Modern fathers' dilemma of work-family reconciliation: Findings from the German Youth Institute Survey AID:A II

Objective: This study investigated how work and family demands and resources relate to fathers’ perceived work-family conflicts. Background: Increasing expectations for family involvement and the lingering centrality of employment in the male life course pose challenges for fathers to combine different life domains. However, most studies on work-family interface continue to focus on mothers and examine work-to-family conflict and family-to-work conflicts separately. Method: First, we used cluster analysis to identify a typology of four groups, each with different manifestations of work-to-family and family-to-work conflict. We then analyzed the relationship between fathers’ group membership in this typology and a number of relevant work and family demands and resources using multinomial logistic regression on a sample of 5,226 German nuclear families with at least one child under 18. Results: Our findings revealed that the greatest proportion of fathers (38.2%) reported being primarily pressured from work (=work-to-family conflict predominates), 19.8% primarily from the family (=family-to-work conflict predominates), but another 13.4% reported feeling conflicted in both directions; only 28.6% of fathers reported being more or less free of conflicts. Results of multinominal logistic regression suggested that long work hours, intrusive work demands, and long commute associated with fathers’ work-to-family conflict or dual conflicts. The higher the fathers’ weekday time investment in childcare and the better the perceived couple and family relationship, the lower the likelihood of fathers’ experience of work-to-family and dual conflict, although the likelihood of family-to-work conflict is unaffected. In addition, a higher family income and having a non-working partner negatively associated with fathers’ perceived work-family conflicts. Conclusion: These findings have strong implications for family-supportive practices and policies that are yet to focus on fathers in their difficult position between work and family obligations.


Introduction
The past few decades have witnessed a drastic surge in men's expected and actual involvement in the family, spurred by increasing female participation in the labor force and growing evidence on the father's pivotal role in child development (Bianchi & Milkie, 2010;Cabrera et al., 2000).At the same time, men continue to see work as part of their core identity (Christiansen & Palkovitz, 2001) and provision as an essential component of positive, responsible fathering, regardless of their financial situation or family structure (Gatrell et al., 2015;Ranson, 2012;Riggs, 1997).International statistics suggest that actual working hours (men and women combined) have remained stable in most developed countries and increased in many developing countries (Lee et al., 2007).Yet parental time spent with children have increased for both fathers and mothers, regardless of parental employment status (Gauthier et al., 2004;Sayer et al., 2004).In this new context, fathers often find themselves juggling work, marriage, and parenthood at the same time, and struggling to fulfill these demanding roles: The dilemma to "have it all" is no longer uniquely female (Perry-Jenkins & Gerstel, 2020;Slaughter, 2012;Winslow, 2005).
The conflicts between work and family, where participating in one role makes it more difficult to carry out the other (Greenhaus & Beutell, 1985), can lead to considerable stress that compromises performance in both domains and in individual well-being (Amstad et al. 2011).Beyond individual outcomes, parents' work-family conflicts have been found to jeopardize couple relationship (Yucel & Latshaw, 2020;Aycan & Eskin 2005;Crouter et al., 2001;Fellows et al. 2016), deteriorate parent-child relationship (Nomaguchi & Milkie, 2020;Dinh et al, 2017;Crouter et al., 1999;Repetti & Wood, 1997), and disrupt the work-family interface of their partners (Fagan & Press, 2008).Given men's shift towards the caregiver-provider dual roles and the pervasive influence of work-family conflict, the prevalence and correlates of the perceived tension between work and family for fathers warrant further research.
To address this under-studied issue, we created a typology of work-family conflicts based on both theoretical inferences and the results of a cluster analysis.This typology identifies fathers who are free from work-family conflicts, fathers who feel primarily work-burdened or family-burdened, and fathers who perceive tension in both directions.We then followed a resources-and-demands approach to systematically investigate factors from both work and family domains that evoke and/or buffer against perceived conflicts.Using unique data from a large-scale survey in Germany, we explore factors that associate with fathers' chances of feeling work-burdened, family-burdened, or to be "sandwiched" between these spheres.Our study thus contributes to the field of work-family interface by focusing on both demands and resources from work and family domains among German fathers, of whom the empirical evidence is still lacking.

