“Low-Income Men and Fathers’ Influences on Children?”

Introduction

Fathers represent an important resource for children, ideally investing the time, money and emotional support that contribute to healthy child development. Over the past four decades, two important—and related—changes have occurred within the post-industrial family, reflecting shifts in family gender roles as well as the division of household labor and market work. First, mothers are increasingly likely to be employed outside of the home, and second, fathers’ roles have expanded from primarily that of a “breadwinner,” who provides economic support, to include more caregiving. As noted in the prior chapter by Berger and Langton, fathering today often includes nurturing and caregiving; engaging in leisure and play activities; providing the child’s mother with emotional or practical support; providing moral guidance and discipline; ensuring the safety of the child; taking responsibility for coordinating the child’s care and activities; and connecting the child to his extended family, community members and other resources (Cabrera, Tamis-LeMonda, Bradley, Hofferth, and Lamb 2000; Marsiglio, Amato, Day, and Lamb 2000; Palkovitz 2002). With the broadening of paternal roles, there has been growing attention to how fathers affect children. Early research focused on the consequences of fathers’ absence for children’s wellbeing. More recent studies have focused on how father involvement may be associated with children’s wellbeing.

While changes in the gender division of labor have occurred across all demographic groups, the shift in fathers’ involvement has played out quite differently by socioeconomic status. Employed fathers in middle-income families have had the opportunity to expand their paternal roles into new areas. In contrast, fathers with limited job prospects in low-income families have been more likely to retreat from the father role altogether (Furstenberg 1988). Although unmarried or low-income fathers sometimes describe their roles in terms similar to those used by married or middle-class fathers (Furstenberg, Sherwood, and Sullivan 1992; Jarrett, Roy, and Burton 2002; Waller 2002), it is likely that a few years after the birth, the former will be living away from their children (Carlson and McLanahan Forthcoming). Once non-resident, fathers are unlikely to be actively involved in raising their children, and many have very limited contact with their children.

In this paper, we review the literature about how low-income men and fathers matter for children. We begin by outlining our conceptual framework about fathers’ roles. Then, we summarize the literature about fathering and child wellbeing, distinguishing between resident fathers and non-resident fathers and highlighting low-income fathers in each section. We briefly discuss social fathering and children’s wellbeing, and we conclude by discussing implications for future research and public policy.

Theoretical Perspectives

The importance of parenting for children’s health, development and wellbeing is well-established (Bornstein 2006; Collins, Maccoby, Steinberg, Hetherington, and Bornstein 2000; Maccoby 2000; Maccoby and Martin 1983). Yet, it is also clear that there remains much to learn about which dimensions of parenting affect children and how such may vary across individuals and contexts (Collins et al. 2000). ‘Authoritative’ parenting, characterized by appropriate warmth and control, has long been thought to contribute to children’s healthy development (Baumrind 1966; Baumrind 1986). Until recently, most theories of parenting either focused on mothering or discussed parenting generally without regard to a particular parent.

Family scholars have long recognized the inter-dependence of family relationships. Within a given family ‘system,’ dyadic relationships affect each other (Cox and Paley 1997; Minuchin 1988) and influence individual-level change (Chase-Lansdale, Kiernan, and Friedman 2004; O’Brien 2005). For this reason, we begin our conceptual discussion of fathering by first providing a heuristic model of the family system and the broader factors that influence child development (see Figure 1). Parents and other family members, jointly, provide the active ingredients that, taken together, promote healthy development. Although clearly not exhaustive, we categorize parents’ contributions into three basic domains: economic resources; socialization and positive interactions; and responsibility and management. We also note that children’s own endowments (i.e., inherited genes and related traits) are important, as these endowments may be attributable, at least in part, to parents and interact with aspects of the family context to influence child outcomes (Collins et al. 2000).

Figure 1.

Heuristic Diagram: Parents’ Contributions to Child Wellbeing

A vast literature points to the importance of economic resources in enhancing child outcomes. Greater economic resources enable parents to purchase the necessary material goods and services, such as medical care, higher quality child care or schools, as well as books and toys that improve developmental processes. Greater economic resources may also reduce parents’ psychological distress, which in turn will reduce harsh parenting, and thus benefit children (McLoyd 1998). Although some debate remains about the magnitude of the associations, accumulated evidence suggests that links between poverty in early childhood and persistent poverty with academic and cognitive skills are particularly pronounced (Duncan and Brooks-Gunn 1997; Magnuson and Votruba-Drzal 2008). Further, recent research suggests that at least some of the effect of income on children’s achievement is likely to be causal (Dahl and Lochner 2008; Milligan and Stabile 2008).

