Narratives of fathering young children in Britain: linking quantitative and qualitative analyses

ABSTRACT The paper examines how in Britain the time fathers and couples spend in employment shifts in the first years of children’s lives, the conditions under which this happens and how fathers feel about and experience time with their families and time in paid work. In order to achieve these aims new longitudinal analysis of the UK Millennium Cohort Study (MCS) is carried out together with secondary analysis of narrative case studies drawn from a qualitative study of Fatherhood across the Generations. By linking these datasets the paper examines the potential for corroboration and complementarity between different types of data. Further, it seeks to show how qualitative cases corroborate, elaborate and expand on the main employment trajectories in the MCS population of fathers and how these extend understandings of fathers’ experience of time within families.

and employment commitments (Dermott, 2008;Fagan, Lyonette, Smith, & Saldaña-Tejeda, 2012;Kelly, Moen, & Tranby, 2011). This paper explores that paradox by considering issues around time: time as in chronological time, time as in the distribution of time and time as it is qualitatively experienced. In particular the paper examines how, in Britain, the time couples spend in employment shifts in the first years of children's lives, the conditions under which these shifts occur and how fathers feel about and experience time with their families and time in paid work.
To achieve these aims, we have conducted new longitudinal analysis of the British Millennium Cohort Study (MCS) together with secondary analysis of narrative case studies drawn from a qualitative study of Fatherhood across the Generations. By linking these datasets we aim to examine the potential for corroboration and complementarity between two different types of data. The MCS longitudinal data provide a nationally representative picture of fatherhood in the early years, particularly in relation to employment trajectories and feelings about time spent with children as children get older while the qualitative case study material captures men's retrospective accounts of fatherhood and their experiences at the time of interview. Further, we show how case studies corroborate, elaborate and expand understanding of the main employment trajectories in the MCS and so of fathers' experience of time within families. Before presenting the analysis we consider the existing evidence on fathers' work-family time.
Fatherhood and work-family time UK time use data and the UK's household panel survey (Understanding Society) suggest that men continue to be less involved in childcare than mothers (Lader, Short, & Gershuny, 2006;, a picture confirmed by qualitative research (Dermott, 2008;Henwood & Procter, 2003). This is typically attributed to fathers' continuing full-time or long employment hours. However, among couple households in the UK, the sole breadwinner model, once the most dominant family configuration, is in decline (accounting for only 22% of all households in 2011). Typically today the most common household pattern is for fathers to work full-time and mothers to work parttime (32% of households) and for both parents to be in employment (29% of households) . Fathers are more likely than mothers to start work early or to finish late and the self-employed are particularly likely to work longer, non-standard hours (Hilbrecht & Lero, 2013).
As to fathers' work-life balance over time, Norman et al.'s longitudinal analysis of the MCS examined the impact of fathers' and mothers' employment hours on paternal involvement in childcare when the child was aged three. They found that the majority of fathers maintained the same caregiver status from when the child was aged nine months to three years old and that, contrary to expectation, among couples who shared childcare, mothers' work schedules were more important than fathers' for fostering greater paternal involvement in the immediate and longer term (Norman, Elliot, & Fagan, 2013). Significant too were fathers' working hours, with fathers who worked moderate full-time hours more likely to be involved in childcare at nine months and three years, than those working more than 45 hours per week. The age of the father had no impact on their involvement but fathers were more likely to care for boys than girls (see also Flouri & Buchanan, 2003). Furthermore, a longitudinal study of 144 Swedish families recruited in 1982 found that while total amount of play and care fathers engaged in declined over a child's lifetime, fathers were more accessible to older children (Tamis-LeMonda, Shannon, Cabrera, & Lamb, 2004) in the belief that their involvement became especially important with older children. Hilbrecht and Lero (2013) argue that seeing paid work and family life as competing for fathers' time is over-simplistic. As Dermott (2005Dermott ( , 2008 found, fathers see work and family spheres as commensurate. Yet, economic provisioning continues to be regarded as integral to 'good fatherhood' (Dermott, 2005;Shirani, Henwood, & Coltart, 2012) while mothers' employment is still seen as detrimental to children (Norman et al., 2013). Concerns about not spending enough time with children abound among mothers, with a wealth of evidence showing how mothers fit their employment around their families (e.g. Hilbrecht & Lero, 2013;O'Hagan, 2014). By contrast, men see part-time work as less rewarding and less helpful for progressing careers (Dermott, 2005). Norman et al. (2013) attribute such differences to pervasive gendered expectations about raising children and employment. It is also the case that work-life balance is related to income, control over work scheduling and job security. Men in higher status occupations are more likely to invest emotionally in their jobs (Ba, 2014), with some having more control over their working time compared with those in lower status jobs. Norman et al.'s MCS analysis and some qualitative research (Brannen & Nilsen, 2006;Mooney, Brannen, Wigfall, & Parutis, 2012) found that men in lower status jobs are more likely to share childcare with their partners than men in higher status occupations.
