Unmarried, Nonresident Fathers' Involvement With Their ... - CiteSeerX

2 downloads 45 Views 80KB Size Report
Keywords: nonresident fathers, risk, resilience, father involvement, fragile families. Almost 25% of ..... logical father involved in your life when growing up?” The.
Journal of Family Psychology 2007, Vol. 21, No. 3, 479 – 489

Copyright 2007 by the American Psychological Association 0893-3200/07/$12.00 DOI: 10.1037/0893-3200.21.3.479

Unmarried, Nonresident Fathers’ Involvement With Their Infants: A Risk and Resilience Perspective Jay Fagan

Rob Palkovitz

Temple University

University of Delaware

The authors used a subsample of fathers (n ⫽ 652) who participated during the 1-year follow-up of the Fragile Families and Child Wellbeing study to assess the influence of risk and resilience factors on unmarried, nonresident fathers’ involvement with their infants. They examined the additive, multiplicative, and moderating models of risk and resilience in relation to paternal involvement. Fathers’ relationship to the child’s mother was conceptualized as a risk or resilience factor. Fathers in acquaintance relationships with the mother and fathers who scored higher on the additive risk index were less involved in child care. Fathers who scored higher on the additive resilience index were more involved in child care. There was a multiplicative effect of relationship status and the risk index on fathers’ involvement. The findings point to the importance of programs that address risk and resilience conditions affecting nonresident fathers in interaction with the quality of relationships they have with their children’s birth mother. Keywords: nonresident fathers, risk, resilience, father involvement, fragile families

Almost 25% of American children under the age of 18 live in mother-only families at any given time, and just under one half of these children live in never-married mother-only households (U.S. Census Bureau, 2006). One critical factor in promoting these children’s well-being is the involvement of nonresidential fathers (Duncan & Brooks-Gunn, 1997; Lamb, 1999; McLanahan & Sandefur, 1994). Research has documented the influence of nonresidential fathers’ quantity and quality of parenting on child outcomes (for review, see Amato, 1998). Recent studies have indicated that nonresident fathers play an “important role in their children’s lives to the extent that they provide authoritative parenting— especially if this occurs within the context of cooperative relationships between the parents” (Marsiglio, Amato, Day, & Lamb, 2000, pp. 1184 –1185). However, evidence reveals that never-married biological fathers are at considerable risk for low levels of involvement with nonresident children (Marsiglio et al., 2000; Stewart, 1999).

Few studies have examined predictors of paternal involvement among this group of fathers (see Carlson, McLanahan, & Brooks-Gunn, 2005). Further, researchers have not yet examined whether the predictors of father involvement with children are affected by the relationship type between father and mother (e.g., romantic involvement, friends). Recent findings from the Early Head Start (EHS) and Fragile Families and Child Wellbeing (FF) studies have provided evidence of decreasing involvement between fathers and young children as the level of closeness in the father–mother relationship decreases (Cabrera et al., 2004; McLanahan & Carlson, 2004). The first year following the child’s birth is critical for the father– child relationship as this is the period of time when unmarried nonresident fathers are most at risk for experiencing a decline in the level of closeness with the mother (Nelson, 2004). At the same time, a number of nonresident fathers manage to stay involved with their children despite their high-risk situations (Mincy & Oliver, 2003; Stier & Tienda, 1993). The present study uses the FF data to examine the extent to which predictors of father involvement are influenced by mother–father relationship status and various risk and resilience variables. FF is a national, longitudinal study designed to address the conditions and capabilities of unmarried parents, especially fathers; the nature of relationships between unmarried parents; the development of children born into these families; and the policies and environmental conditions that affect families and children.

Jay Fagan, School of Social Administration, Temple University; Rob Palkovitz, Individual and Family Studies, University of Delaware. The Fragile Families and Child Wellbeing study was supported by National Institute of Child Health and Human Development Grant R01HD36916. The contents of the article are solely the responsibility of the authors and do not necessarily represent the official views of the National Institute of Child Health and Human Development. Correspondence concerning this article should be addressed to Jay Fagan, School of Social Administration, Ritter Hall Annex, 5th Floor, Temple University, Philadelphia, PA 19122. E-mail: [email protected]

Theoretical Perspective The concept of risk suggests that there are psychological or social factors that increase the likelihood that an individual will experience poor outcomes in behavior, in health, on 479

480

FAGAN AND PALKOVITZ

measures of academic achievement, or in regard to vocational success (Harvey & Delfabbro, 2004). Risk factors are variables that diminish the likelihood of attaining personal goals or goals that society views as important (Specht, Polgar, & King, 2003). In this article, we consider risk factors to encompass conditions that hinder a consistent pattern of fathers’ involvement with children. In contrast, resilience suggests doing well in life despite adversity (Garmenzy & Matsen, 1991; Rutter, 1985, Specht et al., 2003). Resilience connotes that there are factors that protect individuals from being negatively affected by adversity or personal challenges. Risk and resilience perspectives may be particularly appropriate for studying this population of fathers because of observed rates of poverty, unstable work histories, low educational attainment, substance abuse, and legal problems (Meyer, 1998; Sorenson & Wheaton, 2000). Risk and resilience research typically focuses on individual rather than interpersonal variables (Specht et al., 2003). In the present study we suggest that father–mother relationship status represents a central risk or resilience factor in relation to father involvement. This perspective is supported by literature that suggests the importance of the father– mother relationship to paternal involvement with children among all relationship types, including married, cohabiting, divorced, and never-married fathers (Marsiglio, Roy, & Fox, 2005). Three types of relationships are common among nonresidential fathers and the mothers of their children— romantic, friend, and acquaintance. We suggest that nonresidential fathers’ risk for noninvolvement with children increases as the degree of closeness between father and mother decreases. There are a considerable number of high-risk fathers (e.g., acquaintances) that manage to stay involved with their young children. For example, EHS data reveal that 61% of nonresident fathers who are friends with the mother and 47% of fathers in acquaintance relationships provide “a lot of” caregiving help to the child (Cabrera et al., 2004). It is likely that within each relationship group, various risk and resilience factors affect paternal involvement with children.

