Keywords

1 Introduction

The rise in divorce and union dissolution reshaped the social arena of fathering; many fathers no longer live with their children (Andersson et al. 2017). Life in separate households poses substantial challenges for the continuity and quality of non-resident father-child relationships, which, in many cases, are beneficial for both father and children (Amato and James 2010). Children whose fathers are more involved in parenting experience higher academic achievements, life satisfaction, less emotional distress, and fewer behavioral problems (Young et al. 1995; Braver and Lamb 2013). For men, the continuity of these relationships is also relevant in adjusting to divorce (Amato 2000; Amato and Dorius 2010), achieving higher life satisfaction, experiencing fewer depressive symptoms (Eggebeen and Knoester 2001; Kamp Dush 2013; Waldfogel and Ehlert 2016), and negotiating male identity (Collier and Sheldon 2008).

Although there exists a rather large body of research on non-resident fathering (particularly in the US and some European countries), it predominantly concerns the contact frequency and father involvement in child support (Amato and Dorius 2010). Researchers paid substantially less attention to the quality of non-resident father-child relationships that, as it has been proven, is more important for children in the post-divorce environment than the quantity of time spent together (Amato and Gilbreth 1999; King and Sobolewski 2006). Some recent studies focused on these issues examining fathers’ parenting styles (Bastaits et al. 2015), their effect on child well-being (Bastaits et al. 2012, 2014), father-child emotional closeness and child outcomes (Booth et al. 2010). Nonetheless, despite the growing interest in the qualitative side of the father-child bond, this research field remains underdeveloped.

In this chapter, we focus on the quality of non-resident father-child relationships after a parental union dissolution in order to identify the father-related factors that contribute to it. We assess the relationship quality perceived by fathers and focus on both positive elements such as intimacy and approval and negative ones such as conflict and lack of parental authority in relationships. We examine how relationship quality is shaped by fathers’ resources including their personal well-being, socio-economic resources, parenting practices, and family situation. Our research is based on the cross-sectional Fathering after Union Dissolution in Lithuania survey data collected in a country characterized by a long divorce tradition and high divorce rates (Eurostat 2017). The share of children who experience parental union dissolution by the age of 15 is estimated in Lithuania at around 35% (Andersson et al. 2017). The Lithuanian divorce legislation does not entail the joint physical custody thus, after a divorce child’s place of residence is determined with one of the parents, and a non-resident parent receives the visitation rights. As a consequence, an absolute majority of children after the parental union dissolution lives with mother. However, the country has also experienced significant shifts in fatherhood discourse. The nurturing role of father gained some importance and was reinforced by social policies oriented towards more gender equal parenting in families (Maslauskaitė and Tereškinas 2017).

Our chapter contributes to this research field in several ways. First, we focus on the non-resident father-child relationship quality, which remains an under-researched issue despite the growing scholarly interest. Second, existing research predominantly concentrates on paternal relationships in the North American, Northern or Western European context. Our research shifts the perspective to the other side of Europe. We investigate one Baltic country, i.e. Lithuania in which the post-divorce legislation is strongly oriented towards the father’s economic provider role, but it supports, to the very limited extent, his nurturing role and his right to care for children after family dissolution (Tereškinas and Maslauskaitė 2019). Third, we examine the non-resident father and child relationships based on the dataset, which, in a very detailed way, records information of a post-divorce fathering in Lithuania that is, to our knowledge, the only one of this kind in the Baltic and Eastern European countries.

2 Theoretical Background, Previous Research, and Hypotheses

We draw our theoretical background from two perspectives. First, we use Smart’s (1991) conceptualization of care specifically developed to reflect the father-child relationships in a post-divorce context. Second, we rely on the recourse theory (Foa and Foa 1980) that explains the factors shaping the quality of paternal relationships.

While much research on father-child relationship quality after divorce (Bastaits et al. 2015; Bastaits et al. 2012) is based on parental style framework (Baumrind 1968), we argue that this framework has some limitations. Parenting style literature suggests that the amount of support and control provided by a parent determines parenting quality which could consequently result in an authoritative, authoritarian, permissive or uninvolved parenting style (Baumrind 1968). Authoritative fathering combines emotional warmth with rule setting and is the most beneficial for child’s needs (Amato and Sobolewski 2004; Braver and Lamb 2013). High control and low support are characteristic of an authoritarian fathering; permissive parenting style points to high support and low control while uninvolved fathering exhibits low support and low control.