Work-to-family and family-to-work conflicts: The typology
The early research on work-family interface typically measured the inter-role conflict between work and family spheres as one single construct and conceptualized the boundary between the two spheres as equally permeable from either direction (Frone et al. 1992).Later studies explicitly distinguish the work-to-family and family-to-work conflicts as interrelated yet separate constructs, each with its own sets of determinants and outcomes (Byron, 2005;Mesmer-Magnus & Viswesvaran, 2005).The two types of conflicts are also arguably gendered, with work-to-family conflict more likely to affect fathers, and family-to-work conflict more likely to burden mothers (Allen & Finkelstein, 2014;Hill et al., 2003;Keene & Reynolds, 2005).
The differentiation of the direction of work-family conflict adds more conceptual nuance, but this variable-centered, segregated approach also has its limits.The extant literature on the tension between work and family has been asymmetrically dominated by studies on work-to-family conflicts, an imbalance that is often attributed to the relatively more rigid demands from the workplace (Bianchi & Milkie, 2010).Rather recently, the tremendous physical and emotional labor required for childrearing and couple relationship, which can also intrude into the work domain, received more attention (Hays 1996;Nomaguchi & Milkie, 2020;Tausig & Fenwick, 2001).In addition, research indicates that, due to the expansion of digital technologies, the work sphere might become even more permeable by family issues than the other way around (Bernhardt & Zerle-Elsäßer, 2021;Wajcman et al., 2008).The separate investigation of work-tofamily and family-to-work conflicts also overlooks the fact that individuals are whole persons to whom both conflicts may co-exist, which applies to today's men and women who are practicing increasingly fluid gender roles.Thus, an approach that simultaneously considers the influence of work-to-family and family-to-work conflicts yet still differentiates between them is likely to better capture the experiences of today's working parents.However, few studies so far have examined the constellation of work-to-family and familyto-work conflict of the same individual (Leslie et al., 2017;see Winslow, 2005 for an exception).In this study, we took such an approach by creating a typology of work-family conflict constellations to better understand the joint and different influence of both types of conflicts as well as their correlates.

Predictors of work-life conflicts: Demands and resources
Although conflicts and interference continue to be a central focus of the work-family scholarship, research also recognizes the potentially enriching effect of work-family interface.In the early days of work-family scholarship, Crouter (1984) already pointed out that skills learned at home could be transferred to work settings, thereby bringing an "educational spillover".Barnett and Hyde (2001) added in their expansionist theory of work and family that having multiple roles can be beneficial for both men and women through mutual buffering, additional income and social support, expanded frame of reference, and increased selfcomplexity.Greenhaus and Powell (2006) further suggested that participation in multiple roles can bring added happiness and well-being, buffer the distress in one particular role, and lead to positive instrumental and affective transference across the roles.In addition, Carlson and colleagues (2009) argued that managing multiple roles can eventually contribute to improve performance in both domains, thereby leading to workfamily efficiency enrichment.Among the various work-family theories, the demand-resource perspective proposed by Voydanoff (2005) recognizes both the possibility for conflict and the potential for enrichment in work-family interplay and offers a framework for analyzing diverse factors in work and family domains (Bakker & Demerouti, 2007;Grandey & Cropanzano, 1999).Demands are hereby defined as "role requirements, expectations, and norms to which individuals must respond or adapt by exerting physical or mental effort", and resources as "structural or psychological assets that may be used to facilitate performance, reduce demands, or generate additional resources."(Voydanoff, 2005, p. 78).
So far, a small yet growing body of research has investigated how demands and resources from work and family domains influence parents' work-family interface, and which factors are specifically or particularly relevant to fathers (Milkie & Peltola, 1999;Hill, 2005;Parker & Wang, 2013).Among workrelated demands, long work hours have been argued to bring about fatigue, magnify the impact of other stressors at workplace, limit parents' available time for the family and ergo their opportunities to nurture high-quality family relationships (Gray et al., 2004), which all lead to perceived conflicts.The negative impact of long hours has been repeatedly supported by existing research on both parents (Byron, 2005;Major et al., 2002;Roxburgh, 2006), although Milkie and Peltola (1999) found that long work hours are associated with only the fathers' (not the mothers') perceived work-to-family imbalance.Beyond contracted hours, today's workplace sometimes requires additional time investment or accessibility, such as overnight travels, long commute, and phone/email availability during off-work hours.Frequent work travels have