Risk Variables The research literature suggests eight risk variables in addition to relationship status—low-quality mother–father relationship, having biological children in other residences besides the target child’s home, fathers’ legal problems, history of incarceration, lack of paternity establishment, lack of payment of formal and informal child support, drug abuse, and alcohol abuse. Low-quality relationship between mothers and fathers is a critical risk variable (Coley & Chase-Lansdale, 1999). Research has shown that paternal involvement was predicted most strongly by the quality of the parents’ romantic relationship (Gavin et al., 2002). Fagan, Barnett, Bernd, and Whiteman (2003) found that interparental conflict was negatively associated with young nonresident unmarried fathers’ prenatal involvement, regardless of whether the father and mother were still romantically involved with each other. Although relationship quality tends to decrease among couples that have broken

up, there are some fathers and mothers that maintain good quality relationships despite the low level of relationship closeness (McLanahan & Carlson, 2004). Fathers with children from other unions may be at risk for decreased involvement with their new children (Manning & Smock, 1999). These fathers may share their time between caring for other children as well as the target child in data collection. Dividing time between children living in separate residences may decrease involvement in comparison to fathers whose children live in the same household. Fathers’ legal problems, including incarceration or being charged with having committed a crime, also may have a negative influence on paternal involvement. Data reveal that incarceration weakens unmarried fathers’ connections to their children (Western, Lopoo, & McLanahan, 2004). The affects of incarceration on father involvement are complex and should be understood within the context of the relationship between father and mother (Roy & Dyson, 2005). Couples that do not establish legal paternity and that do not pay formal or informal child support may also be at greater risk for low levels of father involvement. Lack of establishment of paternity is associated with reduced contact with the child and having any overnight child visits (Mincy, Garfinkel, & Nepomnyaschy, 2005). Research has also shown a significant association between child-support payments and paternal involvement (Huang, Han, & Garfinkel, 2003). Huang (2006) estimated that improved childsupport enforcement results in a 45% increase in nevermarried fathers’ visitation with children. Finally, drug and alcohol abuse may be associated with low levels of father involvement with children. Substance abuse is comorbid with other mental ailments in more than 50% of men who are users, including passive–aggressive, obsessive– compulsive, and antisocial personality disorders (Brady, Grice, Dustan, & Randall, 1993). Such conditions are deleterious to positive paternal involvement.

Resilience Variables In the present study we suggest four resilience factors— employment in paid work, social network support, religious involvement, and a history of involvement of one’s biological father during childhood. Wilson (1987) hypothesized that employment accounts for increased family roles of the father in disadvantaged communities. Central to this notion is that fathers are more valued when they are able to provide for their families. Findings suggest that fathers who provide financial support to their children also tend to be more involved with them (Seltzer, McLanahan, & Hanson, 1998). Because providing is still central in role prescriptions for fathers, men without the means to provide acceptable levels of material support may avoid contact with their children to evade confrontation with their perceived inadequacies. Alternatively, mothers may encourage the involvement of fathers who are employed. Another potential protective factor is social support (Willoughby, Brown, Polgar, & Havens, 2003). Network support can influence fathers’ involvement by buffering men’s stress (Palkovitz, 2002) and may exert a direct influence on

NONRESIDENT, UNMARRIED FATHERS’ INVOLVEMENT

fathers’ involvement with children (Bunting & McAuley, 2004). For example, network members may actively encourage the nonresident, not-married father to stay involved during times when competing interests entice the man from fulfilling the responsibilities of parenthood. Researchers have also suggested the potential significance of spirituality or attendance at religious services for some individuals (Carlson & McLanahan, 2004; Chapman & Mullis, 2000). Roggman, Boyce, Cook, and Cook (2002) found a positive relationship between low-income fathers’ involvement with infants and both spirituality and active participation in one’s religion. Religious involvement may be particularly significant for fathers in high-risk situations (Dollahite, 2004). Finally, an additional resilience factor may be the positive involvement of one’s own biological father during childhood (Beaton, Doherty, & Rueter, 2003; Palkovitz, 2002). Researchers have suggested that men’s involvement with their children is influenced by what they observed in their family of origin (Amato & Booth, 2001; Snarey, 1993). Accordingly, men whose fathers were positive parental role models during their childhood may be more likely to stay involved with their own children even when faced with significant barriers to maintaining a relationship with the child.

Models of Risk and Resilience Risk and resilience research uses two models to examine predictors of outcomes—the cumulative and moderating (interactive) models (Luthar, Cicchetti, & Becker, 2000; Rutter, 1985). The cumulative model suggests that the accumulation of risk factors is associated with an increment in problematic outcomes (Campbell, 2005). There are three approaches for assessing the relationship between cumulative risk and outcomes—additive, multiplicative, and quadratic approaches. The additive and multiplicative approaches suggest steady linear relationships between risk factors and problematic outcomes (Appleyard, Egeland, & van Dulmen, 2005). The quadratic approach suggests a dramatic increase in problematic outcomes after a certain number of risks (i.e., threshold) have occurred. In the present study we examine both the additive and multiplicative approaches. Although the additive approach to risk research suggests a fairly gradual increase in problematic outcomes as the individual experiences additional risk, the multiplicative approach suggests a steady but more dramatic increase in negative outcomes (Pungello, Kupersmidt, Burchinal, & Patterson, 1996). The essence of this approach is that risk conditions result in increased sensitivity to other risk conditions. We expect fathers’ risk factors to have a more negative influence on fathers’ involvement with children when the father is in a higher risk relationship with the mother. That is, fathers’ additive risk factors will have a dramatically negative influence on fathers’ involvement with the child when the father is an acquaintance or friend relationship with the mother, compared with fathers in a romantic relationship with the mother.

481

The cumulative model can also be applied to examine the relationship between resilience factors and outcomes (Fraser, 2004). This model suggests that individuals with multiple resilience characteristics are likely to experience better outcomes. In the present study we expect that cumulative (additive) resilience factors will have a protective influence on father involvement. Luthar et al. (2000) suggested that resilience factors can have both main and moderating effects on outcomes. According to the moderation model, resilience factors have a relatively potent influence on high-risk individuals and a relatively minor influence on low-risk individuals. Fathers in the low-risk romantic involvement group are expected to benefit less than fathers in the higher risk friend and acquaintance groups from personal resilience factors. The moderation model also suggests that resilience factors may reduce the negative effect of cumulative risk factors. That is, there will be an interactive effect of fathers’ additive resilience score on the association between fathers’ additive risk score and paternal involvement with children.

