* Portions of this paper were previously presented at the 1996 American Sociological Association Annual Meeting, New York and the 1997 National Council on Family Relations’ Annual Meeting, Washington, D.C. We thank William Aquilino, Tamela Eitle, Nadine Marks, and Debra Umberson and anonymous JMF reviewers for their assistance with this work. Address all correspondence to Adam Shapiro, Department of Sociology, University of North Florida, Jacksonville, FL 32224 (ashapiro@unf.edu).

Using panel data from the National Survey of Families and Households (N=844), we examine (a) the impact of divorce on father-child relationship quality and fathers’ psychological well-being, (b) the extent to which the residence of a focal child moderates these associations, and (c) how changes in father-child relationship quality over time affect fathers’ psychological well-being. Results indicate that the effect of divorce on father-child relationship quality and fathers’ psychological well-being is moderated by the residence of children. Divorce is associated with lower relationship quality only for non-resident fathers, and is associated with a decline in happiness for co-resident fathers. Divorced fathers are more depressed than their married counterparts regardless of child residence. Changes in relationship quality do not significantly influence fathers’ psychological well-being.

Key Words: Fatherhood, Divorce, Mental Health, Intergenerational Relationships

Fathers’ relationships with their children have become quite complex as the proportion of marriages which end in divorce nears 60 percent (Norton & Miller, 1992). While a voluminous body of research has emerged in recent years which documents the aftermath of divorce for children (Amato & Booth, 1991; Amato & Keith, 1991) and mothers (Morgan, 1991), relatively little research has addressed the impact of divorce on fathers. The role of the divorced father in the life of his children is a timely research and policy issue, and understanding fathers’ experience of divorce is paramount to this dialogue.

Past research in the divorce and parenthood literatures suffer from three limitations. First, much published research uses data collected from fathers of only non-resident children. While divorced fathers are much less likely to co-reside with a child than divorced mothers, those who do are becoming increasingly more common and are in need of further consideration (Meyer & Garasky, 1993). Second, studies of father-child relationships are frequently based on reports from the child and the current or former spouse. Research is needed which addresses fathers’ perceptions of relationships with children. Third, and perhaps most important, the bulk of these studies are cross-sectional. This is quite problematic because it is plausible to argue that fathers who have positive pre-divorce evaluations of their relationship with their children may be more likely to maintain those relationships after divorce. Similarly, fathers with low pre-divorce psychological well-being may refrain from being involved with their children. The only effective means of accounting for these selection effects is by taking measurements before and after the divorce.

We address these shortcomings by using a nationally representative panel study of fathers with either resident or non-resident children. Informed by a microstructural perspective, we pose three specific questions: (1) How does divorce influence father-child relationship quality and fathers’ psychological well-being over time? (2) Is the effect of divorce on father-child relationship quality and psychological well-being influenced by the residential status of children? (3) Do changes in father-child relationship quality over time affect fathers’ psychological well-being?

Microstructural theory provides a useful framework for analyzing the experience of post-divorce parenting (Risman, 1987; Risman & Park, 1988). This perspective suggests that varying parental experiences can be accounted for by different socially structural opportunities and social expectations which are linked to marital status. When comparing divorced and married fathers, for example, we see that married fathers have far more parental opportunities than do divorced fathers. Relative to divorced fathers, married fathers generally co-reside with their children, have larger kin networks, and have a generally well-defined cultural script to follow as fatherhood is socially legitimated and defined through marriage (Arendell, 1995).

Microstructural theory suggests that the parenting experiences of married and divorced fathers would be similar if they were given identical expectations and positions in society (Risman, 1987; Risman & Park, 1988). This indicates that if a divorced father were to become the primary caretaker of the child, the expectations and opportunities of that status would possibly negate the effects of changing marital status. In fact, single co-residential fathers might be likely to report both greater parental stress and satisfaction (Hughes, 1989; Umberson & Gove, 1989) than married fathers with spouses, as married fathers may be exempt from being the primary caretaker.

One of the more powerful microstructural forces exerted on divorced fathers is cultural expectations for parenting behavior (Risman, 1987). Recent scholarship has pointed to the greater ambiguity of the social and cultural script for fathers, particularly divorced fathers. Griswold (1993) argues that the fatherhood identity which was once tied to breadwinning no longer prevails due to demographic, economic, and political changes. The result, Griswold notes, is that fatherhood has become ambiguous. Daly (1994) suggests that the ambiguity surrounding fatherhood may be associated with an interesting contradiction between conflicting cultural ideologies of fatherhood and social structure. This contradiction may be particularly poignant for divorced fathers. At the same time fathers are being called upon to become more active in raising their children, high divorce rates and custodial arrangements ensure that many children are not residing with their fathers.