been found to positively correlate with perceived work-family conflict (Voydanoff, 2005), and long commute, with perceived time-based conflict (Henz & Mills, 2015) and lower well-being (Stutzer & Frey, 2008) for both sexes.Although the developing information technology (e.g., cellphone, emails) allows greater flexibility in work time and location (Chesley, 2005), the expected constant availability has been found to increase negative work-family conflict and lower family satisfaction (Olson-Buchanan et al., 2017).Furthermore, the expectation for employees to remain available after work hours is not only an apparent intrusion from work into family (i.e., work-to-family conflict) but may also lead to perceived distraction or disruption by the family whilst working from home (i.e., family-to-work conflict).Overall, however, few studies so far have thoroughly investigated the effects of these boundary-spanning work characteristics on parents' work-family interface.
Meanwhile, the workplace can provide work-related resources that may help individuals manage the interface between the professional and family spheres.One such example is job autonomy, defined as individual power and control over how the job should be done (Roxburgh, 2006).High-autonomy jobs typically involve non-routine tasks and schedule control.Job autonomy has been found to boost individual performance in work by promoting empowerment and professionalism at work (Pearson & Moomaw, 2005).High job autonomy has also been associated with lower distress and work-to-family conflicts among females (Grönlund, 2007;Lennon & Rosenfield, 1992), lower perceived parental and spousal time pressure for both fathers and mothers (Roxburgh, 2006), and higher perceived satisfaction, greater positive work-family spillover and lower work-family conflict for both sexes (Thompson & Prottas, 2006).Job autonomy might be a work resource that is particularly relevant to fathers: Fathers self-reported to have jobs with greater autonomy than mothers (Haines et al., 2019); job autonomy has also been found to foster positive, accepting parenting style among fathers, but not mothers (Grimm-Thomas & Perry-Jenkins, 1994;Whitbeck et al., 1997) The associations between various family demands and parents' perceived work-family conflict appear to be more complex.As expected from a scarcity perspective, Byron's (2005) meta-analytic study suggested that time investment in family, housework, childcare or other nonwork-related activities is related to higher family-to-work conflict.Several other studies including both men and women found that the presence of young children increased the likelihood of parents' work-family conflicts (Allen & Finkelstein, 2014;Milkie & Peltola, 1999;Nomaguchi & Fettro, 2019).
Notably, empirical evidence on the impact of family demands on perceived work-family conflicts, which is well validated among mothers, has been inconsistent among fathers.Baruch and Barnett's (1986) research suggested that fathers' participation in childcare did make them feel more competent as a parent, but also less satisfied of their time use and more concerned about the potential interference of their family responsibilities on their career.However, Hill and colleagues (2003) found that fathers who shared more childcare reported higher work-family fit, opposite to the trend among the mothers.Nomaguchi (2009), too, found negative associations between fathers' time spent with children and perceived work-family conflict among two cohorts of American parents (although the direction of such conflict was not clear from the single-item measure used).The different patterns of finding might be attributable to the nature of parentchild engagement, such as interactive "quality time" (often performed by fathers and on parents' off-work days) versus routine care (often shouldered by mothers and constitute a major part of parent-child interaction on weekdays) (Brown et al., 2018;Yeung et al., 2001).Milkie and colleagues (2010) found, for example, that the former fosters a sense of work-family balance whereas the latter leaves (non-college educated) fathers feel imbalanced.
The family life, while demanding, can also offer tangible and intangible support for individuals who are juggling multiple roles.Positive emotions during interactive family events, such as shared meals or leisure activities, could contribute to emotional well-being (Offer, 2014).Good relationships with family members not only reduce the amount of emotional labor required at home, but might also offset, at least in part, the exhaustion and negative affect arising from professional work.Among the few studies that have examined the effect of time investment in couple relationship on work-family conflicts, Voydanoff and Donnelly (1999) found significant association between spousal time and psychological distress among parents with teenagers, suggesting that a moderate amount of time spent with spouse is related to low psychological stress.Moreover, positive family relationship and role quality were found to contribute to positive family-towork spillover for both sexes (Pedersen et al., 2009;Stevens et al., 2007).Good marital quality and spousal support have been found to increase job satisfaction (Rogers & May, 2003) and decrease family-to-work conflict (Aycan & Eskin, 2005) for both fathers and mothers.In summary, both the work and the family domains can exacerbate or buffer the tension perceived by individuals.