Account must be taken of how fathers experience time at particular times of the day, week or season. For example, the fathers in Dermott's (2008) study described mornings and evenings as particularly hectic family times with weekends having a more relaxed and child-centred pace. Evidence on family meal times (Brannen, O'Connell, & Mooney, 2013) concluded that who ate together was affected by the way different temporal domains and schedules intersected and whether children's increasingly busy lives coincided with those of parents. Yeung, Sandberg, Davis-Kean, and Hofferth (2001) found that fathers' working patterns negatively affected time spent with children on weekdays but not weekends, while Sunday working appeared to affect time spent with children more than other atypical work patterns. However, drawing on analysis of the European Union Labour Force survey, Connolly et al. (2013) found that non-standard work patterns are in decline.

Secondary analysis of two data sources
The millennium cohort study The MCS is a UK-wide prospective study of the social, economic and health-related circumstances of British children born in the new century (Smith & Joshi, 2002). All families who were eligible for Child Benefit and resident in the UK when their children were aged nine months were asked to participate. Information was collected directly from each parent about his or her relationship and activities with the cohort child. A stratified clustered sampling design was employed to ensure an adequate representation of all four UK countries, disadvantaged areas and ethnic minority groups (Plewis, 2007). The original cohort (MCS1) comprised 18,818 children whose parents were first interviewed at home. Five home interviews have so far been completed, at ages 9 months, then 3, 5, 7, 11 and 14 years. Interviews are already being planned for 2018 when they will be age 17.
In 2007/2008, when cohort members were age 7, 13,857 families took part and, of these, 11,721 (84.6%) participated in all 4 sweeps (see Appendix, Table A1). For this paper the base MCS sample comprises two-parent families that have been stably partnered throughout the first four MCS sweeps (N = 8302), to be broadly comparable with some cases taken from the Fatherhood Study (see below). For 96% of the families, the mother was the main respondent throughout and the father in only 0.3% of families. This left 3.7% of families, where the main and partner respondent switched between mother and father between sweeps (see Appendix, Table A2).

The fatherhood study
The original Fatherhood Study (2008)(2009)(2010)(2011)(2012) was designed to examine changes in fatherhood across three generations of white British, Irish origin and Polish families, paying particular attention to life course, period and migration effects. The data comprise biographic-interpretative interviews (N = 89) carried out with ten chains of first generation Polish fathers and their fathers and sons; 10 chains of second generation Irish fathers and their fathers (born in Ireland) and sons; 10 chains of white British fathers and their fathers (born in the UK) and their sons aged 5-17. For this article, cases of the current generation of fathers were selected in order to match the families in the MCS cohort analysis (children aged 3-7 in two-parent households). There were only six cases of fathers with children in this age group.
From these we selected four cases for detailed discussion in the paper to reflect the main employment trajectories found among the MCS data and to show the particular conditions in which the MCS trajectories and 'chronicles' of time with children take place. We also included one case which ran counter to the others, exemplifying a case of a father who took on a shift-working job in order to take significant responsibility for his children and was content with his work-family time. The cases also explore the different ways in which fathers experienced and deployed time in their work-family lives.