Research on post-divorce fathering suggests that divorce has a profound effect on father-child relationships (see review by Hetherington & Stanley-Hagan, 1997). However, the residence of the child(ren) plays a crucial role in this association. Several cross-sectional analyses have demonstrated the negative impact of divorce on non-resident father-child relationships (Furstenberg, Nord, Peterson, and Zill, 1983; Seltzer, 1991; Stephens, 1996). Using the 1988 child supplement to the National Longitudinal Survey of Youth (NLSY), King (1994) found that only 27.2% of children over age 4 saw their non-resident fathers at least once a week in the last year, while 30.7 had no contact at all during that same time period. Relationships between divorced fathers and their children continue to be characterized by low levels of contact, affection, and intergenerational exchanges well after the children become adults (Booth & Amato, 1994; Cooney & Uhlenberg, 1990; Rossi & Rossi, 1990).

Non-resident fathers’ involvement with their children falls steadily over time, even for fathers who were highly involved initially (King, 1994; Seltzer, 1991). One important concern centers around the fact that, as fathers decrease contact, they also tend to decrease child support. While the estimates of fathers who paid child support regularly and in full ranges from one-half (Peterson & Nord, 1990) to nearly three-fourths (Peters, Argys, Maccoby, and Mnookin, 1993), approximately one-quarter of non-resident fathers pay nothing at all. It has been estimated that $34 billion more in child support could be paid by non-resident fathers (Sorenson, 1997). Certainly child support noncompliance accounts for some of the negative effects of divorce on the well-being of single mothers and their residential children. It may also be indicative of the post-divorce ex-spousal relationship, as some fathers resent the fact that the mother has control of child support moneys and rationalize noncompliance by citing the mother’s failure to use the money for the children.

Studies of divorced fathers who co-reside with their children provide strikingly different results and suggest that divorced fathers who live with their children report close relationships with them (Grief, 1990). Aquilino (1994a) found that adult children from custodial father families perceive higher relationship quality with their fathers than those children from mother-custody families. Further, few differences exist in a variety of outcomes such as disclosure and affection when single custodial fathers and single custodial mothers are compared (Risman, 1987; Risman & Park, 1988).

A vast literature documents that divorced persons have lower levels of psychological well-being than their married counterparts (Mirowsky & Ross, 1989). This effect exists for many constructs of mental health such as psychological distress (Booth & Amato, 1991) and depression (Shapiro, 1996) as well as physical health (Gove, 1972) and suicide (Stack, 1990). Despite this clear evidence, the extent to which parenting stressors, specifically those associated with the parent-child relationship, contribute to the psychological well-being of divorced fathers is unclear.

Relationships with minor children have been shown to have both positive and negative effects on the psychological well-being of divorced persons (Hughes, 1989). These effects are moderated by whether minor children are present in the household. Divorced fathers with absent children often feel guilt, frustration, anxiety, a sense of longing, sadness, and emptiness due to loss of an ongoing relationship with their children (Gerson, 1993; Hughes, 1989; Kruk, 1992). Many non-resident fathers often express that they feel a loss of control over their children’s lives (Feldman, 1990) and in the divorce process (Arendell, 1995; Umberson & Williams, 1993). This loss of control may be manifestated in the withdrawal of fathers from their children’s lives, a negative assessment of relationships with children, psychological distress (Mirowsky & Ross, 1989), and/or a failure to support children economically. On the other hand, fathers with absent children may be relieved of daily caregiving tasks as well as have greater opportunities to develop new relationships (Hughes, 1989).

Being an unmarried father with co-resident minor children has been shown to be particularly burdensome, particularly if there is little assistance provided by other adults (Hughes, 1989). Additionally, relations with co-resident children may limit fathers’ ability to establish long-term conjugal relationships as children compete for fathers’ affection and commitment. On the other hand, living with minor children adds a sense of meaning to a fathers’ life and serves as a mechanism of social integration and support for divorced fathers (Umberson & Gove, 1989).

Regardless of the number of parenting stressors faced by fathers, recent research indicates that men may be less psychologically reactive to the quality of relationships with children than are women (Umberson, Chen, House, Hopkins, and Slaten, 1996). This perspective suggests that one’s reactivity to parent-child relationship stressors corresponds to the socially structured roles that men and women play in the family, particularly a married-couple household. For example, men may be more reactive to financial or occupational stressors as men’s family roles typically involve financial provision for the family. Similarly, since women’s family roles typically involve responsibility for emotional and physical well-being of family members (Kessler & McLeod, 1984), the parental role is salient in different ways for the psychological well-being of women than it is for men (Scott & Alwin, 1989). In contrast, Simon (1992) found that, while women are exposed to more parental strains, men are equally likely to experience psychological distress as a result of these strains. However, little is known about men’s reactivity to parent-child relationships when men assume primary responsibility for the care of their children. Thus, the degree to which parent-child stressors affect men’s psychological well-being outside the context of marriage and the extent to which changes in these relationships over time influence well-being are not well understood.

