SEARCH

SEARCH BY CITATION

Keywords:

  • family and work;
  • family caregivers;
  • children with disabilities;
  • parent stress;
  • parent mental health

Abstract

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

There is a growing literature examining the impact of work and family responsibilities on the psychological well-being of parents of children with disabilities and other special needs. A number of studies using small, nonprobability samples of mothers find that work provides a respite from the stressful effects of caregiving. Using data from the National Survey of American Families, this study found higher mental health among working mothers of older children with disabilities compared to their nonworking counterparts and mothers of typically developing children, a result consistent with caregiver-specific positive spillover. No significant differences in mental health were found among working and nonworking mothers of younger children with disabilities or among fathers. Results also indicate that caregiver mothers who are experiencing high levels of parent-role stress benefit more from work, and that the beneficial effects from work persist until rather high levels of work (50 or more hours per week).

There is a growing literature examining the impact of work and family responsibilities on the psychological well-being of working parents (Barnett & Baruch, 1985; Barnett & Marshall, 1992; Frone, 2000; Goodman & Crouter, 2009; Grzywacz & Bass, 2003; Melchior, Berkman, Niedhammer, Zins, & Goldberg, 2007). Of particular concern are families with heavier-than-average caregiving responsibilities, including parents of children with disabilities (Brennan & Brannon, 2005; Canning, Harris, & Kelleher, 1996; Rosenzweig, Brennan, & Ogilvie, 2002). Research consistently shows that parents of children with disabilities and other exceptional needs have heavier caregiving duties than parents of typically developing children and that they are more likely to experience mental and emotional distress (Deater-Deckard, 2004; U.S. Department of Health and Human Service, 2008), raising concern about the effects of work–family conflict on the well-being of caregivers.

Research investigating the effect of work on parents of children with disabilities and health conditions is sparse; but, on balance, it suggests that working outside the home may offset the stressful effects of caregiving. The empirical evidence of caregiver-specific mental health benefits from work is, however, based primarily on small, nonprobability samples (Freedman, Litchfield, & Warfield, 1995; Gottlieb, 1997; Lewis, Kagan, Heaton, & Cranshaw, 1999; Parish, 2006; Thyen, Kuhlthau, & Perrin, 1999; Warfield, 2001, 2005). Moreover, though there has been an increase in the study of fathers in work and family research in general (Bianchi & Milkie, 2010; Hill, 2005; Levine & Pittinsky, 1997), most of the studies examining the effects of work on the mental health of parent caregivers have been conducted using mother-only samples (one exception is Warfield, 2005) or nearly mother-only samples (McDonald, Poertner, & Pierpont, 1999).

In an effort to confirm the possibility of a caregiver-specific positive spillover effect, an earlier study conducted by the author of this article (Morris, 2012) used a nationally representative sample of married mothers (N = 43,342) from the National Survey of American Families (NSAF) to examine the relationship between mental health, work, and caregiving. The study compared the relationship between work and mental health among mothers of children who were in poor to fair health, had disabilities, or exhibited behavior problems, and mothers of typically developing children. The findings suggested that mothers of older children with special needs do benefit more than other mothers from working outside the home. The analysis also indicated that the positive mental health effects from work were larger among caregiver mothers who worked full-time, during the day, and in executive, managerial, or professional occupations. The study reported here expands on this earlier work by using a sample of mothers and fathers as well as single and married or partnered parents, and employing a more robust analysis of the relationships between work, caregiving, and mental health. Specifically, the analysis employed a more precise definition of child disability and examined of the relation between parent-role stress and mental health.

Conceptual Framework

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

The primary objective of this study was to investigate whether the relationship between work and mental health was different for parents of children with disabilities and parents of typically developing children. The study also investigated whether the associations between work, caregiver status, and mental well-being differed for mothers and fathers and among married or partnered and single parents, and whether parent-role stress and certain job conditions attenuated the relationship between work and parent mental health. In keeping with other research on the impact of work and family responsibilities on parent well-being, role theory was used to frame the analysis. Role theory seeks to explain how individuals balance multiple roles, such as spouse, parent, caregiver, and employee, and the consequences for role performance and well-being (Barnett & Baruch, 1986; Voyandoff, 2002). While some studies have examined the effects of work-family on a more expansive set of outcomes (see, e.g., Gareis, Barnett, Ertel, & Berkman, 2009), most of the work to date has focused on the impact of work and family on parents' psychological well-being. This report also focuses on psychological well-being, which enhances consistency and comparability.

Role theory conceptualizes the dynamics between work and family roles as “spillover” effects. Spillover can be negative (conflict, strain) or positive (facilitation, enhancement, enrichment) and can occur in either direction: from work to family and from family to work. Psychological distress is a symptom of negative spillover, occurring when family-based stressors undermine work performance or, vice versa, when work-related stressors undermine the quality of parenting and family relationships. Alternatively, conflict can be time based; for example, when time spent at work leaves too little time for home, interfering with the functioning of parenting and family relationships and causing mental and emotional strain (Barnett & Marshall, 1992; Greenhaus & Powell, 2006; Grzywacz, Almeida, & McDonald, 2002; Grzywacz & Bass, 2003; Hanson, Hammer, & Colton, 2006; Hill, 2005; Tiedje et al., 1990; Voyandoff, 2002, 2004).

In contrast, improved psychological well-being is a symptom of positive spillover (Greenhaus & Powell, 2006; Hill, 2005; Voyandoff, 2002). Positive spillover effects are posited to be the result of a transfer of attitudes, skills, or abilities from one role to another role. For example, in a study conducted by Tiedje et al. (1990, p. 70), one mother said “Enjoying my work makes it easier to enjoy my children too. I feel better about myself.” Alternatively, positive spillover occurs when participation in one role buffers the individual from strain resulting from participation in another role. In this case, mental and emotional distress is lowered by work (family) rewards or resources offsetting family (work) stressors. It could also be that simply having some time away from home responsibilities is in itself stress reducing (Greenhaus & Powell, 2006; Hill, 2005; Voyandoff, 2002). For example, in the study by Tiedje et al. (1990), a mother said, “The distance and time away from the children gives me a better perspective about them and their behavior” (p. 70).

Positive psychological outcomes can also be produced in the absence of spillover between family and work domains simply as an additive effect, whereby individuals' participation in and satisfaction with each additional role incrementally increases their overall psychological well-being. Similarly, a decline in psychological well-being can result when the negative effects of work (or family) generate net losses in psychological well-being (Greenhaus & Powell, 2006; Voyandoff, 2002).

Work–family dynamics and spillover effects are also predicted to be influenced by individual characteristics and attitudes, family structure, and work conditions (Voyandoff, 2002). Different types of families will have different resources and abilities for balancing work and family, and these differences will influence the direction and intensity of spillover between work and family. The inclusion of variables capturing family structure and socioeconomic status helps explain the conditions under which specific relationships between work, family, and mental well-being occur (Canning et al., 1989; Minnotte, 2012; Voyandoff, 2002). Family and gender norms and expectations also influence the experiences of caregiving and work as well as work–family dynamics (Barnett & Marshall, 1992; Coltrane, 1996; Hill, 2005; Lewis et al., 1999; Lin, Fee, & Wu, 2012; Minnotte, Minnotte, Pedersen, Mannon, & Kiger, 2010; Usdansky, Gordon, Wang, & Gluzman, 2012; Voyandoff, 2002). Traditional sex role theory proposes that fathers are more invested in work, emotionally and temporally, and mothers are more invested in family and caregiving. Due to these gender differences in role demands and salience, gender moderates the relationships between work, family, and psychological well-being, producing different outcomes for mothers and fathers. Attitudes about work, work-role quality, and specific job conditions are also influential either as rewards and resources or constraints and stressors. Although research consistently demonstrates positive or neutral mental health effects from work (Barnett & Baruch, 1985), studies also find that hours, work schedules, job complexity, and autonomy as well as job satisfaction and job security are related to work–family dynamics and mental health outcomes (Grzywacz & Butler, 2005; Pedersen, Minnotte, Kiger, & Mannon, 2009; Usdansky et al., 2012; Voyandoff, 2002, 2005).

Previous Research

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

Work–family conflict and negative spillover have been the focus of most studies (Barnett & Marshall, 1992; Bianchi & Milkie, 2010; Voyandoff, 2004); however, increasingly researchers are looking for and finding positive psychological effects (Gareis et al., 2009; Greenhaus & Powell, 2006; Grzywacz & Bass, 2003; Grzywacz & Marks, 2000; Hanson et al., 2006). Using a nationally representative sample (N = 1,986) from the National Survey of Midlife Development in the United States, Grzywacz and Bass (2003) examined the impact of work–family conflict and facilitation on the mental well-being (depression, anxiety, and problem drinking) of working parents. As theory predicts, they found a negative relationship between work–family conflict and mental well-being and a positive relationship between work–family facilitation and mental well-being. Interestingly, they also found evidence that work–family facilitation buffers the negative consequences of work–family conflict.