Control Variables We controlled for the following variables: father’s age, race/ethnicity, education, child gender, child temperament, and maternal risk factors. Father’s age and education have been found to correlate with paternal involvement (Pleck, 1997). Several recent studies have found higher levels of paternal child-care involvement among African American fathers compared with European American fathers (Sanderson & Sanders Thompson, 2002). Studies generally support the finding that fathers are more involved with sons than with daughters (Pleck, 1997). McBride, Schoppe, and Rane (2002) found consistent associations between difficult child temperament and father involvement. We also controlled for mothers’ risk factors—nonresident fathers may increase their involvement with children if the mother has many risk factors. On the basis of these theoretical perspectives and empirical findings, we hypothesized the following: Hypothesis 1: The amount of child care provided by fathers will vary by father’s relationship status with the child’s mother. Fathers who are acquaintances with the child’s mother will provide the least amount of child care, fathers who are friends with their child’s mother will provide moderate amounts of care, and fathers in romantic relationships with their child’s mother will provide the greatest amounts of child care. Hypothesis 2: Fathers’ involvement in child care will be negatively associated with an additive index of risk factors across all three relationship groups (romantic, friend, acquaintance). Hypothesis 3: Fathers’ involvement in child care will be positively associated with an additive index of resilience factors across all three relationship groups.

482

FAGAN AND PALKOVITZ

Hypothesis 4: There will be a multiplicative relationship between the risk index and relationship status in relation to father involvement in child care.

Table 1 Demographic Characteristics of the Study Sample

Hypothesis 5: There will be a moderating relationship between the resilience index and relationship status and between the resilience index and risk index in relation to father involvement.

Relationship status Romantic Friends Acquaintance Father’s race White Black Hispanic Asian American Indian Other Father’s education Less than high school Some high school High school graduate or GED Some college or technical training College graduate or higher Father’s age (M, SD)

Method FF follows a cohort of nearly 5,000 children born in the United States between 1998 and 2000 (McLanahan & Garfinkel, 2000). The study over-samples births to unmarried couples; and, when weighted, the data are representative of nonmarital births in large U.S. cities at the turn of the century. The study consists of interviews with both mothers and fathers at birth and again when children are ages 1, 3, and 5. The sample is made up of 3,712 unwed couples and 1,200 married couples. Mothers and biological fathers were interviewed shortly after the child’s birth in the hospital. The response rate for baseline data was 87% for mothers and 75% for fathers. The one-year response rate was 90% for mothers and 70% for fathers who participated at baseline. The data for the present study are based on the sample of unwed and noncohabiting couples, fathers’ and mothers’ one-year follow-up interviews. The number of nonresident biological unmarried fathers that participated in the one-year follow-up was 835. A number of these fathers (n ⫽ 122) did not have any contact with the child during the month preceding the father interview. For these men, we recoded the father child-care variables as providing no child care during this time period. Twenty-two percent of the 835 fathers had substantial missing data that could not be easily imputed. We omitted these fathers from the study. The final sample consisted of 652 fathers with complete data or 78% of noncohabiting, unmarried fathers that participated in the one-year follow-up interview.

Participants Table 1 reveals that the majority of nonresident biological fathers in the sample self-identified as Black (66.3%), and smaller percentages identified themselves as White (8.6%) or Hispanic (15.5%). Fathers’ median education level was high school graduate, and their mean age was approximately 27 years. The majority of fathers were friends with the baby’s mother (54.4%). About equal proportions were romantically involved (22.4%) or in acquaintance relationships (23.2%) with the mother.

Variables Dependent variable. The present study focuses on components of fathers’ direct engagement with children (Lamb, Pleck, Chernov, & Levine, 1987) instead of creating variables that would highlight paternal availability or responsibility because meta-analytic reviews of father involvement

Variable

n

%

146 355 151

22.4 54.4 23.2

56 432 101 9 14 40

8.6 66.3 15.5 1.4 2.1 6.1

16 211 288 122 15 26.86

2.5 32.4 44.2 18.7 2.3 6.99

find moderate associations between indicators of paternal engagement and child outcomes (Marsiglio et al., 2000). Further, sensorimotor children do not have the capacity to meaningfully process paternal responsibility or availability (Hawkins & Palkovtiz, 1999). Only fathers’ self-reports of their involvement with the child are used. We were unable to triangulate by using mother reports because mothers’ reports of fathers’ involvement were not directly comparable in the one-year follow-up survey. All items are based on a scale with responses that ranged from 0 (no days) to 7 (7 days per week). Items include how often the father plays games such as peek-a-boo or gotcha, sings songs or nursery rhymes, reads stories, tells stories, plays inside with toys, and hugs or shows physical affection. An item centered on putting the child to bed was omitted because nonresident fathers are unlikely to have equal opportunity to participate in this activity. An item about taking the child to visit relatives was omitted because it overlaps with the social support measure. A composite father involvement score was created by summing fathers’ responses to the six involvement items. The Cronbach’s alpha for the scale was .92. Relationship status. Fathers were asked, “What is your relationship with (mother) now? Are you romantically involved, just friends, or not in any kind of relationship?” We use the term acquaintances to refer to fathers who said that they were not in any kind of relationship with the child’s mother. In all, 146 noncohabiting fathers stated that they were in romantic relationships, 355 said they were “just friends,” and 151 fathers indicated that they were in acquaintance relationships with the mother. Father risk variables. Low-quality mother–father relationship was measured with one item that asked fathers, “How is your relationship with the child’s mother?” Responses to this item were rated on a five-point scale, ranging from 1 (excellent) to 5 (poor). One item asked the respondent how many nonresident biological children he has other than the target child. One item asked fathers whether they have established legal paternity of the child (0 ⫽ yes or 1 ⫽

NONRESIDENT, UNMARRIED FATHERS’ INVOLVEMENT

no). The FF survey included one item about payment of formal child support and one item about payment of informal child support. These items were combined to form a child support variable (0 ⫽ yes or 1 ⫽ no). One item assessed legal problems: “Have you ever been booked or charged with breaking the law?” Responses were 1 ⫽ yes or 0 ⫽ no. One item asked whether the respondent currently has charges pending against him (1 ⫽ yes, 0 ⫽ no). One item asked whether the father had ever been incarcerated (1 ⫽ yes, 0 ⫽ no). One item asked fathers how many days in the past month they had at least five alcoholic drinks in one day. Responses ranged from 0 to 31. One item addressed fathers’ use of marijuana and one item focused on use of cocaine, crack, speed, LSD, heroin, or other drugs during the last month. The scores on these variables were combined to form a composite illegal drug use variable (0 ⫽ no, 1 ⫽ yes). The nine items were converted to z scores and then summed to create a fathers’ risk index. Father resilience variables. Involvement of the father’s own father was measured with one item, “Was your biological father involved in your life when growing up?” The response format ranged from 1(very involved) to 3 (not involved). The responses were reverse coded to form the resilience index. Attendance at religious services was assessed with one item, “How often do you go to religious services?” Responses ranged from 1 (more than one time per week) to 6 (never). Responses were reversed for the resilience index. Frequency of child visits with paternal grandparents was measured with one item: “How often does your child see your parents?” Responses ranged from 1 (one or more times per week) to 5 (never). Responses were recoded so that a high score suggests frequent visits to paternal grandparents. Finally, items from the baseline and one-year follow-up fathers’ surveys were used to measure father’s employment. The item from the one-year follow-up asked, “Have you worked since your child’s birth?” The item from the baseline survey asked, “Are you currently working?” Fathers who worked at baseline and follow-up were coded as 1. Fathers who did not work at baseline but who worked at follow-up were also coded as 1; we assumed that these fathers were showing movement toward resilience by increasing their ability to engage in economic provision. Fathers who worked during neither time period or during baseline but not during follow-up were coded as 0. After calculating z scores on all items, the resilience index was calculated by summing the recoded scores for each item. Control variables. Fathers were asked to indicate their age in years. The FF eight response categories for education were collapsed into five categories: less than high school, some high school, high school graduate or GED, some college or technical training, and college graduate/graduate school. The recoded variable was treated as interval data in subsequent analyses. Three race dummy variables were created: Black, White, and Hispanic fathers. The omitted category was “others” (Asian, American Indian, and other). Child gender (1 ⫽ boy, 0 ⫽ girl) was also included as a control variable. Mothers’ risk measures were obtained from the mother’s one-year follow-up survey. Though some maternal risk