The quality of parent-child interactions and socialization also contribute to children’s development. High-quality parent-child interactions are characterized by positive engagement and discipline; this involves not only parental warmth, such as being responsive, affectionate, and nurturing and rewarding positive behavior, but also teaching children in a productive and supportive manner (Brooks-Gunn and Markman 2005). Research suggests that such positive engagement is most effective when also coupled with appropriate control and discipline, so that children learn that there are consequences for misbehavior (Baumrind 1986).

The final dimension of parenting that we highlight is responsibility and management, which includes both the supervisory tasks of making sure a child’s activities are appropriately monitored as well as scheduling and completing routine child-related tasks. This domain encompasses such tasks as making sure that a young child goes to a well-baby doctor exam or that someone walks the child home from preschool, as well as making sure that an older child completes their homework or gets to soccer practice. Parents differ in the extent to which—and priorities by which—they manage their children’s lives, and striking differences have been observed by social class. Lareau (2003) argues that middle-class parents, defined by having highly-skilled jobs, engage in “concerted cultivation” by providing stimulating learning activities and social interactions; in contrast, working-class and poor parents, with less skilled jobs, view child development as unfolding “naturally” and believe they need do little more than provide for basic needs, including food, shelter, and comfort.

Whether a child’s biological parents reside together or apart, these parenting tasks require collaboration and coordination—sometimes referred to as coparenting (McHale 1995)—and suggest a variety of ways to share such tasks. Thus, understanding the direct and indirect ‘effects’ of fathering or mothering is difficult, without situating it in the broader family context of interdependent roles and transactional processes (Parke 2004). Moreover, the whole may be more than the sum of individual parts, such that the overall quantity and quality of all family resources that children experience, including interactions among fathers’ and mothers’ parenting, matters the most for children (Cox and Paley 1997). Low-income families may structure the rearing of children in fundamentally different ways from their middle-class counterparts, in particular creating a greater role for non-resident and extended family members (Furstenberg 2007; Lareau 2003; Stack 1974).

During the past two decades, fatherhood scholars have developed conceptual models to identify the key components of fathers’ involvement. One of the first ‘typologies’ of father involvement, developed by Lamb and colleagues, identifies three key components—accessibility, engagement, and responsibility (Lamb, Pleck, Charnov, and Levine 1985): accessibility (or availability) refers to time that fathers are available to children, even if they are not directly interacting; engagement (or interaction) refers to time spent with children, especially doing activities together that are known to contribute to healthy development (e.g., reading); and responsibility refers to fathers’ helping to arrange resources and activities for children.

Building on Lamb’s early work and the broader conceptualization of parenting described above, we rely on a model of fathering that highlights four key domains of fathering, each of which is expected to be important to children’s wellbeing.

First, as described above economic resources enable parents to provide the food, clothing, and shelter requisite for daily living, as well as the material goods and experiences that promote positive child development. In addition, fathers’ financial support may function to increase or improve their non-pecuniary involvement with their children (Nepomnyaschy 2007; Seltzer, Schaeffer, and Charng 1989). Second, father-child interaction provides the opportunity to demonstrate the appropriate warmth, support, control and monitoring, that are intrinsic to authoritative parenting (Baumrind 1986; Maccoby and Martin 1983). Yet, time spent interacting is a necessary, but not sufficient condition for developing close parent-child ties: both the quantity and quality of interaction are important. Third, responsibility captures the aspects of parenting related to managing and coordinating the activities of childhood. Although paternal responsibility is important, it is often neglected in surveys (Cabrera et al. 2000; Lamb 1986), and there is limited research evaluating its effects. As a result, we do not focus on it within this review. Finally, co-parenting—defined as the extent to which parents can effectively work together to rear a common child—is distinct from both couple relationship quality and parenting behavior, and may have unique implications for child wellbeing (Hayden, Schiller, Dickstein, Seifer, Sameroff, Miller, Keitner, and Rasmussen 1998; McHale 1995; McHale, Kuersten-Hogan, Lauretti, and Rasmussen 2000).

Since the influence of fathering may differ across contexts, below we summarize three key factors that may moderate the effects of father involvement on child and adolescent wellbeing—child age or developmental period, child gender, and race and ethnicity.

Developmental Perspective

Parenting tasks and activities evolve and adapt as children grow and develop. During infancy, children are dependent on their caregivers to meet their physical, social and emotional needs. During toddlerhood and the preschool years, children express unique personalities and acquire language and self-regulation skills, which in turn promotes their greater independence and social integration, which continues to grow during childhood. During middle childhood, children continue to develop a sense of self and a range of important competencies as they engage in an increasingly wide array of activities and social relationships. During adolescence, youth seek even greater autonomy from their family as they establish their own identities.