Based on the analysis of a random sample of 158 married women college professors, Tiedje et al. (1990) found that those experiencing high work–family enhancement and low work–family conflict scored higher on measures of mental health, whereas those experiencing low enhancement and high conflict scored lower. In another study, Hanson et al. (2006) examined the relationship between three dimensions of work–family positive spillover: behavior-based instrumental positive spillover, value-based instrumental positive spillover, and affective positive spillover. Using a sample of 193 employees from two companies, they found significant positive correlations between mental well-being and work-to-family value- and behavior-based instrumental positive spillover and family-to-work behavior-based instrumental positive spillover.

Although most of the research has been conducted on parents of typically developing children, there are a handful of studies examining the impact of work–family on the psychological well-being of parents of children with disabilities and special health care needs. Although some studies found that working outside the home is detrimental to the well-being of parent caregivers (McDonald, Poertner, & Pierpont, 1999), a number of other studies reported positive spillover effects. A study by Thyen et al. (1999) collected information on work and mental well-being from 70 mothers of children with chronic disease (study group) and 58 mothers of children with acute health conditions (comparison group). Using linear regression models with maternal mental health scores as the dependent variable and group membership (study vs. comparison), employment status, and the interaction of group and employment as independent variables, they found a statistically significant and positive interaction effect, suggesting that employment moderated parent stress specifically for mothers of children with chronic disease.

A number of qualitative studies also report positive mental health outcomes from work. Based on four focus groups of working parents of children with developmental disabilities, Freedman et al. (1995) found evidence of positive spillover, from work to family and family to work, especially among mothers of children with behavioral problems. Many of the parents said that having a child with disabilities positively influenced their attitudes toward work and enhanced their performance. Parents cited psychological benefits from working, with some specifically describing work as a “respite” or “a welcome distraction” (p. 510). Other parents stated that their experiences raising children with disabilities made them better able to handle difficult work situations. In a focus group study (N = 8) conducted by Parish (2006), mothers of adolescents with developmental disabilities reported emotional and psychological benefits from working outside the home as well as boosts to self-esteem, despite facing considerable difficulties balancing work and caregiving and suffering from feelings of isolation and depression. Lewis et al. (1999) reported that almost all the mothers in their sample of 40 families with children with disabilities specifically highlighted the positive psychological benefits from employment, even when balancing work and family was stressful. Several stated specifically that work provided respite: one mother said, “I went back to work in desperation to get out … it saves my sanity,” and another said, “It's not just earning money, it's giving me some space of my own, and it gives me something to concentrate on, and while I'm at work I'm busy … not wallowing in self pity … it keeps you sane” (p. 565).

Two studies by Warfield (2001, 2005) produced evidence consistent with positive spillover among parents with high levels of work interest. In her 2001 study, Warfield used a sample (N = 56) of employed mothers of children with behavior problems obtained from an early intervention program to examine the impact of work and caregiving demands on parent stress, a measure that encompassed the parent's mental and emotional well-being and the parent's perception of parenting role quality and marital satisfaction. Using linear regression methods and interactions between caregiving demands and work variables (hours, workload intensity, and level of interest in work) as independent variables, she found that the level of work interest moderated the negative influence of increased caregiving on parent stress, but only at lower levels of caregiving (i.e., high caregiving difficulties resulted in higher stress no matter what the level of work interest). In a second study, using a sample (N = 51) of working married couples with children with disabilities obtained through the same intervention program and the same analytic approach, Warfield (2005) found a similar relationship between parent stress and work, moderated by work interest, among mothers and fathers, although the results were statistically significant only for mothers.

Although intriguing, these studies were conducted on small, nonrandom samples. As described earlier, the current author conducted an study (Morris, 2012) using a large, nationally representative sample of married mothers. The analysis used linear regression models with interactions between child special needs by type (disabilities, poor health, and behavior problems) and maternal employment status. To gauge the extent to which bias caused by simultaneity between the dependent variable (parental mental health) and employment variables might be driving the results, logistic regression models using dichotomized versions of caregiving responsibilities (children with behavior problems vs. children with only positive behaviors; children in fair to poor health vs. children in good or excellent health) and parent mental health (high anxiety and depression vs. low to no anxiety and depression) were also estimated, and the results compared to those obtained using linear models with interactions. Results produced by both approaches were consistent with a positive spillover effect from work among mothers of older children (age 6–17) with disabilities, health conditions, and behavior problems. Using a subsample of working caregiver mothers, the study also found evidence that the positive mental health effects from working outside the home may be limited to caregiver mothers who work full-time, during day-time hours (between 6 am and 6 pm), and in executive, managerial, or professional occupations.

This Study

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

The study reported here expands on this earlier research (Morris, 2012) by examining the relationship between work and psychological well-being among mothers and fathers as well as single and partnered/married parents. In addition, this study also addresses a limitation of the earlier one by employing a more precise operationalization of child disability with which to identify caregiver parents. Specifically, restricting the definition to include only parents of children with functional limitations improves the validity of the findings and enables a more reliable comparison to other work–family studies. This study also tests whether parent-role stress moderates the relationship between work and parent mental health, thus providing a more nuanced analysis of the work–family dynamics of caregiver parents. Last, the potentially biasing effects caused by simultaneity between mental health and employment are addressed differently, this time by employing two-stage least squares instrumental variables estimation techniques and comparing the results to those obtained using the standard, single equation linear models with interaction effects.

The primary question guiding this study examined whether the relationship between work and parent mental health is different for caregiver parents (those with children with disabilities) and parents of typically developing children. Based on role theory and the balance of findings from the existing research, it was expected that work would have larger mental health benefits for parents of children with disabilities compared to other parents. Due to gender differences in role demands and salience, the dynamics between work, caregiving, and mental health were expected to be different for mothers and fathers. Specifically, it was expected that beneficial mental health effects from work, if present, would be smaller for fathers than mothers because fathers typically engage in less caregiving. Research shows that despite substantial increases in labor force participation among women with children and some smaller increases in fathers' involvement in child-raising, women continue to be the primary person in charge of child and family caregiving (Bianchi & Milkie, 2010; Coltrane, 1996; Harrington, Deusen, & Humberd, 2011; Parish, 2006). Moreover, though most mothers, even those with very young children, work outside the home, fathers still tend to work more hours (U.S. Bureau of Labor Statistics, 2012). In addition, fathers may face workplaces that are less family friendly compared to mothers (Harrington et al., 2011; Hill, 2005; Williams, 2010). That said, this expectation was tempered by research showing that, despite high rates of employment among mothers, sociocultural ambivalence regarding the appropriateness of working mothers (Usdansky et al., 2012; Williams, 2000, 2010) persists, especially for mothers of children with disabilities (Lewis et al., 1999).

A second question examined whether the effects of work and family caregiving are different for partnered and single parents. Although it was expected that the effects of work would differ significantly for partnered and single parents, the direction of difference could not be proposed a priori. Although research shows that among married couples, mothers do relatively more caregiving, there is still some sharing of family and home responsibilities within parenting couples (Coltrane, 1996; Essex & Hong, 2005; Lewis et al., 2000; Parish, 2006). Therefore, it might be expected that psychological benefits from work would be smaller for single parents because of greater work–family conflict offsetting the mental health benefits of work. However, it may be that heavier caregiving responsibilities among single parents lead to greater parent-role stress, which in turn may create more of a need for a break from caregiving and, hence, larger positive spillover effects.

A third question investigated whether the relationship between work and parent mental health differed according to specific job conditions. Specifically, it was hypothesized that any caregiver-specific mental health benefits from work would be offset by greater work–family conflict at higher levels of work (hours per week). Because working irregular shifts has been found to be associated with increased stress and depression (Campione, 2008), it was also predicted that parents of children with disabilities who work daytime shifts would benefit more from work than caregivers who work shifts outside 6 am to 6 pm. Studies also show that women in higher-level jobs have greater job and life satisfaction (Campione, 2008) and lower work–family strain (Swanberg, 2005), and that women employed in entry-level, nonprofessional occupations have more rigid schedules (Golden, 2008; McCrate, 2002). A study by Wellington (2006) showed that women with greater family-care burdens were more likely to be self-employed, presumably because of greater control over workload, hours, and schedule. Therefore, the analysis also tested whether the relationship between work and parent mental health differed according to parent's occupation and self-employment status.

The final question investigated whether parent-role stress moderates the relationship between work and mental health among parents of children with disabilities (Lin et al., 2012; Warfield, 2001, 2005). Although it was expected that the effect of working outside the home would be different for caregiver parents reporting high levels of stress, the direction of effect was not clear. On the one hand, high levels of parent-role stress may reflect heavier caregiving duties and greater work–family conflict, which in turn may offset the beneficial effects from working outside the home. On the other hand, parents experiencing higher levels of parent-role stress may have a greater need for a break from caregiving, and, hence, experience larger respite effects.