483

items were the same as the father risk items, the surveys were not identical, necessitating the use of several different mother risk items. FF included one item about mother’s general health. Responses to this question ranged from 1 (excellent) to 5 (poor). Mothers’ illegal drug and alcohol use were coded in a fashion directly parallel to fathers’. We recoded mother’s age into a dichotomous variable that indicated whether the respondent was a teenage mother (age 19 years and below) or not (0 ⫽ adult, 1 ⫽ teen). Finally one item asked the mother whether she could count on someone to help with child care in an emergency (0 ⫽ available support, 1 ⫽ no available support). These five items were first converted to z scores and then summed to create a mothers’ risk index. We also controlled for children’s difficult temperaments by using data obtained from the mother’s follow-up survey. Three items were available: child often fusses and cries, child gets easily upset, and child reacts strongly when upset. Responses ranged from 1 (not at all) to 5 (very much). The items were added together to form a difficult temperament variable. The Cronbach’s alpha for the scale was .60.

Results Preliminary and Missing Data Analyses A series of statistical tests were conducted to determine whether unmarried nonresident fathers differed from married, coresidential; separated; or unmarried, cohabiting fathers in the FF study. There were no divorced fathers at the time of the one-year follow-up. There was a significant association between relationship status and father’s age, F(5, 3356) ⫽ 75.30, p ⬍ .001. Scheffe’s method for multiple comparisons revealed that married, coresidential fathers were significantly older than fathers in unmarried, cohabiting; romantic; friend; or acquaintance relationships with the child’s mother. There was also a significant association between relationship status and fathers’ education, F(5, 3356) ⫽ 8.06, p ⬍ .001. Scheffe tests revealed that married, coresidential fathers had completed a higher level of education than fathers in unmarried, cohabiting; romantic; or friend relationships. Analyses were performed to determine whether participants that were omitted from the study because of missing data differed from those that remained in the study (i.e., no missing data on the study variables; see Acock, 2005). Participants with missing data did not differ from those without missing data on father’s age, race/ethnicity, or education; child gender; child temperament; or mothers’ risk index. Participants with missing data scored higher, however, on the fathers’ risk index, t(833) ⫽ 2.14, p ⬍ .05, and they scored higher on the fathers’ resilience index, t(833) ⫽ 2.13, p ⬍ .05. Participants with missing data reported lower quality relationships with the child’s mother than those with no missing data, t(833) ⫽ 3.28, p ⬍ .001. Those with missing data reported higher levels of social support from family members, t(833) ⫽ 2.66, p ⬍ .01, and greater likelihood of engaging in continuous work, ␹2(1, N ⫽ 835) ⫽ 7.36, p ⬍ .01.

484

FAGAN AND PALKOVITZ

Descriptive and Bivariate Analyses

Table 2 also shows the descriptive statistics (z scores) for the risk and resilience composite scores and individual variables across all relationship groups. There was a significant univariate relationship between relationship status and the additive risk index, F(2, 649) ⫽ 6.54, p ⬍ .01. Scheffe tests revealed that the mean risk index was higher for acquaintance fathers in comparison with romantic fathers and friends. Acquaintance fathers had poorer relationships with the mother than romantically involved or friend-only fathers, and friend-only fathers had poorer relationships than romantically involved fathers. There was also a significant univariate relationship between relationship status and the additive resilience index, F(2, 649) ⫽ 7.15, p ⬍ .001. The mean resilience index was greater for romantically involved and friend-only fathers compared with acquaintance fathers. Fathers’ social support was the only resilience variable that was associated with relationship status, F(2, 649) ⫽ 15.70, p ⬍ .001. Romantically involved and friendonly fathers were more likely than acquaintance fathers to take their child to see grandparents.

The average score for father involvement for all fathers suggests that men provide care to their children about 2.23 days per week (SD ⫽ 2.0). The first study hypothesis suggested that fathers in acquaintance relationships would be significantly less involved with their children than friends or romantically involved fathers, and friends would be less involved than romantically involved fathers. The findings partially supported this hypothesis (see Table 2). There was a significant univariate relationship between relationship status and father involvement, F(2, 649) ⫽ 13.70, p ⬍ .001. Scheffe’s method for multiple comparisons revealed that the mean paternal involvement score for acquaintance fathers was significantly lower than the mean involvement scores of romantic and friend-only fathers. Descriptive analyses revealed that fathers experience many risk factors (these data are not shown in a table). Most fathers (60%) reported that their relationship status with the mother declined in level of closeness (i.e., from romantic to friend or acquaintance) from the baseline survey to the 1-year follow-up. On the average, 48% of fathers had been booked or charged with breaking the law, 29% had spent time in jail, 9% had charges pending, 30% paid no formal or informal child support, 12% drank five or more alcoholic beverages per day more than two times per month, 11% used illegal drugs in the last month, 40% had additional biological children not residing with the target child and not residing with the father, and 48% reported fair to poor relationships with the mother. In regard to the resilience factors, 65% said their own biological father was somewhat or very involved while growing up, 83% worked continuously in paid jobs, 72% took the child to see paternal grandparents a few times per month or more, and 44% attended religious services a few times per month or more.