Changes in children’s developmental needs and tasks suggest that the effects of parenting are likely to differ, both in magnitude and nature, over the course of childhood and adolescence. For example, some scholars argue that fathering may become especially important during adolescence (Forehand and Nousiainen 1993). Studying and comparing the effects of parenting across childhood is challenging because of its changing nature, and few studies have systematically investigated whether the effects of fathering differ as children develop. With respect to non-resident fathers, divorced fathers with offspring in middle childhood or adolescence are most often studied, so less is known about father involvement in early childhood. Although it is important to consider developmental processes, to date the research base is too thin to make any conclusions about how fathering and its effects may differ across particular periods of childhood and adolescence.

Gender Perspective

Social learning theory suggests that fathers may be more involved with and have a greater influence on sons than daughters because of the comparative advantage, greater internal rewards, and/or higher external expectations for socializing same-gender children (Harris and Morgan 1991; Rossi and Rossi 1990). Some (but not all) studies show that fathers are more involved with sons than with daughters, particularly at older child ages (Anderson 1989; Cooksey and Fondell 1996; Harris and Morgan 1991). Gender differences in father involvement may depend on the type of activity, with some types of involvement being more ‘gendered’ than others (Cooksey and Craig 1998). Of particular interest, however, is the extent to which the effects of father involvement differ by child gender; the few studies that have examined this issue, have found that father involvement is as beneficial for daughters as for sons (Amato and Gilbreth 1999; King, Harris, and Heard 2004). Whether this pattern of findings is equally true among low-income children is unclear.

Race and Ethnicity Perspective

Scholars of the family have long recognized racial, ethnic, and cultural differences in family roles and functioning, including parenting (Parke 2004; Stack 1974). The sources of such differences are complicated, including limited prospects for male employment and economic mobility (Mincy 2006), adaptation to other structural and historical factors, most notably slavery (Patterson 1999), as well as cultural contexts (Edin, Tach, and Mincy 2009; Furstenberg 2007; Mincy and Pouncy 2007). This implies that fathering may have differing effects on child wellbeing, depending on a family’s cultural background, and that the interpretative meaning of fathering and the processes by which it affects children may vary (Livingston and McAdoo 2007; Nelson 2004). Studies have documented differences across broadly-defined racial or ethnic groups in terms of patterns of fathering and father involvement (see Berger and Langton this volume; Hofferth 2003; Hofferth, Cabrera, Carlson, Coley, Day, and Schindler 2007). Yet, few studies have systematically considered whether and how the effects of fathering differ across racially, ethnically, or culturally-defined populations, and those that have yield mixed results (Coley 1998; King 1994b). As a result, variation by race/ethnicity (and by immigration status and origin) is an important avenue for future research.

Methodological Challenges

As with all non-experimental research, the research on fathering (and parenting more generally) is largely descriptive and faces considerable challenges to establishing causal effects, even if there is strong theoretical support for such. We describe several key challenges here. The first relates to social selection, that omitted variables may bias the estimates of the effects of father involvement on child wellbeing. For example, non-resident fathers that pay higher levels of father support and are more advantaged along a number of dimensions compared with those that provide less support (Berger & Langton, this volume); if differences in these characteristics are not fully accounted for, then studies may attribute to father involvement what is really due to some other unobserved characteristic(s). Studies differ in the rigor of their efforts to reduce such omitted variable bias and often face the unenviable position of over- or under-controlling for confounding covariates. In the absence of experimental studies, these issues are unlikely to be easily resolved, yet researchers’ continued attention to ruling out alternative explanations for observed associations is an important effort in moving the field forward.

A related selection issue is how to untangle the effects of multiple dimensions of fathers’ involvement on child wellbeing. Despite theoretical reasons for distinguishing between dimensions of father involvement, such distinctions are rarely made in empirical research. Most studies characterize father involvement in general terms, describing levels of involvement (i.e., high vs. low), as measured by contact, activity frequency, or father-child closeness (or some combination). Rarely do studies evaluate the unique, additive, or multiplicative effects of several aspects of father involvement.