However, the parental style framework has some drawbacks. First, it was developed for the analysis of married couples (Baumrind 1968) and it does not consider various constraints that shape the relationship opportunities for non-resident father. Time restrictions make it problematic to maintain relationships combining emotional warmth, support and effective control (Amato and Sobolewski 2004). In addition, visitation arrangements are often limited to weekends and thus, encourage fathers to engage in leisure activities (Pasley and Braver 2004). Consequently, many fathers develop a style of “recreational fathering” characterized by the leisure or “so-called Disneyland activities” (Amato and Dorius 2010; Stewart 1999). Second, the parental style framework does not capture tensions and contradictions of post-divorce fathering in contemporary society. In the past several decades, many developed countries experienced the normative and legal turn towards a more active role of the father after the divorce. The previously widespread pattern of “absent fathers” (Bradshaw et al. 2002) or “deadbeat dads” (Cassiman 2008), when men restrained from contacts with their children after divorce, was replaced by a more involved pattern of fathering. This has been reflected in the overall increase in the frequency of non-resident father-child contact over divorce cohorts reported for many countries (Amato et al. 2009; Westphal et al. 2014). The shift is related to the broader societal changes: a generally growing emphasis on the caring role of the father, normative shifts towards more egalitarian gender roles in the family and the introduction of more gender-neutral child custody legislation implemented across the EU and North America. However, some argue that these developments produce tensions between fathers’ rights to care established in the legislation and supported by the norms of involved fatherhood, on the one hand, and fathers’ capability to care, on the other hand.

Reflecting upon this tension, Smart (1991) applies the concepts of “caring about” and “caring for” to the fathering in a post-divorce family context. She argues that “caring for” relationships reflect the everyday activity of meeting the child’s needs and these activities include both emotional and practical work. “Caring about” expresses an intellectual concern and the abstract notion of feelings of care (Smart 1991). It has to be stressed that Smart’s (1991) conceptualization of care differs from the one suggested by other authors who confine “caring for” to practical care, while “caring about” – to emotional care (Calasanti and Slevin 2001). Thus, fathers might be inclined to “care about” their children and this inclination is driven by legal and cultural shifts related to post-divorce fathering. Nevertheless, men might lack the capacities, skills and personal resources to “care for”, because it is mothers who are predominantly involved in this type of care before separation (Smart 1991). “Caring about” might mean more frequent meetings with non-resident children that are not accompanied by “caring for” practices in relationships.

Although Smart’s (1991) concepts have not yet been applied to the quantitative empirical research, they offer significant insights and complement the parenting style approach (Baumrind 1968). The parallels could be drawn between the “caring for” fathering and parenting styles beneficial to the children. Parents’ emotional and practical involvement in the child’s everyday life contributes to the higher levels of intimacy in relationships, paternal approval and, thus, the higher levels of support. In contrast, the “caring about” fathering distinguished by a low personal and emotional involvement results in lower levels of paternal authority and effective control. We have to recognize that even if “caring about” relationships are less beneficial to the child, they might still be ultimately more advantageous than no father’s contact with the child. However, due to the empirical limitations, this type of relationships is beyond the scope of this research. Thus, hereinafter we will use the notions of “caring for” and “caring about” father-child relationships. The first notion indicates the higher levels of intimacy and support as an outcome of higher personal and emotional involvement in childrearing. The second notion points to lower paternal authority, which in the everyday life might manifest in a loss of control over the child and more conflicts in the relationships.

The other framework, which shapes the theoretical background of the research, is the resource theory that uncovers factors associated with the quality of paternal relationships. The general underlying assumption suggests that the higher the amount of resources possessed by a person, the more likely they are to be shared with others (Foa and Foa 1980, p. 93). In addition, the more resources shared, the better are non-custodial father-child relationships. The resources encompass the father’s personal well-being, perceived economic and social status resources, communication with the mother and co-parenting (Rettig et al. 1999). In the following, we will formulate our research hypotheses based on the theoretical assumptions and the existing research evidence.

Firstly, the research on divorced fathers’ personal well-being indicates that divorced parents generally experience a higher risk of depression, unhappiness, and health problems (Braver and Lamb 2013). Lower emotional and psychological well-being is associated with the negative parenting strategies and lower level of responsiveness to the child’s needs (Pruett et al. 2003). Hence, we may expect that higher levels of non-resident fathers’ personal well-being are positively associated with the “caring for” type of relationships and negatively with the “caring about” type of relationships (Hypothesis 1).

