COVID-19 and the relationships and involvement of nonresident fathers.

Objective: To support nonresident fathers in maintaining involvement and relationships with their children during the COVID-19 pandemic, we must understand how such fathers have been impacted thus far by the pandemic. Background: All families have been impacted by the pandemic, but fathers who do not reside with their children are particularly likely to be negatively impacted. Social distancing, restrictions on travel, job loss/economic downturn, family court closures, and numerous other effects of the COVID-19 crisis impact the ability of nonresident fathers to maintain relationships and involvement with their children. Method: The current study analyzed retrospective data from 373 nonresident U.S. fathers to assess perceived parenting and coparenting changes during the pandemic, as well as whether coparenting relationships and mental health were associated with their involvement and relationships with their children.
Results: On average, fathers' involvement, father-child relationship quality, and coparenting support declined, but wide variability also existed, with a substantial minority of fathers reporting increased involvement. Coparenting support was positively associated with current levels of involvement and relationship quality as well as changes to both since the pandemic, but mental health was inconsistently associated with outcomes. Conclusion and Implications: To keep these average declines from becoming permanent, future research should investigate what factors foster higher involvement and improved relationships. Practitioners should offer additional support to nonresident fathers to promote their involvement and relationships in ways compatible with evolving restrictions.

The COVID‐19 pandemic has created innumerable challenges for individuals and families. Lockdowns, quarantines, stay‐at‐home orders, and travel restrictions have had a significant impact on the ability of people and families to stay connected with one another and have continued long past initial estimates, and the economic fallout in a variety of industries likely will continue for years. This is particularly true for nonresident parents, and although the number of single fathers has increased, the vast majority of nonresident parents (about 80%; Grall, 2020; U.S. Census Bureau, 2010) are fathers. Even among “joint” parenting arrangements in the United States, time typically is not split exactly 50–50 between the parents; instead, although every state has its own language and regulations around shared parenting, the language in statutes typically includes phrases such as “significant, but not necessarily equal, periods of time” (Domestic Relations, Missouri State Statute 452.375, 2018), resulting in mothers still having more time with children than fathers. Therefore, in a time when traveling between households was restricted and family courts were closed (leaving families unable to seek legal recourse to their parenting disputes), nonresident fathers likely were particularly affected by the COVID‐19 crisis. Research is needed to investigate the ongoing impacts of the COVID‐19 pandemic on nonresident fathers across the United States, in states with varying restrictions and resources and with varying rates of infection by the disease. Given the importance of routines and patterns of interaction in families, it is necessary to understand what parenting patterns families have been implementing to best plan services for these families moving forward.

LITERATURE REVIEW AND THEORETICAL FRAMEWORK

Nonresident fathering

It is well established that nonresident fathers tend to have more strained relationships with their children and lower levels of involvement than resident fathers (Peters & Ehrenberg, 2008). Not surprisingly, more time spent with children is associated with greater knowledge and understanding of one another (Ashbourne et al., 2011) and better father–child relationships (Aquilino, 2006; Bauserman, 2012), and when children perceive that fathers are not invested in the relationship, father–child relationship quality is adversely affected (Nixon et al., 2012). In addition, a substantial body of research demonstrates the importance of father involvement (including nonresident father involvement) for a variety of child outcomes, whether academic, behavioral, social, physical, or psychological, and even across generations (Adamsons & Johnson, 2013; Cabrera et al., 2000; Jessee & Adamsons, 2018; Lamb & Lewis, 2010). As such, the particular impediments that the COVID‐19 pandemic presents to relationship maintenance across households point to a need for timely research regarding how these factors influence nonresident fathers' ability to maintain involvement and relationships with their children.

COVID‐19 pandemic and parenting

Limited literature is available on the impact of the pandemic, as this is an unprecedented sociohistorical event. However, existing research indicates variability in patterns of family adjustment during the pandemic. In some two‐parent households, gendered divisions of labor deepened during the pandemic, whereas in others, fathers who were working from home took on greater shares of housework and childrearing responsibilities (Shafer et al., 2020; Yerkes et al., 2020). Increases in parents' stress and precipitous declines in mental health also have been reported as a result of the pandemic and concurrent lockdowns, which adversely affect child outcomes as well as parent–child relationships (Russell et al., 2020).

One study of nonresident parents in Hungary found multiple patterns of change in visitation with children, with approximately one third of parents continuing their usual schedules of in‐person visits with their child, about one quarter describing “significant” changes in visitation (most reflecting a reduction), and 40% reporting a complete suspension of visitation during the pandemic (Szalma & Rekai, 2020). Parents who reported complete suspension of visitation frequently reported increasing their use of virtual communication (e.g., video chats) to make up for the lack of in‐person time (Szalma & Rekai, 2020). Finally, a study of Turkish parents of young children found that parent–child interactions and relationships changed for both mothers and fathers (in different ways) as a result of the pandemic (Uzun et al., 2021). Therefore, the limited extant research suggests numerous changes for parents and families as a result of the pandemic and lockdowns and a need for more research examining the impact on nonresident fathers and their children.

Family systems theory

Family systems theory (FST; Cox & Paley, 2003) is an ideal theoretical framework to use when studying nonresident parenting during a global pandemic. Key to FST is the idea that families act as systems and, therefore, are interdependent, with members mutually influencing one another and generally attempting to maintain homeostasis, or stability, within the family system (White et al., 2019). Families differ in terms of their level of boundary permeability, meaning how easily new members are allowed in or old members are removed. Even if family members live in different households due to divorce, coparents ideally still remain part of an overall family system for their children, with the focus shifting to what Ahrons termed the binuclear family (see, e.g., Ahrons, 1980, 1981; Ahrons & Rodgers, 1987). However, after divorce, many nonresident fathers struggle with feeling like “secondary” or optional parents, and prior legal terminology that granted them “visitation” with their children perpetuated this. With the recent increased emphasis on shared parenting rather than “primary” or “secondary” custody and parenting time rather than visitation, the boundaries of divorced family systems generally have increased in permeability, encouraging fathers to remain a part of their children's family system. COVID‐19 restrictions and accompanying concerns for children's safety, however, might have encouraged some families to “close” their boundaries, restricting or eliminating the movement of children between physical households despite the importance of psychological connections (Szalma & Rekai, 2020). FST will help scholars, practitioners, and policymakers understand the varying ways that families have addressed the crisis of COVID‐19 and how best to promote the involvement and relationships of nonresident fathers. Therefore, I turn our attention next to key influences on fathering.