Father's work-family conflicts: The German case
Among the world's major economies, Germany presents an interesting mixture of relatively strong familyfriendly policies and moderate conservative-familialistic welfare regime and work culture (Adler et al., 2016).To encourage father involvement, the German federal government implemented a generously paid statutory parental leave in 2007, which offers up to 12 months of leave for one parent, or up to 14 months if both parents share the leave (Reimer et al., 2018).During the leave, parents also receive a benefit that amounts to 65% to 67% of their pre-leave income.These policies indeed raised fathers' leave uptake from 3.5% by 2006 to around 36% by 2015(Federal Statistical Office of Germany, 2017).At the same time, the male-breadwinner norm remains strong.Germany has the third highest gender pay gap in Europe (Federal Statistical Office of Germany, 2021), a tax and social security system that favors single-earner families (Jentsch & Schier, 2019;Jurczyk et al., 2019), and a shortage of public childcare resources especially for families with pre-kindergarten and school-age children (Hüsken & Alt, 2017).In addition, the German workplace has a strong emphasis on physical presence, which leaves little flexibility to combine work and family demands, especially for fathers (Possinger, 2013).As a result, most German fathers resign from their right to parental leave or only take up to two months off (the so-called 'daddy-months') around childbirth (OECD, 2017) before resuming to work full-time.German mothers, on the other hand, tend to re-enter the labor force in shorter part-time (i.e., around 20 hours per week) after parental leave (OECD, 2017), although longer part-time work gained significance among German mothers during the last years (BMFSFJ, 2020).In this way, mothers continue to shoulder most child-and house-care duties, leaving the typical "male breadwinner, female caregiver" model unchallenged.
Consequently, fathers' "agency gap" (Hobson & Fahlén, 2009) between the wish to actively participate in childrearing and their actual engagement in childcare is much higher in Germany than in other countries with similar or different welfare schemes (Adler & Lenz, 2016).Although the lack of work-family reconciliation -especially for men -is commonly blamed for German father's low involvement, workfamily conflicts are traditionally classified as a women's issue in social and scholarly discourses in the German context.Apart from a few recent qualitative (Oechsle & Reimer, 2016;von Alemann et al., 2017) or quantitative studies (Bernhardt & Bünning, 2020;Bünning, 2016;Hipp et al., 2017), there is no nation-wide survey so far that investigates German fathers' work-family conflicts or their contributing and protective factors.
The rising significance of men's work-family interface and the limited evidence on its correlates in the work and family domains call for more research on father's work-family conflicts.These sociopolitical and cultural contexts of Germany make it a particularly fascinating case to investigate fathers' challenges and resources to reconcile work-and family duties.The present study thus addresses the conceptual and empirical gaps in the work-family and fatherhood literature by investigating the perceived work-family conflicts of today's German fathers and the relevant work and family demands and resources.In light of the existing literature and considering both work-to-family and family-to-work conflicts, we hypothesize that, for work-related characteristics: 1. Fathers with long work hours or frequent overnight travels would be more likely to experience work-to-family conflict (but not family-to-work conflict).
2. Fathers with long commute would be more likely to experience work-to-family or family-to-work conflict, or conflicts from both directions.(This is because long commutes may be a workplace characteristic but can also result from the family's living preference, such as in Haines et al., 2018, or a combination of both) 3. Fathers who are expected to remain available for work responsibilities during off-work hours would be more likely to experience work-to-family or family-to-work conflict, or conflicts from both directions.
4. Higher job autonomy would be associated with decreased work-to-family or family-to-work conflict, or conflicts from both directions.
For family-related characteristics, we expected that: 5. Fathers' involvement in childcare would relate to fathers' perceived work-family conflicts, and the association depends on the timing of involvement.Specifically, fathers' childcare involvement on Sundays would relate to reduced work-to-family or family-to-work conflict, or both, whereas fathers' childcare involvement on weekdays would relate to fathers' elevated family-to-work conflict (but not work-to-family conflict).
6. Fathers' time spent with their partners would relate to lower work-to-family or family-to-work conflict, or both.
7. Fathers' enjoyment to spend time with the family and satisfaction with the couple relationship would relate to lower work-to-family or family-to-work conflict, or both.