Linking MCS and the fatherhood study
In undertaking this analysis we moved iteratively between the two studies, using findings from the Fatherhood study to contextualise and shape the MCS analysis and vice versa. This approach helps to go beyond existing hypotheses 'requir[ing] more than the mapping of data which are already there' (Brannen & Moss, 2012). On the other hand, the longitudinal summaries provided by the MCS analysis can also be read as a type of narrative, akin to 'chronicles' (White, 1987). However, they lack 'that summing up of the meaning of a chain of events … that we normally expect from a well-made story' (White, 1987, 16) that is provided by fathers' personal narratives, which have more of a sense of resolution.
As stated above, the paper aims to examine how findings from the largescale MCS longitudinal data (that plots longitudinal household employment patterns and men's reported feelings about time spent with their children over time) are complemented by case study material, based on a narrative life history method that addresses how fathers describe the experience of fatherhood. However, it is important to stress that, in bringing these studies together for secondary analysis, these analyses ask new questions of studies that were not designed with this purpose in mind. Unlike the MCS, the Fatherhood study did not track parenting practices over real time. Further, it asked questions concerning satisfaction with time spent with children only when these questions were relevant in an interview, rather than routinely as in the MCS. Because the children in the study were of varied ages at interview this made comparisons of fathers' accounts provisional. There are large differences also in the size of the populations studied and some difference in the timing of the studies. The Fatherhood data were collected in 2009/2010, whereas the MCS data were collected in 2004 (age 3), 2006 (age 5) and 2008 (age 7). The age seven MCS and the Fatherhood data were thus collected at similar time points, but the MCS included longitudinal data, while the Fatherhood study collected contemporaneous and retrospective accounts.
In analysing the MCS data we first examine fathers' employment patterns at different sweeps. Next we examine men's satisfaction with time with their children at different sweeps. Lastly, we consider how the different longitudinal chronicles of satisfaction with time relate to the different employment trajectories identified.

Findings
Fathers' employment patterns: MCS Table 1 gives the working status of fathers and mothers at the time of interview in the three MCS sweeps from when the child was age three. 1 The information was available for 99.8% of the families. In the initial categorisation of families, we do not distinguish between full-time, part-time or a more finely graded grouping based on specific hours worked, opting instead for a simple binary 'in work' 'no work' at each age. The more detailed approach was too cumbersome to group families longitudinally in any meaningful way. When looking at fathers who were consistently in work over the three sweeps, we are able to take account of the specific hours they worked at each point in time.
We can see that 88.1% of fathers were in work at all three sweeps and a small minority 3.5% was consistently not in work. This left a sizeable minority of 8.3% of fathers moving into and out of employment. Although fewer mothers than fathers were in work at all three sweeps, it is important to note that just over half -50.5%were working at each of the three sweeps. 2 Although nearly a quarter (22.9%) of mothers were never in work, a sizeable proportion returned to work as their child got older. For example, 8.8% were working from the time their child was aged five and a further 7.7% from the time their child was aged seven. Any movements out of work as the cohort child got older would likely be strongly related to the birth of younger siblings (Neuburger, Joshi, & Dex, 2011).
Moving to the combined working status of fathers and mothers (Table 2), there are four dominant patterns which account for 88.2% of the sample: father and mother in work at all three ages (47.3%); father only in work at all three ages (17%); father only in work when child aged three, mother and father subsequently in work (14.5%); father, mother or both in various work patterns over the three ages (9.4%).
Father's views on time with their children: MCS The questions on how parents feel about time spent with their child were identical in the MCS at ages 5 and 7, but differed slightly from age 3. Table 3 shows the distribution of fathers' responses and how the 'age 3 variable' was 'aligned' to the information collected later on.
The largest response categories were that fathers had 'not quite enough time' or 'just enough' time with their children, with fewer reporting 'nowhere near enough time'. Where fathers felt they did not have enough time with their children the reasons were overwhelmingly work-related. 3 This finding suggests that, for a sizeable minority of men, work and family life were in tension.