Gatekeeping, coparenting, and fathering

Research has long supported the influence of mothers on fathers' involvement and relationships with children, commonly termed maternal gatekeeping—namely, mothers can serve as either gate‐openers (promoting fathers' engagement and relationships) or gate‐closers (discouraging or preventing fathers' engagement and relationships). When mothers support father involvement, father involvement tends to be higher, and vice versa; this is particularly true when fathers are nonresident and dependent on mothers' cooperation to see their children (Adamsons, 2010; Allen & Hawkins, 1999; DeLuccie, 1995, 1996; Fagan & Barnett, 2003; Hamer, 1998; Seery & Crowley, 2001). It is anticipated that safety concerns about travel and exposure during the pandemic, coupled with court closures that render nonresident fathers unable to seek legal recourse if mothers heighten their restrictions on fathers' parenting time, likely will enhance mothers' roles as gatekeepers of the father–child relationship. Although gatekeeping attitudes are not explicitly measured in the present study, some items that address gatekeeping behaviors are included in the measures of coparenting support and difficulties.

Relatedly, associations between the quality of the coparenting relationship and father involvement also are relatively well established, especially for nonresident fathers (Carlson et al., 2008; Fagan & Palkovitz, 2011, 2019). One study, however, found no relationship between fathers' caregiving and coparenting quality for low‐income nonresidential fathers (Lee et al., 2020). Most of the research related to nonresident fathers has examined coparenting as a unidimensional construct, being either “high/low quality,” “high/low in support,” or “high/low conflict.” Few studies have examined coparenting support and coparenting difficulties or conflict as separate, independent variables, despite the fact that research supports such a distinction and that at least a small proportion of families have high levels of both support/cooperation and difficulties/conflict (Maccoby et al., 1990; Waller, 2012). In addition, little research has examined predictors of father–child relationship quality, focusing primarily on involvement or specific parenting behaviors.

Given these gaps in the research, here I assess both coparenting support and coparenting difficulties and their associations with father involvement and father–child relationship quality. In addition, to assess the importance of particular combinations of coparenting support and difficulties, I investigated the associations of involvement and relationship quality with a variable capturing the interaction between coparenting support and coparenting difficulties.

Father mental health and father involvement

Although several studies have demonstrated associations between mental health and parenting for mothers (e.g., Dietz et al., 2009; Letourneau et al., 2010; Schellinger et al., 2012), there is a relative dearth of work that examines similar associations for fathers. Studies have noted that depressed fathers tend to be less strongly bonded to their children (Dudley et al., 2001), have lower ratings of father–child relationship quality, and exhibit less warmth and expressive parenting and more harsh parenting (Shafer & Renick, 2020; Trahan & Shafer, 2019). Trahan and Shafer (2019) also found that self‐efficacy moderated the association between fathers' depression and paternal warmth (but not harsh parenting), indicating that self‐efficacy might serve as at least a partial buffer for children with depressed fathers, but its impacts on relationship quality or involvement remain unstudied. Even fewer studies have examined the role of fathers' anxiety. One study (Johnson et al., 2004) found that high‐anxiety fathers exhibited negative parenting behaviors such as low affection and poor communication with the child, but this was conducted with resident fathers.

In addition to its potential direct influences on involvement with children, fathers' mental health also can influence the coparenting relationship. Turney and Hardie (2021) analyzed prepandemic data from mothers and fathers (both resident and nonresident) and found that paternal depression was associated with lower levels of coparental cooperation, across both mothers' and fathers' reports of coparenting and across both resident and nonresident parents; in fact, the coefficient for fathers' depression was nearly twice as large as that for mothers' depression. In addition, these associations were mediated by relationship quality and status; depressed fathers were likely to have lower quality overall relationships with and more likely to be separated from their children's mothers, which was then associated with a less cooperative coparenting relationship.

Mental health is likely to be particularly salient to parents during the COVID‐19 pandemic and associated lockdowns. At the time of data collection in May 2020, there was an initial relaxing of restrictions when peaks in infection began to decline (although restrictions were implemented again in the summer to late fall of 2020 in many states, when infection rates rose). Recent research supports that parents experienced an increase in parental stress, depression, and anxiety, particularly during the height of the initial pandemic restrictions in spring 2020 (Johnson et al., 2021; Kerr et al., 2021; Martins et al., 2021). One study directly compared mothers and fathers in the fall of 2020, well after many initial restrictions had lifted, and found that, despite a popular gendered narrative that mothers have been more strongly affected by the pandemic than fathers, fathers' mental health had been similarly adversely impacted compared with mothers, “suggesting that the ways parenting during the pandemic is affecting women and men are largely similar” (Elder & Greene, 2021, p. 8). Although nonresident fathers have not been specifically assessed, it is likely that their precarious position vis‐à‐vis their children and their lack of ability to contest unsatisfactory parenting time through the courts (due to lockdowns and restricted court schedules) contributed to even further mental health challenges. Given the limited and mixed findings available in the literature, here I examined the associations between negative mental health (symptoms of anxiety and depression) and self‐efficacy, and father involvement and father–child relationship quality. Investigating these factors is especially important during the pandemic, a time of particularly high levels of stress for many fathers and families that likely has given rise to higher levels of depression and anxiety, as well as challenges to self‐efficacy in the form of feeling “out of control” in the current situation.

CURRENT STUDY

In view of the foregoing, the present study collected descriptive data to address the following research questions:In addition, regression analyses were conducted to investigate predictors of fathers' involvement and relationships with children and changes in these variables since the pandemic. Specifically, associations with demographic variables, fathers' mental health, and the coparenting relationship were tested to assess the following hypotheses:

How have fathers' perceptions of their involvement, time spent, and relationships with their children changed during the pandemic?

How have fathers' perceptions of their coparenting relationships with their children's other parent changed during the pandemic?