Sample
This study uses data from Growing Up in Germany II: Everyday Life (Aufwachsen in Deutschland: Alltagswelten II; www.dji.de/aida;AID:A II thereafter), a nationwide German panel study on the organization of everyday family life.The dataset contains rich information on work and family demands and resources thanks to its multi-actor design and its interdisciplinary approach, and offers unique insight in German mother's and father's work-family issues.The survey sampling followed a "target person" approach that randomly drew persons under 33 years from resident registry data.For minors under 18 years of age, their parents (typically the mother) were contacted as the main informant of the family.In Wave 2 of the survey (2013)(2014), the other parent of the child (typically the father) who lives in the same household as the target child was contacted as the secondary informant.The primary informant (typically the mother) answered all questions and provided demographic data, the secondary informant (typically the father) answered questions on couple relationship, family life (e.g., family climate and routines), division of paid and household labor.A subsample of secondary informants (typically the father) received additional questions on their time use, gender role, and employment and work conditions.
This study analyzed families from AID:A II that 1) have at least one child under 18 years of age; 2) where both biological parents live together with the child; 3) in which the father belongs to the subsample that received all questions.These criteria were used to include only families with childcare needs and to avoid the confounding influences of the family structure.The final sample consists of 5,226 nuclear families with at least one child under 18 years of age, whose basic demographic information is presented in Table 1.This sample included primarily upper-middle class families, as seen in the relatively high parental education (65.2% fathers and 65.5% mothers educated to the equivalence of academic high school or above).In this sense, this sample over-represents the better-educated fraction of the German population, among which only 33.5% have received education equal to academic high school or above (Federal Statistical Office of Germany, 2022).Therefore, parental education level was included as a control variable in later analyses.Despite the high education level of the mother, the father remains the primary breadwinner, with 47.7% families having a full-time working father and a part-time working mother and another 31.3%families having a full-time working father and a stay-at-home mother.In only 13.1% of the families, both parents work fulltime.This distribution differs only slightly from the general German population with a maternal employment rate of 66.8% and nearly the half of the employed mothers with children under 15 years of age working part-time (OECD, 2017, p. 46).Most families have two children (52.2%), although this sample has an above-average proportion of families with three or more children (30.7%) because it was the child, not the family, that has been randomly selected in public registers.