To look at longitudinal trajectories of reported time with the cohort child, we simplified the response options further to two categories: Enough Time combined too much time, more than enough time and just enough time; Not Enough Time combined not quite enough and nowhere near enough. Table 4 shows the distribution of fathers across the simplified two-categories over time. We can see that nearly a third (31.9%) of fathers were  17.0 Father in work, mother returned to work: father in work age 3, both in work age 5 and 7, or father in work age 3 and 5, both in work age 7 14.5 Father or both parents in work, various patterns over the three ages a 9.4 Mother only in work all ages 0.6 Mother, father or both parents in work all ages 3.7 Both parents not in work all ages 2.6 Both parents not in work, two of three ages 1.7 Both parents not in work, one of three ages 3.4 N (100%) 8287 a For example, father in work age 3, both in work age 5, father in work age 7; both in work age 3 and 5, mother in work age 7; both in work age 3, father in work ages 5 and 7.
consistently dissatisfied with the amount of time they spend with their child whereas nearly a quarter (22.9%) consistently said they had enough time. This information was only available for 73.8% of fathers, as a result of the lower response rates of fathers completing the partner questionnaire. It is noteworthy that 18% of fathers who, as their child grows older, move from reporting having enough time to not having enough time (trajectories 2 and 5, Table 4), whereas 14.3% of fathers change from reporting having not enough time to having enough time (trajectories 4 and 7). Although this suggests that, on average, fathers are more likely to feel they do not have enough time with their child as the child grows older the difference is not large. Perhaps more striking is the consistency of responses, with over half of fathers reporting consistently either enough or not enough time.
The next step was to examine links between longitudinal patterns of reported time with the cohort child and the longitudinal working patterns of families. We concentrate here on the four main working patterns: consistent dual earner households, consistent male breadwinner families, father in work and mother returned to work and father, mother or both parents working (termed other irregular patterns). Table 5 shows that very similar levels of enough and not enough time were expressed by fathers in the four family types. Nearly 1 in 5 fathers consistently (that is, across all three sweeps) felt they spent enough time with their child, and at the other end of the scale, more than a third of all fathers felt they never spent enough time with their child.  What is perhaps most striking here is that the two largest and most contrasting groups, namely the 'dual earner' and 'traditional male breadwinner' models have almost identical chronicles of satisfaction with time spent with their child. This suggests that, in keeping with the literature, satisfaction with time spent with children is more complex than chronological time. Subjective feelings about work-life balancefor example how happy or dissatisfied fathers are with the quality of working and home lifeare also important (Milkie, Kendig, Nomaguchi, & Denny, 2010).

Fathers' working hours and time with child: MCS
We divided the 5428 fathers in MCS who were always in work and supplied information on the number of hours worked, by whether they worked 'short hours' (1-39 hours per week) or 'long hours' (40+ hours per week). The majority of fathers worked more than 40 hours a week at each sweep (68% at sweep 2, 60% at sweeps 3 and 4). Taking a longitudinal view, 44% of fathers always worked long hours, with a further 16.7% moving to working long hours. Just 1 in 5 (19.9%) always worked less than 40 hours per week. Table 6 shows that over three sweeps fathers who have consistently worked long hours, those whose working hours have been inconsistent or who have increased their hours are most likely consistently to feel they have not spent enough time with their child. Those most likely to feel they consistently spend enough time with their child are fathers who have over three sweeps worked less than 40 hours per week.
There is therefore an association between long working hours and evaluations of time spent with the study child; those consistently working long hours are most likely to state that they do not have enough time with their child. Amongst our sample, long working hours were associated with the self-employed (67%) and among those whose highest qualification was at GCSE level D-G or A*-C, or vocational equivalent (51%). 4

The qualitative case
We now examine in detail four case study fathers with children who roughly match the age of children in the MCS and exemplify the main patterns of employment in the MCS. We analyse the ways in which the fathers talked about their experiences of spending time with young children. By drawing on these cases we can see why particular patterns of everyday practices are pursued as well as what these are. Note: A chi-square test showed that there was a significant difference between fathers reporting they spent 'consistently not enough' time with their child by their working hours (X 2 = 85.27, p < .001).