Fathers with more perceived coparenting support and fewer perceived coparenting difficulties will be more involved and report more positive relationships with their children and will be more likely to have increased their involvement and improved their relationships with their children during the pandemic.

Regarding the interaction between coparenting support and difficulties, it is hypothesized that, relative to fathers with low perceived levels of both coparenting support and difficulties, fathers with higher support (but low difficulties) will be more involved and have better relationships with their children and fathers with higher difficulties (and low support) will be less involved and have poorer relationships.

Fathers who report better mental health and who have higher levels of self‐efficacy will report higher levels of involvement and more positive relationships with their children and will be more likely to report having increased their involvement and improved their relationships with their children during the pandemic.

METHODS

The present study draws on data from a nationwide U.S. survey of 373 self‐identified nonresident fathers regarding the short‐term impact of the COVID‐19 pandemic on their involvement and relationships with their children, parenting roles, employment and financial status, and mental/emotional well‐being. Survey data were collected via Amazon MTurk in mid‐ to late May 2020. During this period, all states in the United States were experiencing locked down orders to varying degrees (i.e., stay‐at‐home orders, travel prohibitions, all nonessential businesses closed) from mid‐March to early May. Most states began “reopening” in May 2020.

Therefore, the current data capture patterns from the height of the initial period of lockdowns and restrictions in the United States. MTurk (Amazon Mechanical Turk) is an online survey publication platform sponsored by Amazon. This platform links “requesters” (i.e., researchers who set eligibility criteria) with “workers” (individuals who sign up to take surveys). Workers are presented with lists of surveys for which they are eligible based on criteria set by the researcher. Each worker chooses the surveys to which they will respond. Researchers set the level of payment to be received by each participant with amounts depending on the degree of involvement required. Participants in the current study were compensated for their responses at rates comparable to other 20‐ to 30‐minute surveys. To be eligible for the study, participants had to be men over age 18 years, had to have a child aged 18 years or younger who did not live with them all or most of the time before the pandemic, had to be able to respond to the survey questions in English, and had to reside in the United States. Being nonresident was defined for participants as “not living with your child all or most of the time prior to COVID‐19” to capture those fathers who might have previously been nonresident but had children move in with them as a result of the pandemic. Data collection procedures were approved and exempt from ongoing review by the institutional review board of the University of Connecticut due to the anonymous nature of the online survey.

Data originally were collected from 417 individuals. Data were deleted for participants who did not complete at least 75% of the survey, took less than 5 minutes to complete the survey (the average time of completion was just under 23 minutes), responded incorrectly to attention check questions, or had multiple open‐ended answers that were nonsensical (e.g., cut and pasted from a page of engineering instructions, reported an age of “OK”). A few fathers were removed who reported that their youngest nonresident child was 19 or older or who reported living outside the United States. These restrictions resulted in a final sample of 373 nonresident fathers.

Although not without limitations, prior research demonstrates the overall validity of samples obtained from MTurk, as they are typically fairly representative of the United States and more diverse than those obtained via typical college student populations or other online recruitment modalities (e.g., Bartneck et al., 2015; Sheehan & Pittman, 2016). Although the present sample was not necessarily representative (see more description of participants below), particularly during the present pandemic when in‐person data collection is infeasible, MTurk is a valuable method for timely online data collection from a national sample.

Study participants

Nonresident fathers in the present sample resided in 45 U.S. states (the most common were California and Florida, with 12% of participants living in each of these states). Fathers on average were 37 years old, and two thirds of their youngest nonresident children (the target child in this study) were male; target children were approximately 8 years old on average (child age ranged from less than 1 year to 18 years old; SD = 4.60). Fathers lived an average of 112 miles from their children, although the median was 15 miles (some fathers lived more than 2000 miles away, which skewed the mean). Eighty‐six percent of fathers said they were answering the questions about a biological child, but 14% were responding about a foster child, adopted child, or child to whom they are a father figure.

Eighty‐eight percent of fathers stated that they were employed full time; however, 44% reported that their hours had been reduced, and 11% reported losing their job due to the COVID‐19 pandemic. It is unclear from these conflicting percentages whether some fathers had been working more than 40 hours per week and so, even with a reduction, still were working full‐time hours, or whether some fathers were working less than full‐time hours due to reduced hours or furloughs but still reported being employed full‐time. It might be that fathers in the latter group viewed any such reductions as temporary and so still considered themselves employed “full‐time.” Seventy percent of participants had a college degree or higher, which is twice as high as the general population (32%; U.S. Census Bureau, 2019). In addition, although the majority of participants were White Americans (61%), 9% of participants were Black or African American, 11% of participants were Latinx, and 17% of participants reported identifying as Asian or Asian American. This differs from demographics reported in the United States (U.S. Census Bureau, 2019), which identifies 76% of persons as White American, 13%, Black or African American, 18.5%, Hispanic or Latinx, and 6%, Asian or Asian American. Therefore, the current sample is not representative of the overall U.S. population in terms of its race/ethnicity or education. Of the participants, 32% knew someone who had been diagnosed with or who they suspected had had COVID‐19. See Tables 1 and 2 for additional demographic descriptive statistics.

TABLE 1

Descriptive statistics for family‐related demographic variables (N = 373)

Variables	M	SD
Father age (years)	36.76	8.06
Target child age (years)	8.15	4.60
Distance from child (in miles)	111.78a	540.86
	n	%
Sex of target child
Male	243	65.1
Female	128	34.3
Nonbinary/genderqueer	2	0.5
Target child was a
biological child	318	85.5
adopted child	13	3.5
foster child	12	3.2
child to whom they are a father figure	19	5.1
Other	10	2.7
Knows someone with COVID‐19
Yes	120	32.3
No	251	67.7

Note. aMedian distance = 15 miles.