Multivariate Analyses We used a hierarchical multiple regression procedure to test the multivariate association between father involvement and the risk index, resilience index, and control variables (see Table 3). The control variables were entered simultaneously in Model 1; the risk and control variables were entered in Model 2; and the resilience, risk, and control variables were entered in Model 3. Model 4 includes all independent variables plus significant multiplicative and moderating terms. The control variables explained 3% of the variance in the dependent variable (p ⬍ .01). Older nonresident fathers were significantly less likely to provide caregiving to their

Table 2 Descriptive Statistics for the Major Study Variables by Relationship Group Romantic (n ⫽ 146) Variable Father involvement Risk index Low quality relationship No. other children Legal problems Charges pending Drug use No legal paternity History of incarceration Alcohol abuse Does not pay child support Resilience index Steady employment Social support Own biological father inv. Attends religious serv.

M

SD a

15.03 ⫺0.60 ⫺0.74 ⫺0.04 0.02 0.10 ⫺0.05 ⫺0.11 0.13 ⫺0.06 ⫺0.01 0.37 0.01 0.24 0.07 0.05

12.39 3.57 0.84 0.86 1.00 1.13 0.94 0.95 1.06 0.86 0.90 1.83 0.99 0.89 0.96 0.98

Friends (n ⫽ 355) M 14.59 ⫺0.14 ⫺0.08 0.05 ⫺0.02 ⫺0.04 ⫺0.03 0.00 ⫺0.06 0.01 ⫺0.06 0.08 0.04 0.06 0.00 ⫺0.02

SD 11.74 3.70 0.88 1.09 1.00 0.94 0.96 1.00 0.97 0.04 0.88 2.10 0.96 0.96 1.00 1.00

Acquaintance (n ⫽ 151) M 9.01 0.91 0.90 ⫺0.08 0.03 0.00 0.13 0.09 0.01 0.04 0.14 ⫺0.54 ⫺0.12 ⫺0.37 ⫺0.06 0.00

SD

␹2

F ***

11.23 13.70 4.05 6.54** 0.70 ⫺147.49*** 0.90 1.00 1.00 1.14 1.04 1.01 0.00 0.80 1.03 2.53 7.15*** 1.09 1.09 15.70*** 1.04 1.27 1.03 0.58

1.04 0.28 2.11 3.25 2.98 3.86 2.48† 2.73

Scheffe R, F ⬎ A A ⬎ R, F A ⬎ R, F; F ⬎ R

R, F ⬎ A R, F ⬎ A

Note. Means and standard deviations are z scores for all variables except father involvement. R ⫽ romantic relationship; F ⫽ friends; A ⫽ acquaintance; inv. ⫽ involved; serv. ⫽ service. df for F ⫽ 2, 651. df for ␹2 ⫽ 2, N ⫽ 652. a 15.03 is equivalent to 2.51 days per week. † p ⬍ .10. ** p ⬍ .01. *** p ⬍ .001.

*

p ⬍ .05.

**

p ⬍ .01.

Change in R F for all variables Total R2 N

***

p ⬍ .001.

.03 2.28** .03 652

0.07 0.58 1.59 2.16 1.90 0.94 0.15 0.20

⫺0.21 0.91 3.23 3.13 5.64 0.52 ⫺0.10 0.05

Father’s age Father’s education African American White Hispanic Child gender Child temperament Mother risk index Father Risk Acquaintance relationship Risk index Father Resilience Romantic relationship Resilience index Multiplicative relationship Acquaintance ⫻ Risk Index

2

SE

B

Variable

Model 1 **

⫺.12 .06 .13* .07 .17** .02 ⫺.03 .01



1.07 0.12

⫺4.91 ⫺0.80

.10 8.81*** .13 652

0.07 0.56 1.51 2.05 1.82 0.90 0.14 0.19

SE

Model 2 ⫺0.21 0.47 2.36 2.37 3.87 0.38 ⫺0.13 0.03

B **

⫺.17*** ⫺.25***

⫺.13 .03 .09 .06 .12* .02 ⫺.03 .01



Table 3 Hierarchical Regression of Father Involvement on Risk, Resilience, and Control Variables

1.09 0.21

⫺0.17 1.27

.05 10.67*** .18 652

1.09 0.12

0.06 0.54 1.48 2.00 1.78 0.87 0.14 0.19

SE

Model 3

⫺4.26 ⫺0.63

⫺0.23 0.22 2.29 2.33 2.90 0.29 ⫺0.11 0.04

B ***

⫺.01 .23***

⫺.15*** ⫺.20***

⫺.13 .02 .09 .05 .09 .01 ⫺.03 .01



0.26

0.68

.01 10.49*** .19 652

1.08 0.21

1.10 0.14

0.06 0.54 1.47 1.99 1.77 0.87 0.14 0.19

SE

Model 4

⫺0.24 1.27

⫺4.69 ⫺0.82

⫺0.23 0.23 2.38 2.36 2.93 0.19 ⫺0.08 0.04

B

0.11**

⫺.01 .23***

⫺.17*** ⫺.26***

⫺.13*** .02 .09 .06 .09 .01 ⫺.02 .01



NONRESIDENT, UNMARRIED FATHERS’ INVOLVEMENT 485

486

FAGAN AND PALKOVITZ

infants. African American and Hispanic fathers were significantly more likely to provide caregiving. The findings supported the study’s second hypothesis (see Model 2): The additive risk index was negatively related to paternal involvement. There was also a significant negative relationship between acquaintance fathers and father involvement. These risk variables explained an additional 10% of the variance in father involvement. The third study hypothesis was supported (see Model 3): The resilience index was positively associated with father involvement. However, there was no significant relationship between romantic involvement with the mother and the dependent variable. The resilience variables explained an additional 5% of the variance in the dependent variable. In support of the study’s fourth hypothesis, the results revealed a significant multiplicative relationship between the risk index and acquaintance fathers (see Model 4). A post hoc analysis was conducted to further examine the multiplicative relationship. Model 4 was run first with acquaintance fathers only in the sample and no interactive term, and then with romantic and friends in the sample and no interactive term (not shown in a table). The risk index had a greater negative influence on paternal involvement in caregiving among fathers in the acquaintance group (␤ ⫽ ⫺.28, p ⬍ .001) than among fathers in the other relationship groups (␤ ⫽ ⫺.16, p ⬍ .001). There was no support for the study’s fifth hypothesis: The resilience index did not moderate the relationship between risk and fathers’ caregiving.