Alec: father in work consistently/mother various patterns/father not quite enough time Alec typifies the pattern found in the MCS analysis of the 9.4% of cases (Table 2) in which fathers or both parents were in work over the first seven years of their children's lives. Alec is a corporate senior executive, married with four children born within seven years, a period when Alec was undertaking part-time training for an alternative less lucrative career and which he describes as 'hairy'. Alec's wife moved in and out of employment between her first two maternity leaves in a similarly high paying job but was not working at the time of the interview. Alec works very long hours which he took for granted in his high flying job. He expected that his wife would shortly return to work once he commenced his new job. Alec's answer to a question about how happy he is about the amount of time he spends with his children is equivocal, resonating with the group of fathers in the MCS who said they did not have 'quite enough time' with their children and Dermott's (2008) suggestion that father's time with their children is differently patterned in the week and at weekends. 'I'm sure I would like to spend more time (pause) I'd like to spend more time during the week.' Central to the way Alec's family manage time constraints is a high income that enables the couple to employ a nanny. Alec's routine involves getting the children ready for school and getting home around 8 pm, in time for a special bedtime routine with them. With a nanny's help, his wife currently undertakes the childcare and manages the children's social lives. Mornings and evenings are narrated as flashpoints in family life. Alec's role is to diffuse difficulty though discipline and distraction. 'I can come in when they've had a terrible day and (my wife) is at the end of her tether and I can actually just grab one of them and say 'let's go and do some piano practice'. He also describes how he supports his wife with disciplining 'from a distance' via his mobile phone. Time is spent differently at weekends compared with during the week. At weekends, Alec helps his children with homework and joins in with football and computer games. 'I just enjoy doing what they enjoy doing with them.' Alec has considerable emotional investment in both his current career and in the new career he is training for. An enduring theme of the interview is finding a balance between his demanding job and his new vocation in the context of a busy family life. He notes the tensions in fathers' roles, between financial provisioning and high emotional and practical involvement in the family.
… a real issue for men at the moment that their traditional identity has been rewritten (pause) but men haven't quite caught up with it yet. So the expectation that you are the breadwinner … has sort of been slightly washed away, and in fact what you're supposed to be is … the perfect partner and … be a wonderful dad … and earn the money.
When asked about his work-life balance, Alec positions himself as a senior manager. He points out gendered practices around part-time work and expectations within his company about employee 'presenteeism' which are at odds with UK parental leave legislation. He suggests that these expectations are internalised by employees themselves.
In this particular environment … where it's all about absolute commitment … you've got to be available when the clients want youon the whole the people who thrive on that sort of atmosphere and incentive … don't want to work part time. We've got a [few men in support roles] who work part time, but most … are women.
Discourses around temporality in this case also relate to the children's life course. At interview, Alec's oldest sons were seven, an age where Alec says he can talk to and 'muck around' with them. However he is aware that his children will be less interested in spending time with him when they are older, while he considers he will need to be around to notice any problems they have in later life. He evaluates the present from the standpoint of an imagined future by imagining his current parenting being scrutinised by a future self, 'You know often at … 12, 13, particularly with boys, you know you're just out of the frame altogether, and I'd hate not to have felt that I ever had that time with them.'

Jason: father in work consistently/enough time with child
Jason is a self-employed builder who still lives in the rural area where he grew up, with his wife and two sons, aged five and four at the time of interview. Jason's wife returned to work briefly after the birth of her first child but has not been in paid work since. She had no plans to return to work. Jason's employment pattern is closest to the 17% of the MCS cases (Table 2) where only the father was in paid work at each sweep. Jason's job is central to his identity. He equates it with providing and also giving him a purpose in life.
[Work is] important because it enables us to do everything that we can do doesn't it? But I mean if I had enough money not to I wouldn't. Like everybody really I suppose. But you need to do something don't you, it doesn't matter how much money you've got -I'm always going to be a doer, you know. Jason did not reflect on his work hours and was content with the time he had with his children. His narrative suggests that sometimes he was in the 'long working hours' category. His work was governed by the weather and involved co-ordination with other workers' timetables. He took steps to stop his work encroaching on family time at the weekends. He enjoys the time he spends with his children, 'I mean that's the enjoyable bit isn't, it, looking after them, I mean if you don't put it in, you're not going to get it out are you'. Jason spent a good deal of his own childhood outdoors and enjoyed relative freedom. He is keen for his children to enjoy the outdoors too and spends time with them in the woods, at the beach and playing sports.