TABLE 2

Descriptive statistics for general demographic variables (N = 373)

Variables	n	%
Father education
Less than high school	5	1.4
High school or equiv.	37	10.0
Some college or associate degree	67	18.1
4‐year college degree	161	43.4
Some graduate education or higher	101	27.2
Father race/ethnicity
White	229	61.4
Black or African American	33	8.8
Hispanic or Latinx	40	10.7
Asian/Asian American	64	17.2
Other	24	6.5
Father employment status
Full‐time	327	87.7
Part‐time	23	6.2
Unemployed	10	5.1
Retired	4	1.1
Employment hours
Hours are the same	168	45.0
Reduced	165	44.2
Lost job	32	8.6
Lost job but found a new one	8	2.1
Pre‐COVID‐19 monthly household income
<$2000	55	14.7
$2000–2999	92	24.6
$3000–3999	68	18.3
$4000–4999	49	13.1
≥$5000	109	29.2
Current monthly household income
<$2000	95	25.5
$2000–2999	83	22.3
$3000–3999	69	18.5
$4000–4999	41	11.0
≥$5000	85	22.7
Most common states of residence
California	46	12.3
Florida	46	12.3
New York	33	8.8
Texas	25	6.7

Measures

Fathers were asked to respond to questions regarding demographic information, their parenting time before and during the pandemic, relationship and involvement with their children, relationship with their children's other parent, mental health, and challenges and opportunities created by the pandemic for them as fathers. For all questions regarding their children, fathers were asked to think about their youngest child who did not reside with them all or most of the time because younger children would have less ability to independently seek (or avoid) time with fathers and would be more influenced by contextual factors. Fathers were instructed that “Before COVID‐19” should refer to the time before quarantines, lockdowns, and stay‐at‐home orders were implemented (e.g., in February) and “Since or During COVID‐19” should refer to the time after quarantines, lockdowns, and stay‐at‐home orders were implemented (typically starting in March and continuing through the time of data collection in May).

Father involvement

Time spent

Fathers were asked the average number of days per week they saw their child before and since COVID‐19 (responses ranged from 0 to 5 or more days per week) and the average number of overnights they had with their child per week before and since COVID‐19 (responses ranged from 0 to 4 or more nights per week). They also were asked “Overall, how has your time spent with your child changed since COVID‐19?”

In addition to the last question, if an increase or decrease in time was reported, fathers were asked whether several reasons accounted for that change. For an increase, fathers were asked to check all that apply from the following: “The child's mother has allowed me to see my child more,” “My child has been living with me since COVID‐19,” “I have made more of an effort to see my child since COVID‐19,” “I have been working less/laid off, which gives me more time to spend with my child,” and “I have been working from home, which gives me more time to spend with my child.” For a reported decrease in time spent, fathers were asked to check all that applied from the following: “The child's mother has imposed greater restrictions on my time with the child,” “I have been unable to obtain more parenting time due to the courts being closed,” “Travel restrictions have prevented me from seeing my child,” “Stay‐at‐home orders have prevented me from seeing my child,” “Myself or a member of my household was diagnosed with COVID‐19 and I didn't want to expose my child,” and “My child or a member of the household where my child primarily lives was diagnosed with COVID‐19 and I didn't want to expose myself or my household members.”

Fathers also were asked how many times per week they typically talked to their child virtually “by telephone, videoconference (e.g., Zoom, FaceTime), email, text, or social media” before and since COVID‐19. Responses ranged from 0 to “multiple times a day.” Finally, fathers were asked whether how often they talked to their child virtually increased, decreased, or had stayed the same since COVID‐19.

Father involvement and behaviors

Fathers were asked to report the number of days in an average week that they engaged in 14 activities or behaviors before and since COVID‐19. Items were adapted from father involvement items used in Years 1 and 9 of the Fragile Families and Child Well‐Being Study (FFCWBS, Bendheim–Thoman Center for Research on Child Wellbeing, 2003), to capture involvement with both older and younger children. The FFCWBS was chosen due to its participation in the DADS (Developing a Daddy Survey) project, which was a consortium of four national projects in the United States focused on improved measures for and data collection from fathers and because the sample of the FFCWBS intentionally has an overrepresentation of unmarried and nonresident fathers.

Sample items included “Hug or show physical affection to my child,” “Talk with my child about his/her day,” and “Watch tv or videos together.” Scores were averaged to provide average levels of involvement both before and since COVID‐19. Higher scores indicated greater involvement; Cronbach's alpha was .96 for the Before COVID‐19 Involvement scale and .98 for the During COVID‐19 Involvement scale. A change score also was calculated for each father by subtracting his previous level of involvement from his current level of involvement for each item; positive scores therefore indicated greater involvement since COVID‐19 and negative scores indicated a decrease in involvement. Item‐level change scores then were summed to create an overall Change in Involvement score. The Involvement During COVID‐19 score and the Change in Involvement scores were used as outcomes in regression analyses. Fathers also were asked “How challenging have you found it to maintain involvement and a relationship with your child during COVID‐19” with responses ranging from 1 Not at all challenging to 5 Very challenging.

Father–child and coparenting relationships

Father–child relationship

Fathers were asked to evaluate the overall quality of their relationship with their child both before and during the COVID‐19 pandemic. Responses for each question ranged from 1 I did/do not have a relationship with this child to 5 Excellent. A Change in Quality score also was computed by subtracting fathers' rating of relationship quality before COVID‐19 from their current rating of relationship quality; positive scores indicated improvements in relationship quality and negative scores indicated declines in quality. The Relationship During COVID‐19 item and the Change in Quality scores were used as outcomes in regression analyses.

Coparenting support

Five items were utilized from the FFCWBS that assessed the degree of coparenting support exhibited by both parents. Sample items include “My child's other parent supports me in the way I want to raise our child” and “We usually agree on issues involving our child.” In addition, two items were added that related specifically to coparenting during COVID‐19, “My child's other parent has encouraged me to be involved with our child during COVID‐19” and “My relationship with my child's other parent has gotten stronger since COVID‐19.” Responses ranged from 1 to 6 (strongly disagree to strongly agree), and higher scores indicated greater support; Cronbach's alpha was .89.

