Connecting Fathers: Fathers’ Impact on Adult Children’s Social Networks

Abstract

We examine the relationship between having an emotionally close and active father in an adult child’s social network compared to having a father who is not close, or a father who was not named. We hypothesize that fathers provide both essential and important contributions to their children’s psychosocial development, and those contributions continue into active adulthood. Using the 2015 UC Berkeley Social Networks Study (UCNets), we find that adult children who name an emotionally close father in their network tend to have more males as social ties, but not more female ties. We conclude that fathers continue to play an important and active role in their children’s lives long after childhood.

Introduction

Fathers are important because they can enable the successful launching of children into being autonomous and socially-integrated adults. Studying fathers’ involvement in their children’s lives is an area of research recently highlighted by Hofferth and Goldscheider (2015) as crucial. Nevertheless, fathers have often been sidelined because of the more common emphasis on mothers and family structure, such that the father’s absence rather than his presence is the focus. The body of literature that does exist on the father’s presence tends to examine three aspects of his impact on: (1) young children; (2) early life inputs, e.g., parenting time or parental divorce, and later life outcomes for adult children; and (3) the father’s social capital as reflected in the resources he provides given his education, occupation, and income. The goal of this paper is to redirect the focus to the child/father dyad in adulthood, and to its relationship with the adult children’s social support network. The research utilizes rich network data provided by the UC Berkeley Social Networks Study (UCNets).

Fathers are both an important factor in a person’s psychosocial development as well as an important contributor to their offspring’s social capital. The initial contribution of fathers emerges from attachment theory: children need to form a secure connection with a protective and nurturing significant other, and parents are the first and most common source (Bowlby, 1988). As they age, the importance of parents for the norms of reciprocity and social exchange come into play, such that the developing child learns that behavior is evaluated and has consequences. Positive actions beget similar results, and over time the child develops an understanding of successful relationships. The markers of a successful adult life include stable and rewarding social relationships and a generalized norm for repaying the caring (or negative behavior) to parents in later life. Paternal contributions to social capital include skills and knowledge, network connections, financial footing, and advice derived from experience. These factors are known to be critical in childhood. In this paper we explore how the trust developed through attachment will also affect the ability to form relationships as adults. We ask whether the valuable presence of a father continues well into adulthood.

Two hypotheses frame the understanding of the value of fathers in a child’s life: ‘essential’ father and ‘important father.’ According to the essential father hypothesis, a father, because of his gender, makes unique contributions that a mother cannot, such as stimuli for psychosocial development. An important father, on the other hand, can significantly affect child outcomes, but these contributions can be provided by other parental figures such as grandparents or stepfathers (King, Thorsen & Amato, 2014). These contributions include household income, a breadwinning role, shared discipline, quality time, and more. For example, two mothers would produce the same outcomes in children as would a traditional family. When a father is absent, the child lacks a second parent, but it is the loss of the contributions, not necessarily his ‘fatherness’ that is important (Lamb, 2010). Having a second parent involved with young children is associated with positive outcomes such as fewer psychological problems (Amato and Rivera, 1999) and greater stability (Hofferth & Goldscheider, 2015).

The effects of fathers do not stop at childhood. A central area of this article examines parental divorce and later-life outcomes. The voluminous body of research consistently indicates that growing up in two-parent stable households is the gold standard for launching children successfully into adulthood in terms of successful employment and marriage, as well as improved health (Amato, 2001). Some of this success owes to a greater stability of financial resources, with its concomitant reductions in stress and risk, e.g., greater access to health care, less frequent moves, greater household ownership, and higher income for other inputs into childhood. In contrast, parental divorce weakens access to those resources. In addition, when only one parent is present, typically the mothers, she now shoulders the bulk of the child-raising responsibility, often while working, and thus children receive less parental attention. This loss is particularly salient if the absent father ceases to invest in his children’s well-being, for example, due to incarceration, or to punish the mother. Beyond that, there is simply the absence or reduced presence of the father in the child’s life. Whether ‘essential’ or ‘important’, fathers contribute to their children’s outcomes.