A second key methodological issue is that most studies focus on estimating models in which father involvement affects children’s development, rather than modeling more complicated processes by which fathers’ involvement affects children (both directly and indirectly), and child wellbeing also affects fathering. Family scholars have become increasingly mindful of the extent to which family relationships are reciprocal, and within the dyadic parent-child relationship, children influence parents just as parents also influence children (Crouter and Booth 2003). If so-called ‘child effects’ are prevalent, as is suggested by some recent research on non-resident fathers’ involvement and adolescents (Coley and Medeiros 2007; Hawkins, Amato, and King 2007), then many studies may be overstating the importance of father involvement in determining children’s wellbeing. For example, if non-resident fathers become less involved because children are having trouble, then estimates of how fathering and child outcomes are linked reflect, at least in part, this reverse causality and should not be interpreted as evidence that father involvement entirely drives child wellbeing.

Third, scholars have increasingly recognized the need to evaluate fathers’ involvement in the context of mothers’ involvement (Marsiglio, Amato, Day, and Lamb 2000). Within families, levels of mothers’ and fathers’ parenting are positively correlated (Aldous, Mulligan, and Bjarnason 1998; Harris and Ryan 2004), or as described by Marsiglio and colleagues (2000), “fathering is often a co-constructed accomplishment” (p. 1185). Therefore, it is important to use data and methods that enable evaluation of the direct, indirect and conjoint influence of fathers along with mothers on child wellbeing. Similarly, who provides the information about the nature, content and quality of fathers’ involvement with children also matters. Fathers, especially non-resident or low-income fathers, are typically under-represented in national surveys (Garfinkel, McLanahan, and Hanson 1998; Nelson 2004), so early studies often relied on mothers’ reports about fathers. However, mothers typically report lower levels of father involvement than fathers themselves, especially for non-resident fathers (Coley and Morris 2002). Finding modest correlations across individual reporters’ perceptions about the same family relationship is more often the norm than the exception (Tein, Roosa, and Michaels 1994), so researchers would be well-served to by using methods that can incorporate information from multiple reporters (Marsiglio et al. 2000).

Research on Residential Fathers and Child Wellbeing

Before turning to the research on particular types of father involvement, it is important to note that studies on family structure provide insight into the effects of fathers’ co-residence with their children, even if it does not identify how and why fathers matter. An extensive literature suggests that children fare better on a host of social-psychological, behavioral, and cognitive outcomes when they spend their entire childhood living with both of their biological parents (McLanahan and Sandefur 1994; Sigle-Rushton and McLanahan 2004). This research (using observational data) suffers from the concerns about social selection described above, and indeed some of the ‘effect’ of stable family structures is certainly due to the characteristics of parents that remain in stable unions (Cherlin 1999). Yet, a recent review of the literature suggests that beyond selection factors, there is also likely some (perhaps quite small) causal effect of living in a two-parent family (versus single-parent family) on at least some aspects of child wellbeing (Sigle-Rushton and McLanahan 2004). This, suggests that fathers commitment to ‘being there’ for children, and all that it entails, promotes children’s wellbeing.

Not surprisingly, co-resident fathers typically spend more time with their children than non-resident fathers, simply because sharing a household affords greater opportunity for contact and day-to-day interactions (Amato and Gilbreth 1999). Beyond the higher level of father involvement typical among resident fathers, father involvement in two-parent families may also yield greater benefit for children, because it reinforces mothers’ parenting and thereby strengthens the cohesiveness of the family as an institution (Gable, Belsky, and Crnic 1994; Harris, Furstenberg, and Marmer 1998). On average, father involvement across the domains outlined above is associated with better child outcomes when provided by a co-residential biological father (Amato 1998; Harris, Furstenberg, and Marmer 1998; Lamb 2004; Marsiglio, Amato, Day, and Lamb 2000).

With respect to economic resources, it is difficult to evaluate specifically how co-resident fathers’ economic contributions benefit children, since studies cannot easily isolate the unique effects of financial resources from other aspects of co-resident fathers’ involvement; as fathers’ education and income are strong predictors of fathers’ engagement and involvement (Amato and Rivera 1999). Two-parent families do have significantly higher economic resources than single-parent families; according to U.S. Census data, in 2006, average family cash income for married couples with children was $89,096, compared to only $28,865 for single-mother families with children (U.S. House of Representatives 2008). Economic resources, in turn, have been shown to account for about half of the gap in wellbeing between children living in two-parent versus single-parent families (McLanahan and Sandefur 1994), leading scholars to conclude that at least in the U.S., fathers’ economic contributions are fundamental to their children’s wellbeing.