Our purpose in this paper is to understand how the process of divorce impacts fathers’ relationships with their children and their own psychological well-being employing data from a nationally representative panel study. Combining the extant literature and microstructural theory, we have developed the following hypotheses:

This study is based on panel data from two waves of the National Survey of Families and Households (NSFH), a national probability sample (Sweet, Bumpass, and Call, 1988). The first wave was collected during 1987 and 1988 (T1) and the second wave was collected between 1992 and 1994 (T2). The first wave contains data from 13,017 respondents who represent the non-institutionalized United States population aged 19 and over. The NSFH oversamples for single parents and racial/ethnic minorities. To reduce interviewer bias, an effort was made to assign Black interviewers to predominantly Black areas and bi-lingual interviewers to predominantly Spanish-speaking areas. Data were collected through face-to-face interviews. At various segments of the interviews, respondents were asked to complete a self-administered questionnaire in order to facilitate the collection of sensitive information. The second wave of the NSFH includes computer assisted re-interviews with 10,008 of its original respondents. The re-interview rates were slightly lower among Blacks (77.4%) and other racial/ethnic minorities (i.e. Puerto Ricans, 68%) than Whites (84.5%). Re-interviews were conducted in similar fashion to the first wave.

The subsample for this study contains 844 fathers who have at least one biological child under age 19 at T2. All respondents had at least one biological child under age 19 on the full-time (more than 50% of the time) household roster at T1. Among this group, one focal child meeting these characteristics was selected at random for each respondent and all questions regarding parent-child relationship quality were asked with regard to this child.

All respondents were married with at least one minor child at home at T1. At T2, some had remained married with their focal child still living at home (N=729), some had divorced or separated and had the focal child living with them (N=33), and some had divorced or separated and the focal child resided elsewhere (N=82). Determination of residential status is based on respondents’ full-time household roster. A child is considered by the NSFH to live with the respondent if he/she resides in the household for more than 50% of the time. Thus, it is plausible that some fathers who report having a co-resident child share physical custody of their child with the former spouse. In fact, of divorced co-resident fathers in this sample who had a legal custody agreement, 27% had joint custody. Similarly, among divorced fathers who claimed that they did not reside with their focal child and who had a legal custody agreement, 10% had joint custody. These findings suggest that legal custody is a poor indication of the proportion of the time children spend in their parents’ homes. Thus, our interest in this analysis is the physical presence or absence of the focal child in the father’s household regardless of the legal arrangements regarding custody.

We eliminated all fathers who divorced and remarried between waves due to the confounding issues of marital statuses and due to the extremely few cases of men who met these criteria. We also eliminated continuously married fathers whose focal child was not living with them at T2. This was done for several reasons. First, this configuration was a relatively rare event in our data, where roughly 3% of all continuously married fathers (N=25) had a nonresident focal child at T2. Second, we believe reasons for home-leaving to be qualitatively different for children of married fathers than for children of divorced fathers, which makes attempts to draw comparisons between divorced and married fathers more difficult. Ancillary analyses suggest that of the 25 continuously married fathers whose focal child left the home between waves, roughly 60% were 17 or 18 years of age. This configuration of the subsample allows us to analyze father-child relationships in a specific dyad including several sociodemographic characteristics of both father and child.

Relationship Quality. A single indicator was used to assess the global relationship quality of fathers with children who resided with them at T1. The wording of the two questions and the number of response categories were slightly different between the two waves. At T1, fathers were asked the following question with reference to the focal child: "How would you describe your relationship with each of these children?" The response categories were Likert-type and ranged from 1 "very poor" to 7 "Excellent." At T2, fathers were asked the following question with reference to the focal child: "Taking things all together, on a scale from 0 to 10, where 0 is a really bad and 10 is absolutely perfect, how would you describe your relationship with (focal child)?" The two measures are modestly correlated (r=.30). Although there is some discordance in the two measures of relationship quality we decided to include these variables in the analysis because the wording of the measures is not drastically different and has face validity. In order to account for the variance in response categories of the relationship quality measure between waves, individual observations at both waves were transformed to Z-scores which permit standard comparison across unequal distributions. We acknowledge the limitations of using a single-item to assess a potential spectrum of components within father-child relationships. We advise caution in interpreting the results.