Discussion As in other analyses of at-risk fathers (Cabrera et al., 2004; McLanahan & Carlson, 2004), the present study confirms that unmarried nonresidential fathers maintain patterns of involvement with their children one year after the birth. Overall, the analyses provided evidence that the nature of the father’s relationship with the child’s birth mother predicts variability in levels of fathers’ involvement with their children. In general, the closer the relationship between the father and his child’s birth mother, the higher the level of paternal involvement in child care. Fathers in romantic and friendship relationships with their child’s mother maintain significantly higher levels of engagement with their infants than fathers in acquaintance relationships with the mother. The multivariate data demonstrate that additive risk and resilience factors influence paternal involvement with infants. These data further reveal that risk factors influence fathers’ involvement in different ways for fathers in different kinds of relationships with their children’s birth mothers. The findings do not show that the additive resilience index moderates the association between relationship status and paternal involvement. The resilience index also does not moderate the relationship between the additive risk index and father involvement. Hypothesis 1 was partially supported. Specifically, data analyses confirmed that fathers’ relationship classification (romantic, friend, acquaintance) was a significant predictor of father’s composite involvement score. There were not significant differences in involvement scores between all

groups of relationship classifications. Fathers who had either romantic or friendship relationships with their child’s biological mother had higher levels of father involvement with their children than men with acquaintance relationships. However, romantically involved fathers and friends did not differ significantly in their involvement scores with their children. This finding points to the importance of fathers maintaining some form of relationship with the child’s mother in order to facilitate ongoing involvement in the child’s life. Hypothesis 2 found support as well. Scores on the additive risk index were negatively related to paternal involvement across all relationship classifications. Although previous studies have shown significant relationships between risk variables (e.g., legal problems) and paternal involvement (Smith, Krohn, Chu, & Best, 2005), this study revealed that accumulating risk factors have a negative influence on involvement. To the best of our knowledge, prior research has not examined the influence of additive risk factors on involvement. Future research should focus on the mechanisms by which risk variables influence fathering. For example, it is not clear whether additive risk factors result in reduced involvement because fathers are less available to provide care to their children or whether risk factors are consistent with a pattern of lower overall competency in all areas of psychosocial functioning on the part of the father. It would be beneficial to engage in exploratory qualitative work to explore underlying processes, meanings, and motives that diminish paternal involvement in at-risk fathers. Hypothesis 3 was supported. The additive resilience index was positively associated with paternal involvement. It is interesting to note that the relationship between the resilience index and paternal child-care involvement was about equal in size to the relationship between the risk index and the dependent variable. This finding suggests the importance of including measures of additive resilience in studies of at-risk fathers. We noted several limitations with the resilience index obtained from FF. The data set included a limited number of variables that appeared to measure resilience. For example, FF did not include measures of fathers’ resilient personality characteristics, which have been found to correlate with father involvement in high-risk populations (Fagan et al., 2003). Our analyses may therefore underestimate the effect of resilience factors on paternal involvement. Hypothesis 4 was supported. Regression results reflected a multiplicative relationship between the risk index and relationship status, as risk had its greatest influence in impeding father involvement in the acquaintance group. These findings are noteworthy because they suggest that the relationship between risk factors and paternal involvement in child care is both additive and multiplicative. That additive risk factors operate differently and independently across relationship categories to predict patterns of father involvement raises important questions about underlying processes. Acquaintance fathers appear to be more sensitive than fathers in other relationship groups to the adverse influence of risk factors. It is possible that mothers are more willing to overlook fathers’ risk characteristics when the couple is still in a close relationship, such as friendship or

NONRESIDENT, UNMARRIED FATHERS’ INVOLVEMENT

romantic relationship. On the other hand, mothers in acquaintance relationships may have little motivation to encourage fathers’ involvement with children when the man has multiple risk characteristics. Hypothesis 5 was not supported. There was no significant moderating relationship between the additive resilience index and relationship status or between the additive risk and additive resilience indexes. Many resilience studies find relatively small interaction effects (Roosa, 2000). The lack of a significant interaction effect may be a reflection of the fact that more robust measures of resilience are needed. Although secondary analysis of large data sets affords many analytic advantages, it is not the ideal context for creating fully elaborated measures to rigorously test theoretical relationships that were not centrally represented in the original measures. The descriptive findings of this study were also noteworthy because they suggest that fathers in the FF study experience many risk factors. The first year following the child’s birth also appears to be critical for examining the effect of risk factors. The majority of fathers experienced a decline in relationship closeness with the child’s mother between the child’s birth and first birthday. Nearly half of the fathers in this study had experience with the criminal justice system, and almost half of the fathers have additional nonresident biological children not living with the target child. Many fathers reported poor relationships with the target child’s mother. More than one third of fathers reported little to no involvement with their own biological father while growing up. The descriptive analyses also revealed differences in risk and resilience factors across relationship categories. Acquaintance fathers scored significantly higher on the risk index than fathers in romantic or friend relationships with the mother. Acquaintance fathers were far more likely than their counterparts to report having a lowquality relationship with the mother. Romantic and friend fathers, on the other hand, scored significantly higher than acquaintance fathers on the resilience index. Most of the variance was explained by one resilience variable— social support. Extended family members may be more motivated to pursue involvement with the child when the couple has some level of close relationship with each other. On the other hand, mothers may restrict the degree to which they allow their children to visit paternal grandparents when the couple relationship has dissolved. This explanation suggests that maternal gatekeeping (Allen & Hawkins, 1999) is applied differently to fathers with different relationships with the mother. We think these findings are of sufficient interest to warrant future examination by researchers. The involvement of extended family members may be critical for the continued involvement of nonresidential fathers with their children. The potential for fathers in acquaintance relationships to stay in touch with their children may be diminished if their parenting is not embedded in a network of caring individuals from the onset.

487

Limitations This study is limited by the factors inherent in research that uses self-report measures. Previous studies have shown that fathers overestimate involvement with children (Wical & Doherty, 2005). Because of incomparable data, we were not able to use a multimethod approach of using both mother and father reports of paternal involvement with children. Further limitations are associated with conducting secondary analyses of large data sets. Creating additive indexes of risk and resilience factors from available items in the data set produced variables that are unevenly developed, and the limited number of risk factors examined in the data set may have muted the real effects of risk on paternal involvement. These factors may explain the relatively moderate amount of variance in the dependent variable accounted for by the risk and resilience indexes. The data are also limited because of the amount of missing data in the one-year follow-up interviews with fathers. Approximately 22% of fathers who participated in the one-year follow-up had missing data on at least one of the study variables. Fathers with missing data scored higher on both the additive risk index and the additive resilience index, suggesting that these fathers are qualitatively different from fathers that answered all questions in the survey.