Secondly, following the resource theory, we argue that socio-economic resources are relevant in maintaining the higher quality of nurturing relationships after a partnership dissolution. Higher educated fathers will be more aware of the child’s developmental needs thus, they will be more successful in pursuing the relationships with the higher level of support (Bastaits et al. 2015). Higher educated fathers also have better communication skills and abilities to manage the relationships and better conflict solving skills (Amato and James 2010) beneficial for the non-resident father-child relationship quality. Fathers of a higher socio-economic status might be also more conscious of the adverse effects of divorce on child development and thus might put more effort to sustain the close paternal bond with the child. Moreover, educated fathers are also more involved in childrearing tasks prior to divorce (Hook and Wolfe 2012), therefore, they are most likely to have stronger dedication to actively participate in the child’s upbringing after a partnership dissolution. The father’s economic resources are also relevant to child maintenance duties that are an important marker of the father’s involvement in a child’s life (Carlson and McLanahan 2006; Kalmijn 2015). This leads to our second hypothesis according to which higher socio-economic resources will positively contribute to the “caring for” type of relationships and will reduce the occurrence of “caring about” type of relationships (Hypothesis 2).

Thirdly, in reflecting on the relationship quality one should also consider fathering practices, which are routine action men perform to exchange resources and to fulfill their role as non-resident fathers. By spending time with their children non-resident fathers manifest their availability to them (Lamb 2004). Visitation for non-resident fathers is the only opportunity to engage with children and to know their needs and worries as well as to exercise their paternal authority and control. Child alimony payments might also have an impact on the relationship quality because financial contributions reflect fathers’ responsibility for the child’s material living conditions and the continuity of the fathers’ role as economic providers. Besides, the child support and contact frequency are interconnected, thus fathers who pay child support meet their children more often (Nepomnyaschy 2007). Additionally, the exchange of resources in post-divorce context is embedded in co-parenting relationships. Parental conflict leads to low levels of co-parenting and strengthens the maternal gatekeeping (Allen and Hawkins 1999) that sets serious limits to the development of the “caring for” relationships. Therefore, our third research hypothesis suggests that involved fathering practices will lead to “caring for” relationships, while uninvolved fathering practices will be positively associated with the “caring about” relationships (Hypothesis 3).

Fourthly, father’s re-partnering and new children might negatively affect the sharing of resources with children from his previous unions. Men with children from different partnerships might experience competing time, financial, and emotional demands, therefore, it could be more difficult for them to build nourishing relationships with children from their previous partnerships (Swiss and Le Bourdais 2009; Manning et al. 2003). Re-partnered fathers “swap” old children for the new ones; they invest in new children, while their offsprings from previous unions receive low paternal support and control (Furstenberg Jr and Nord 1985). Evidence suggests that “swapping” occurs only when men have to choose between new biological children and non-resident biological children (Manning and Smock 2000). Moreover, some argue that remarriage could be beneficial for father-child relationships. A new partner might encourage the paternal involvement with non-resident children supporting father’s responsibilities and taking care of the household duties (Hetherington 2006). In addition, re-partnered fathers might be more involved, because they have more economic resources compared to single divorced men (Seltzer 1991). Furthermore, re-marriage signals father’s attachment to the traditional family form and it encourages his paternal commitments to the non-resident children (Cooksey and Craig 1998). Thus, we expect that fathers’ family transition will in a negative way affect “caring for” relationships only for re-partnered fathers with new biological children; moreover, we expect that these fathers will experience more “caring about” type of relationships (Hypothesis 4).

3 Data and Methods

Our analysis is based on the representative Fathering after Union Dissolution in Lithuania survey of non-resident fathers with under-aged children in Lithuania (N = 1500). The survey was conducted in 2016. The sample was obtained by using a stratified sampling method. The respondents were men with non-resident children under 18 years of age from dissolved cohabitations or marriages. Face-to-face interviews were carried out with respondents in their homes by using a standardized questionnaire. The survey recorded a wide range of themes related to the men’s life course events, including partnership and fertility histories, divorce process and post-divorce relationships with child’s mother, respondents subjective and psychological well-being, child support payments, father-child contacts and the types of contacts, father-child relationship characteristics, men’s current partnerships, and socio-demographic, socio-economic and well-being indicators.