Coparenting difficulties

One item regarding coparenting difficulties was used from the FFCWB survey (“I am critical of the things our child's other parent does with our child”) because it did not load well with the coparenting support items (factor loading = .09, and it was the sole item on a second factor with a loading of .95), and three other items were created for this survey specifically to assess coparenting difficulties during the COVID‐19 pandemic: “COVID‐19 has added extra challenges to my relationship with my child's other parent,” “I am worried about my child's other parent taking appropriate measures to keep my child safe during COVID‐19,” and “My child's other parent has limited my involvement with our child since COVID‐19.” Responses again ranged from 1 strongly disagree to 5 strongly agree with higher scores indicating greater coparenting difficulties; Cronbach's alpha was .77.

Coparenting interaction

As noted, support and difficulties can both exist within coparenting relationships, and relationships characterized by high support and high difficulties likely function differently from those with high support and low difficulties, for example. Therefore, I created an interaction term that combined both coparenting support and coparenting difficulties. Each variable was centered, and then support and difficulties scores were multiplied together to create a “Combined Coparenting” measure that incorporated both support and difficulties simultaneously.

Fathers' mental health

Negative mental health

Fathers' anxiety and depression were assessed via the GAD‐7 (Generalized Anxiety Disorder brief scale; Spitzer et al., 2006) and the PHQ‐9 (Patient Health Questionnaire–9; Kroenke et al., 2001), minus the question relating to suicidal ideation (which was removed because of the inability to follow up with suicide prevention protocols for anyone responding yes, given that this was an anonymous survey). For both scales, responses ranged from 0 not at all to 3 nearly every day; higher scores indicated greater depression and anxiety symptoms. Initially, items were averaged to create two variables: an overall average anxiety score and an overall average depression score. However, bivariate correlations revealed that the two scales were correlated at r = .89 (p < .001). Therefore, to avoid collinearity and because the scales appeared to be tapping a single overall construct, the two scores were averaged to create a composite score of overall negative mental health. Cronbach's alpha for the resultant 15‐item measure of overall negative mental health was .97.

Self‐efficacy

Fathers' self‐efficacy was assessed with the General Self‐Efficacy Scale (GSE; Schwarzer & Jerusalem, 1995). This scale includes 10 items that assess an individual's perceived ability to cope with everyday challenges. Sample items include “I can always manage to solve difficult problems if I try hard enough” and “When I am confronted with a problem, I can usually find several solutions.” Responses ranged from 1 (not at all true) to 4 (exactly true); higher scores indicated greater self‐efficacy. Cronbach's alpha in the present sample was .91.

RESULTS

Descriptive results

Descriptive statistics were analyzed for all variables of interest; tables of full descriptive statistics are available in online supplementary materials. Key findings for each descriptive research question are described in this section.

Father involvement

Time spent

Overall, fathers perceived a significant decline in their time spent with children during the COVID‐19 pandemic shutdown. The number of days that fathers saw their children declined (pre‐pandemic = 3.85, post‐pandemic = 3.05; t = 9.83, p < .001), as did their number of overnight visits (pre‐pandemic = 3.17, post‐pandemic = 2.57; t = 9.19, p < .001). Before the pandemic, 5% of the fathers did not see their children at all each week, and 8% had no overnight visits in an average week. After the pandemic, the percentages quadrupled in each category, to 26% who did not see their child at all during each week and 35% who had no overnight visits. Fifty‐two percent reported that they had spent less time with their child since the pandemic. However, it is important to note that 31% reported spending more time with their child. Among fathers who reported less time spent with their children, 26% reported that travel restrictions have prevented them from seeing their child, 37% reported that stay at home orders had prevented them from seeing their child, and 23% reported that the child's mother had imposed greater restrictions on their time with their child. Eight percent reported that someone having COVID‐19 kept them from seeing their child. Among those who reported an increase in time spent, 12% reported that their child has started living with them since the pandemic, 17% stated that working less or working from home has allowed them to see their child more, and 15% stated that the child's mother has allowed them to see their children more. More than half (57%) of all fathers reported spending more time talking with their child virtually (by phone, videoconference, email, text, or social media) since the pandemic.

Involvement

On average, fathers perceived a significant decrease in involvement since the pandemic (pre‐pandemic = 4.96, post‐pandemic = 4.36; t = 5.18, p < .001), but there was wide variability: 57% declined in their involvement, 11% maintained the same level of involvement, and 32% increased their overall level of involvement. Fathers' overall rating of the quality of their relationships with their children also significantly declined (pre = 4.05, post = 3.90; t = 3.11, p < .01). The number of fathers rating their relationship with their child as poor increased from 3% to 8% (20 additional fathers), and the number of fathers reporting that their relationship was either good or excellent decreased slightly, from 75% to 68%. One third of the participants reported that it was challenging or very challenging to maintain involvement and a relationship with their child during the pandemic and 17% stated that it was Not at all challenging.

Coparenting

The average level of perceived coparenting support in the current sample was 4.41 (on a 1 to 6 scale), indicating they generally somewhat agreed (4) to agreed (5) that their coparenting relationship was mutually supportive. Fathers averaged 3.72 on a 1 to 6 scale for coparenting difficulties, indicating that they somewhat agreed (4) with the coparenting difficulty items. Half of the fathers reported that their relationship with their child's other parent had stayed the same, about 20% stated that it had gotten a little bit worse or much worse, and about 30% reported that it had improved a little bit or a lot. Promisingly, 75% of the fathers rated their current relationship with their coparent as good or excellent.

Predictive results

Initial bivariate correlations (see Table 3) revealed that fathers' education, whether his employment hours had been reduced during the pandemic, and the target child's gender and age were significantly associated with at least one of the outcomes of interest. Therefore, these demographic variables were included as controls in the later regressions. It should be noted that, given differences in restrictions and infection rates across states, it would have been interesting to test for differences by geographic region. However, creating meaningful “regions” or clusters of states would require rating each state's infection rates, regulations, and response to the pandemic. These calculations were beyond the scope of this article (geographic regions could not be used because different states within geographic regions varied drastically in infection rates and restrictions; for example, in the five states that are considered the Northeast Region: Connecticut and New York were drastically different from New Hampshire and Vermont, which in turn were different from Maine), and there was insufficient power and sample size to test for differences at the state level.