Method

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

Data and Sample

Data used in the analysis came from the National Survey of American Families (NSAF), a representative sample of the U.S. civilian population of children and adults younger than age 65 (see Urban Institute and Child Trends, 2002). The survey was administered by the Urban Institute in 1997, 1999, and 2002 to three cross-sectional samples of approximately 40,000 households each. In households with children younger than age 18, up to two children were sampled for in-depth data collection: one younger than age 6 and one between ages 6 and 17. Interviews were conducted with “the most knowledgeable adult” (MKA), defined as the adult in the household who was most knowledgeable about the sampled child's well-being, health, and education. The parent respondent was most often, though not always, the mother or the father of the child (78% of parent respondents were mothers, 17% were fathers, and 5% were some other adult in the household). In addition to answering the questions about the sampled child(ren), questions were also asked about the parent respondent and his or her spouse or partner if the spouse or partner lived in the same household. Data were extracted from all three waves of the NSAF and merged to create a pooled, cross-sectional sample of families. The sample used here was restricted to biological, custodial parents identified as the MKA to the sampled child(ren), including 63,075 mothers and 14,599 fathers. (Note: the married/partnered mothers and fathers used in this study are not married to each other; they are the MKAs from different families.)

Measures

Dependent variable

Parent mental health was based on a scale rating the severity of anxiety and depression symptoms using a 5-item mental health instrument that asked how often in the past month (1 = all of the time, 2 = most of the time, 3 = some of the time and 4 = none of the time) the respondent felt (a) very nervous, (b) calm or peaceful, (c) downhearted and blue, (d) happy, and (e) so down in the dumps that nothing could cheer him or her up. Responses to the questions about being nervous, sad, and downhearted were reverse coded and summed to create a score ranging from 5 to 20, with higher scores indicating greater mental and emotional well-being. The NSAF measure, adapted from a 38-item scale used in the Medical Outcomes Study (MOS), performed well on psychometric assessments of internal reliability and construct validity as well as benchmark comparisons using the MOS sample and a general population sample (Ehrle & Anderson Moore, 1997). The NSAF also benchmarks poor mental health. A measure was created by summing the responses to the five mental health items and multiplying by 5 to create a score (ranging from 25–100, with higher scores indicating better mental health) calibrated to a 100-point scale used by the MOS. The MOS defined a cutoff for “poor mental health” as scores 67 and below, based on the lowest 19% (coming as close as possible to the bottom quintile) of a general population sample. When using the same cutoff point in the NSAF sample, the lowest 17% of the sample were identified as having poor mental health, a result similar to the MOS result based on a general population sample (see Ehrle & Anderson-Moore, 1997).

Independent variables

Employment status was based on the parent's answer (yes or no) to the NSAF question: “Are you currently working? Hours per week was based on the question: “Considering all the jobs you have right now (including self-employment), how many hours per week on average do you work?” A shift variable was also included to compare the mental health of parents who worked regular day-time hours between 6 am and 6 pm to those who worked nonregular shifts. A variable identifying those respondents who worked in executive, supervisory and professional occupations enabled the comparison to those who work in other occupations. A variable flagging self-employed parents was used to test whether these parents experienced greater mental health benefits from work compared to other working parents.

Child disability status

All NSAF parent respondents were asked of each randomly selected child (one younger than age 6 and one between ages 6–17) the following question: “Does (CHILD) have a physical, learning, or mental health condition that limits (his/her) participation in the usual kinds of activities done by most children (his/her) age/or limits (his/her) ability to do regular school work?” The question reflects the federal definition of disability that requires that the condition result in significant functional limitations. This was used to create two dummy variables identifying parents who have a young child (age 0–5) with a disability and parents who have an older child (age 6–17) with a disability to compare them to parents with typical children.

Parent-role stress

The NSAF measured parent-role stress created by summing 4-point scale responses (1 = all of the time, 2 = most of the time, 3 = some of the time, 4 = none of the time) to the following four items: “How much during the past month have you felt (a) your (child/children) much harder to care for than most, (b) your (child/children) does/do things that really bother you a lot, (c) you are giving up more of your life to meet your (child's/children's) needs than you expected, and (d) angry with your (child/children)?” Responses were summed to create a scale score ranging from 4 to 16, with higher scores indicating lower levels of parenting stress and aggravation. The NSAF also benchmarks high parent-role stress. A score less than or equal to 11 identifies parents experiencing high levels of parent-role stress. The continuous measure and this indicator variable were used to test whether the relationship between work and mental well-being was influenced by the level of parent-role stress. The NSAF adapted the scale from a measure used in the National Evaluation of Welfare-to-Work Strategies (NEWWS) project and assessed the psychometric properties by comparing outcomes in the NSAF sample with the control group used in the NEWWS evaluation. Given the substantial differences between the NSAF and the NEWWS samples and how the scales were used in each survey, point estimates were not comparable; however, NSAF staff did examine patterns among socioeconomic subgroups and the outcomes were similar across both samples: parent-role stress increased with poverty, single parenthood, and lower parent education. Based on the strength of the sociodemographic subgroup patterns in the NSAF data and their comparability with patterns in the Jobs Opportunity and Basic Skills Program/NEWWS control group data, the measure's psychometric properties were deemed sufficiently valid and reliable (Ehrle & Anderson Moore, 1997).

Control variables

A number of demographic variables were included as controls and to examine the conditions under which specific relationships between work, family, and mental well-being occur. Demographic variables included the parent's relationship status, education, age, and race, as well as household income and the number of children age 0 to 5 and the number age 6 to 17. The NSAF asks respondents about their marital and partner status, thus enabling the identification of parents whose spouse or partner lives with them. Income was controlled using a categorical variable measuring annual household income relative to the federal poverty line based on money income received in the prior year by each person in the household age 15 and older, including nonmarried partners. Sources of income counted include money wages or salary, net income from self-employment, Social Security, Supplemental Security Income, interest and dividends, alimony and child support, Veteran's benefits, and unemployment insurance and worker' compensation payments. A variable flagging those parents who reported their current health as excellent or good (as opposed to fair or poor) was also included in all multivariate models. Last, a variable indicating whether the parent receives employer-provided health insurance was included as a control. Previous research found a significant association between employer-provided health benefits and higher mental well-being among mothers of children with disabilities, health conditions, and behavior problems, although the link was largely due to the strong correlation between full-time hours and the receipt of health benefits (Morris, 2012).

Analysis Plan

A series of regression equations with interactions were estimated to examine the relationships between work, caregiving, and parent mental health and address the research questions. To investigate whether work impacted the psychological well-being of parents of children with disabilities differently than parents of typically developing children, interactions between the child's disability status by age (0–5 years and 6–17 years) and the parent's employment status and job conditions were tested. Three-way interactions between the parent's gender, caregiver status (as indicated by their child's disability status), and employment status were used to assess whether work and caregiving affected the mental health of mothers and fathers differently. Regression models with three-way interactions between the parent's relationship status, caregiver status, and employment status were estimated on subsamples of mothers and fathers to assess if work and caregiving affected married or partnered and single parents differently. To test the hypothesis that the relationship between work and mental health is moderated by parent-role stress, regressions using interactions between the parent's employment status and a dummy variable identifying those parents experiencing high levels of parent-role stress were estimated. Formal statistical testing (incremental F test, Wald test) was conducted to confirm that interaction effects were statistically significant.

Although the intent of the analysis was to examine the effect of work on parent mental health, it is likely that there is a reverse relationship between work and parent mental health (i.e., anxiety and depression causing the parent to cut back on hours or leave the workforce altogether). A strong reciprocal relationship between mental health and employment would inflate positive spillover effects and deflate negative spillover effects (for a more extensive discussion, see Morris, 2012). To gauge the bias caused by simultaneity between parental mental health and employment, instrumental variables estimation was employed, and the results compared to those obtained using the basic interaction models described earlier. Employment status (for both main and interaction effects) was instrumented using the NSAF variable “number of weeks worked last year.” The variable was highly correlated with current employment status but only very weakly correlated with current mental health status. Moreover, the number of weeks worked in the prior year was highly correlated with the endogenous variable (current employment status) even after all the other independent variables were controlled for. The adjusted R2 for the first stage equation regressing employment on all the independent variables was 0.115, while the adjusted R2for the equation with weeks worked last year to the model was 0.540, indicating the variable was a suitable instrument for current employment status. Other instrumental variables were tested including current state unemployment rate and variables measuring the parent's attitudes toward welfare and work. Using Stata's instrumental variables diagnostics (estat endogenous, estat firststage, and estat overid) it was determined that the variable “number of weeks worked last year” performed better than the others in terms of strength and validity.

All models were estimated using survey commands in Stata (StataCorp, 2007) to adjust for the effects of NSAF's complex sample design. Like most large surveys, the NSAF is based on a stratified cluster sample rather than a pure random sample (Brick, Ferraro, & Strickler, 2004). Taylor-series linearization was used to compute standard errors adjusted for the sample design effects. Stata survey data analysis options are useable for regular linear regression as well as instrumental variables regression.