Because parents are at the heart of one’s successful juvenile psychosocial adjustment, another area of inquiry extends to children’s social networks in adult life, critical for social support and expanded opportunities. Parents tend to have a positive influence on the quality of their young children’s social networks. They do so directly through arranging activities or do so indirectly through, say, choice of neighborhoods, friendship opportunities, and role-modeling (Rubin & Sloman, 1984). Fathers, particularly those with higher socioeconomic status, provide insights through experience to inform their children regarding networks for work, expertise in job-seeking and successful employment, and financial strategies, as well as parenting mentorship (Adamic & Filiz, 2017. O’Regan & Quigley, 1993). Each of those guiding contributions is correlated with successful psychosocial adjustments, so it follows that there should be positive effects of paternal contributions for their adult children’s social networks from having a present and contributing father, beyond the effect of contributions made in childhood.

Despite that foundation, few studies examine the presence and importance of fathers in the daily lives of adult children (Sharabany, Scher & GalKruz, 2006). Available studies suggest that paternal presence and closeness continue to be important for psychological well-being into young adulthood (Amato, 1994; Barnett, Marshall & Pleck, 1992). For example, adult sons who describe their relationship with their fathers as “positive” were more likely to adopt similar personalities to their fathers (Long et al. 2014). Other research found that college students tended to report higher interpersonal competence and more relationship satisfaction when they had “secure attachment” with their parents (Mattanah, Lopez & Govern, 2011). Evidence of the continued presence of parents can be found in recent work documenting that young adults tended to list their parents as sources of emotional support through confiding and advice more often than other social ties, with more exchange between children and their mothers (Offer & Fischer 2018).

It is further likely that the continuing effect of fathers on adult children’s lives would vary by gender of the child. Fathers tend to interact differently with their daughters compared to their sons. Fathers tend respond more attentively to their daughters and respond more strongly to their emotions (Mascaro et al., 2017). In contrast, fathers of sons were more likely to engage in “rough and tumble play” and have stronger responses to neutral face reactions. Others have found that fathers react more positively to their daughters’ submissive behavior (Chaplin, et al, 2005), spend money differently on their sons compared to their daughters (Nikiforidis, et al., 2017), and even decide to work more or fewer hours depending on the gender of their newborns (Lundberg & Rose, 2002). Others have reported surprisingly few differences in the way fathers treat sons versus daughters (Endendijk, 2016). The direction of effect is therefore unclear, but the bulk of research indicates that while fathers may favor a daughter, they invest in their sons.

There is also evidence to suggest that sons and daughters respond differently to the quality of presence from their father. Sons with absent fathers, for example, are more likely to be suspended from school, have more trouble paying attention, and generally perform worse than girls in academia (Lundberg, 2017). Teenage girls with absent fathers tend to exhibit behavior for riskier behavior, such as smoking and sex at an earlier age, compared to girls with both parents present (Boothroyd & Cross, 2017). There is little research on how adult sons respond to the quality of the presence of a father compared to daughters, although the research that exists suggests that fathers continue to play a role well into adulthood (Rostad, Silverman & McDonald, 2014, Quinlan, 2003; Ellis, et al., 2003; TenEyck, Knox & Sayed, 2021).

The rationale for our conceptual model is that the active presence of a close father in an adult child’s personal network is associated with a larger network of friends and family. Specifically, we address the following hypotheses:

1. Adult children will have more social ties when fathers are present and emotionally close, compared to those whose fathers are merely named, or not named at all.

2. Adult children with fathers who live proximally will have more social ties.

3. Adult children whose parents are divorced, will have fewer social ties than those who did not.

Data

The UC Berkeley Social Networks Study, or “UCNets” is a five-year panel study funded by the National Institute on Aging (http://ucnets.berkeley.edu), created with the goal of drawing an egocentric map of respondent networks and collecting information about their social connections. Data collection occurred in 2015, with the first wave of the longitudinal panel survey recruited through address-based sampling and Facebook advertisement sampling. The survey collected data on 1159 people in two cohorts: 21–30 year-olds and 50–70 year-olds living in the San Francisco Bay Area. The surveys were conducted face-to-face and online. In addition to network data, the survey also collected detailed information about participants’ socioeconomic and health status. See Fischer and Lawton (2020) and the UCNets website for the questionnaire and documentation on the methodology.