From the original dataset, we excluded men who never lived with their children and those fathers who did not contact their non-resident children in the 12 months preceding the interview. Following the questionnaire, these fathers did not report on the quality of the relationships with their children. As the consequence, our effective sample included 1225 non-resident fathers.

3.1 Dependent Variables

We measure the quality of paternal relationships by using the Network of Relationships Inventory – Relationship Qualities Version (NRI – RQV) (Buhrmester and Furman 2008). A short version of the NRI-RQV was used, which was developed in pairfam – The German Family Panel (Scales and Instruments Manual 2018). The NRI – RQV measures positive and negative dimensions in parent-child relationships and it is a self-reported instrument. The positive dimension subscale consists of 3 items and measures paternal approval and intimate disclosure. The items are “Your child tells you what he/she is thinking,” “You show recognition for the things your child does” and “You show your child that you respect and like him/her.” Negative dimension subscale measures child’s dominance and father-child conflict, which signals a lack of paternal authority and effective relationship control. The subscale includes 2 items: “Your child gets his/her way when you can’t agree on something” and “You and your child disagree and quarrel.” Each item uses a 5-point frequency scale ranging from 1 = always to 5 = never. 3 positive items were inversed thus, higher values indicate more frequent paternal approval and intimate disclosure. For 2 negative items, lower values indicate lower quality.

The dependent variables were developed in the two-stage procedure. First, for the exploratory purposes we applied factor analysis including all items of the NRI-RQV scale. Two factors were extracted: one included the items on paternal support (approval and intimate disclosure) and the other – the items on the lack of paternal authority (conflict and dominance). Based on the factor analysis summary index variables were calculated for the items with the loadings above 0.4. Two dependent variables were composed. The first indicates “caring for” paternal relationships and measures approval for child’s achievements, intimate disclosure, and respect. The second dependent variable subsumed the items on the conflict and dominance; it reflects the lack of effective control and paternal authority and indicates “caring about” relationships. Both dependent variables were standardized for further analysis.

3.2 Independent Variables

Based on the theoretical considerations and the research hypotheses the first set of independent variables measures fathers’ emotional and psychological well-being. First, fathers’ depressive feelings were measured by using the Center for Epidemiologic Studies depression scale (CES-D 8) (Radloff 1977; UNECE 2005). The CES-D 8 records the absence or frequency of depressive feelings experienced during the last week and includes seven items. The summary index variable was composed with the lower values indicating the absence of depressive feelings or lower occurrence and higher values – the more exposure to depressive feelings. Second, we included the variable of the General Life Satisfaction ranging from 1 = not at all satisfied to 10 = completely satisfied. Third, the variables of emotional and social loneliness were used. The Survey measured self-perceived loneliness with the shortened version of De Jong Gierveld Loneliness scale (de Jong-Gierveld and Kamphuls 1985). Based on the suggested technique it has been transformed into the Emotional and Social Loneliness variables (de Jong-Gierveld and Kamphuls 1985), ranging from 0 = absence of loneliness to 3 = intensely lonely. Fourth, the variable of the locus of control was incorporated. It was measured with the Locus of Control Scale (UNECE 2005) developed to assess an individual perception of the level of control in five life domains (financial situation, work, housing, health, and family life). Each item measured on a 5-point scale. Summary index variable was composed. The above-mentioned variables of well-being were standardized for multivariate regression analysis.

Paternal socio-economic resources were assessed with three variables. First, we included a variable of education with the three categories: below upper secondary (ISCED 0–2), upper secondary (ISECD 3–4), and tertiary (ISCED 6–8).Footnote 1 Second, fathers’ financial resources were assessed by using the indicator of the self-perceived financial conditions. The scale is widely used in the national surveys in order to overcome the very high non-response to questions related to the personal or household incomes. The scale ranges from 1 to 5 (1 = severe financial deprivation and 5 = financial sufficiency). The variable was standardized for the multivariate regression analysis. Third, the father’s employment status also included three broad categories: unemployed/inactive, blue-collar worker, and professional.