Psychological Well-Being. Three indicators of psychological well-being are used in this analysis. Depression is measured by a 12-item version of the CES-D (Center for Epidemiological Studies) scale. Respondents were asked how often they experienced the following feelings during the past week: "Felt bothered by things that usually don’t bother me," "Not feel like eating," "Feel that you could not shake off the blues," "Have trouble keeping your mind on what you were doing," "Feel depressed," "Feel that everything you did was an effort," "Feel fearful," "Sleep restlessly," "Talk less than usual," "Feel lonely," "Feel sad," "Feel you could not get going." The index is summed and each observation is transformed to standard scores (a _T1=.92; a _T2=.93). Happiness is a single-item global happiness measure and is gauged on a scale from 1 (unhappy) to 5 (happy). While we acknowledge the limitations of a single-item measure of happiness, we feel confident in making conclusions about overall psychological well-being based on the use of multiple items in this analysis. Self-Esteem is measured using an index of three Rosenberg (1965) self-esteem items (a _T1=.64; a _T2=.60). Respondents were asked the extent to which they agree with the following statements: "On the whole, I am satisfied with myself," "I am able to do things as well as other people," and "I feel that I’m a person of worth, at least on an equal plane with others." The index was created by taking the average of the three responses. Normal probability plots indicate that the error distributions of all three variables are appropriate for Ordinary Least Squares regression.

Marital/child residence status. In order to ascertain the effects of fathers’ marital status changes by the residence of the focal child, we use three separate categories of fathers’ marital and child residence status. The first category consists of fathers who were continuously married to the same person between T1 and T2 and whose focal child resided with them at T2. The second category includes those fathers who were married at T1 but divorced or separated due to marital problems at T2 and whose focal child resided with them at T2. The third category includes fathers who were married at T1 but divorced or separated due to marital problems at T2 and whose focal child resided elsewhere at T2. Separated and divorced persons were combined into one category for two reasons. First, there were too few cases of separated and divorced individuals (N=38) to perform separate analyses for both groups. Second, we are primarily interested in the separation of fathers from their wives and children not in legal divorce per se.

Fathers’ Sociodemographic Characteristics. Several sociodemographic characteristics of fathers were examined at T2. Previous research suggests that stage of the life course (Rossi & Rossi, 1990; Umberson & Gove, 1989), race or ethnicity (Taylor, Chatters, Ticker, and Lewis, 1990) and socioeconomic status (Aquilino, 1994a) influence the nature of and reaction to parent-child relationships. Age (in years), education (in highest years completed), and personal income (logged in regression analyses) are measured as continuous variables. Race was measured as a four category variable including Whites, Blacks, Hispanics, and other races. Whites are the omitted category in regression models. Additionally, because fathers’ marital and childbearing histories may impinge on parenting, we include number of respondents’ marriages and total number of biological or adopted children as controls.

Characteristics of the Focal Child: Previous research also suggests that characteristics of children such as sex (Lamb, 1986; Rossi & Rossi, 1990) and age (Pleck, 1983) are important in the parent-child relationship. We control for these characteristics (each measured at T2). Age is measured in years and sex is a dichotomous variable (1=Male, 0=Female)

Time since separation. Previous studies find a negative association between time since separation and father-child contact (Furstenberg et al., 1983; Seltzer, 1991). In order to test if father-child relationship quality changes as time since separation increases, we include a measure of time since separation (in months). This variable refers to the date of physical separation of the respondent and his spouse, not the final date of the divorce or legal separation. We chose this operationalization because our primary interest is in the separation of fathers from their wives and children, which usually precedes legal divorce.

Relationship with Former Spouse. Fathers’ relationships with children have been shown to be moderated by relationships with the former spouse (Kurdek, 1986). A single item was used to assess divorced fathers’ relationships with their former spouse at T2. The question is worded "How would you describe your current relationship with your former wife?" Response categories range from 1 (very unfriendly) to 5 (very friendly). Again, we advise caution in interpreting result from a single-item measure.

RESULTS

To test our first and second hypotheses, we present means and standard deviations of the variables used in the analysis by marital and child residence status in Table 1. The data clearly support both hypotheses and show marital status differences in relationship quality at T2 which are moderated by child residence. Divorced non-resident fathers report significantly worse relationship quality than co-resident fathers regardless of marital status. Contrary to expectations, divorced co-resident fathers report significantly better relationship quality than continuously married co-resident fathers. These data suggest that co-residence with the focal child moderates the relationship between divorce and relationship quality in that co-residence may alleviate the hypothesized negative effects of divorce on relationship quality. Table 1 also suggests that the lack of any significant marital status difference in relationship quality at baseline provides little evidence that levels of relationship quality were different at T1 for fathers who eventually would divorce.

--INSERT TABLE 1 HERE--

In order to test whether marital status differences in relationship quality hold following the addition of controls, two OLS regression models are presented in Table 2 which assess the net effect of divorce on relationship quality while controlling for relationship quality at T1 and a series of sociodemographic variables. Model 1 of Table 2 shows that, compared to the continuously married, being a divorced non-resident father is negatively associated with relationship quality. This finding is consistent with other studies of non-resident divorced fathers which find negative associations with father-child relationship outcomes (Seltzer, 1991; Stephens, 1996). Results in this model also indicate that there are no significant differences in relationship quality between divorced and continuously married co-resident fathers. This supports microstructural arguments in that fathers in similar structural positions, namely co-residence with the focal child, will have similar parental experiences. Education, child age, and being Black emerge as significant predictors of relationship quality. Fathers’ education negatively predicts relationship quality, and black fathers are more likely than white fathers to perceive positive parent-child relationship quality. The age of the focal child is negatively associated with relationship quality. Separate analyses by race (not shown) were also conducted to reduce the threat of higher attrition rates among Blacks. These analyses suggest that there are no racial differences in the effect of divorce or child residence on fathers’ perceived relationship quality.