TABLE 3

Correlations among study variables

Variable	1	2	3	4	5	6	7	8	9	10	11	12
1. Hours reduced	—
2. Education	−.08	—
3. Child age	−.11*	−.10*	—
4. Child gender	.01	.03	−.02	—
5. Coparenting support	.06	.11*	−.08	−.09	—
6. Coparenting difficulties	.29**	.08	−.10	−.03	−.10	—
7. Negative MH	.39**	.04	−.19**	−.11*	.06	.53**	—
8. Self‐efficacy	−.17**	.10	.05	.01	.32**	−.21**	−.24**	—
9. PC involvement	.18**	.17**	−.20**	−.21**	.42**	.40**	.38**	−.08	—
10. Involvement change	−.11	.16*	−.06	−.01	.21**	.00	.03	.02	.57**	—
11. RQ	−.12*	.07	−.04	.03	.38**	−.18**	−.23**	.31**	.28**	.30**	—
12. RQ change	.03	.09	−.02	.01	.21**	−.03	−.03	.18**	.25**	.38**	.58**	—

Note. PC = post‐COVID‐19; MH = mental health; RQ = relationship quality.

p < .05.

p < .01.

Predictors of fathers' perceived involvement, changes in involvement, relationship quality, and changes in relationship quality were tested via four hierarchical linear regressions (one for each outcome). Demographic variables were entered in the first block, coparenting support and difficulties were entered in the second block, and fathers' mental health variables were entered in the third block. Finally, in the last block, the combined coparenting interaction term was entered to test whether particular combinations of coparenting support and difficulties were associated with the outcomes of interest. See Table 4 for all regression results.

TABLE 4

Regression estimates for predictors of post‐COVID‐19 father involvement, relationship quality, and change in relationship quality (final models with all variables)

	PC involvement		RQ		Change in RQ
	β	B (SE)	β	B (SE)	β	B (SE)
Employment hours reduced	.00	−.01 (.17)	−.04	−.05 (.07)	.05	.07 (.07)
Education	.07	.10 (.07)	.03	.02 (.03)	.06	.03 (.03)
Child gender	−.11 †	−.38 (.19)	.05	.08 (.09)	−.00	.00 (.09)
Child age	−.17**	−.09 (.03)	−.06	−.01 (.01)	−.01	−.00 (.01)
Coparenting support	.40**	.89 (.12)	.31**	.30 (.05)	.14*	.12 (.05)
Coparenting difficulties	.31**	.56 (.11)	−.04	−.03 (.05)	−.00	−.00 (.05)
Negative mental health	.10	.28 (.18)	.18**	.22 (.08)	−.02	−.02 (.08)
Self–efficacy	−.11*	−.04 (.02)	.17**	.03 (.01)	.14*	.02 (.01)
Support × difficulties interaction	.11 †	.17 (.09)	.05	.03 (.04)	.01	.01 (.04)
Adjusted R 2	.40		.20		.03

Note. PC = post‐COVID‐19; RQ = relationship quality.

Note.

p < .05.

p < .01.

p < .06.

Post‐COVID father involvement

The first three blocks of variables were significant in the model predicting fathers' average current level of involvement (during the pandemic) and the last block with the interaction term was marginally significant (p = .052). The overall model explained 40% of the variance (see Table 4 for full results of final model). The first block of demographic variables explained 13% of the variation in involvement and all variables were significant. Fathers with more education (β = .19), whose hours of employment had not been reduced (β = .19), and who had younger sons (β = −.20; β = −.19, respectively) tended to be more involved with their children.

The second block of coparenting variables explained an additional 24% of the variance and both coparenting support (β = .37) and coparenting difficulties (β = .39) were positively associated with involvement since the pandemic. Adding these variables reduced fathers' reduced hours of employment and education to nonsignificance. Adding the third block of mental health variables only explained an additional 2% of the variance, and only self‐efficacy was significant; interestingly, fathers who reported lower self‐efficacy (β = −.13) were more involved with their children. Negative mental health was unrelated to involvement, and both coparenting variables and child gender and age remained significant.

The final block with the interaction term combining coparenting support and coparenting difficulties was marginally significant (β = .11, p = .052). To interpret the interaction, I examined scatterplots of the association between perceived coparenting support and involvement across three groups of coparenting difficulties (less than −1 SD, −1 to +1 SD, and greater than 1 SD; available from author upon request). These revealed that as coparenting difficulties increased, the association between coparenting support and involvement also increased. That is, the amount of perceived support in the coparenting relationship was particularly influential to fathers' involvement when they also perceived high levels of difficulties.

Change in involvement

The predictors of changes in involvement since the pandemic were quite different from those predicting overall levels of involvement. Neither the overall model nor any of the individual blocks of variables were significant, although perceived coparenting support was associated with changes in involvement (β = .18). Fathers who reported supportive coparenting relationships also reported that their involvement had increased during the pandemic. However, given that the overall model and blocks of variables were not significant, this finding should be interpreted with caution, particularly because it explained only 2% of the variation in changes in involvement (full results of this analysis available from author upon request).

Father–child relationship quality

When predicting current father–child relationship quality, the first block of demographic variables was not significant and the interaction term was not significant, but the second and third blocks were significant and explained a total of 20% of the variance in relationship quality (see Table 4 for full results of final model). Although overall the block of variables was nonsignificant, a reduction in employment hours was negatively associated with relationship quality (β = −.12) and explained 1% of the variance. Fathers whose hours of employment had been reduced due to the pandemic reported lower quality relationships with their children since the pandemic. Adding coparenting variables explained an additional 14% of the variation in relationship quality, and both perceived support and difficulties were significant in opposing directions. More coparenting support (β = .34) and fewer coparenting difficulties (β = −.14) were associated with better current father–child relationships, and reduced hours of employment became nonsignificant. Adding mental health variables explained an additional 5% of the variance, and both negative mental health and self‐efficacy were associated with relationship quality. Having better mental health (β = −.18) and greater self‐efficacy (β = .17) were associated with better current father–child relationships.