Like Alec, Jason distinguishes between being a dad at the weekend compared with during the week when his time is limited; the children are ready for bed by the time he gets home. He avoids working on Saturdays and pays others to do odd jobs around the house.
Jason identifies a tension between the rhythm of his working day and his son's school day. In his job, he needs to get most of the work done in the morning, which can clash with school appointments. Running his own business and being dependent on the schedules of other contractors and clients, he is rarely in a position to take time off at short notice, for example, if his sons are sick. Time lost has to be made up later, usually at the weekends when he wants to spend time with his children.
You know you've got to ring and say 'Look I'm not coming, the kids are ill. And if the people you're working for they've said 'Oh we've taken the day off work for this' you know what I mean? And … . so you'll try and grab [time] it back by working on a Saturday, and then you'll only see them on a Sunday. Thus the rhythms of work and family life are at times out of kilter even though Jason's working patterns are to a certain extent within his control and he is much less well paid compared with Alec. Despite commenting that he would like his wife to be the main breadwinner so that he could have 'five years off' his preference is for her to be the main carer for the foreseeable future. He does not want more time with the childrenonly the 'precious' weekendsand thinks he would be temperamentally unsuited to being a 'house husband'. For Jason, fathering is gendered in terms of skills, spheres of influence and amount of time each parent spends with the children.
… what I do will influence their more practical and their more playful sides, whereas the more intellectual sides is probably what she's going to be doing. 'Cos more often than not she's going to have more time to sit there with them … , and then the weekends you know, the practical bits. And obviously the dark evenings don't helpwhen it's the light evenings you can get out.
As in Alec's account, preferences for spending time with children were related to the children's life course. Thus, Jason enjoys being with his children at their current age and envisages that, as they get older, his time with them will be enjoyable and valuable. However, when they were born, he took no time off. As a self-employed worker he was not entitled to paternity leave but he did enjoy some flexibility.

Liam: both parents consistently in work/not enough time
Liam is married with three children under eight. He works in the voluntary sector while his partner worked part-time throughout the births of three children (apart from maternity leave) including some nights. Their employment trajectory is close to the 47.3% where both parents are in work in the MCS (Table 2). Liam often worked long hours and, while he felt he lacked sufficient time with his children, he was ambivalent because he felt driven by his strong work ethic. Asked whether he achieved a balance between work and family life, he appears to agree but then says … like I'm determined to get certain work done. And I also think sometimes certain jobs, you have to go the extra mile, you have to do some extra timeand that's my sort of philosophy, if I need to stay I stay. And if I need to do occasionally a bit at the weekends I will … I'm not expecting any more money for it … that's my sort of background really.
Liam had recently reduced his hours to a 40-hour week in part because his job has changed, but also because he wants to spend time with his children and not be overly stressed. While he describes his employer as flexible he has not put this to the test beyond working from home one day a week. Liam could not recall taking time off when his children were born. However the overriding impression he gave was an inability to be able to relax with his children.
And I think one of the key things for me, having come from a working class Irish background is I enjoy working hard … I want to do my best. I want my kids to get on in life, … to have more opportunities than I do, … I absolutely adore my kids, … probably I wish that sometimes I could be probably more relaxed with them, and wouldn't sort of get annoyed.
Unlike Alec, Liam has mixed feelings about his wife's employment, believing that she should have some 'independence' while also thinking that her being at home more would ease the pressures on family life. Unlike Alec, Liam's family has no nanny. The pace of their family lives, Liam complains, is also 'ruled by our kids'. Liam emphasises the importance of relational time with his children 'because it helps that bonding and love and looking up'. However, he gives little sense of being highly involved with his six-year-old son. He does not get home from work until 7 pm or later. His task it to take his son to tennis lessons and occasionally swimming. Like Jason, Liam is more motivated to do things when the evenings are lighter and the days warmer. In this case there is a strong disconnect between working time, the temporalities of everyday family life and leisure. Liam also sees a family skiing holiday as an opportunity for extending his children's experiences, a pattern resonant with 'intensive parenting', but only employed by Liam when he has the time (Shirani et al., 2012). Furthermore, Liam's ambivalence about creating a better work-family balance relates to his aspirations for his children and to wanting to provide a strong, hardworking role model as his own father had done.