Previous research has suggested that certain divorce-specific variables, such as time since separation (e.g., Seltzer, 1991) and relationships with the former spouse (e.g., Umberson & Williams, 1993), may be important factors in relationships between divorced fathers and their children. To examine this, we estimate a separate model for divorced fathers by adding time since separation (in months) and fathers’ relationship with the ex-spouse to Model 2 (excluding marital status). Similar to results from the pooled sample, being a non-resident divorced father is negatively associated with relationship quality. Contrary to the findings of other studies, there is no appreciable effect of time since separation or relationship with ex-spouse on father-child relationship quality. Ancillary analyses show no interactions between father-child co-residence and time separated or relationship with the ex-spouse. Taken together, these results suggest that the loss of co-residence following divorce has adverse consequences for father-child relationship quality. Moreover, co-residence may serve as a buffer of divorce as it is highly significant to positive relations between fathers and their children.

--INSERT TABLE 2 HERE—

We begin to test our third hypothesis by analyzing results from Table 1 which provides evidence for the influence of divorce and child residence on fathers' psychological well-being. First, consistent with prior research, divorce is associated with greater depression among fathers. Regardless of child residence, divorced fathers have higher mean depression scores at T2 than continuously married co-resident fathers. Second, these data show that a selection effect may partly explain marital status differences in mean levels of depression at T2. Fathers who eventually divorced had higher depression levels at T1 than did fathers who remained married. Thus, depression may not be a causal result of divorce or separation itself, rather that fathers who are more depressed are "selected out" of the pool of married persons by divorce. While other longitudinal studies (e.g., Booth & Amato, 1991; Menaghan, 1985) do not contest the partial selection effects, these studies have provided more support for social causation explanations -- that depression results from the experience of marital disolution.

Mean differences in happiness are also examined in Table 1. These results show that divorced co-resident fathers have significantly lower levels of happiness at T2 than any other category. Although divorced non-resident fathers have lower levels of happiness than the continuously married at T1, no other selection effects are apparent. There are no significant differences by marital and child residence status on self-esteem.

The next stage of the analysis tests the effect of divorce on three indicators of fathers’ psychological well-being net of T1 well-being and sociodemographic controls. Models are estimated separately for the pooled sample and for the divorced sample, and are presented in Table 3. Results suggest that the effect of divorce on psychological well-being varies by residency status of the focal child as well as the specific outcome that is measured. When examining fathers’ depression in the pooled sample, we find that divorced non-resident fathers experience greater depression than continuously married co-resident fathers, leading us to partially reject our third hypothesis.

In the divorced sample, there are no significant differences in depression between co-resident and non-resident fathers. A negative association between months separated and depression is also evident. Taken together, these findings suggest that divorced fathers have higher levels of depression than their continuously married counterparts. Although divorced co-resident fathers are not statistically more likely to experience depression than their married counterparts, the differences in depression between divorced fathers in different residential arrangements are small.

The models which analyze fathers’ happiness similarly show that the effect of divorce on fathers is contingent upon the specific outcome and the residential status of the child. As seen in the pooled sample, divorced fathers who reside with their focal child are less happy than fathers who remain married. When analyzing divorced fathers separately, however, the coefficient for residential status is insignificant. It is quite plausible that the insignificance of this coefficient is due to the relatively small sample size. Despite the insignificance of the coefficient, its large size and negative effect (b= -.508) support the conclusion that the process of divorce has negative consequences for happiness only for co-resident fathers. Furthermore, controlling for fathers’ and children’s characteristics cannot explain differences in fathers’ happiness. This finding suggests that parenting for divorced co-resident fathers in this sample is likely to be more stressful due to factors other than those measured in this analysis.

It should be noted that item nonresponse for happiness was roughly 20% of the sample. To inspect for systematic nonresponse bias we compared the respondents who had and had not completed this question on all of the variables examined in this study. Three significant differences were observed between the two groups. Fathers who had missing data were more likely to be Black, have lower levels of education, and more depressed than those who did respond to that item.