Conclusion The findings of this study suggest that the status of the father’s relationship with his child’s birth mother serves an important influence on father involvement. Men in different kinds of relationships with their children’s mother experience different levels and patterns of influence from risk factors. Men in the most tenuous relationships also have the fewest resilience factors and the greatest number of risk factors. These patterns may indicate a general lack of social capital (Coleman, 1988) on the part of nonresidential unmarried fathers without ongoing relationships with their children’s mother. Though they manifest the lowest level of involvement with their infants, it is remarkable that they can navigate the deterrents to engagement to maintain some level of involvement with their children. These findings demonstrate the importance of situating fathering within the larger framework of family relationships and the personal risk and resilience factors influencing those men. We think it also worth noting that risk and resilience factors pertinent for nonresidential fathers may also be predictive of involvement of residential fathers with their infants. There are policy implications to be gleaned from these findings. Programs serving nonresident unmarried fathers should place greater emphasis on fathers and mothers maintaining a meaningful relationship with each other, even if that relationship is one based in friendship rather than in romantic involvement. Nonresident fathers are at high risk for becoming disconnected to their children over time. Lacking a minimally close relationship, as is the case when couples become acquaintances, is likely to result in lower levels of paternal engagement of children. Our findings also point to the importance of programs that address the risk

488

FAGAN AND PALKOVITZ

and resilience conditions affecting those fathers. Although we cannot conclude from our data that reductions in cumulative risk factors will lead to greater father involvement, we think that policy makers should consider the development of demonstration projects and experimental research studies that examine the effects of reducing fathers’ risks and increasing their resilience in addition to encouraging responsible fatherhood and healthy couple relationships. We note the importance of conducting such studies with fathers and families during the first year or two following the child’s birth, a time period when families are at considerable risk for losing touch with their children. We further suggest that focused qualitative work in regard to men’s experiences of risk and resilience factors would inform our ability to design interventions and to formulate policies that may facilitate continued father involvement in fragile families.

References Acock, A. C. (2005). Working with missing values. Journal of Marriage and Family, 67, 1012–1028. Allen, S. M., & Hawkins, A. H. (1999). Maternal gatekeeping: Mothers’ beliefs and behaviors that inhibit greater father involvement in family work. Journal of Marriage and the Family, 61, 199 –212. Amato, P. R. (1998). More than money? Men’s contributions to their children’s lives. In A. Booth & A. Crouter (Eds.), Men in families: When do they get involved? What difference does it make? (pp. 241–278). Mahwah, NJ: Erlbaum. Amato, P., & Booth, A. (2001). The legacy of parents’ marital discord: Consequences for children’s marital quality. Journal of Personality and Social Psychology, 81, 627– 638. Appleyard, K., Egeland, B., & van Dulmen, M. H. M. (2005). When more is not better: The role of cumulative risk in child behavior outcomes. Journal of Child Psychology and Psychiatry, 46, 235–245. Beaton, J. M., Doherty, W. J., & Rueter, M. A. (2003). Family of origin processes and attitudes of expectant fathers. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 1, 149 –168. Brady, K. T., Grice, D. E., Dustan, L., & Randall, C. (1993). Gender differences in substance use disorders. American Journal of Psychiatry, 150, 1707–1711. Bunting, L., & McAuley, C. (2004). Research review: Teenage pregnancy and parenthood: The role of fathers. Child and Family Social Work, 9, 295–303. Cabrera, N. J., Ryan, R. M., Shannon, J. D., Brooks-Gunn, J., Vogel, C., Raikes, H., et al. (2004). Low-income fathers’ involvement in their toddlers’ lives: Biological fathers from the Early Head Start Research and Evaluation Study. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 2, 5–30. Campbell, S. (2005). Determining overall risk. Journal of Risk Research, 8, 569 –581. Carlson, M. J., & McLanahan, S. S. (2004). Early father involvement in fragile families. In R. Day & M. Lamb (Eds.), Conceptualizing and measuring father involvement (pp. 241–271). Mahwah, NJ: Erlbaum. Carlson, M. J., McLanahan, S. S., & Brooks-Gunn, J. (2005). Unmarried but not absent: Fathers’ involvement with children after a nonmarital birth. Center for Research on Child Wellbeing (working paper #05– 07-FF). Retrieved April 4, 2007, from http:// crcw.princeton.edu/workingpapers/WP05– 07-FF-carlson.pdf

Chapman, P. L., & Mullis, R. L. (2000). Racial differences in adolescent coping and self-esteem. Journal of Genetic Psychology, 161, 152–160. Coleman, J. H. (1988). Social capital in the creation of human capital. American Journal of Sociology, 94, 95–120. Coley, R. L., & Chase-Lansdale, P. L. (1999). Stability and change in paternal involvement among urban African American fathers. Journal of Family Psychology, 13, 1–20. Dollahite, D. C. (2004). A narrative approach to exploring responsible involvement of fathers with their special-needs children. In R. D. Day & M. E. Lamb (Eds.), Conceptualizing and measuring father involvement (pp. 109 –127). Mahwah, NJ: Erlbaum. Duncan, G. J., & Brooks-Gunn, J. (1997). Income effects across the life span: Integration and interpretation. In G. J. Duncan & J. Brooks-Gunn (Eds.), Consequences of growing up poor (pp. 596 – 610). New York: Russell Sage Foundation. Fagan, J., Barnett, M., Bernd, E., & Whiteman, V. (2003). Prenatal involvement of adolescent unmarried fathers. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 1, 283–302. Fraser, M. (2004). Risk and resilience in childhood. Washington, DC: National Association of Social Workers. Garmenzy, N., & Matsen, A. S. (1991). The protective role of competence indicators in children at risk. In E. A. Cummings, A. L. Green, & K. H. Karraker (Eds.), Life-span developmental psychology: Perspectives on stress and coping (pp. 151–174). Hillsdale, NJ: Erlbaum. Gavin, L. E., Black, M. M., Minor, S., Abel, Y., Papas, M. A., & Bentley, M. E. (2002). Young, disadvantaged fathers’ involvement with their infants: An ecological perspective. Journal of Adolescent Health, 31, 266 –276. Harvey, J., & Delfabbro, P. H. (2004). Psychological resilience in disadvantaged youth: A critical overview. Australian Psychologist, 39, 3–13. Hawkins, A. J., & Palkovitz, R. (1999). Beyond ticks and clicks: The need for more diverse and broader conceptualizations and measures of father involvement. Journal of Men’s Studies, 8, 11–32. Huang, C. (2006). Child support enforcement and father involvement for children in never-married families. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 4, 97–111. Huang, C., Han, W., & Garfinkel, I. (2003). Child support enforcement, joint legal custody, and parental involvement. Social Service Review, 77, 255–278. Lamb, M. E. (1999). Noncustodial fathers and their impact on children of divorce. In R. A. Thompson & P. Amato (Eds.), The postdivorce family: Research and policy issues (pp 105–125). Thousand Oaks, CA: Sage. Lamb, M. E., Pleck, J., Charnov, E., & Levine, J. A. (1987). A biosocial perspective on paternal behavior and involvement. In J. Lancaster, J. Altmann, A. Rossi, & L. Sherrod (Eds.), Parenting across the lifespan: Biosocial dimensions (pp. 111–142). New York: Aldine de Gruyter. Luthar, S. S., Cicchetti, D., & Becker, B. (2000). Research on resilience: Response to commentaries. Child Development, 71, 573–575. Manning, W. D., & Smock, P. J. (1999). New families and nonresident father– child visitation. Social Forces, 78, 87–116. Marsiglio, W., Amato, P., Day, R. D., & Lamb, M. E. (2000). Scholarship on fatherhood in the 1990s and beyond. Journal of Marriage and Family, 62, 1173–1191. Marsiglio, W., Roy, K., & Fox, G. L. (Eds.). (2005). Situated