The post-separation fathering practices were examined by including three variables. First, the categorical variable of child alimony payments during the 12 months prior to the interview (0 = non-payment, 1 = payment). Second, the variable of face-to-face contact frequency with the child in the year preceding the interview (1 = contact once in less than 6 months, 5 = contact at least once a week). An indicator for co-parenting is father’s assessment of relationship with a child’s mother, ranging from 1 = very bad to 5 = very good. The literature suggests that mother-related characteristics might be also a significant predictor of the father’s parenting style (Bastaits et al. 2015), however, our dataset provides only a very inaccurate measure of mothers’ education and thus, it was not included into the analysis.

Fathers’ family context was measured with the categorical variable indicating the states of living single, living with a new partner without new biological children and living with a new partner and new biological children. The variable was transformed into dummy variables.

3.3 Control Variables

Child’s age and sex were incorporated into the analysis. Older children have more contact with their fathers than younger ones (Aquilino 2006). However, some argue that the child’s transition into early adolescence increases the significance of peer groups over parents and contacts recede (Amato et al. 2009). Child’s age at the parental union dissolution is also relevant predictor because more time spent in one household provides more opportunities to develop a closer emotional bond (McLanahan and Sandefur 1994). Fathers tend to be more involved with sons than daughters (Hetherington 2003), yet, the findings related to gender are inconsistent (Swiss and Le Bourdais 2009).

The geographical distance between a non-resident father and child living areas was also included as a control variable. Men living farther from their children will have fewer opportunities to interact with them and to develop high-quality relationships (Smyth et al. 2001). The variable of geographic distance comprises three categories: same locality (city, town, village), different locality, and different country.

Time elapsed after divorce is also a relevant control variable because it indicates the time available to men to adjust to a union dissolution (Amato 2000). The variable was composed of the information recorded in the partnership calendar and measured in months elapsed after union dissolution.

The descriptive statistics are presented in Tables 14.1 and 14.2. We conducted a multivariate analysis based on the stepwise linear regression models for each of the relationship quality dimensions. The baseline model considers personal well-being variables and the following models include additional sets of variables related to the research hypotheses. The final model presents all sets of independent variables and control variables. There is no collinearity of predictors in all models, we examined this by using VIF.

Table 14.1 Descriptive statistics of unstandardized continuous dependent, independent and control variables (N = 1225)
Full size table
Table 14.2 Descriptive statistics for categorical independent and control variables (N = 1225)
Full size table

4 Results

Table 14.3 presents the results of the multivariate regression analysis for the “caring for” dependent variable (intimacy and approval). Model 1 suggests that statistically significant predictors of “caring for” relationships are social loneliness, depressive feelings and the locus of control, while general life satisfaction and emotional loneliness do not contribute to the model. Lower levels of intimacy and approval are associated with higher levels of social loneliness and depressive feelings (b = −0.19 and b = −0.10). Fathers with a higher sense of control over their lives also exhibit a higher level of “caring for” paternal relationships (higher intimacy and approval).

Table 14.3 Multivariate regression results for “caring for” relationships (intimacy and approval), unstandardized B, Std. errors
Full size table

In Model 2, we added variables of the father’s socio-economic resources. The significant effects associated with the predictors of paternal personal well-being remained in place, although the effect sizes decreased slightly. In addition, the father’s employment status is associated in a positive way with relationship quality. Professionals demonstrate higher levels of “caring for” relations compared to the unemployed (b = 0.43). The same direction of the association is observed for blue-collar workers however, the effect size is smaller (b = 0.24). Contrary to our expectation, there is no link between subjectively assessed financial situation and education, and “caring for” type of relationships.

In the next step, we included variables of the fathering practices (Model 3). The frequency of contacts, child support payments and co-parenting are beneficial to the level of intimacy and approval in father-child relationships. Thus, fathers who pay child maintenance exhibit more “caring for” relations compared to fathers who restrain from this obligation (b = 0.25). The contact frequency also positively affects the level of intimacy and approval in relationships (b = 0.17). In addition, higher quality relationships between a father and mother positively contribute to “caring for” bond (b = 0.17). Model 3 also shows that after adding the variables of parenting practices the effect sizes for the variables of personal well-being (social loneliness and locus of control) slightly decreased and the predictor of socio-economic resources (employment status) became insignificant. Thus, it could be that the association between the socio-economic resources and relationship quality is transmitted through fathering practices. Employment and earnings enable fathers to pay child support, negotiate better with the mother, and spend time with children that positively affects higher relationship quality.