--INSERT TABLE 3 HERE--

Given our finding that some divorced fathers experience a decline in relationship quality following divorce, we test our fourth hypothesis and examine whether these changes influence fathers’ psychological well-being. Contrary to our hypothesis, it is clearly evident from Table 3 that there is no appreciable effect of a change in relationship quality on fathers’ psychological well-being. This supports findings of Umberson et al. (1996) who find men to be less psychologically reactive to strained relationships with children. In sum, the results suggest that post-divorce fathering is characterized by both support and strain. The conditions under which individual fathers’ perceive support or strain is dependent upon the specific outcome variable under study and the social structural position of the father. Additionally, despite divorced fathers’ reports of lessened post-divorce relationship quality, factors other than changes in relationship quality explain the susceptibility of some divorced fathers to psychological distress.

DISCUSSION

This study was designed to analyze changes in father-child relationship quality and psychological well-being among fathers who divorced between waves of data collection compared to fathers who remained stably married. It was also designed to assess the relative impact of child residence which may account for marital status differences in father-child relationship quality and psychological well-being. This study represents a significant improvement over previous studies in that it assesses change over time, addresses the issue of fathers’ psychological well-being which has been almost absent in the research literature, and includes divorced fathers with both co-resident and non-resident children.

We find that fathers who divorced between interviews perceive substantially lower levels of relationship quality with their young children than do fathers who are continuously married. This finding is consistent with and extends upon cross-sectional research which documents divorced fathers’ lower levels of contact and affect with children than married fathers (e.g., King, 1994; Seltzer, 1991). We do not suggest, however, that divorce, in and of itself, is wholly responsible for negative changes in father-child relationship quality. We find that the connection between divorce and relationship quality is moderated by the residential status of the focal child. Divorced fathers who co-reside with their minor focal child perceive similar levels of relationship quality than continuously married fathers who co-reside with their focal child. Thus, it is divorced fathers who discontinue residence with their focal child who perceive the largest decrease in relationship quality over time.

There are several reasons why co-residence with children is important to fathers’ evaluations of parent-child relationships. First, the nature of single fatherhood may be a contributing factor, in that most fathers have a choice to fight for custody of their child(ren) (Greif, 1990). Single fathers with custody may value their relationship with their child(ren) more due to their struggle for custody. Additionally, fathers may be likely to seek custody of children with whom they have a good relationship (Grief, 1990). Second, the presence of children in the home permits the father a greater sense of control over the lives of his children Studies of non-resident fathers frequently cite fathers’ perceived loss of control over their children’s lives (Umberson & Williams, 1993). Co-residence affords fathers time to spend with their children, allows them to make economic provisions, and to make decisions which may be crucial to the lives of their children (Seltzer, 1991). Third, co-residence may also represent the freedom of the father from having to negotiate visitation with their former spouses. Divorced fathers who report strained relationships with children often charge that they feel their former spouses are the "gatekeepers" of the father-child relationship and that their former spouses allow them to play a small and insignificant role in the daily lives of their children (Arendell, 1995; Kurdek, 1986; Umberson & Williams, 1993). Finally, the dimension of physical proximity which is inherent in co-residential arrangements may encourage fathers to become more involved in their children’s lives. Studies of fathers with nonresident children suggest that physical distance between fathers and children is a notable predictor of paying child support, visitation, and interest in decision making (Seltzer, 1991).

In sum, this study supports microstructural theory in that the negative effects of life course events can be limited by the continuity of social structural arrangements. Since family and father identities are often tied to the roles of marriage (Arendell, 1995), the continuity of co-residence among divorced fathers may reflect a connection with normative expectations associated with fatherhood in the context of marriage.

Years of study have shown that divorce is consistently and negatively associated with psychological well-being. We find that this association is dependent upon both the specific psychological well-being outcome under study and the residential status of the focal child. Divorced fathers are more depressed than continuously married fathers. The difference between divorced fathers with non-resident children and co-resident children is modest at best, suggesting similarly positive associations between divorce and depression.

Our study also suggests that divorced fathers experience psychological costs associated with having co-resident children. Fathers who reside with their focal child are less happy than both divorced non-resident fathers and continuously married fathers. There are two possible explanations for this finding. First, children in the home impose demands of time and energy and limit social and personal choices especially among the divorced (Hughes, 1989; Umberson & Gove, 1989). Second, single parenthood may be more emotionally taxing for males than for females. Men have smaller confidant networks than females (Kendig, Coles, Pittelkow, and Wilson, 1988) and the parental role may not be well integrated into the male role as a result of differential gender socialization (Scott & Alwin, 1989). Further, characteristics of men who become co-resident fathers may predispose them to stress. Our national data suggest that co-resident divorced fathers are more socioeconomically disadvantaged than are continuously married or non-resident fathers. Research which examines the mechanisms by which fathers maintain residence with their children following divorce is needed.

Do declines in father-child relationship quality affect fathers’ psychological well-being? Our findings indicate that while divorce has a negative effect on fathers’ relationships with children, changes in the evaluation of these relationships are not associated with fathers’ psychological well-being. This finding supports the school of thought which suggests that men are less psychologically sensitive to strained relationships with children than are women. Men’s lessened sensitivity to relationship strain may be the result of the differential salience of the parental role to the psychological well-being of women regardless of marital status (Scott & Alwin, 1989; Simon, 1992). It is plausible that these findings may be very different if other measures of the father-child relationship were used.