With respect to father-child interactions, an extensive literature has demonstrated the benefits of father involvement for child wellbeing by resident fathers, although much of this work has focused on samples of middle-income, school-aged children or adolescents (Shannon, Tamis-LeMonda, London, and Cabrera 2002). Greater interaction with fathers has been significantly linked to decreases in delinquency and behavioral problems and to increases in cognitive development, educational attainment, and psychological wellbeing (Amato and Rivera 1999; Harris, Furstenberg, and Marmer 1998; Hofferth 2006; Marsiglio, Amato, Day, and Lamb 2000; Tamis-LeMonda, Shannon, Cabrera, and Lamb 2004). A 2000 Journal of Marriage and Family review of fatherhood scholarship noted that 55 studies in the 1990s explored the link between paternal involvement and outcomes for children ages 0-19 in two-parent families; the authors conclude that the results “are consistent with the belief that positive father involvement is generally beneficial to children” (p. 1183), with no discernible differences by child age or race/ethnicity (Marsiglio, Amato, Day, and Lamb 2000).

It is important to note that many studies are correlational in nature, do not include extensive control variables, and do not use longitudinal data with techniques that are best suited to causal inference by ruling out omitted variable bias and child ‘effects.’ Moreover, nearly all of the research on co-resident father involvement is based on married fathers, so we have little evidence about how children may fare in long-term cohabiting unions. However, since cohabiting unions with children in the U.S. are much more likely to dissolve than marital unions with children (Osborne, Manning, and Smock 2007), it is the rare child that will experience a stable cohabiting family. Still, there are good theoretical reasons to expect co-resident biological fathers’ involvement to positively influence children, so it is likely that the association is positive for cohabitors; more at issue is the magnitude of the association.

Co-parenting

We know that mothers play an important role in shaping father-child interactions. Research on resident, and especially married, fathers has underscored that men’s roles as fathers are strongly linked to men’s roles as partners in the so-called “package deal” (Furstenberg and Cherlin 1991; Schoppe-Sullivan, Brown, Cannon, Mangelsdorf, and Sokolowski 2008; Townsend 2002). Better co-parenting among married parents is linked to better child development, net of marital quality and parent-child relationships (Gable, Belsky, and Crnic 1994; Schoppe, Mangelsdorf, and Frosch 2001).

Low-Income Fathers

Most of the literature about co-resident father involvement has focused on middle-income (mostly married) families, leaving uncertain whether the nature and strength of associations may differ among low-income populations. However, several recent studies made use of data from the Early Head Start Study to examine low-income, co-resident biological fathers. Positive father-child interactions and parenting (sensitivity, positive regard, and cognitive stimulation) is linked to children’s language development, cognitive functioning, and socio-emotional behavior at ages 2 and 3 (Cabrera, Shannon, and Tamis-LeMonda 2007; Shannon, Tamis-LeMonda, London, and Cabrera 2002; Tamis-LeMonda, Shannon, Cabrera, and Lamb 2004) and children’s math and language abilities at age 5 (Martin, Ryan, and Brooks-Gunn 2007). These findings suggest that the quality of co-resident fathers’ parenting is likely to matter for their children’s development across the economic spectrum. However, it is important to note that this research is based on a small and select sample of men who resided with their biological child and agreed to participate in the study. As the authors of this work recognize, these findings cannot be generalized to low-income co-resident fathers more generally (Cabrera, Ryan, Shannon, Vogel, Raikes, Tamis-LeMonda, and Cohen 2004).

Research on Non-residential Fathers and Child Wellbeing

After increases in divorce rates in the 1970s and 1980s, researchers began to examine the nature and consequences of fathers’ involvement following a divorce. Therefore, much of the early literature on non-resident fathers focused on samples of only (or largely) divorced fathers. Today, another common pathway into being a non-resident father is nonmarital childbearing. Due to the high dissolution rates of unmarried romantic relationships, unmarried fathers, who are disproportionately low-income, are likely to live away from a new child just a few years after the baby’s birth (Carlson and McLanahan Forthcoming). Yet, this is not a new phenomenon, especially among black families. A study of children born to young (mostly disadvantaged) mothers between 1979-83 found that over half of African American children never lived with their biological father over their first four years of life (based on reports from annual surveys), compared to 9% of non-black children (Mott 1990).

Economic resources (child support)

Although family roles have changed over time, so that mothers and fathers now often share breadwinning and caregiving roles, fathers’ important contributions to family economic wellbeing cannot be underestimated. Unmarried fathers are less likely to pay formal child support (and at lower amounts when they do pay) than previously-married fathers (Seltzer 1991). Informal financial support (i.e., outside the legal child support enforcement system), especially the purchase of goods and services for the child, is quite common among unmarried fathers, especially around the time of a new baby’s birth (Edin and Lein 1997; Marsiglio and Day 1997; Waller 2002). In contrast, formal child support orders are rare just after a birth, in part because many unwed couples are still romantically involved and coresiding. By five years after the focal birth, 47% of mothers had a child support order in place (and 27% received a formal payment), while 45% received in-kind support (and 32% received informal financial support) (Nepomnyaschy and Garfinkel 2008). As explained in more detail by Berger and Langton (this volume), there are several explanations for the relatively low levels of formal support and the higher prevalence of informal support among low-income fathers, and Cancian, Meyer and Han (this volume) highlight the challenges in securing child support from non-resident fathers.