Adam: both parents in work/satisfied with time
Adam is married with three young children aged eight years and under. He gave up a job in sales with its culture of long-hours work once he had children. Adam had downshifted to a job with flexible hours. His wife is the higher earner in the household with a full-time professional, public sector job. This couple's employment trajectory is also close to the 47.3% in the MCS, who are consistent dual earners (Table 2). Adam's shift pattern allows him to work flexibly and care for his children three days a week. He organises his work around their routines and mentioned the many caring tasks and activities he takes charge of, describing himself as a 'hands-on' dad. He states that time is the most important commodity for families: 'my best times are when I spend times with the family. They're my most enjoyable times'. Adam was content with the time he spends with his children. Critical of the poor public policy for British fathers in general, he was highly appreciative of the conditions available to him in his job. 'The support you need is immense, and I think it's absolutely disgusting that men get a few days and then they all take their basic rate'. That Adam takes a full part in family life is made possible by his higher earning partner and by the couple prioritising the mother's investment in her employment career.

Discussion and conclusion
The MCS analysis indicates that the most prevalent household configuration when children were 3, 5 and 7 years was for both parents to be working. At any one MCS sweep between a half and 45% of fathers said they felt they had enough/more than enough or just enough time with their children, while between 50% and 55% reported not quite enough or nowhere near enough time with their children, with more fathers reporting not quite enough time as children got older. However, by carrying out longitudinal analysis it emerges that men's satisfaction with time spent with their children is more complex than these findings suggest and does not reflect household employment trajectories in the ways one might expect. Fathers who were sole breadwinners throughout the period covered have the same chronicle of satisfaction with time as those in dual earner households: in each group one third report having consistently not enough time and a fifth say they had enough time throughout. However this analysis also confirms the negative effects of long working hours: fathers who work long hours were more likely to feel consistently that they lacked enough time with their children. However, while just over half of fathers felt this way, dissatisfaction did not increase as their children got older.
The value of the MCS data is that they provided 'the big picture', summarising experiences over time for a large and broadly representative sample of fathers living in couple relationships in contemporary Britain. The nuances of how time is experienced are not sought in surveys such as the MCS, given the need for pre-coded responses to a limited number of questions compared with a qualitative study. While the qualitative Fatherhood study was not designed to represent the range of work-family trajectories, the cases presented point to the ways in which particular fathers (and their partners) corroborate and help us to understand the employment trajectories identified in the MCS analysis and the meanings that men attribute to time with their children. Thus Jason, a working class father, represents a father who was the sole breadwinner almost for the entire early years of his children's lives. His narrative was one of contentment with the time he spent with his children and he regarded everyday care of children as the responsibility of an 'at home' mother. A fifth case (not discussed) concerned a Polish migrant manual worker whose wife had not worked since having children; this father was happy with the time spent with his children and with his working time mainly, he said, because he had more time at home than when they lived in Poland. By contrast, Adam who downshifted to a low paid, low status job that provided flexibility in organising his work hours, did so to enable him to take major responsibility for his three young children. The condition under which this occurred was the prioritisation of his wife's career and the subordination of his own employment especially in terms of earnings. Adam's narrative was of having achieved a satisfactory work-family balance that involved prioritising the care of his children. It also meant enjoying time with them.
The middle class fathers were also varied in terms of their feelings about time. Alec's narrative suggested that, in theory, he wanted more time at home with his children. However, he was investing his spare time in training for a new career. Alec shared breadwinning with his partner who had an interrupted employment pattern in between having four children. His children's care depended on his wife working part-time and on the help of a nanny. In a sixth case, not discussed, the wife also had an interrupted work career but was working full-time. The father, a stockbroker, said that work was central to his identity and that he was happy with the time spent with his children. By contrast, Liam, an administrator, admitted to wishing his wife was at home (she had worked part-time since the births of three children apart from periods of maternity leaves) and produced an ambivalent narrative: feeling torn between wanting more time with his three young children and being very work driven and seeking to model and pass on a strong work ethic to his children.