Results

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

Descriptive Statistics

Table 1 displays (weighted) sample characteristics and the descriptive statistics for the variables used in the regression analyses. Fathers were, on average, older, less poor, more educated, and more likely to be married or partnered, compared to mothers. Overall employment rates were higher for fathers compared to mothers, and fathers worked more hours. Mothers and fathers of children with disabilities were somewhat less likely to be working, but among employed parents, the effect of having a child with disabilities on hours worked was relatively small, and insignificantly so among mothers. Mothers reported higher levels of depression and anxiety, compared to fathers; they also ranked higher on the parent stress scale. As expected, parents of children with disabilities, especially parents of older children with disabilities, reported higher levels of depression and anxiety and parent stress.

Table 1. Sample Description by Parent's Gender and Employment Status and Child Disability Status
VariableMothers
All mothersMothers of typically developing childrenMothers of children with disabilitiesWorking mothers of typically developing childrenNonworking mothers of typically developing childrenWorking mothers of children with disabilitiesNonworking mothers of children with disabilities
% or M (SD)% or M (SD)% or M (SD)% or M (SD)% or M (SD)% or M (SD)% or M (SD)
Parent and household characteristics
Mental health score (5–20, higher score means less depression and anxiety)15.8115.9514.7216.0315.7715.0414.20
 (2.66)(2.52)(3.10)(2.44)(2.79)(2.89)(3.33)
Poor mental healtha19%17%33%15%21%28%40%
Age (years)35.5735.4736.5036.2333.937.6034.90
 (7.60)(7.84)(7.43)(7.56)(8.16)(7.05)(7.71)
Race (1 = White, 0 = all other)80%80%79%79%82%81%77%
Has bachelor's degree22%23%14%26%17%17%9%
Has no high school diploma29%28%38%23%39%31%47%
Partnered/married61%68%57%67%72%58%55%
Number of children age 0 to 5 (number)0.660.680.520.550.940.370.73
 (0.81)(0.81)(0.77)(0.73)(0.91)(0.65)(0.88)
Number of children 6 to 17 (number)1.321.281.611.301.251.591.65
 (1.07)(1.07)(1.01)(1.00)(1.19)(0.87)(1.20)
Parent in good to excellent health (1 = yes, 0 = fair or poor health)90%91%77%93%88%83%67%
Family income (relative to poverty line)b2.742.782.383.032.302.741.84
 (1.27)(1.26)(1.28)(1.15)(1.32)(1.19)(1.23)
Has child age 0 to 5 with a disability2%na17%nana12%25%
Has child age 6 to 17 with a disability9%na0.83nana0.880.75
Parent-role stress score (4 to 16, higher score means more stress)6.055.956.925.935.996.876.99
(1.86)(1.76)(2.30)(1.72)(1.84)(2.26)(2.37)
Highly stressed (1 = high parent-role stress, 0 = moderate to low or no stress)9%8%21%7%9%20%21%
Currently employed66%67%60%    
Employed parents' job conditions
Self-employed9%10%9%10% 9% 
Typical hours per week 36.78  36.70 37.52 
(average)(12.25)  (12.30) (12.63) 
Works 1 to 34 hours/week29%30%27%30% 28% 
Works 35 to 49 hours/week58%58%58%58% 57% 
Works 50 plus hours/week12%12%15%12% 15% 
Works during day (between 6 am to 6 pm)83%83%79%83% 79% 
Executive, managerial, or professional occupation33%33%28%34% 28% 
Receives health insurance benefits from employer77%77%71%77% 71% 
Sample size63,04055,8127,04238,16017,6024,3722,667
 Fathers
All fathersFathers of typically developing childrenFathers of children with disabilitiesWorking fathers of typically developing childrenNonworking fathers of typically developing childrenWorking fathers of children with disabilitiesNonworking fathers of children with disabilities
  1. Source: National Survey of American Families (1997, 1999, 2002 waves).

  2. Notes: Data adjusted for complex sampling design of the NSAF.

  3. a

    Parents in poor health have raw mental health scores of below 13.4. The cutoff is based on the mental health scores of the bottom approximately 20% of the U.S. population (see Erhle & Anderson-Moore, 1997).

  4. b

    Family income as a percentage of poverty: 0.5 (less than 50%), 1 (between 50% and 100%), 1.5 (between 100% and 150%), 2 (between 150% and 200%), 3 (between 200% and 300%), 4 (300% or more). All mother-father differences are statistically significant to at least the p < .10 level except for number of children age 6 to 17 years, percent White, percent who work day shift, percent who work 35 to 49 hours per week, and percent who work in executive, managerial, or professional occupations. Among mothers, all differences between those with and without a child with disabilities are statistically significant to at least the p < .10 level except for percent White, percent self-employed, and percent who typically work 35 to 49 hours per week. Among mothers of typically developing children, all differences between those working and those not working are statistically significant to at least the p < .10 level. All differences between working and nonworking mothers of children with disabilities are statistically significant to at least the p < .10 level except the percent experiencing high levels of parent-role stress. Among fathers, all differences between those with and without a child with disabilities are statistically significant to at least the p < .10 level except for percent self-employed. Among fathers of typical children, all differences between those working and those not working are statistically significant to at least the p < .10 level except for the number of children age 6 to 17. All differences between working and nonworking fathers of children with disabilities are statistically significant to at least the p < .10 level except for the number of children age 0 to 5 and the number of children age 6 to 17.

Parent and household characteristics
Mental health score (5 to 20, higher score means less depression and anxiety)16.6516.7015.8816.8215.4716.1014.08
(2.45)(2.40)(2.97)(2.29)(3.06)(2.76)(3.47)
Poor mental healtha12%11%21%9%29%18%36%
Age (years)39.1839.0840.5638.9540.4540.2243.53
 (8.09)(8.13)(7.26)(7.96)(9.66)(7.09)(9.29)
Race (1 = White, 0 = all other)81%80%87%81%72%87%81%
Has bachelor's degree33%34%27%35%20%30%11%
Has no high school diploma22%22%29%21%32%25%48%
Partnered/married78%79%70%80%62%74%55%
Number of children age 0 to 5 (number)0.570.590.440.590.490.430.41
 (0.73)(0.73)(0.70)(0.74)(0.71)(0.70)(0.73)
Number of children 6 to 17 (number)1.291.271.631.271.231.661.57
 (1.04)(1.03)(1.01)(1.03)(1.06)(1.03)(0.97)
Parent in good to excellent health (1 = yes, 0 = fair or poor health)93%94%85%95%79%89%61%
Family income (relative to poverty line)b3.183.212.833.292.363.031.87
 (1.07)(1.06)(1.19)(0.99)(1.30)(1.07)(1.19)
Has child age 0 to 5 with a disability1%na14%nana14%11%
Has child age 6 to 17 with a disability5%na86%nana86%89%
Parent-role stress score (4 to 16, higher score means more stress)5.725.676.475.665.836.406.96
 (1.67)(1.63)(2.02)(1.60)(1.94)(1.93)(2.22)
Highly stressed (1 = high parent-role stress, 0 = moderate to low or no stress)6%6%15%5%10%13%21%
Currently employed91%91%84%    
Employed parents' job conditions
Self-employed12%12%12%12% 12% 
Typical hours per week 46.92  47.03 44.91 
(average)(11.57)  (11.60) (10.79) 
Works 1 to 34 hours/week4%4%7%4% 7% 
Works 35 to 49 hours/week57%57%61%57% 61% 
Works 50 plus hours/week39%39%32%39% 32% 
Works during day (between 6 am to 6 pm)83%83%78%83% 78% 
Executive, managerial, or professional occupation34%34%27%35% 27% 
Receives health insurance benefits from employer83%83%79%83% 83% 
Sample size14,58613,5061,05612,2011,294889166

As discussed, the mental health measures contained in the NSAF are adapted from those used in the MOS. Compared to MOS results produced using a general population survey, the overall mean mental health score among mothers in the NSAF (15.8) is slightly above the average of the U.S. population (15.6) and the percentage in poor mental health is the same (19%). Fathers have, on average, slightly higher average mental health scores (16.7) than the general population (15.6) and are also less likely to be experiencing poor mental health (12% compared to 19%)1. The finding of lower mental health among mothers in general, relative to fathers, is consistent with other research (Voyandoff, 2002).

As predicted, parents of children with disabilities and nonworking parents have lower mental health scores than the general population average (15.6) and rates of poor mental health that are significantly higher than the general population (19%). Mothers of typically developing children scored, on average, 15.9, compared to mothers of children with disabilities, whose mean score is 14.7. Fathers of typically developing children scored, on average, 16.7, compared to fathers of children with disabilities, whose mean score is 15.9, which is slightly higher than the mean score (15.6) found among the general population. Nonworking mothers and fathers of children with disabilities have mental health scores close to the cutoff (13.4) designating poor mental health. Nonworking mothers of children with disabilities are twice as likely as the general population to be in poor mental health (40% compared to 19%). Nonworking fathers of children with disabilities are also much more likely to be in poor mental health (36%).