Two cautionary notes regarding our study must be delivered. First, we acknowledge the multi-dimensionality of both the parent-child and ex-spousal relationship, and the potential problems that a single-item measure of these relationships may produce. With respect to parent-child relationship quality, it is quite plausible that men may be differentially psychologically reactive to various aspects of parent-child relationships (i.e. child strain vs. child social support) (Umberson et al., 1996). A systematic bias among groups of fathers, by race and social class, for example, may also exist. Unfortunately, the NSFH panel data contains no other assessments of parent-child relationships for all minor children regardless of the focal child’s residence status. Nevertheless, this research as a whole explores new areas of fathers’ and children’s lives that adds to the existing research. With respect to fathers’ relationship with the ex-spouse, studies of fathers and divorce (e.g., Arendell, 1995; Umberson & Williams, 1993) invariably show the complex and multifaceted nature of ex-spousal relations. In no way can a single item capture this complexity. Future research should explore this relationship in order to have a better understanding of the mothers’ contribution to father-child relations after divorce.

Second, recent research suggests that conventional indicators of psychological well-being (i.e. depression, happiness, self-esteem) cannot adequately gauge men’s well-being. Specifically, researchers argue that gendered expressions of well-being, such as alcohol and drug use, are more sensitive to men’s symptoms of distress (Horwitz, White, and Raskin-White, 1996; Umberson et al., 1996). Although these same studies suggest that conventional indicators do gauge well-being to some extent, measures such as alcohol and drug use should be used in addition to these traditional indicators in future research.

Despite its limitations, our study has significantly advanced our knowledge of the process of divorce for men The use of panel data is highly instructive and supports much of the extant cross-sectional research. However, our study suggests that many negative effects of divorce on fathers can be explained by examining the socially structured opportunities and expectations of divorced fathers. While the results of this study represent an important step toward understanding the longitudinal process of divorce for men, research which specifically analyzes change in the meaning and salience of, and commitment to fathers’ parenting identities is sorely needed. Specifically, we need to better understand the interplay between structural and psychosocial forces in order to determine both fathers’ motivations for and barriers to meaningful relationships with children as well as their own psychological well-being.

Amato, P., & Booth, A. (1991). The consequences of parental divorce and marital unhappiness for adult well-being. Social Forces, 69, 895-914.

Amato, P., & Keith, B. (1991). Parental divorce and the well-being of children: A meta-analysis. Psychological Bulletin, 110, 26-46.

Aquilino, W. (1994a). Impact of childhood family disruption on young adults relationships with parents. Journal of Marriage and the Family, 56, 295-313.

Aquilino, W. (1994b). Later life parental divorce and widowhood: Impact on young adults’ assessment of parent-child relations. Journal of Marriage and the Family, 56, 908-922.

Arendell, T. (1995). Fathers and divorce. Newbury Park, CA: Sage.

Booth, A. & Amato, P. (1991). Divorce and psychological stress. Journal of Health and Social Behavior, 32, 396-407.

Booth, A., & Amato, P. (1994). Parental marital quality, divorce, and relations with offspring in young adulthood. Journal of Marriage and the Family, 56, 21-34.

Cooney, T. M., & Uhlenberg, P. (1990). The role of divorce in men’s relations with their adult children after mid-life. Journal of Marriage and the Family, 52, 677-688.

Daly, K. (1994). Uncertain terms: The social construction of fatherhood. In M. L. Dietz, R. Prus, & W. Shaffir (Eds.), Doing everyday life: Ethnography as human lived experience (pp. 170-185). Toronto, Canada: Copp Clark Pittman.

Furstenberg, F. F., Jr., Nord, C. W., Peterson, J. L., & N. Zill. (1983). The life course of children of divorce: Marital disruption and parental contact. American Sociological Review, 48, 656-668.

Gerson, K. (1993). No man’s land. New York: Basic Books.

Gove, W. (1972). The relationship between sex roles, marital roles, and mental illness. Social Forces, 51, 34-44.

Grief, G. L. (1990). The daddy track and the single father. Lexington, MA: Lexington Books.

Griswald, R. (1993). Fatherhood in America: A history. New York: Basic Books.

Hetherington, E. M., Stanley-Hagan, M. M. (1997). The effects of divorce on fathers and their children. In M. Lamb (Ed.), The role of the father in child development (pp. 191-211). New York: John Wiley & Sons.

Horwitz, A., White, H. R., & Howell-White, S. (1996). The use of multiple outcomes in stress research: A case study of gender differences in responses to marital dissolution. Journal of Health and Social Behavior, 37, 278-291.