Outcome Variable: Fathers' Work-Life Conflicts
Fathers' work-life conflicts were measured by two questions that ask the fathers about their perceived workto-family and family-to-work conflicts on a six-point Likert-type scale.In relation to work-family conflicts, people were asked to (dis)agree with the statement 'I find it difficult to fulfil my family responsibilities because of my work.'The family-work conflict was measured by (dis)agreement with the statement 'I find it difficult to fulfil demands at work because of my family responsibilities.'The response scale ranges from 1 ('I totally agree') to 6 ('I do not agree at all').To emphasize that fathers may have conflicts not only in one direction, but also in both directions, we subjected the fathers' answers to a two-step cluster analysis that aims to identify the potential groups.In a cluster analysis, similar cases are grouped together in relation to several variables of interest (in this study, two variables on fathers' work-to-family conflict on the one hand and fathers' family-to-work conflict on the other) to generate clusters with greatest possible heterogeneity.Cluster analysis is therefore well suited for identifying conflict types based on both work-to-family and family-to-work conflicts in this study.
After examining the cluster characteristics of cluster solutions ranging from two to five clusters, a fourcluster solution was found to be optimal.The identified clusters included: "the Balanced" (low work-tofamily and low family-to-work conflicts); "the Work-burdened" (medium to high work-to-family conflict but low family-to-work conflict); "the Family-burdened" (medium family-to-work and work-to-family conflict, but as the only case across all clusters, with the average family-to-work conflict significantly greater than the average work-to-family conflict,); and "the Sandwiched" (high work-to-family and high family-to-work conflict).The fathers' membership in a particular cluster was then used as the dependent variable and the descriptive statistics of the clusters are presented in Table 2. Intrusive work demands.The extent to which the father's work intrudes into family life was measured by fathers' self-reports of the frequency at which they are expected to 1) work in their own time, and 2) stay available for work beyond contracted hours (via phone, emails and such) (1 = never, 4 = very often).Their initial responses were aggregated to form a composite indicator, with higher score indicating greater intrusiveness.
Mobility demands.The mobility demands of the father's work were measured by maternal reports of the father's daily commuting time and frequency of overnight work travels.The number of nights the father needs to spend away from home due to work-related travels was later categorized into "never" (= 1), "seldom" (= 2; 1-3 nights per month), "sometimes" (= 3; 4-10 nights per month) and "often" (= 4; 11 nights or more per month).
Autonomy of father's job.Autonomy of the father's job was measured by a predefined classification system which categorizes jobs on a five-point scale according to the task complexity, independence, leadership and decision-making power required (1 = low autonomy; 5 = high autonomy) (Hoffmeyer-Zlotnik, 2003).

Predictors: Family demands and resources
Paternal involvement in childcare.Fathers' involvement in childcare was measured by fathers' reports of the absolute time they spend on interacting with the children on weekdays and not-working days as well as their relative share of various childcare tasks.Specifically, fathers' relative share of childcare was measured whether each childcare task was mostly done by mother (= 0), equally shared (= 1), or mostly done by fathers' (= 2).Answers to these questions were aggregated, yielding a total score between zero and eight.
Couple and family time.Fathers' time investment in couple and family activities was measured by father reports on how often they spend time with their spouses alone without the children, and how often they spend time as a family (1 = never; 6 = every day).
Family relationships.Fathers' family relationships were measured by fathers' self-reported satisfaction with couple relationship (1=not at all satisfied; 4 = very satisfied) and their enjoyment of spending time with the family (1 = never; 6 = always).

Control variables
Families vary in the amount of family work required and the available resources to meet such demands.Having young children and/or many children at home have been consistently found to be burdensome for parents (Adkins & Premeaux 2012;Dilworth 2004).Spousal employment might either render a negative cross-over effect through contagious stress (Yucel & Latshaw, 2020;Fagan & Press, 2008) or promote individual's sense of successfully managing work and family by encouraging greater involvement in family life (Milkie & Peltola, 1999;Roeters et al., 2009).Therefore, the participating families were asked how many children they have in the household and the age of every child.Their answers were categorized by the number of children in the household (1, 2, 3 or above) and the age of the youngest child (0-2, 3-5, 6-11, and 12-18).The age of the youngest child was categorized according to the common cut-off point of children's developmental stage, which typically associates with their need for care.The family socioeconomic status was measured the adjusted family monthly income, which was calculated according to a standard OECD scale and further categorized into five bands (1 = less than €1,000 Euro, 2 = €1,000-1,499, 3 = €1,500-1,999, 4 = €2,000-2,999, 5 = €3,000 or more).Parental education in the German education system which is very special was categorized into three bands (equivalents of 1 = middle school, 2 = secondary school 1 , and 3 = academic high school or above).Parental employment constellation was categorized into four types (1 = father and mother full-time; 2 = father full-time, mother part-time; 3 = father full-time, mother not working; 4 = other combinations).

Analytic strategy
The descriptive statistics were summarized in Table 1.The predictor variables were then entered into a multinominal logistic regression model, with "the Balanced" type as the reference group.For each categorical predictor variable, the category that is most common in this sample was used as the reference category (i.e., for parental education, academic high school graduate; for parental employment, father full- The middle school and secondary school are different in Germany's tripartite education system.The middle school is the lowest tier of lower secondary education and provides a basic general education from Grade 5 up to Grade 9 or 10, whereas the secondary school, representing the intermediate track in secondary education, provides vocational training in trade, technical and administrative professions up to Grade 10.See Phillips, 2000 andSchneider, 2008.time/mother part-time; for family income, €1,500-1,999/month; for child number, 2 children; for age of youngest child, 6-12 years; for father's actual work hours, 36-40 hpw; for father's commuting time, 0-0.5 hours).