Table 1 provides preliminary evidence that working outside the home provides additional mental health benefits for parents of children with disabilities: though working parents in general scored higher on the mental health scale, the difference in mean mental health scores between working and nonworking parents of children with disabilities was larger than between working and nonworking parents of children who were typically developing. Among working mothers of typically developing children, the mean mental health score is 16.0, compared to 15.8 among those not working, whereas mean mental health scores are 15.0 and 14.2 for working and nonworking mothers of children with disabilities, respectively. Among working fathers of typically developing children, the mean mental health score is 16.8, compared to 15.5 among those not working, and mean mental health scores are 16.1 and 14.1 for working and nonworking fathers of children with disabilities, respectively. The same pattern emerges when observing rates of poor mental health. For example, the difference in the likelihood of being in poor mental health (a raw score of 13.4 or below) between working (15%) and nonworking (20%) mothers of typically developing children is 5 percentage points, compared to a difference of 12% between working (28%) and nonworking (40%) mothers of children with disabilities. A similar pattern is found among fathers, although the difference is smaller. In addition to providing preliminary evidence of positive spillover from work, altogether the findings indicate that the mental health of fathers suffers more from nonwork than the presence of a child with disabilities, a finding consistent with traditional gender norms and family dynamics.

Bivariate analysis (not included here in the interest of space but available from the author on request) also showed a significant and positive relationship between work and parent mental health (r = .10, p < .001), and weaker but still positive relationships between parent mental health and executive, managerial, or professional occupation (r = .07, p < .001); works 35 to 49 hours per week (r = .02, p < .001); works more than 50 hours per week (r = .03, p < .001); and day shift (r = .04, p < .001). Bivariate correlations also revealed a relatively strong negative relationship between parent mental health and parent-role stress (r = −0.36, p < .001), and a weak but significant negative relationship between parent-role stress and currently employed (r = –0.04, p < .001). Taken together, these results are consistent with the hypothesis that the relationship between work and mental health is influenced by the level of parent-role stress, and that caregiver parents who are highly stressed benefit more from working.

Regression Results

Regression results are displayed in Table 2. The study's main hypothesis predicted that parents of children with disabilities would experience greater mental health benefits (reduced depression and anxiety) from working outside the home than would parents of typically developing children. It was also expected that caregiver-specific positive effects would be larger for mothers compared to fathers. To test these hypotheses, three-way interactions between parent's employment status, gender, and variables identifying parents of children with disabilities (age 0–5–6–17) were added to regression models. The significant positive coefficient for the three-way interaction between employment status, female gender, and the presence of a child age 6 to 17 with a disability (ß = .943, p < .10, 95% confidence interval [CI]: [0.004, 1.881]) indicates that mothers of older children with disabilities benefit more from work than do other parents. Specifically, it is estimated that the increase in the mean mental health score among caregiver mothers of older children with disabilities as a result of going to work is about 1 point (or, based on the 95% CI estimates, between 0.004 and 1.881 higher). Results produced using the instrumental variables approach were largely similar to those produced by the linear interaction models, indicating that the findings suggestive of a positive spillover effect from work were not merely an artifact of simultaneity between parent mental health and employment status. For the sake of simplicity and in the interest of space, these results are not shown here.

Table 2. Ordinary Least Squares Estimates of the Association between Work, Caregiving, and Parent's Mental Health
VariablesFull sample: Mothers and fathersFull sample: Mothers and fathersMothersFathersEmployed mothersEmployed fathersParents of children w/disabilities
  1. Source: National Survey of American Families (1997, 1999, 2002 waves).

  2. Notes: Data adjusted for the complex sampling design of the NSAF. Beta coefficients are not standardized (standardized coefficients not available using Stata's survey commands). Standard errors are presented in parentheses. Sample excludes parents with more than one child with disabilities.

  3. a

    Family income as a percentage of poverty: 0.5 (less than 50%), 1 (between 50% and 100%), 1.5 (between 100% and 150%), 2 (between 150% and 200%), 3 (between 200% and 300%), 4 (300% or more).

  4. b

    Parent-role stress scale score (4–16) is coded such that higher scores are indicative of greater levels of stress/aggravation.

  5. c

    Reference category includes those who work less than 35 hours per week.

  6. d

    p < .10,

  7. e

    p < .05,

  8. f

    p < .01

Parent's age0.004e0.005e0.005e0.0020.002–0.001–0.013
(0.002)(0.002)(0.003)(0.005)(0.003)(0.005)(0.008)
Parent's race (White = 1)–0.398f–0.406f–0.443f–0.307e–0.496f–0.253e–0.588f
(0.045)(0.045)(0.052)(0.095)(0.058)(0.102)(0.168)
Parent's sex (female = 1)–0.432f0.216nananana–0.687e
(0.037)(0.135)    (0.190)
Parent has bachelor's degree0.140f0.139f0.217f–0.0580.116e–0.0220.183
(0.034)(0.034)(0.038)(0.073)(0.049)(0.086)(0.143)
Parent partnered/married0.430f0.419f0.783f–0.0370.289f0.387f0.401e
(0.041)(0.041)(0.087)(0.278)(0.055)(0.092)(0.141)
Parent in good to excellent health1.504f1.483f1.484f1.264f1.303f1.239f1.856f
(0.062)(0.062)(0.067)(0.169)(0.087)(0.189)(0.1573)
Household incomea0.200f0.199f0.1898f0.184f0.147f0.114f0.369f
(0.016)(0.016)(0.018)(0.037)(0.025)(0.042)(0.351)
Number of children age 0 to 50.201f0.193f0.196f0.164f0.139f0.158f0.153
(0.024)(0.024)(0.027)(0.059)(0.037)(0.058)(0.100)
Number of children age 6 to 170.102f0.010f0.099f0.090e0.094f0.113***0.005
(0.017)(0.017)(0.019)(0.038)(0.025)(0.039)(0.073)
Has disabled child age 0 to 5–0.327e0.166–0.0370.784–1.04e1.614dna
(0.132)(0.492)(0.473)(1.072)(0.407)(0.736) 
Has disabled child age 6 to 17–0.455f0.127–0.690e–0.864–0.374#*–0.273na
(0.060)(0.425)(0.187)(0.565)(0.231)(0.545) 
Parent-role stressb–0.466f–0.466f–0.476f–0.417f–0.456f–0.409fsee below
(0.010)(0.009)(0.011)(0.021)(0.012)(0.022) 
Currently employed0.165f0.770f0.406f0.625fnana0.325e
(0.038)(0.135)(0.084)(0.244)  (0.147)
Interactions between parent employment status and child disability status
Female e Employed –0.712f     
 (0.141)     
Female e Disabled child 0 to 5 –0.477     
 (0.536)     
Female e Disabled child 6 to 17 –0.999d     
 (0.438)     
Employed e Disabled 0 to 5 –0.017–0.447–1.048   
 (0.599)(0.600)(1.690)   
Employed e Disabled 6 to 17 –0.3580.417e0.164   
 (0.460)(0.223)(0.727)   
Female e Employed e Disabled 0 to 5 –0.142     
 (0.656)     
Female e Employed e Disabled 6 to 17 0.943e     
 (0.478)     
Interactions between parent employment status, relationship status, and child disability status
Married/Partnered e Employed  –0.488f0.382   
  (0.093)(0.291)   
Married/Partnered e Disabled 0 to 5  –0.067–1.048   
  (0.5158)(1.328)   
Married/Partnered e Disabled 6 to 17  –0.0731.515d   
  (0.222)(0.855)   
Married/Partnered e Disabled 0 to 5 e Employed  0.2011.472   
  (0.624)(1.898)   
Married/Partnered e  0.077–0.989   
Disabled 6 to 17 e Employed  (0.274)(0.987)   
Parent employment conditions
Self-employed    0.237f–0.243d 
    (0.063)(0.126) 
Executive, managerial, professional occupation    0.077–0.039 
    (0.048)(0.086) 
Day shift (between 6 am and 6 pm)    0.106d–0.006 
    (0.059)(0.097) 
Has employer-provided health insurance    0.270f0.216d 
    (0.059)(0.115) 
Works 35 to 49 hours/weekc    –0.0260.168 
    (0.048)(0.169) 
Works 50 plus hours/weekc    –0.143e0.001 
    (0.071)(0.173) 
Interactions between parent employment conditions and child disability status
Self-employed e Disabled 0 to 5    0.1760.412 
    (0.489)(0.886) 
Self-employed e Disabled 6 to 17    0.0130.971 
    (0.286)(0.670) 
Executive/management/professional e Disabled 0 to 5    0.2270.515 
    (0.308)(0.637) 
Executive/management/professional e Disabled 6 to 17    0.153–0.509 
    (0.162)(0.349) 
Day shift e Disabled 0 to 5    0.133–1.359e 
    (0.304)(0.628) 
Day shift e Disabled 6 to 17    0.073–0.798e 
    (0.223)(0.367) 
Employer-provided health insurance e Disabled 0 to 5    0.323–0.257 
    (0.373)(0.887) 
Employer-provided health insurance e Disabled 6 to 17    0.2130.632 
    (0.176)(0.653) 
Works 35 to 49 hours/week e Disabled 0 to 5    0.377–0.531 
    (0.339)(0.994) 
Works 35 to 49 hours/week e Disabled 6 to 17    –0.1830.045 
    (0.177)(0.510) 
Works 50plus hours/week e Disabled 0 to 5    0.236–0.539 
    (0.452)(1.048) 
Works 50plus hours/week e Disabled 6 to 17    –0.631e0.446 
    (0.252)(0.484) 
Interactions between parenting stress and employment status
High parent stress      –0.216f
      (0.895)
High parent stresse employed      0.508d
      (0.315)
Sample size77,62677,62662,81414,56142,61713,1076,792
R2 (adjusted)0.21000.21390.21080.15110.18090.12630.2318

To assess whether the combined effects of work and caregiving are different for married or partnered and single parents, models with three-way interactions between parent's relationship status, employment status, and child disability status were estimated on subsamples of mothers and fathers.2 Results displayed in columns 3 (mothers) and 4 (fathers) of Table 2 indicate that no significant additional benefits from work accrued to married/partnered mothers of children with disabilities.