Hughes, M. (1989). Parenthood and psychological well-being among the formerly married: Are children the primary source of psychological distress? Journal of Family Issues, 10, 463-481.

Kendig, H. L., Coles, R., Pittelkow, Y., & Wilson, S. (1988). Confidants and family structure in old age. Journal of Gerontology: Social Sciences, 43, S31-S40.

Kessler, R. C., & McLeod, J. C. (1984). Sex differences in vulnerability to undersirable life events. American Sociological Review, 49, 620-631.

King, V. (1994). Nonresident father involvement and child well being. Journal of Family Issues, 15, 78-96.

Kruk, E. (1992). Psychological and structural factors contributing to the disengagement of noncustodial fathers after divorce. Family and Conciliation Courts Review, 30, 81-101.

Kurdek, L. (1986). Custodial mothers perceptions of visitation and payment of child support by noncustodial fathers and families with low and high levels of preseparation interparental conflict. Journal of Applied Developmental Psychology, 7, 307-322.

Lamb, M. E. (1986). The changing roles of fathers. In M.E. Lamb (Ed.), The father’s role: Applied perspectives (pp. 3-27). New York: John Wiley & Sons.

Meyer, D. R., & Garasky, S. (1993). Custodial fathers: Myths, realities, and child support policy. Journal of Marriage and the Family, 55, 73-89.

Mirowsky, J., & Ross, C. E. (1989). The social causes of psychological distress. New York: Aldyne de Gruyter.

Morgan, L. A. (1991). After marriage ends: Economic consequences for mid-life women.

Newbury Park, CA: Sage.

Norton, A. J., & Miller, L. F. (1992). Marriage, divorce and remarriage in the 1990s. Current Population Reports (Bureau of the Census Publication No. 23-180). Washington, DC: U.S. Government Printing Office.

Peters, E., Argys, L., Maccoby, E., & Mnookin, R. (1993). Enforcing divorce settlements: Evidence from child support compliance and award modification. Demography, 30, 19-35.

Peterson, J., & Winquist Nord, C. (1990). The regular receipt of child support: A multistep process. Journal of Marriage and the Family, 52, 539-557.

Pleck, J. H. (1983). Husbands paid work and family roles: Current research issues. In H. Lopata & J. H. Pleck (Eds.), Research in the interweave of social roles (Vol. 3), Families and Jobs. Greenwich, CT: JAI Press.

Risman, B. (1987). Intimate relationships from a micro-structural perspective: Men who mother. Gender and Society, 1, 6-32.

Risman, B., & Park, K. (1988). Just the two of us: Parent-child relations in single parent

homes. Journal of Marriage and the Family, 50, 1049-1062.

Rosenberg, M. (1965). Society and the adolescent self-image. Princeton, NJ: Princeton

University Press.

Rossi, A., & Rossi, P. (1990). Of human bonding: Parent-child relations across the life course. New York: Aldyne de Gruyter.

Scott, J., & Alwin, D. F. (1989). Gender differences in parental strain: Parental role or gender role? Journal of Family Issues, 10, 482-503.

Seltzer, J. A. (1991). Relationships between fathers and children who live apart: The fathers role after separation. Journal of Marriage and the Family, 53, 79-101.

Shapiro, A. D. (1996). Explaining psychological distress in a sample of remarried and divorced persons: The influence of economic distress. Journal of Family Issues, 17, 186-203.

Simon, R. (1992). Parental role strains, salience of parental identity, and gender differences in psychological distress. Journal of Health and Social Behavior, 33, 25-35.

Stack, S. (1990). The impact of divorce on suicide: New micro level data. Journal of Marriage and the Family, 52, 119-127.

Stephens, L. S. (1996). Will Johnny see daddy this week?: An empirical test of three theoretical perspectives of postdivorce contact. Journal of Family Issues, 17, 466-494.

Sweet, J., Bumpass, L., & Call, V. (1988). The design and content of the National Survey of Families and Households. Working paper NSFH-1, Center for Demography and Ecology, University of Wisconsin-Madison.

Taylor, R., Chatters, R. Ticker, M., & Lewis, E. (1990). Developments in research on Black families: A decade review. Journal of Marriage and the Family, 52, 993-1014.

Umberson, D., Chen, M. D., House, J. S., Hopkins, K., & Slaten, E. (1996). The effect of social relationships on psychological well-being: Are men and women really so different? American Sociological Review, 61, 837-857.

Umberson, D., & Gove, W. (1989). Parenthood and psychological well-being: Theory,

measurement, and stage in the life course. Journal of Family Issues, 10, 440-462.

Umberson, D., & Williams, C. L. (1993). Divorced fathers: Parental role strain and

psychological distress. Journal of Family Issues, 14, 378-400.

White, L. (1992). The effect of parental divorce and remarriage on parental support for adult children. Journal of Family Issues, 13, 234-250.