Descriptive statistics
In this sample, 28.6% fathers reported little or low work-family conflict in either direction, constituting the group of "the Balanced" (see Table 2 for descriptive statistics).The negative interference was clearly more severe from work to family: As much as 38.2% fathers indicated that they suffered from medium or great work-to-family conflict but little family-to-work conflict ("the Work-burdened"), whereas 19.8% were on a medium conflict-level in both directions.This group was named "the Family-burdened" because it is the only cluster that experienced greater family-to-work conflict than work-to-family conflict.The rest of the fathers (13.4%) self-reported to be under great stress in both directions ("the Sandwiched"), with higher perceived conflict in both directions compared to all other groups.Fathers in our sample typically worked full time in jobs that belong to band three (30.8%),four (45.7%), and five (12.4%) on the autonomy scale.Over half of the fathers (52.0%) worked over 40 hours per week and 30.7% work 46 hours or more.Although most fathers were not burdened by work travels and commuting demands, 41.2% fathers "often" or "very often" had to complete work-related tasks in their own time, and 43.2% fathers "often" or "very often" had to remain available via phone and email beyond contracted hours.
The surveyed fathers devoted some but limited time to childrearing, spending more time interacting with their children on not-working days (75.1% fathers three or more hours) than on weekdays (53.1% fathers between one and three hours, and 25.0% less than an hour).The mothers still shouldered the majority of childcare, doing more than the father on all tasks except playing and activities with the child, which was "equally shared" among 74.2% families according to fathers' report.A similar trend was observed for housework, with the mothers shouldering most cooking, shopping, cleaning, and laundry (46.2% to 78.9%).Fathers had even less time for quality couple life: 35.0%fathers reportedly "almost never" spending time alone with their partners without children.Instead, time was prioritized for family activities, with approximately three-fourth (74.7%) of the fathers spending time on family activities at least once or twice a week.Despite the time shortage, most fathers were rather satisfied (43.4%) or very satisfied (42.1%) with their couple relationship and were "always" (61.2%) or "often" (37.8%) happy to spend time with their families.
Fathers' work-to-family and family-to-work conflict were significantly intercorrelated (r = .312,p < .05),but there was no collinearity (r ≥ .75)among the predictor variables, or between predictor and outcome variables.The correlations among all continuous variables were summarized in Table 2.

Multinominal Logistic Regression Analyses
The multinominal logistic regression model was significant (Goodness-of-Fit = 411; pseudo R 2 = 26.1%).The unstandardized logistic coefficients and the odds ratio of all predictors of the final model with all predictors are presented in Table 3.
Fathers with a middle school education were less likely to belong to "the Work-burdened" group than "the Balanced" group, compared to their academic high school or better-educated counterparts (β = .708,p < .05).Secondary school-educated fathers were also less likely to feel pressured on both the work and the family fronts (β = .685,p < .05).Unexpectedly, fathers with three or more children were less likely to feel work-burdened (β = .736,p < .01).Fathers whose youngest children were below six years of age were more likely to be sandwiched (β = 1.463 -1.486, p < .05)than their counterparts with 6-to-11-year-olds, whereas fathers whose youngest children were already between 12 to 18 years of age were less likely to fall into "the Family-burdened" (β = .489,p < .001) or "the Sandwiched" groups (β = .324,p < .001).Having a stay-athome spouse appeared to have alleviated fathers' perceived conflict, as fully-employed fathers with nonworking spouses had a slightly yet significantly lower likelihood of falling into "the Work-burdened" or "the Family-burdened" categories (β = .724-.765, p < .05).Compared to families with a 1,500-to-2,000 Euro monthly income, fathers from families in the lowest income bracket were more likely to feel work-burdened (β = 1.629, p < .05)or sandwiched (β = 1.929, p < .05),whereas the relatively affluent ones (€2,000/month or more) were less likely to experience the double bind of work and family (β = .428to .687,p < .05)..261% Father in the "Balanced" Group 28.6% Note: 3,439 cases in the final model.Controls: Fathers' educational attainment (s.), mothers' educational attainment (n.s.), number of children in the household (s.), age of the youngest child (n.s.), family income (s.), parental employment pattern (s.).
Father's work hours were closely related to their perceived work-family conflicts: The father working more than 40 hours per week were significantly more likely to belong to "the Work-burdened" group than to "the Balanced" (β = 1.284 to 2.032, p < .05),compared to those who work full-time but without the extra hours (36 -40 hpw).Extreme long hours (more than 46 hours per week) were also associated with a higher likelihood of being under pressure from both sides and the same time (β = 2.075, p < .001).Fathers' time investment around work such as availability for work after hours or working on holidays, frequent overnight work trips and commuting time were similarly related to fathers' perceived work-family conflicts: Fathers who need to respond to work demands after scheduled hours or to go on frequent overnight travels were more likely to fall into "the Work-burdened" (β = 1.299-1.516,p < .05),"the Family-burdened" (β = 1.603 -1.682, p < .01),or "the Sandwiched" (β = 1.453 -2.449, p < .05)categories as opposed to "the Balanced" group.Furthermore, fathers who need to commute for more than half an hour to work were significantly more likely to fall into "the Work-Burdened" or "the Sandwiched" group than to "the Balanced" (β = 1.410 -3.688, p < .01),compared to those whose workplace was within a 30-minute trip.Surprisingly, the irregularity of father's work location was unrelated to father's experience of being burdened by work or by the family (β = .719-.831, p > .05).The prestige of the job did not have a significant association with fathers' chance to be "the Family-burdened" or "the Work-burdened", although it made the fathers marginally less likely to fall into "the Sandwiched" group (β = .867,p < .10).
Corroborating the correlation results, fathers' involvement in the family life did not seem to interfere with their work: More time spent on childcare, couple life, or family activities, or having greater share of childcare or housework, did not make it more likely for fathers to be in "the Family-Burdened" group.On the contrary, the more the father was engaged in childcare on weekdays, the less likely was it for the father to experience strong family-to-work conflict (β = .845,p < .001).Father's workday involvement in childcare, father's share of childcare responsibilities, and fathers' time investment in couple time and family activities also made them less likely to be "the Work-Burdened" or "the Sandwiched" as opposed to "the Balanced" (β = .759-.848, p < .05).Notably, father's engagement in childcare on not-working days did not have significant correlation with their perceived work-family conflicts.Finally, fathers' perceived family relationship provided a powerful buffer against work-family conflicts: Better couple relationship made it more likely for fathers to be in "the Balanced" group than in any other group (β =.575 -.802, p < .01).Similarly, the more the father enjoyed spending time with the family, the less likely he would belong to "the Family-burdened" or "the Sandwiched" group (β = .651-.662, p < .01).