Table 2 also displays the results from ordinary least squares regressions examining the relationship between different job conditions (hours, occupation, shift) and parent mental health using subsamples of employed mothers and fathers. To test the hypothesis that mental health benefits from work disappear at high levels of work, a number of different hours per week specifications were estimated to determine at what point mental health benefits from work might be offset by time constraints and work–family conflict. Among mothers, none of the interactions between job conditions and child disability status was found to be statistically significant, except for the interaction identifying mothers of older children with disabilities who typically work 50 or more hours per week, which was negatively associated with mental well-being. Among fathers, none of the interactions between job conditions and child disability status were significant, except those identifying fathers of disabled children (age 0–5 and 6–17) who worked during the day, which were negatively associated with mental well-being.

The final column in Table 2 displays regression results investigating the relationship between parent's employment status and level of parent-role stress, using a subsample of parents of children with disabilities. The significant positive coefficient on the two-way interaction between parent's employment status and high parent-role stress (ß = .508, p < .10, 95% CI [–0.114, 1.130]) indicates that these parents experience additional beneficial effects from working outside the home. Specifically, working caregiver parents who are experiencing high levels of parent-role stress score about a half point higher on the mental health scale (or, according to the 95% CI estimates, between 0.01 point lower and 1.13 points higher).

Discussion

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

This study adds to the existing work–family literature in two important ways. First, it examined the relationships between caregiving, work, and mental health among mothers and fathers using a nationally representative sample. Second, by investigating whether parent-role stress attenuates the relationship between work and parent mental health, it provided a more thorough analysis of the work–family dynamics of caregiver parents. In general, the results showed some evidence of a positive effect from work, at least among mothers of older children with disabilities. The results also suggested that the psychologically harmful effects of high parent-role stress may be offset, at least somewhat, by working outside the home.

Using three-way interactions between parent's employment status, gender, and variables identifying parents of disabled children (age 0–5 and 6–17), the analysis confirmed the hypothesis for mothers of older children with disabilities that work provides additional mental health benefits. This result is consistent with respite effects from work found in earlier, smaller studies. Importantly, the psychological benefits from work do not appear to be due to mothers of children with disabilities working fewer hours. Mothers of children with disabilities were somewhat less likely to be employed; when employed, however, they worked the same number of hours, on average, as mothers of children who are typically developing (see Table 1). Moreover, the results of this study indicate that the beneficial effects from work for mothers of older children with disabilities disappear only at high levels of work (50 or more hours per week).

The analysis also investigated whether the relationship between work and mental health was moderated by parent-role stress. Not surprisingly, the results showed a strong negative relationship between parent-role stress and mental well-being. However, work does not appear to make things worse. In fact, the significant two-way interaction between parent's employment status and the variable signaling high levels of parent-role stress indicates that highly stressed parents experience larger benefits from working outside the home, compared to caregiver parents who are experiencing less stress. This is consistent with the hypothesis that parents experiencing higher levels of parent-role stress have a greater need for a break from caregiving, and, hence, experience larger respite effects.

The finding that only mothers of older children with disabilities benefit from work may reflect a greater need for respite, and/or sources of satisfaction outside the home, despite higher levels of work–family conflict (Brennan & Brannan, 2005; Parish, 2006). This result might also reflect concerns about leaving younger children, especially those with disabilities, in the care of others (Lewis et al., 1999). Another possibility is that younger children with disabilities require more therapeutic or medical treatments, doctor's visits, and other appointments, thus increasing work–family conflict. In other words, work itself might be respite, but the care coordination aspects of the child's condition may offset any positive spillover.

The finding that fathers of children with disabilities do not experience additional psychological benefits from work is consistent with less involvement in caregiving, however, this finding might also reflect higher work–family conflict. Table 1 shows that fathers work more hours than mothers, and that though fathers of children with disabilities work fewer hours than other fathers, they still work, on average, 45 hours per week. In comparison, working mothers of children with disabilities worked, on average, 37 hours per week. It could also be that fathers face a less family-friendly workplace, with employer expectations of the “ideal worker,” a person unencumbered by family responsibilities, greater for fathers compared to mothers (Harrington et al., 2010; Hill, 2005; Williams, 2010).

It was tentatively proposed that married and partnered parents of children with disabilities would benefit more than single parents of children with disabilities from working outside the home because sharing of home and family responsibilities would reduce work–family conflict. Indeed, bivariate analysis did show that married/partnered parents experienced lower levels of parent-role stress. However, the regression results did not support the hypothesis: the interaction between parent's relationship status, parent's employment status, and child disability status was not found to be significantly different from zero. In other words, though mothers of older children in general experience positive mental health effects from work, there does not appear to be a significant difference in the benefits accruing to married/partnered and single mothers. The absence of difference here might reflect a lack of significant sharing of home and family duties among married and partnered parents. All of the mothers in the focus group study by Parish (2006) reported that they bore a disproportionate amount of the caregiving responsibilities, even those who were married or partnered. Using a nationally representative sample of parents, Minnotte (2012) did not find reduced levels of work–family conflict among partnered parents. She speculated that competing work–family responsibilities and unmet expectations about sharing home and family responsibilities created stress and frustration that canceled out the potential benefits of “an extra pair of helping hands.”

Altogether, the results of this study indicate that work is not necessarily harmful to the mental well-being of parents of children with disabilities and may even produce significant positive spillover effects. It should also be noted that the overall mean mental health score among nonworking parents of children with disabilities was above the “poor mental health” cutoff found in the general population (13.4), indicating a special resilience among these parents. If there is a mental health-enhancing spillover effect from work, supporting parents of children with disabilities in work and reducing work–family conflict has importance beyond the obvious short-term and long-term economic impacts (Porterfield, 2002; Powers, 2002; Stabile & Allin, 2012). In addition to compromising the mental health of parents, caregiver mental and emotional distress can interfere with the parent–child relationship and undermine children's healthy emotional, cognitive, and physical development (Deater-Deckard, 2005; Downey & Coyne, 1990; Ehrle & Anderson-Moore, 1997; Perry-Jenkins & Gillman, 2000; Walker, Ortiz-Valdes, & Newbrough, 1989). Public and employer support to assist working parents of children with disabilities to help them maintain employment may be justified (Boushey, 2011; Hill, Martinson, Ferris, & Baker, 2004; Leiter, Krauss, Anderson, & Wells, 2004; Williams, 2000). Compensatory time, alternative scheduling, and compressed work weeks as well as schedule predictability help working parents and caregivers resolve work–family conflicts (Boushey, 2011). Of course, such benefits will be supportive only to the extent that the parents feel that their commitment to work will not be questioned if they avail themselves of such family-friendly options. Right-to-request legislation is needed to protect working caregivers who request alternative schedules and to encourage employers to accommodate such requests (Boushey, 2011). In addition, research finds that the work–family problems facing parents of children with disabilities are often related to difficulties accessing child care. Mothers of children with disabilities express concern about the costs (Lukemeyer, Meyers, & Smeeding, 2000) and quality of child care available to their child (Lewis et al., 1999; Parish, 2006). For families with older children with disabilities, schools are an important component of care (Epstein et al., 2005). Therefore, support might also include grants to schools to increase the supply of care options and subsidies to offset higher costing child care for older children with exceptional needs (Lukemeyer et al., 2000).

Positive spillover is a plausible explanation for the results reported here. Nonetheless, the study is subject to a number of limitations. First, I speculate that working outside the home provides a break from caregiving, which in turn has a positive impact on the caregiver's psychological well-being; however, there are alternative or additive explanations. For example, Greenhaus and Powell (2006) posit that a parent's involvement in work enhances her caregiving through the acquisition of resources (e.g., skills, knowledge, material resources, social support, etc.), which in turn improves her mental and emotional well-being. More research is needed to determine whether gains in psychological well-being are direct (i.e., work as respite from caregiving reduces stress) or indirect (i.e., improved mood comes from enhanced efficacy in caregiving, which in turn comes from developmental gains from work). The policy and practice implications differ considerably depending on the mechanism by which participation in work produces positive mental health impacts, and the mechanisms may not be universal.