Change in relationship quality

Finally, when predicting changes in father–child relationship quality since the pandemic, the second block of coparenting variables was significant and the third block was marginally significant, but only predicted 3% of the change in quality (see Table 4 for full results of final model). No demographic variables were associated with changes in relationship quality. In the second block, only perceived coparenting support was positively associated with changes in relationship quality (β = .18) and explained 2% of the variance in relationship quality changes. Fathers who reported more supportive relationships were more likely to report that their relationships with their children had improved since the pandemic. Finally, although the overall third block of variables was not significant, higher levels of self‐efficacy were associated with father–child relationships having improved since the pandemic (β = .13) and explained an additional 1% in the variation in relationship quality change. The interaction term was not significant.

DISCUSSION

The descriptive statistics paint a picture of the diverse impact that the COVID‐19 pandemic has had on nonresident fathers, with many experiencing obstacles to their parenting and coparenting but others experiencing increased opportunities for parenting and improved relationships. Fathers ranged widely in both their ability to maintain involvement and their relationships with their children, as well as their relationships with their children's other parent. It was therefore important to investigate the predictors of this variability. Support for the predictive hypotheses of the current study was mixed. Although some hypotheses were supported, others were not.

Overall, the present findings support the interdependent nature of family systems even when spread across separate households; the perceptions of fathers regarding how they were coparenting during the pandemic was crucial to fathers' involvement and relationships with their children during the pandemic. These findings also highlight the variability of families' abilities to adapt to challenging events and abrupt changes in circumstances, with some coparenting partnerships and father–child relationships improving and others suffering when faced with the challenges, and sometimes opportunities, brought about by the pandemic.

Coparenting

Consistent with the first hypothesis, having higher perceived levels of coparenting support was consistently associated with higher and increased involvement and with better and improving father–child relationships. However, having higher levels of perceived coparenting difficulties was associated only with having more current involvement. It appears that the greater the difficulties, the more strongly support influences father involvement. The current findings replicate previous research that supportive coparenting is associated with more involvement among nonresident fathers (Fagan & Palkovitz, 2011, 2019) and extends this finding to the quality of father–child relationships as well.

Research has demonstrated the role of father involvement in creating high quality father–child relationships (e.g., Jessee & Adamsons, 2018), but more involvement does not automatically equate to close father–child relationships and they are not necessarily impacted by the same processes or factors (something also replicated here). For example, coparenting difficulties only were associated with current involvement, not changes in involvement or either relationship quality variable, and interestingly, coparenting difficulties were positively associated with involvement. Given that the current findings are cross‐sectional in nature, this replicates previous work that has noted that increased involvement by nonresident fathers sometimes is associated with greater conflict between parents because it brings hostile couples into more contact with one another. As well, given that most of the measure of coparenting difficulties was specific to the pandemic, these findings seem to indicate that coparents may have been successfully working through these difficulties rather than being stymied or stalemated by them. Difficulties, after all, are risks rather than certainties.

It is interesting that fathers' relationships with children were less dependent on the quality of the coparenting relationship. Although support promoted their relationships, difficulties did not impact father–child relationships one way or the other. It seems, therefore, that coparenting difficulties might enable mothers to limit the contact between fathers and children, but fathers are able to find alternate ways of maintaining their relationships with their children. It should be noted, however, that these were father reports of involvement and relationships. As such, future research would benefit from exploring children's perspectives, to see if children similarly feel that their relationships remain positive in the face of limited involvement.

The findings regarding the interaction between coparenting difficulties and supportiveness were somewhat intuitive. At low levels of difficulties, support is helpful but not necessarily critical for fathers to stay involved; however, when difficulties increase, fathers appear to benefit from higher levels of concurrent support to stay involved with their children rather than disengaging. Reexamining the items that comprised the coparenting difficulties scale provides a possible explanation for such findings. Items for this scale included bring critical of things the other parent does with the child, being worried about the other parent keeping the child safe during the pandemic, the pandemic adding extra difficulties to their relationship with the other parent, and the other parent limiting their involvement with their child since the pandemic. With the exception of the item regarding restricted involvement, the remaining items reflected concerns that might motivate fathers to increase their involvement with and closeness to their children, particularly in the context of a coparenting relationship that is also high in supportiveness. If concerned that their child is unsafe, or if they are worried about or critical of what the other parent is doing with the child, fathers might make extra efforts to “keep an eye on things,” particularly when the coparenting relationship generally is supportive and promotes access. That is, because these fathers are high in both supportiveness and difficulties, these fathers both are concerned and challenged during the pandemic, and also “can talk with my child's other parent about problems” and “usually agree on issues involving our child” (for example). Thus, although difficulties exist in their relationships with the other parent as a result of the pandemic, for this particular group of fathers, such difficulties served to enhance their engagement and relationships with their children rather than detracting from them. In the absence of such difficulties, however, fathers' involvement was less affected by support, possibly because mothers engaged in less interference with fathers' involvement even if they weren't actively supportive. Overall, this finding warrants additional exploration and is an important reminder that coparenting relationships are complex and multidimensional—both in their nature and in their very practical influence on fathers' involvement with their children.

Mental health

Mental health was inconsistently associated with involvement and relationships. Negative mental health (depression and anxiety) was negatively associated only with current father–child relationship quality. Fathers with more anxiety and depression symptoms reported poorer current relationships with their children. This might relate to the point made earlier about involvement not being synonymous with relationship quality. For example, fathers who are more depressed and anxious might still have regular contact and engagement with their children (as evidenced by the lack of an association between mental health and involvement). However, they also might be less available emotionally to their children, adversely influencing the quality of their relationships.

It was surprising that changes in involvement or relationship quality were unrelated to mental health. It was expected that fathers who were struggling psychologically, particularly during the pandemic, might see their children less or drift apart from their children emotionally. Alternatively, if fathers saw their children less or felt that their relationship was declining as a result of the pandemic, they might be expected to become more depressed and anxious as a result. However, the current cross‐sectional data did not support either expectation. Thus, while the adverse associations with current relationship quality warrant attention, it is encouraging that these fathers apparently can maintain active engagement with their children even when struggling psychologically.