NONRESIDENT, UNMARRIED FATHERS’ INVOLVEMENT fathering: A focus on physical and social spaces. New York: Rowman & Littlefield. McBride, B. A., Schoppe, S. J., & Rane, T. R. (2002). Child characteristics, parenting stress, and parental involvement: Fathers versus mothers. Journal of Marriage and Family, 64, 998 –1011. McLanahan, S., & Carlson, M. S. (2004). Fathers in fragile families. In M. E. Lamb (Ed.), The role of the father in child development (4th ed., pp. 368 –396). Hoboken, NJ: Wiley. McLanahan, S., & Garfinkel, I. (2000). The Fragile Families and Child Wellbeing Study: Questions, design, and a few preliminary results (Center for Research on Child Wellbeing, Working Paper # 00 – 07). Retrieved June 19, 2006, from Princeton University, Center for Research on Child Wellbeing: http://crcw.princeton.edu/papers.html McLanahan, S., & Sandefur, G. (1994). Growing up with a single parent: What hurts, what helps. Cambridge, MA: Harvard University Press. Meyer, D. R. (1998). The effect of child support on the economic status of nonresident fathers. In I. Garfinkel, S. McLanahan, D. R. Meyer, & J. A. Seltzer (Eds.), Fathers under fire: The revolution in child support enforcement (pp. 94 –127). New York: Russell Sage Foundation. Mincy, R., Garfinkel, I., & Nepomnyaschy, L. (2005). In-hospital paternity establishment and father involvement in fragile families. Journal of Marriage and Family, 67, 611– 626. Mincy, R., & Oliver, H. (2003). Age, race, and children’s living arrangements: Implications for TANF reauthorization. Washington, DC: The Urban Institute Press. Nelson, T. J. (2004). Low-income fathers. Annual Review of Sociology, 30, 427– 451. Palkovitz, R. (2002). Involved fathering and men’s adult development: Provisional balances. Mahwah, NJ: Erlbaum. Pleck, J. H. (1997). Paternal involvement: Levels, sources, and consequences. In M. E. Lamb (Ed.), The role of the father in child development (3rd ed., pp. 33– 48). New York: Wiley. Pungello, E. P., Kupersmidt, J. B., Burchinal, M. R., & Patterson, C. J. (1996). Environmental risk factors and children’s achievement from middle childhood to adolescence. Developmental Psychology, 32, 755–767. Roggman, L. A., Boyce, L. K., Cook, G. A., & Cook, J. (2002). Getting dads involved: Predictors of father involvement in Early Head Start and with their children. Infant Mental Health Journal, 23, 62–78. Roosa, M. W. (2000). Some thoughts about resilience versus positive development, main effects versus interactions, and the value of resilience. Child Development, 71, 567–569. Roy, K. M., & Dyson, O. L. (2005). Gatekeeping in context: Babymama drama and the involvement of incarcerated fathers. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 3, 289 –310. Rutter, M. (1985). Resilience in the face of adversity: Protective

489

factors and resistance to psychiatric disorders. British Journal of Psychiatry, 147, 598 – 611. Sanderson, S., & Sanders Thompson, V. L. (2002). Factors associated with perceived paternal involvement in childrearing. Sex Roles, 46, 99 –111. Seltzer, J. A., McLanahan, S., & Hanson, T. L. (1998). Will child support enforcement increase father– child contact and parental conflict after separation? In I. Garfinkel, S. McLanahan, D. R. Meyer, & J. A. Seltzer (Eds.), Fathers under fire: The revolution in child support enforcement (pp. 157–190). New York: Russell Sage Foundation. Smith, C. A., Krohn, M. D., Chu, R., & Best, O. (2005). African American fathers. Journal of Family Issues, 26, 975–1001. Snarey, J. (1993). How fathers care for the next generation: A four decade study. Cambridge, MA: Harvard University Press. Sorenson, E., & Wheaton, L. (2000). Income and demographic characteristics of nonresident fathers in 1993. Washington, DC: Department of Health and Human Services. Specht, J., Polgar, J. M., & King, G. A. (2003). How we got here. In G. A. King, E. G. Brown, & L. K. Smith (Eds.), Resilience: Learning from people with disabilities and the turning points in their lives (pp. 7–29). Westport, CT: Praeger. Stewart, S. D. (1999). Nonresident mothers’ and fathers’ social contact with children. Journal of Marriage and the Family, 61, 894 –907. Stier, H., & Tienda, M. (1993). Are men marginal to the family? Insights from Chicago’s inner city. In J. C. Hood (Ed.), Men, work, and family (pp. 23– 44). Newbury Park, CA: Sage. U.S. Census Bureau. (2006). Statistical abstract of the United States: 2006. Retrieved on September 20, 2006, from http:// www.census.gov/compendia/statab/population/pop.pdf Western, B., Lopoo, L., & McLanahan, S. (2004). Incarceration and the bonds among parents in fragile families. In M. Patillo, D. Weiman, & B. Western (Eds.), The impact of incarceration on families and communities. New York: Russell Sage. Wical, K. A., & Doherty, W. J. (2005). How reliable are fathers’ reports of involvement with their children? A methodological report. Fathering: A Journal of Theory, Research, and Practice About Men as Fathers, 3, 81–91. Willoughby, C., Brown, E. G., Polgar, J. M., & Havens, L. (2003). The resilient self—What helps and what hinders. In G. A. King, E. G. Brown, & L. K. Smith (Eds.), Resilience: Learning from people with disabilities and the turning points in their lives (pp. 7–29). Westport, CT: Praeger. Wilson, W. (1987). The truly disadvantaged. Chicago: University of Chicago Press.

Received April 7, 2006 Revision received October 1, 2006 Accepted October 2, 2006