Consonant with the broader literature about how family income affects child wellbeing, there is evidence (though not entirely consistent) that child support income is associated with better child outcomes. In their meta-analysis of research about non-resident father involvement and child well-being, Amato and Gilbreth (1999) identified 14 studies that examined child support and children’s achievement and behavioral outcomes. The average (weighted) effect size for child support predicting children’s academic success was .09 and for externalizing behavior problems was −.08, although there was no significant link between child support and children’s internalizing behavior problems. These effects did not appear to differ by child gender. Several of the studies included in the meta-analysis indicated that an additional dollar of child support had significantly larger effects than other sources of family income for at least some demographic groups (Argys, Peters, Brooks-Gunn, and Smith 1988; Knox and Bane 1994).

Recent studies also suggest that the amount of support that mothers receive matters. Put another way, whether fathers pay any support appears to less important than how much they pay. Two recent studies found no association between child support receipt and a range of adolescent outcomes (Hawkins, Amato, and King 2007; King and Sobolewski 2006); however, these studies used a binary indicator of whether the father generally paid support, rather than more detailed information about actual payments. An additional study of African American welfare clients and their young children, found that after controlling for father contact and the quality of relationship with the mother, formal and informal child support predicted better child behavior, and informal child support predicted higher-quality home environments (Greene and Moore 2000). Neither formal nor informal support predicted children’s school readiness. By federal law, welfare recipients today only receive a small portion of any formal child support paid ($50); thus, it is not surprising that formal child support shows little evidence of improving children’s wellbeing.

Father-child interaction

Although the literature about non-resident father involvement has grown over the last two decades, the evidence for the effects of fathers’ interaction and time spent with children is more limited and less consistent than that for resident fathers (Amato and Gilbreth 1999; King 1994a; King and Sobolewski 2006). In 2000, a review article by Cabrera and colleagues characterized the state of fatherhood research in this way “Little is known about the effects of non-resident fathers’ involvement on children’s development” (Cabrera et al. 2000).

Early studies focused on the frequency of contact or visitation by non-resident fathers (among samples of primarily divorced fathers) and found that greater father-child contact was not associated with child wellbeing regardless of regardless of the child’s race, gender, mother’s education, or marital status at birth (Furstenberg, Morgan, and Allison 1987; Hawkins and Eggebeen 1991; King 1994a; King 1994b; Seltzer and Bianchi 1988). Marsiglio and colleagues’ (2000) review of 38 studies published in the 1990s on non-resident fathers and child wellbeing noted that, “Taken together, these studies suggest that the frequency of visitation and children’s feeling about their fathers are not good predictors of children’s development or adjustment” (p. 1184).

The fact that more father-child interaction is not linked to better outcomes for children of non-resident fathers may be due, at least in part, to the characteristics of men who end up living away from their children (Jaffee, Caspi, Moffitt, Taylor, and Dickson 2001). Non-resident fathers are more likely to be depressed, have problems with drugs or alcohol, to have engaged in criminal behavior, and been incarcerated than resident fathers (DeKlyen, Brooks-Gunn, McLanahan, and Knab 2006; Garfinkel, McLanahan, and Hanson 1998; Lerman 1993), and this is all the more true for young, low-income fathers. Time spent with fathers who display such characteristics, particularly antisocial behavior, may detract from children’s healthy development (Jaffee, Moffitt, Caspi, and Taylor 2003).