Discussion
The present study is one of the few empirical investigations that specifically address the work-family dilemma of modern fathers, whose everyday struggles during the balancing act has rarely been discussed.Our findings, based on a large, relatively representative sample from Germany, revealed the severe workfamily conflicts that many fathers suffer in a context that demands high performance in both work and family domains, and identified correlates to fathers' work-family conflict among a wide range of workrelated and family-related characteristics.Taken together, these findings demonstrate the profound ambivalence and confusion that today's men have as they readjust their work and family roles, and call for further inclusion of men in family research.
Our study, first and foremost, reveals fathers' challenge to "have it all" or to "keep the balance": More than 70% fathers reported experiencing medium to great work-family conflict in at least one direction.This finding echoes other existing studies in the German population which found that German fathers now highly prize their role as a caregiving parent in addition to their role as a breadwinner (Hofmeister & Baur, 2015).The wide prevalence of work-family conflict is also similar to observations from other industrialized countries (Parker & Wang, 2013).The tension among today's fathers who yearn for further participation in family life but also cherish their professional or provider role sends a strong call for further research at a time of drastic change of men's family roles.
Work hours -both contracted time and the time invested in work-related activities such as work travels, commuting time, and the availability during after-work hours -make it more likely for fathers to experience work-family conflicts.Specifically, the association between long work hours (> 41 hpw) and fathers' workfamily conflicts aligns well with other studies that revealed the detrimental effect of overwork (Major et al., 2002;Milkie & Peltola, 1999;Roxburgh, 2006).Given the fact that a large group of fathers reported working overtime, this seems to be a leading cause of work-family conflicts for German fathers.Overnight travels, long commute to work, and the expected availability beyond contracted hours added to fathers' challenge of balancing their professional and family lives, which also echo previous studies on the impact of work schedules and conditions (Henz & Mills, 2015;Hook, 2006;Presser, 2000).The consistent association between off-hour availability and all types of work-family conflicts, which has not been explored in the extant literature, is particularly striking: The communicative technology appeared to be "time thieves and space invaders" (Towers et al., 2006) rather than the effective tool to help integrate work and family domains as it is often believed to be.While the mechanism and conditions for the impact of communicative technology on parents' well-being needs further research, these findings provide preliminary support for regulations in corporate and family policies regarding not only contracted hours, but also time for work travel, commuting, and phone/email responses.
In our sample, fathers' time investment in family activities and relationships such as childcare and couple relationship was associated with lower, not higher, likelihood of experiencing work-family conflicts.These seemingly counterintuitive findings were in fact congruent with previous studies (Aycan & Eskin, 2005;Brummelhuis et al., 2008;Hill et al., 2003).Such a pattern might have reflected German fathers' aspiration to become engaged parents and supportive partners.Participating in family work allows them to pursue this ideal in practice and experience less cognitive dissonance brought by the agency gap.It is also possible that the fathers' family involvement, at the current level, is perhaps more "rejuvenating" than demanding (Hill et al., 2003, p. 258;Milkie et al., 2010).More than 75% fathers in our sample engaged with their children for "less than one hour" and "one to three hours" on weekdays, and few fathers actually spent abundant time with their partners.Fathers also participated primarily in the relatively intermittent and interesting play activities that grant them interactive "quality time" with their children, whereas the more repetitive, exhausting childcare or housework were still left to their partners.Future studies focusing on highly involved fathers might yield interesting findings on the relations between men's actual work and family practices and work-family interface.
One noteworthy twist to the positive, "balancing" effect of paternal contribution to family work was that fathers' engagement with children on not-working days did not significantly lower their likelihood of falling into the "burdened" groups, unlike workday father involvement.The lack of beneficial effect of paternal engagement on not-working days is inconsistent with the findings from the study of Brown and colleagues (2018).It is possible that today's fathers expect from themselves something more than just a Sunday playmate who only intermittently interacts with the children when it is convenient for him.Rather, they may pursuit a fatherhood ideal that involves a full participation in day-to-day childcare.Should we entertain this optimistic possibility, our findings would serve as a reminder for today's workplace to implement more flexible policies above and beyond the perinatal paternity leave that allow fathers to fulfill their family roles on weekdays throughout their children's infancy and childhood.
Our prediction notwithstanding, job autonomy had relatively little protective effect, which could be due to the limited variance in job autonomy in this sample.This paradoxical finding might also be explained by Schieman and Reid's (2009) "stress of high status" hypothesis which argues that the benefits of highautonomy jobs might be countered by the "total absorption" required, as well as the high interpersonal conflicts and frustration involved (Bianchi & Milkie, 2010, p. 709).Lang and Nomaguchi (2016) found that professional men and women experience greater work-to-family conflict than their nonprofessional counterparts.Drobnič and Rodríguez (2011) also found among German and Spanish parents that job autonomy increases, not alleviates, work-family interference.Indeed, job autonomy significantly and positively related to fathers' work hours, work intrusiveness, and the frequency of overnights travel and having a longer commute in our sample.In fact, fathers in our sample with a middle school or secondary school degree, who tend to have lower-autonomy jobs, were less likely to belong to the work-burdened or sandwiched group, respectively.Further research using samples with diverse level of autonomy and more nuanced measures of job experiences might help yield conclusive findings on the power of job autonomy.
Are there particular "demographic profiles" for the four types of fathers?Unsurprisingly, fathers with non-working partners were less likely to perceive severe work-to-family or family-to-work conflict, and those with three or more children (which also significantly correlates with having a non-working spouse) were less likely to belong to the work-burdened group.Our results also suggested that fathers from relatively affluent families were less likely to experience work-family conflicts, whereas fathers of families in the lowest income bracket were especially likely to experience strong work-to-family conflict or feel trapped between pressures from both domains, even when other demands and resources were controlled for.The association between family income and perceived work-family conflicts is not surprising given the ample evidence for the Family Stress Model, which explains how economic hardship can be a generic stressor beyond its direct implications on work and family demands and resources (Conger et al., 2010;Hammen, 2005;Heintz-Martin & Langmeyer, 2019;Santiago et al., 2011).Family financial stress might be particularly difficult to cope for fathers who see themselves as the family provider.So far, few studies have investigated how family socioeconomic status influences parental work-family interface.Future research exploring how family socioeconomic status impacts fathers' work-family integration will be both promising and necessary.
Like others, our study also has its limitations.Firstly, our results come from the social, cultural, and institutional particularities of Germany and can only be generalized within Germany or to similar contexts.Our sample has a high proportion of well-educated, affluent families.Research on demographically more diverse samples is needed to better understand fathers' work-life interface in different socioeconomic settings.Methodologically, the dataset used one single question each for work-to-family and family-to-work conflict, which do not fully capture the complex, possibly mutually overlapping nature of the two types of stress that the respondent experienced.We also investigated fathers' individual perceptions of work-life conflicts without attending to the dyad-level effects such as how the couple's respective work and family conditions influence each other, an approach that can better reveal how work-family issues impact the family as a system (Minnotte et al., 2013;Yucel & Fan, 2019;Yucel & Latshaw, 2020).Conceptually, this study has focused on work-family conflicts without fully exploring work-family balance, or "satisfaction and good functioning at work and at home" (Clark, 2000, p. 751), which goes beyond the absence of conflicts.These issues, we believe, are potentially fruitful directions for future research.
Despite these limitations, our study still has strong practical and policy implications, especially given the economic austerity as well as the increasing job mobility in recent years, both of which can make the workplace even more demanding.Our findings highlight the harmful effects of excessively long working hours and family-unfriendly work demands, which make it significantly more likely for fathers to experience at least one type, if not both types of work-family conflict.As such, our findings call for better workplace practices and family policies that support men in their family involvement without sacrificing their career prospective.Preventative strategies for overwork (e.g., labor law and workplace culture campaigns), less intrusive work conditions (e.g., regulations limiting off-hour work-related communications), and temporal and spatial flexibilities on workdays (e.g., work-from-home options) could be considered as first steps to alleviate fathers' work-family conflicts.Such family-friendly measures may also increase the attractiveness of potential employers, as new generations of fathers place more emphasis on the reconciliation of work and family to fulfil their aspired family responsibilities.

Table 1 :
Fathers' work-family conflicts, and work and family demands and resources: Descriptive statistics(N = 5, 226)

Table 2 :
Two-step cluster analysis solution for fathers' work-family conflicts

Table 3 :
Summary of multinomial logistic regression analysis for correlates of fathers' work-family conflicts