As well, the field has yet to sufficiently identify how these differences might vary among subgroups of families and/or according to community characteristics, family resources, and the presence and quality of supports. Controlling for differences in coping skills and social support, informal (family, relatives, friends) and formal (social services), would improve the analysis (McDonald et al., 1999; Perry-Jenkins, Repetti, & Crouter, 2000). More precise measurement of the actual division of labor within the household and across extended family and noncustodial parents as well as satisfaction with that division of labor (Essex & Hong, 2005) is also needed. It should also be noted that the regression results produced here show a strong, independent association between mental well-being and income, with poorer parents experiencing significantly higher rates of anxiety and depression. There is also a strong link between parent-role stress and income, with poor parents experiencing significantly higher stress (Ehrle & Anderson-Moore, 1997). Because the challenges facing low-income parents of children with disabilities are likely to be considerably greater than those facing other parents, the impacts of work and the types of work–family supports needed by these families warrants particular attention by researchers, policy analysts, and practitioners.

The analysis would also benefit from control of child care quality and cost. Concern about the quality of child care may have a moderating effect on the mental health benefits from work (Gordon, Usdansky, Wang, & Gluzman, 2011). Controlling for child care expenses would also improve the analysis. The NSAF did not query parent's perceptions of child care quality but did ask parents about child care use, type, and expense, so these variables were not included in the current analysis. However, the questions vary somewhat across waves (1997, 1999, and 2002) and too many cases were missing information, especially about child care expense. Control of household income relative to poverty and the categorization of child disability status according to age (0–5 years and 6–17 years) should control for the effect of child care expenses to some degree, and to a lesser, more indirect way, child care quality.

Caution must also be taken when interpreting and using these findings due to the reciprocal relationship between work and parent mental health. Measures were taken to confirm mental health benefits from work separate from the endogenous effects between work and parent mental health, but the methodology is still limited by the fact that the data are cross sectional. Additionally, the disability measure available in the NSAF is relatively crude. It is based on only one question and does not account for the type or severity of the child's disability, detailed information about functional status, or the absence or presence of behavior problems. Additional research is needed to capture the complexity of child needs, the extent of parent caregiving involved, and the nature of interactions among these factors and sources of support. The pooling of samples across the three waves of the NSAF produced larger samples than used in previous studies; however, the number of single parents, especially fathers, with a child with special needs was small, making it difficult to quantitatively verify whether work has beneficial or harmful mental health effects.

Last, it should also be noted that estimated effect of going to work on the mental health of caregiver mothers is modest. The point estimate on the three-way interaction between employment status, female gender, and child-disability status indicates that participation in work increases the mean mental health score among mothers of disabled children by just under 1 point (ß = .943) whereas the 95% CI indicates the effect of work on average mental health scores could be nearly nonexistent or nearly 2 points. The average mental health score among the nonworking mothers of children with disabilities in this sample was 14.2; a 1-point increase to about 15 brings the mean score among caregiver mothers closer to the population average of 15.6 and reduces by about 5% the number in poor mental health. Although these predicted improvements in maternal mental health are certainly hopeful, the role of work as a solution to caregiver stress should not be overstated.

Parents of children with disabilities may benefit from going to work, but work–family conflict has an independent impact on parent mental health that should not be ignored (Grzywacz & Bass, 2003). Indeed, though this study found evidence suggesting that working outside the home may somewhat offset parent-role stress, stress was higher among parents of children with disabilities, compared to parents of typically developing children, and its negative relationship with parent mental health was strong. Mothers of older children with disabilities participating in Parish's (2006) study reported psychological benefits from work but also reported feeling isolated and depressed and overwhelmed by the stress of balancing work and caregiving. Freedman et al. (1995) found evidence of positive spillover, from work to family and family to work, especially among mothers of children with behavioral problems, but that parents (usually mothers) reduced their work hours because of difficulty finding appropriate and affordable child care and balancing work and caregiving responsibilities. Additionally, parents in the Freedman et al. study commented on working extra hard and “bending over backwards” or being “super achievers” to make up for the fact that they might need to ask for concessions or work–family accommodations (p. 510). This calls in to question the sustainability of positive spillover effects and highlights the potential limits to workplace policy changes; there is a corresponding need to examine the attitudinal and cultural influences on parent, employer, and coworker perceptions of dual roles and the strain they may produce. Workplace policies are but one of several contextual factors that combine to influence how individuals ultimately experience and manage their complex roles as parents, spouses/partners, employees, colleagues, and extended family members. There remain complex social expectations, gendered perceptions of caregiving and employment, and the pervasive influence of socioeconomic status on adaptation. Taken together, these factors influence employee productivity, parent stress, and the mental health of family members, and there are economic and human costs associated with each. Given the centrality of work–family balance in the lives of families, and the national commitment to family the fundamental social unit, it behooves us to better understand how policy and practice can support the optimal function of parents who have children with extraordinary caregiving needs.

Acknowledgment

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References

This research on which this paper is based was supported by funding from the Association for Public Policy Analysis and Management/NSAF Small Research Grants Program in collaboration with the Annie E. Casey Foundation.