In Model 4 we included the independent variables of the father’s family context (partnership status, new children, and time after union dissolution). However, none of the variables seems to have an effect in predicting the level of intimacy and approval. In the last step (Model 5), we added control variables. All the above discussed significant predictors remained. In addition, the child’s age at a union dissolution contributes in a positive way to the “caring for” type of fathering. Child’s sex is not relevant. Surprisingly, the variable of geographical distance shows no difference in relationship quality when a child lives farther away compared to the one who lives closer. However, fathers with children living abroad report higher levels of intimacy and approval (b = 0.52).

Summing up, intimacy and approval (“caring for” relationships) are associated with some aspects of personal resources (well-being) and fathering practices, while socio-economic recourses and father’s family situation are not significant. Social loneliness and depressive feelings are inversely linked with “caring for” type of fathering, while fathers with a higher sense of being in charge of their lives report more intimacy and approval in relationships. In addition, a more intimate bond is manifest when fathers pay child support, have good relationships with the child’s mother and see their children more often. In talking about the association between the father’s well-being and relationship quality we have to admit that our data do not allow us to examine the direction of causality, thus, it could be that the lower personal well-being is an outcome of unsatisfying relationships with the child.

Table 14.4 presents the results of the multivariate regression analysis for the “caring about” type of relationships (conflict and child’s dominance). “Caring about” variable coded with lower values expressing lower quality, i.e. higher conflict and child dominance. Our modeling strategy replicated the one discussed above. In the first step (Model 1), we included only variables indicating the father’s personal well-being. The results show the negative association between “caring about” fathering and emotional loneliness, social loneliness, life satisfaction, while the association proves to be positive for the locus of control. Thus, fathers experiencing higher levels of emotional loneliness (b = −0.17) and social loneliness (b = −0.08) also experience more conflict and lack of authority in their relations with children. Men scoring higher on the locus of control also experience less conflict and dominance of child (b = 0.11) (less “caring about” type of relationships). Interestingly, we observe that higher life satisfaction is negatively associated with more frequent conflict and dominance (b = −0.08).

Table 14.4 Multivariate regression results for “caring about” relationships (conflict and child’s dominance), unstandardized B, Std. errors
Full size table

In the next step (Model 2), we added the variables related to the father’s socio-economic resources; nonetheless, father’s education, financial conditions and employment status show not to be significant predictors. In addition to the above-mentioned variables, Model 3 included the variables of fathering practices. Child support payment and “caring about” relationships are negatively associated. Thus, fathers paying child support experience more tensions linked to paternal control compared to fathers who withdraw from child support payment. The frequency of contacts and co-parenting are not relevant characteristics in predicting this dimension of relationship quality. It could be noted that after adding fathering practices the effects for predictors of paternal well-being remained stable.

In Model 4, we added the variables of father’s family situation and time after the union dissolution. Both regressors are statistically significant. Re-partnered fathers with new children have better relationships (lower level of conflict and less child dominance) compared to fathers in other family contexts (b = 0.27). In addition, the relationship quality decreases with the time elapsed after paternal divorce. Model 5 presents the results for all independent and control variables. A father’s well-being (emotional and social loneliness, the locus of control, and life satisfaction) anticipates the relationship quality linked to paternal authority and control. Child’s age at paternal union dissolution is significantly associated with the relationship quality. There are less conflict and authority-related problems when children are younger. In addition, the above-discussed effects of child support and time elapsed disappear after adding the child-related characteristics. The same is true for the previously observed significant association between time after divorce and relationship quality. The geographical distance between a father and a child also affects relationship quality. Children living farther from their fathers have worse relationships compared to children living in the same locality.

5 Conclusion

The chapter explored the relationship quality between non-resident fathers and their children after the parental separation. Although it has been acknowledged that relationship quality is among main moderating factors in children and fathers’ adjustment to divorce (King and Sobolewski 2006; Amato and James 2010; Waldfogel and Ehlert 2016), the issue is still under-researched (Amato and Dorius 2010). By providing evidence on perceived quality of relationships by father and on various types of father-related factors at play, our study contributes to this expanding field. Following Smart’s (1991) distinction between “caring for” and “caring about” fathering after parental union dissolution and the resource theory (Rettig et al. 1999) we argue that the father’s higher personal, socio-economic, and parenting resources contribute positively to the “caring for” or a more nurturing type of relationship. On the contrary, lower resources are associated with non-effective parenting and result in lower relationship quality and more intergenerational conflict that reflect the “caring about” type of fathering. Our study is based on the data from a recent survey of non-resident fathers in Lithuania. The dataset is the only one of this kind in the region; it includes an extensive list of indicators and thus provides the unique opportunity to investigate non-resident fatherhood in this part of Europe.