In contrast, when non-resident fathers have the motivation, skills, and capacities to engage in positive interactions with their children, their presence may facilitate healthy child outcomes. Thus, scholars argue that the quality of non-resident fathers’ involvement may be more important than the mere quantity (Amato and Gilbreth 1999; Cabrera et al. 2000). A number of studies have suggested that positive interaction and engagement by non-resident fathers (characterized by close and supportive father-child relationships and responsive parenting) can, in fact, promote child wellbeing (Amato and Gilbreth 1999; Marsiglio, Amato, Day, and Lamb 2000), including reducing child and adolescents’ behavioral problems (Carlson 2006; Coley and Medeiros 2007; King and Sobolewski 2006), facilitating adolescent adjustment post divorce (Stewart 2003), improving adolescents’ eating habits (Stewart and Menning 2009), decreasing the likelihood that adolescents will start smoking (Menning 2006b) or drop out of school (Menning 2006a). At the same time, although such associations are statistically significant, they are typically substantively small. Also, there is little evidence that the benefits of positive father involvement are moderated by child age, gender, race or socioeconomic disadvantage (Amato and Gilbreth 1999; King and Sobolewski 2006). As noted earlier, we must also be mindful that some of the positive association between non-resident father involvement and child outcomes might arise from bidirectional influences in the father-child relationship, particularly for adolescents (Coley and Medeiros 2007; Hawkins, Amato, and King 2007); thus, we must be cautious in interpreting the estimates as reflecting only causal effects of fathering.

Co-parenting

For fathers living away from their children, mothers’ play an important ‘gatekeeping’ role in encouraging or preventing fathers’ access to children (Ahrons and Miller 1993; Allen and Hawkins 1999), and fathers’ ability to coordinate parenting with mothers, sometimes referred to as co-parenting, is a predictor of fathers’ continued involvement with their children (Carlson, McLanahan, and Brooks-Gunn 2008; Sobolewski and King 2005). To our knowledge, beyond studies of children following divorce (e.g., Maccoby and Mnookin 1992), there is no existing research about whether or how cooperative co-parenting between non-resident fathers and custodial mothers directly benefits children, although this is a fruitful topic for future research.

Low-Income Fathers

Although low-income fathers are included in most studies of non-resident fathers, they are typically under-represented because they are less strongly attached to households as compared to higher-income men; they are more likely to be in the military or incarcerated, (Garfinkel, McLanahan, and Hanson 1998; Nelson 2004). A small, but growing literature has focused specifically on low-income non-resident fathers, in order to better understand whether greater involvement by such fathers is beneficial for children’s development. A study of low-income African American families found that fathers’ nurturing behavior is correlated with children’s language skills at age 3 (Black, Dubowitz, and Starr 1999). In addition, non-resident fathers’ consistent contact with their children predicted fewer behavioral problems and better cognitive outcomes among a small sample of school-age children born to teen-age mothers (Howard, Lefever, Borkowski, and Whitman 2006). Finally, increases in non-resident-father contact and parenting responsibility led to reductions in delinquency among young adolescents (ages 10-14) living in low-income neighborhoods (Coley and Medeiros 2007). By contrast, one study of young children of welfare recipients found no association between the frequency of father visitation and child wellbeing, although this study had limited information about father involvement (Greene and Moore 2000). Also, analysis of the Fragile Families data shows that the quantity and quality of non-resident father involvement has no overall association with children’s behavior problems at age 5, although there are beneficial effects when involvement occurs in the context of a strong co-parenting relationship with the child’s mother (Carlson, McLanahan, and Brooks-Gunn 2009).

Taken together with the more general literature on non-resident fathers, this research suggests that there are circumstances under which low-income fathers can have a positive influence on children’s development and wellbeing, but that, on average, it does not. Further, many of these studies include only the ‘best’ fathers that are willing to participate in surveys about father roles, so estimates of the beneficial effects of fathering cannot be generalized to the larger population of low-income fathers, many of whom are much less involved with their children and may have other characteristics that limit the potential for positive influence.

Social Fathers and Child Wellbeing

Biological fathers are not the only father-figures in the lives of low-income children, and it is important not to overlook non-biological father figures, or so-called “social” fathers. With high rates of divorce and nonmarital union dissolution, many children are likely to live with the romantic partner of one of their biological parents at some point during childhood. About one-third of children will spend time in a (married) stepfamily before age 18 (Seltzer 1994), and recent estimates suggest that fully two-fifths of children will spend time living with their mother and a cohabiting partner by age 12 (Kennedy and Bumpass 2008). The chance of living with a non-biological social parent is far greater for children born outside of marriage. Given the instability in co-resident romantic relationships, men in low-income communities will often fulfill multiple father roles over time—biological co-resident father, biological father to their own children who live away, and social co-resident father to the children of a romantic partner (Nelson and Edin Forthcoming). Yet, maternal re-partnering likely complicates and possibly deters the involvement of non-resident biological fathers with their children (Tach, Mincy, and Edin Forthcoming).