References

  1. Top of page
  2. Abstract
  3. Conceptual Framework
  4. Previous Research
  5. This Study
  6. Method
  7. Results
  8. Discussion
  9. Acknowledgment
  10. References
  • Barnett, R. C., & Baruch, G. K. (1985). Women's involvement in multiple roles and psychological distress. Journal of Personality and Social Psychology, 49, 135145.
  • Barnett, R. C., & Marshall, L. N. (1992). Worker and mother roles, spillover effects, and psychological distress. Women Health, 18, 940.
  • Bianchi, S. E., & Milkie, M. A. (2010). Work and family research in the first decade of the 21st century. Journal of Marriage and Family, 72, 705725.
  • Boushey, H. (2011). The role of government in work-family conflict. The Future of Children: Work and Family, 21, 163190.
  • Brennan, E. M., & Brannan, A. M. (2005). Participation in the paid labor force by caregivers of children with emotional and behavioral disorders. Journal of Emotional and Behavioral Disorders, 13, 237246.
  • Brick, J. M., Ferraro, D., & Strickler, T. (2004). No 4: Variance estimation for the 2002 NSAF. Washington, DC: Urban Institute Press.
  • Campione, W. (2008). Employed women's well-being: The global and daily impact of work. Journal of Family and Economic Issues, 29, 346361.
  • Canning, R. D., Harris, E. S., & Kelleher, K. J. (1996). Factors predicting stress among caregivers to children with chronic medical conditions. Journal of Pediatric Psychology, 21, 735749.
  • Coltrane, S. (1996). Family man: Fatherhood, housework, and gender equity. New York: Oxford University Press.
  • Deater-Deckard, K. (2004). Parenting stress. New Haven, CT: Yale University Press.
  • Downey, G., & Coyne, J. C. (1990). Children of depressed parents: An integrative review. Psychological Bulletin, 108, 5076.
  • Ehrle, J., & Anderson Moore, K. (1997). NSAF benchmarking measures of child and family well-being. Report No. 6. Assessing the new federalism. Washington, DC: Urban Institute Press.
  • Epstein, M. H., Nordness, P. D., Gallagher, K., Nelson, J. R., Lewis, L., & Schrepf, S. (2005). School as the entry point: Assessing adherence to the basic tenets of the wraparound approach. Behavioral Disorders, 30, 8593.
  • Essex, L. E., & Hong, J. (2005). Older caregiving parents: Division of household labor, marital satisfaction and caregiving burden. Family Relations, 54, 448460.
  • Freedman, R. I., Litchfield, L., & Warfield, M. E. (1995). Balancing work and family: Perspectives of parents of children with developmental disabilities. Families in Society: The Journal of Contemporary Human Services, 76, 507514.
  • Frone, M. R. (2000). Work-family conflict and employee psychiatric disorders: The National Comorbidity Survey. Journal of Applied Psychology, 85, 888895.
  • Gareis, K. C., Barnett, R. C., Ertel, K. A., & Berkman, L. F. (2009). Work-family enrichment and conflict: Additive effects, buffering, or balance? Journal of Marriage and Family, 71, 696707.
  • Golden, L. (2008). Limited access: Disparities in flexible work schedules and work-at-home. Journal of Family and Economic Issues, 29, 86109.
  • Goodman, W. B., & Crouter, A. C. (2009). Longitudinal associations between maternal work stress, negative work-family spillover, and depressive symptoms. Family Relations, 58, 245258. doi:10.1111/j.1741-3729.2009.00550.x
  • Gordon, R. A., Usdansky, M. L., Wang, X., & Gluzman, A. (2011). Child care and mother's mental health: Is high-quality care associated with fewer depressive symptoms? Family Relations, 60, 446460.
  • Gottlieb, A. S. (1997). Single mothers of children with developmental disabilities: The impact of multiple roles. Family Relations, 46, 512.
  • Greehaus, J. H., & Powell, G. N. (2006). When work and family are allies: A theory of work-family enrichment. Academy of Management Review, 31, 7292.
  • Grzywacz, J. G., Almeida, D. M., & McDonald, D. A. (2002). Work-family spillover and daily reports of work and family stress in the adult labor force. Family Relations, 51, 2636.
  • Grzywacz, J. G., & Bass, B. L. (2003). Work, family, and mental health: Testing different models of work-family fit. Journal of Marriage and Family, 65, 248262.
  • Grzywacz, J. G., & Butler, A. B. (2005). The impact of job characteristics on work-to-family facilitation: Testing a theory and distinguishing a construct. Journal of Occupational Health Psychology, 10, 97109.
  • Grzywacz, J. G., & Marks, N. F. (2000). Reconceptualizing the work-family interface: An ecological perspective on the correlates of positive and negative spillover between work and family. Journal of Occupational Health Psychology, 5, 111126.
  • Hanson, G. C., Hammer, L. B., & Colton, C. L. (2006). Development and validation of a multidimensional scale of perceived work-family positive spillover. Journal of Occupational Health Psychology, 11, 249265.
  • Harrington, B., Van Deusen, F., & Humberd, B. (2011). The new dad: Caring, committed, and conflicted. Boston, MA: Boston College Center for Work and Family.
  • Hill, E. J. (2005). Work-family facilitation and conflict, working fathers and mothers, work-family stressors and support. Journal of Family Issues, 26, 793819. doi:10.1177/0192513X05277542
  • Hill, E. J., Martinson, V. K., Ferris, M., & Baker, R. Z. (2004). Beyond the mommy track: The influence of new-concept part-time work for professional women on work and family. Journal of Family and Economic Issues, 25, 121136.
  • Leiter, V., Krauss, M. W., Anderson, B., & Wells, N. (2004). The consequences of caring: Effects of mothering a child with special needs. Journal of Family Issues, 25, 379403.
  • Levine, J. A., & Pittinsky, T. L. (1997). Working fathers: New strategies for balancing work and families. Reading, MA: Addison-Wesley.
  • Lewis, S., Kagan, C., & Heaton, P. (2000). Dual-earner parents with disabled children: Family patterns for working and caring. Journal of Family Issues, 21, 10311061.
  • Lewis, S., Kagan, C., Heaton, P., & Cranshaw, M. (1999). Economic and psychological benefits from employment: The experiences and perspectives of mothers of disabled children. Disability and Society, 14, 561575.
  • Lin, I. F., Fee, H. R., & Wu, H.-S. (2012). Negative and positive caregiving experiences: A closer look at the intersection of gender and relationship. Family Relations, 61, 343358. doi:10.1111/j.1741-3729.2011.00692x
  • Lukemeyer, A., Meyers, M. K., & Smeeding, T. (2000). Expensive children in poor families: Out-of-pocket expenditures for the care of disabled and chronically ill children in welfare families. Journal of Marriage and Family, 62, 399415.
  • McCrate, E. (2002). Working mothers in a double bind: Working mothers, minorities have the most rigid schedules, and are paid less for the sacrifice (Briefing Paper). Washington, DC: Economic Policy Institute.
  • McDonald, T. P., Poertner, J., & Pierpont, J. (1999). Predicting caregiver stress: An ecological perspective. American Journal of Orthopsychiatry, 69, 100109.
  • Melchior, M., Berkman, L., Niedhammer, I., Zins, M., & Goldberg, M. (2007). The mental health effects of multiple work and family demands. Social Psychiatry and Psychiatric Epidemiology, 42, 573582.
  • Minnotte, K. L. (2012). Family structure, gender, and the work-family interface: Work-to-family conflict among single and partnered parents. Journal of Family and Economic Issues, 33, 95107.
  • Minnotte, K. L., Minnotte, M. C., Pedersen, D. E., Mannon, S., & Kiger, G. (2010). His and her perspectives: Gender ideology, work-to-family conflict and marital satisfaction. Sex Roles, 63, 425438.
  • Morris, L. A. (2012). Testing respite effect of work on stress among mothers of children with special needs. Journal of Family and Economic Issues, 33, 2440.
  • Parish, S. L. (2006). Juggling and struggling: A preliminary work-life study of mothers with adolescents who have developmental disabilities. Mental Retardation, 44, 393404.
  • Pedersen, D. E., Minnotte, K. L., Kiger, G., & Mannon, S. (2009). Workplace policy and environment, family role quality, and positive family-to-work spillover. Journal of Family and Economic Issues, 30, 8090.
  • Perry-Jenkins, M., & Gillman, S. (2000). Parental job experiences and children's well-being: The case of working-class two- and single-parent families. Journal of Family and Economic Issues, 21, 123147.
  • Perry-Jenkins, M., Repetti, R. L., & Crouter, A. C. (2000). Work and family in the 1990s. Journal of Marriage and Family, 62, 981998.
  • Porterfield, S. L. (2002). Work choices of mothers in families with children with disabilities. Journal of Marriage and the Family, 64, 972981.
  • Powers, E. (2002). New estimates of the impact of child disability on maternal employment. American Economic Review, 91, 135139.
  • Rosenzweig, J. M., Brennan, E. M., & Ogilvie, A.M. (2002). Work-family fit: Voices of parents of children with emotional and behavioral disorders. Social Work, 47, 415424.
  • Stabile, M., & Allin, S. (2012). The economics costs of childhood disability. Future of Children: Children with Disabilities, 22, 6596.
  • StataCorp. (2007). Stata statistical software: Release 10. College Station, TX: StataCorp LP.
  • Swanberg, J. (2005). Job-family role strain among low wage workers. Journal of Family and Economic Issues, 26, 143157.
  • Thyen, U., Kuhlthau, K., & Perrin, J. M. (1999). Employment, child care, and mental health of mothers caring for children assisted by technology. Pediatrics, 103, 12351242.
  • Tiedje, L. B., Wortman, C. B., Downey, G., Emmons, C., Biernat, M., & Lang, E. (1990). Women with multiple roles: Role-compatibility perceptions, satisfaction, and mental health. Journal of Marriage and Family, 52, 6372.
  • Urban Institute, and Child Trends. (2002). National Survey of America's Families (NSAF) (ICPSR04582-v1). Ann Arbor, MI: Inter-University Consortium for Political and Social Research3. doi:10.3886/ICPSR04582.v1
  • U.S. Bureau of Labor Statistics. (2012). Economic News Release - Table 4: Families with own children: Employment status of parents by age of youngest child and family type, 2011-2012, annual averages. Retrieved from http://www.bls.gov/news.release/famee.t04.htm
  • U.S. Department of Health and Human Services, Health Resources and Services Administration. (2008). National survey of children with special health care needs chartbook 2005-2006. Retrieved from from http://mchb.hrsa.gov/cshcn05/
  • Usdansky, M. L., Gordon, R. A., Wang, X., & Gluzman, A. (2012). Depression risk among mothers of young children: The role of employment preferences, labor force status, and job quality. Journal of Family and Economic Issues, 33, 8394.
  • Voyandoff, P. (2002). Linkages between work-family interface and work, family, and individual outcomes: An integrative model. Journal of Family Issues, 23, 138164.
  • Voyandoff, P. (2004). The effects of work demands and resources on work-to-family conflict and facilitation. Journal of Marriage and Family, 66, 398412.
  • Walker, L. S., Ortiz-Valdes, J. A., & Newbrough, J. R. (1989). The role of maternal employment and depression in the psychological adjustment of chronically ill, mentally retarded, and well children. Journal of Pediatric Psychology, 14, 357370.
  • Warfield, M. E. (2001). Employment, parenting, and well-being among mothers of children with disabilities. Mental Retardation, 39, 297309.
  • Warfield, M. E. (2005). Family and work predictors of parenting role stress among two-earner families of children with disabilities. Infant and Child Development, 14, 155176.
  • Wellington, A. (2006). Self-employment: The new solution for balancing family and career? Labour Economics, 13, 357386.
  • Williams, J. C. (2000). Unbending gender: Why family and work conflict and what to do about it. London, UK: Oxford University Press.
  • Williams, J. C. (2010). Reshaping the work-family debate: Why men and class matter. Cambridge, MA: Harvard University Press.
  1. 1

    The higher levels of mental health, particularly among fathers, probably reflects the fact that the NSAF is a parent-only sample as well as differences between the NSAF and MOS survey item construction and response categories and the way the two surveys were administered (Ehrle & Anderson-Moore, 1997).

  2. 2

    Higher-order interaction models can, in general, be unreliable and the effects difficult to interpret, but because there were a small number of single fathers with a child with disabilities (especially children younger than age 6) analysis of models including four-way interactions between parent's employment, gender, marital/partner status, and child disability status by age were particularly unreliable.