The research was guided by four hypotheses. According to our first hypothesis, higher levels of fathers’ personal well-being are positively associated with the “caring for” type of relationships and negatively, with the “caring about” type of relationships (Hypothesis 1). The hypothesis was supported. The “caring for” bond between a separately living father and a child is most likely to be maintained when men feel generally more in control of their own lives (the locus of control), they are more socially integrated (lower social loneliness) and they confront fewer depressive feelings. The “caring about” relationships characterized by conflict and lower parental authority are associated with a higher level of emotional and social loneliness and the weaker feeling of control over life. Thus, in line with the previous research, our results suggest that personal psychological well-being is the resource used in adopting positive parenting strategies and applying positive parenting skills (Braver and Lamb 2013).

Our second hypothesis stated that higher socio-economic resources positively contribute to the “caring for” type of relationships and reduce the occurrence of “caring about” type of relationships (Hypothesis 2). Surprisingly, the second hypothesis was not confirmed and this contradicts the previous findings (Kalmijn 2015; Bastaits et al. 2015). Men’s education, employment status or subjective assessment of financial living conditions are not associated with the “caring for” or “caring about” type of relationships. Even though the effect of higher employment status on “caring for” relationships was positive, it disappeared after adding the variables related to the parenting practices. Thus, it seems that the association between socio-economic resources and relationship quality is transmitted through the parenting practices. Higher socio-economic resources facilitate effective fathering practices (paying child support, more frequent visits to the child, and co-parenting) that positively affect higher relationship quality. In addition, we did not find any significant associations between socio-economic resources and “caring about” type of relationships.

In our third hypothesis, we suggest that involved fathering practices lead to “caring for” relationships, while uninvolved fathering practices will be positively associated with the “caring about” relationships (Hypothesis 3). The hypothesis was partially supported. Fathers’ report higher levels of “caring for” relationships if they are more engaged (spend more time with children) and responsible for children’s material well-being and for developing effective co-parenting relation with a child’s mother. Thus, our findings are consistent with the ones reported previously (Carlson and Turner 2010) and they support Lamb’s (2004) assumption on the role of engagement and responsibility in effective fathering. On the other hand, fathering practices do not explain the “caring about” type of relationships. The multivariate regression model that did not consider child-related characteristics indicated the significant association between fathering practice (alimony payments) and “caring about” relationships. Fathers’ paying child support reported more conflict and problems in paternal control (more “caring about” relationships). It could signal that fathers tend to exchange money for the power over the children and this could lead to more conflict. However, the effect became insignificant after adding the child’s age at a paternal union dissolution and the child’s sex. Thus, more conflict and tensions related to paternal control are associated with the child’s age. Older children might have autonomy aspirations that lead to confrontation and conflict with the non-resident father.

Finally, our fourth hypothesis assumed that fathers’ family transition negatively affects “caring for” relationships only for re-partnered fathers with new biological children; moreover, we expected that these fathers experience more “caring about” type of relationships (Hypothesis 4). The hypothesis was not confirmed. Our results do not show any significant effects of fathers’ family transitions on the “caring for” relationships. Nonetheless, re-partnered fathers with new children report less conflict and paternal authority problems compared to re-partnered fathers without new biological children. This finding contradicts not only the argument of the “family swapping” (Furstenberg Jr and Nord 1985) but also the evidence on the negative effect of new biological children (Manning and Smock 2000). We could possibly argue that new partners not only encourage men’s involvement with children in taking care of household duties (Hetherington 2006) but also provide support in solving the father-child relationship problems. In addition, it could be that re-partnered fathers with new biological children are more attached to the traditional family forms and obligations (Cooksey and Craig 1998) and thus, they invest more in successfully managing the conflicts with children.

An important limitation of our study is the inability to identify the causal direction between the factors studied and father-child relationship quality. Perhaps these links are bidirectional or simultaneous. However, considering the limited evidence on post-separation fathering in the region, we believe that our contribution highlighting the factors associated with relationship quality is particularly relevant.