Dependent Variable

Data on social networks were collected by asking respondents 9 name-eliciting questions about how often they engaged in social activities, received advice, confided, and other interpersonal interactions, and then asking them to provide a list of names with whom they engaged in that activity. One of these questions asked respondents (ego) to name people (alters) they “go out to concerts, plays, clubs, sports, or other events with…” This social activities question is the focus of our study as it represents the ability to form friendships. We tallied the number of names, i.e., their alters, listed for each respondent. We limited the results to exclude parents from this list of names, resulting in the dependent variable of the respondents’ number of ‘Social Interaction Ties.’ Three respondents in this study refused to provide any information about their social network and were removed from the sample.

Table 1 shows the distribution of all ‘Social Interaction’ alters named in a network by role relationship and gender. Because we were interested in the association between a close parent and the number of people the respondent interacted with beyond them, we removed parents from the data set if they were named from the list. We then created three distinct sets of alters for social interaction ties: the total number, male-only, and female-only. Table 2 features cross-tabulations of weighted case means of ‘Social Interaction Ties’ by type of father relationship alongside N counts. Generally, those with an ‘especially close’ father report a higher average amount of ‘Social Interaction Ties.’

Table 1:

Cross-Tabulation by Number of Alters Named as Social Ties

Category	Social Ties
All	6,539
Males	2,622
Females	3,906
Fathers	70
Mothers	131
Total Beyond Parents
All	6,338
Males	2,552
Females	3,775

Note: Fathers and Mothers in this sample are anyone the respondent considers to be a ‘Father’ or ‘Mother,’ stepparents are included.

Table 2:

Cross-Tabulations Number of Respondents in Father Named and/or Close Categories (Sons and Daughters) by Average Amount of Social Interaction Ties

	N	Social Interaction Ties
Father Not Named	247	4.70
Daughters	165	4.72
Sons	82	4.67
Father Named/Not Close	147	4.83
Daughters	104	5.40
Sons	43	4.83
Father Named/Close	198	6.05
Daughters	128	6.00
Sons	70	6.10
All Respondents	592	5.19
Daughters	391	5.25
Sons	758	5.13

Note: This sample contains all respondents who reported having a father who was alive and/or present in their lives in some capacity. The above presents unweighted N counts alongside weighted case means.

Explanatory Variables

Variables describing the relationship with fathers were constructed based on a question asking respondents to identify their father in their network as, “anyone they consider to be a father.” This father variable includes adoptive and a small number of stepfathers (n=8) as well as biological fathers. People who reported more than one father or mother in this study were excluded (these were often coding errors). Each name in the list of alters was described by the ego (respondent) regarding age, sex, relationship, and geographic and emotional closeness, among other descriptors. Using this information, we created a set of dummy variables for fathers who are ‘Named and close,’ ‘Named but not close’ and ‘Not named.’ In the analyses, we excluded egos for whom fathers are no longer living. Parallel ‘close’ mother variables were also created, but the focus of this research is on relationships with fathers. Accordingly, we selected respondents who reported having a father who was alive; the same was not done for mothers.

Covariates

We controlled for respondent’s personal income, gender (sons, daughters), education, age and age groups (21–30 years old, 50–70 years old at time of recruitment), race (White, Black, Asian), ethnicity (Hispanic), and whether the father was living within one hour from the ego. We also controlled for certain life events that could affect the size of personal networks (Gerstel & Sarkisian, 2006; Kalmijn & Graaf, 2012; Guiaux, et al., 2007), notably whether the respondent had married within the past year, had a new baby, became widowed or got divorced, or got a new job (Wrzus, et al, 2013). Finally, we controlled for the mode of interview (web versus face-to-face) and whether the respondent was recruited by Facebook versus addressed-based sampling.

Methods

The number of social ties ranged from 1 to 9, so for ease of interpretation, we utilized an ordinary least squares linear regression model. There are three models: all social ties, male-only ties, and female-only ties. We conducted a sensitivity analysis (results available by request) by using a General Linear Model (GLM), which can incorporate categorical, ordinal, or continuous dependent variables, and received the same pattern of results.

