes.

Father-child relationships appear to be more sensitive to the overall family environment than mother-child relationships, emphasising a need to view fathering as embedded in the whole family system. The quality of father-child relationships seems to depend on the quality of family interactions more generally, including the partner relationship. Targets of a family systems approach to support good fathering could therefore include measures to boost a cohesive family ethos and improve marital relations. In addition, advice and support for co-parenting, where couples learn to communicate better, establish trust and work together harmoniously when parenting the child may be particularly helpful.

Compared to children living with a biological father, children are less likely to perceive a non- biological father figure as being supportive. The GUS sample of families with a non-biological father figure is too small to permit a detailed quantitative study of factors promoting more supportive father figure-child relations within families of this type. A qualitative follow-up study might strengthen our understanding of the particular difficulties faced by father figures, both within the family and in relation to the child's non-resident biological father, and what could help them in their fathering role. Again, a family systems approach might encourage mothers to find ways of facilitating the child's acceptance of a new father figure.