Berger and Langton provide greater detail about levels of social father involvement by low-income men. The extant literature suggests that among co-resident parents, married stepfathers tend to exhibit lower quality parenting practices toward stepchildren than married biological fathers do toward biological children (Amato and Sobolewski 2004; Coleman, Ganong, and Fine 2000; Hetherington and Stanley-Hagan 1999; Nelson 2004). Married stepfathers are less likely than married biological fathers to participate in activities with resident children, as well as to express support and positive feelings toward them, or to provide monitoring and supervision (Thomson, McLanahan, and Curtin 1992). In contrast, analysis of the Fragile Families data indicate that unmarried resident social fathers are often involved with (non-biological) young children to the same extent as are unmarried resident biological fathers (Berger, Carlson, Bzostek, and Osborne 2008).

Although research on how social fathers affect children’s development is limited, there is some evidence that social fathers’ involvement is beneficial for young children’s behavior and health status (Bzostek 2008), although again, the magnitudes of the associations are typically quite small. Also, having a close relationship with a stepfather is shown to reduce adolescents’ depressive symptoms and delinquent behaviors (Yuan and Hamilton 2006) and externalizing and internalizing behavior problems (White and Gilbreth 2001). Other work that differentiates the types of social fathers involved with low-income children suggests that involvement by male relatives may be more beneficial than involvement by mothers’ romantic partners (Jayakody and Kalil 2002). At the same time, research on married social fathers finds that children in such stepfamilies fare no better than children in single-parent families, despite their higher incomes (McLanahan and Sandefur 1994). This suggests that either stepfamilies spend their money differently than two biological families, or that stepfamily relationships do not confer the same benefits as biological family relationships (for various reasons, including the possible role of step-siblings). Future research should consider the variation in patterns of involvement by fathers and father-figures, especially as family structure changes over time, as well as the conjoint influence of involvement by biological and social fathers on children’s wellbeing.

Conclusions

This paper has summarized research about how low-income men and fathers matter for children. Sociological, developmental, and family system theories all point to fathering as an important influence on children, and research has found that indeed, some measures of fathering appear to be associated with at least some measures of child and adolescent wellbeing. In particular, fathers’ financial contributions to their children’s households and the quality of father-child interactions, including authoritative parenting, appear to be linked to better outcomes for children. Focusing specifically on low-income fathers, however, the evidence base becomes quite thin. This is perhaps not surprising, given how difficult it may be to study fathering among a population of men that often has comparatively little involvement in their children’s lives, experiences unstable living arrangements, and has a relatively high likelihood of spending time in jail or prison.

Scholars have increasingly called for better and more research on fatherhood, in general, and on low-income and non-resident fathers in particular (e.g., Nelson 2004). Our review reaches the same conclusion: There is still much to learn about how fathers affect their children’s lives, both directly and indirectly. We suggest that several important areas in need of research include: 1) understanding how biological and social fathers (independently and jointly) matter for children, considering the full range of involvement domains and situating analyses within theoretical models; 2) evaluating how fathering processes and their effects may differ by race, ethnicity and cultural background, as well as child age and gender; and 3) analyzing the nature and implications of fathering in the context of contemporary, highly complex patterns of union formation and fertility among low-income populations (e.g., multi-partnered fertility). We also urge researchers to improve their study designs with particular attention to issues of social selection, measurement, bidirectional effects, and indirect effects.

With respect to implications for public policy, we are reticent to draw strong conclusions, given the limited evidence available about low-income fathers to date. As we’ve described, existing research suggests that some low-income men and fathers can and do have a positive influence on children, with the most consistent findings about the payment of child support (by non-resident fathers). This suggests that increasing child support payments is a worthy policy goal. As noted by Sum (this volume), many low-income men are unemployed and have little hope of improving their economic position through the formal labor market. Thus, child support enforcement policy should strive both to collect support from fathers who can afford to pay and to improve the labor market prospects of low-income men so that more men are able to pay.

Cumulative evidence also points to the benefits of high-quality interaction and authoritative parenting. But, a number of studies find that fathers’ greater involvement (especially frequency of contact) has essentially no effect on children’s outcomes. Thus, it is unclear whether fatherhood programs that encouraged paternal involvement generally would, in fact, benefit children. Also, to the extent that unobserved heterogeneity plays a role in why high-quality fathering is shown to improve child wellbeing, one may worry that the fathers not currently involved—but who would be incentivized to become involved by an intervention—might differ in key characteristics that are related to both involvement and children’s wellbeing. As a result, we conclude that interventions and programs should strive to encourage fathers’ positive and engaged parenting as opposed to increasing the mere quantity of time. We hope that future research will shed light on the most promising avenues for enhancing fathers’ positive roles in family life and their contributions to children’s wellbeing.

Figure 2.

Conceptual Diagram: Father Involvement and Child Wellbeing