Results

The results for the regression models are in Table 3. Consistent with the hypotheses, people who report a father in their network of activities with whom they are emotionally close have nearly one more overall tie (b=.894; p< .001; CI .387, 1.401) and male tie (b=.730; p<.001; CI .381, 1.079), but not more female ties. However, those who name a father in their network who is not also regarded as close do not have significantly more social ties compared to those who do not name a father in their social activities network. A change in reference category (not shown) indicates that those who have a ‘close’ father are likely to name more social ties relative to those who name him but don’t consider him ‘close.’ There was no distinction between having a father in the network who was not designated as ‘close’ versus no named father, indicating that presence alone is not sufficient for a positive impact. Comparatively, none of the comparable variables for mothers were significant. The male respondents, that is the sons, had significantly more male ties than did female respondents, or daughters (b = .1284, p<.001; CI: 1.008, 1.559) but significantly fewer female ties (b=−1.656; p < .001; CI: −1.938, −1.355). We tested the interaction between sons and close fathers and named fathers but found no significant results.

Table 3:

OLS Regression Models for the Number of Social Activity Ties

	All Social Ties			Male Only Social Ties			Female Only Social Ties
Variable	B	SE	Beta	B	SE	Beta	B	SE	Beta
(Constant)	4.429	1.035		2.050	0.713		2.382	0.752
Father Named and Close	0.894***	0.258	0.159	0.730***	0.178	0.179	0.173	0.188	0.040
Father Named and Not Close	0.383	0.276	0.066	0.157	0.190	0.037	0.237	0.201	0.053
Mother Named and Close	0.152	0.260	0.030	0.041	0.179	0.011	0.105	0.189	0.027
Mother Named and Not Close	0.199	0.309	0.034	−0.042	0.213	−0.010	0.245	0.225	0.054
Sons	−0.376*	0.204	−0.074	1.284***	0.140	0.348	−1.646***	0.148	−0.420
Father resides < 1 hr	0.443**	0.220	0.086	0.076	0.151	0.020	0.365**	0.160	0.092
Parents divorced	−0.488**	0.220	−0.091	−0.394***	0.151	−0.102	−0.093	0.160	−0.022
Married	−0.630***	0.231	−0.120	−0.454***	0.159	−0.119	−0.182	0.168	−0.045
New Baby	−0.094	0.298	−0.013	0.014	0.205	0.003	−0.111	0.217	−0.020
New Job	0.481**	0.243	0.082	0.251	0.167	0.059	0.238	0.177	0.052
Personal Income	0.159***	0.039	0.184	0.041	0.027	0.066	0.117***	0.029	0.176
Older Age Group	0.940***	0.276	0.170	−0.096	0.190	−0.024	1.036***	0.201	0.243
Education	0.101	0.064	0.067	0.010	0.044	0.009	0.091**	0.046	0.078
Black	−0.516	0.369	−0.055	−0.536**	0.254	−0.079	−0.011	0.268	−0.002
Asian	−0.473*	0.244	−0.079	−0.241	0.168	−0.056	−0.224	0.177	−0.049
Hispanic	−0.433	0.251	−0.072	−0.057	0.173	−0.013	−0.375**	0.182	−0.080
Web Mode	−0.902***	0.275	−0.177	−0.692***	0.190	−0.188	−0.216	0.200	−0.055
Facebook Recruitment	0.071	0.308	0.013	−0.365*	0.212	−0.094	0.415*	0.224	0.100

Note: P Values

^***p<0.01,

^**p<0.05,

^*p<0.1

When a father is physically accessible, that is, lives within one hour, then the number of all ties is higher (b= .443; p< .05; CI: .012, .445) as is the number of female ties (b=.365; p<.05; CI: .051, .678). This result may be due to a higher propensity for married parents to live near children, whereas divorced fathers may live farther away (Lawton et al., 1994) and mothers are likely more associated with managing all social relationships compared to fathers.

The impact of attachment is also seen in the variable for whether parents were divorced. Here we find a negative effect on the number of all ties (b= −.488; p< .05; CI: −027, −.919) and male ties (b= −.394; p< .01; CI: −691, −.097). Life course status also has an effect on total ties: married egos have fewer total ties (b=−.630; < .01; CI: −1.084, −.176) and male ties (b= −.454; p<.01; CI: −.766, −.142). There is no significant effect of having a new baby, but getting a new job is associated with higher total ties (b= .481; p<.05; CI: .003, .958).