The MCS reported similar levels of satisfaction in the amount of time spent with children across social class while the qualitative cases show variation between households within as well as between socio-economic groups. On the other hand, qualitative research in Sweden (Plantin, 2007) argues that working class men found fatherhood predictable and familiar, while middle class men viewed fatherhood as a reflexive project of the self and an opportunity for greater self-understanding. It seems further research on class and work-life balance is needed.
The cases show the equivocal nature of men's feelings about work and family time as they struggle with their identities as workers and carers, with the former continuing to have the upper hand. Thus we argue that understanding contemporary fatherhood requires consideration of how fathers experience time as well as how they divide their time. The cases augment the MCS analysis, demonstrating the tensions between the rhythms of working and family life: the daily working patterns in the case of manual self-employment (Jason); the requirement of 24 hour availability for those in the corporate sector (Alec) that limits fathers' capacity to attend school events or take time off when children are ill; the need to stay late at work which leads to fathers missing children's bedtimes (Liam) and the favourable form of flexibility that Adam enjoyed enabling him to take care of his children. On the other hand, Jason, Alec and Liam were happy with their long hours because they found their jobs satisfying.
The cases also point to the way family time was experienced at different points of the day, the week and the year (La Valle, Arthur, Millward, Scott, & Clayden, 2002;Yeung et al., 2001). For example, there were 'flashpoints' in the family day (e.g. in Alec's case, evening bedtimes) which men missed or could be involved in depending on their work schedules. Men's involvement in family life was described as greater and different at the weekend and holidays (Dermott, 2008). Jason and Liam talked about the opportunities for spending and enjoying time with their children on the 'long evenings' of summertime. Time with children was also valued in terms of opportunities for developing relational time with children  but also as occasions for 'concerted cultivation' of their children (Lareau, 2011).
There are considerable limitations in bringing these two very different datasets together because they were designed for different purposes and addressed different research questions. They have different research designs (longitudinal versus biographical narrative accounts at one point in time), very different sampling frames and sizes (a large nationally representative sample compared with small purposive samples of different ethnic groups), and different research methods (surveys compared with in depth life story and semi-structured interviews). However, this exercise in data linkage demonstrates fruitful ways in which quantitative and qualitative data can be brought together (Brannen, 2015;Elliott, 2005Elliott, , 2013. In particular, the analysis of fathers' and household longitudinal employment patterns and men's chronicles of time spent with children has allowed us to situate particular cases within the wider population of fathers and to contextualise the cases in particular historical periods (Elliott, 2005). It shows how qualitative case studies can be used in a number of ways in combination with the MCS data (Greene, Caracelli, & Graham, 1989) to analyse the patterns and conditions of employment and care (Dermott & Miller, 2015). They corroborate patterns found in the MCS. They complement the picture by suggesting the processes involved, notably the conditions under which households with young children maintain or change their employment trajectories and how a father may take on significant childcare responsibility. They add to the picture of fatherhood by demonstrating that time in work and family life may be valued differently and that not all family time is valued to the same extent or in the same ways. Notes 1. We did not include employment status at nine months as we did not have information on how parents felt about the time they spent with their child at nine months old.
2. We cannot, however, be sure that these mothers were continuously in work during the time between each of the three interviews. 3. When fathers responded that they did not have enough time with their child, they were asked why this was in a multi-coded follow-on question but were also asked to select the main reason for this. Answer categories included 'work long hours', 'work away from home', 'demands of other children', 'demands of domestic duties', etc. At least 85% of fathers at each age selected 'work long hours' and 16% that they 'worked away from home'. 4. The National Qualification Framework aligns academic and vocational qualifications into levels. Level 1 equates to GCSE grades D-G, level 2 to GCSE grades A*-C.

Disclosure statement
No potential conflict of interest was reported by the authors.  Note: Although two biological parents are slightly over-represented and lone mothers are slightly under-represented in the longitudinal sample, the composition of families is very similar.