With regard to self‐efficacy, the results again were inconsistent and somewhat surprising based on the proposed hypotheses and existing literature. Greater self‐efficacy was associated with better current relationship quality and with improving relationships, which is consistent with previous research (Trahan & Shafer, 2019) and supports the hypothesis that fathers who feel more able to cope with adverse situations have been able to maintain and even improve their relationships with their children during the pandemic. However, self‐efficacy was unrelated to changes in involvement, and was negatively associated with current involvement, such that fathers with lower self‐efficacy had higher levels of involvement during the pandemic. The latter finding is surprising, as one might expect that during a crisis such as the pandemic, it would be the efficacious fathers who would be better able to engage with their children. However, it is possible that it is the involvement driving the association, not the efficacy. In other words, it might be that fathers who are more involved with their children during the pandemic, feel less efficacious as parents due to being unable to “solve” the problems that the pandemic likely poses for their children (struggling with distance learning, inability to see friends and other family members, changes in routines, possible losses or family illnesses). Thus, it is not feeling out of control that influences involvement, but that greater involvement during such uncertain times leads fathers to feel less in control.

It is also worth noting that a reduction in fathers' employment hours was negatively associated with efficacy, so it might be that although feeling “out of control”, working less meant that these fathers spent more time with their children. In the present sample, 17% reported that working less or working from home had allowed them to spend more time with their child and 12% also reported that their child had been living with them since the pandemic. Both changes might adversely impact fathers' sense of self‐efficacy but increase their involvement. Overall, it is likely that there is more to these opposing associations between self‐efficacy and fathers' involvement and relationships than it first appears.

Limitations and future directions

One obvious limitation of the current data is that they were collected at a single point in time, asking fathers to look back retrospectively on the way things were prior to the onset of the pandemic, rather than longitudinally assessing their involvement prior to and during the pandemic. Therefore, the current data likely are influenced by retrospective bias and fathers viewing their prior experiences through the lens of current events and experiences. However, it was impossible to predict that such an event was about to occur to collect such “pre‐pandemic” data. Too, although relatively diverse in some respects and a nationwide sample of fathers, the sample was of relatively high socioeconomic status, with more than 80% of the sample having a college degree or higher and 88% being employed full time. Therefore, the present findings tell us little about the impact of the pandemic on low‐income fathers, who are likely to have been employed in sectors that were the hardest hit during the shutdowns and therefore are also likely to have been much more adversely financially affected, as well as being less likely to work remotely. An inability to work remotely could particularly impact fathering, and some fathers cited their “essential” jobs and resultant ongoing potential exposure to COVID as a reason for not seeing their children during the pandemic (to keep their children safe).

Finally, the measures used here, in some cases, were designed specifically for this study and to tap into aspects unique to the pandemic restrictions. In the case of coparenting difficulties, for example, the measure used in the present study is not consistent with the way coparenting has been traditionally measured in other studies. The focus of items in this study were primarily focused on conflict or gate‐closing behaviors. This makes it difficult to directly compare findings with the existing coparenting literature.

Given the diverse regulations and restrictions across different states during the pandemic, as well as differences in prevalence of the disease in different regions and areas of the country at different times, it would be interesting to cross‐reference the current findings with disease prevalence and restrictions. Unfortunately, this was beyond the scope of the current study. Both prevalence and restrictions could influence fathers' ability to spend time with their children. In the current sample, 32% reported knowing someone who had COVID‐19, and 8% reported not seeing their child because someone in either their or the child's household had COVID‐19; 26% reported that travel restrictions had prevented them from seeing their child, and 37% reported that time with their children had been limited by stay‐at‐home orders.

It also will be necessary to collect additional, more recent data about how parenting and coparenting have evolved as the pandemic has shifted from what was thought to be a short‐term problem to a long‐term one that will continue for an as‐yet‐undetermined amount of time. As the pandemic continues to wax and wane, even post‐vaccines and 2 years later, different states continue to be impacted by the pandemic differently and have had differing levels of restrictions (or not) in response. Availability of vaccines has led many to call the current pandemic a “pandemic of the unvaccinated,” and so additional information about families with members who are or are not vaccinated will be valuable. Gathering additional information will be necessary to find the best ways to support father engagement and high‐quality father‐child relationships moving forward.

Implications

Overall, the current findings support the utility of viewing families as interconnected systems, whether or not they reside in the same household. This is perhaps not surprising because this is not the first study to use an FST lens to explore the reactions of families to crisis situations. For example, Lester and Flake (2013) and Paley et al. (2013) both applied FST to examinations of how children and families adjust to a parent's wartime combat deployment, as well as their return home. Regardless of the source of influence, when family boundaries and routines are altered, all members of the family system are affected and must renegotiate the adjustment of those boundaries and routines. When such adjustments are made unilaterally, nonresident fathers (and children) tend to suffer. In contrast, when adjustments are negotiated jointly by both parents, the potential exists for all components of the family system to thrive, even in the face of adversity. As the pandemic continues, it is important for court systems, therapists, and others who work with nonresident parents to recognize and work to find creative solutions that support shared parenting and supportive relationships across all households and family members.

The kinds of restrictions, policies, and practices that will be in place in the long‐term remain largely unknown at present, as are the long‐term impacts of prior and future policies and economic ramifications. However, the current findings reinforce the importance of policymakers and practitioners being aware and supportive of nonresident fathers as a population at strong risk of being adversely affected by the current pandemic. In addition, the implications of the present findings likely can extend beyond the current pandemic to include other global crises during which parents and children might be forcibly separated, such as during wartime or during environmental catastrophes such as wildfires or hurricanes, which are only increasing in frequency with changing climate conditions. In such cases, the rights and interests of nonresident parents often are superseded in favor of efficiency and broader concerns about public safety. Although circumstances may sometimes dictate that this is a necessity, it is nonetheless important for scholars, policymakers, and practitioners to be aware of such impacts and the long‐term adverse implications they can have for both fathers and children.

CONCLUSION

The importance of fathers for children has been well‐established over decades of research, and research from the extensive literature on divorce emphasizes the importance of establishing positive routines of care and involvement by fathers early on if such involvement is desired in the long term. Similarly, we must take steps now to recognize and ameliorate the adverse impacts of the pandemic on nonresident fathers and families and to build on the strengths and resilience of fathers in the face of this adversity. Otherwise, we face the likelihood that these temporary negative trends will become permanent, harming both fathers and children now and well into the future.

Appendix S1. Supporting Information

Click here for additional data file.