Discussion

Having a father who is emotionally close and active in one’s personal network is associated with having a higher number of total ties and male ties, but not female ties. This finding highlights the growing awareness of fathers as important figures in a person’s life, not just in childhood, but over the course of adulthood. Having an active network of friends and families for social activities is important for social support and mental health (Child & Lawton, 2017). The presence of a father is one of the factors that contributes to the ability to form this active network.

Although the cross-sectional nature of our findings limits our conclusions in terms of causality, we can clearly see a positive association between the subjective closeness of a father in a respondent’s reported life and a higher count of other people available for social activities, even when we control for the father’s presence and geographic closeness to the respondent. The results suggest that the presence of a close father in adult children’s lives has an influence on their social network by providing tangible and intangible skills and resources (Coleman, 1988). The presence of a father in an adult’s life as an influencer of social network formation is consistent with research carried out on children (Rubin & Sloman, 1984; Lavenda & Kestler-Peleg, 2017).

The father may also be passing down ‘gender specific’ social capital by introducing his children to his social network, which will more often be male than female (McPherson, Smith-Lovin & Cook, 2001). If it were simply the father introducing his adult children to his homophilous social network, then physical closeness and presence would be associated with ties of both genders. This result suggests that fathers, as positive role models for male interaction, pass down social capital in the form of social skills that are useful for generating more male social ties as well as for generating the desire for more male social ties.

Further, that the positive presence of a father is strongly associated with a higher count of social ties is consistent with the predictions of attachment theory (Bowlby, 1969, 1980; Singh, 1988), and yet this confidence derived from earlier and ongoing attachment is apparently important for developing new emotional connections with men specifically. It’s likely that a positive father relationship is conducive for a father introducing his children to more people in his social network (social capital), leading to more opportunities for adults to build a more robust social network. The smaller number of ties for those with divorced parents similarly indicates a less robust social support system, and is also another long-term outcome of parental divorce. Also consistent with attachment theory, these results point to social network development that benefitted from the trust and confidence derived from the reliable presence of parents.

The above discussion supports the concept of the ‘important father.’ In support of the ‘essential father’ hypothesis, father closeness in this model produced distinct outcomes relative to mother closeness. For example, a close father is associated with a higher number of male social ties whereas a close mother is not.¹ In fact, the models repeatedly uncovered that respondents of both sexes will tend to have a higher count of male ties in their networks even after all controls and parental variables are included. If fathers were not ‘essential,’ associations between a ‘close’ father and social network size and composition would be the same as a ‘close’ mother. Fathers may generate unique forms of social capital due to their gendered parental status. That is, gender roles may dictate the acceptability of how ‘close’ a male should normatively get to others and how that closeness should manifest itself (Floyd & Morman, 1997; Swain, 1989; Wood & Inman, 1993). Furthermore, previous research has identified that men are often unwilling or unable to communicate affection with their own children even when they’re emotionally close (Morman, Floyd, 1999). In other words, that fathers communicate closeness in a different way could, potentially, explain why their closeness results in unique outcomes relative to mother closeness.

Limitations

The UCNets data are from the San Francisco Bay Area which may have its own distinctive features, so we do not claim that the results are generalizable. While the sample is representative in age and race of the region, it tends to be more highly educated than the overall population. The social ties measurement included family and friends in order to cast a broader net for different kinds of ties as both kith and kin are important for psychosocial health, but it is possible that distinctive patterns would emerge when looking at only friends. We did not exclude respondents who have no living mother in order to avoid reducing sample size, so the results for mothers may have been attenuated.

In closing, the association between a close and present father with a more robust social network in adults, much like in children, exists even when controlling for education, income, certain life events, and the mother’s presence. However, further research is necessary to identify the direction of this association more concretely and to further disentangle contributors to patterns of network formation. This analysis used just one of the three waves of data. It is possible in a subsequent analysis to use all three waves in a fixed-effects model for a better understanding of causality. Another possible analysis is to expand the frame of analysis to additional measures of networks, such as the availability of confidants and advisors, and satisfaction with the network. This future work is motivated by the present study’s findings: that fathers are important in adult lives, not just figures in a person’s past.