Which of the following is true of how parents work affects the development of their children Group of answer choices?

J Fam Psychol. Author manuscript; available in PMC 2021 Apr 1.

Published in final edited form as:

PMCID: PMC7021583

NIHMSID: NIHMS1037937

Abstract

The current study examines the relationship between working-class mothers’ and fathers’ job autonomy, across the first year of parenthood, and their children’s behavior problems and adaptive skills in the first grade. Data came from a longitudinal study of 120 couples interviewed five times across the transition to parenthood and again when the target child entered the first grade. Mothers’ job autonomy and fathers’ work hours during the child’s first year of life directly predicted fewer behavior problems and more adaptive skills in their children at 6–7 years of age. For all parents a mediated relationship emerged such that greater job autonomy predicted less parenting overreactivity, which in turn predicted better child outcomes. Parent involvement was also a significant mediator linking job autonomy to children’s adaptive skills, but not behavior problems.

Keywords: Dual-earners, parenting, child development, low-income jobs, job autonomy, transition to parenthood

The ecological model of child development (Bronfenbrenner & Crouter, 1982) holds that parents’ roles and experiences outside of the family, like in the workplace, can have a significant impact on child development. By and large, research on parental employment has tended to focus on the effects of work hours and schedules on children’s social and cognitive development, often overlooking how parents’ experiences at work (e.g., autonomy, self-direction) may influence their children (Perry-Jenkins, Goldberg, Smith & Logan, 2011). A focus on parental work conditions, as opposed to work hours alone, is important if we are to discover key areas of intervention at the workplace that can support employed parents.

There are a number of gaps in the current knowledge base linking parental employment and children’s development. First, looking past the simple fact of whether parents work or not, we know little about how the work experiences of employed parents, particularly low-income workers who have limited financial resources and fewer family-friendly policies than their middle-class counterparts, shape family life and child development (Yoshikawa, Weisner, & Lower, 2009). Second, the majority of studies in this area focus on mothers’ employment, overlooking how conditions of fathers’ jobs are related to children’s development, even as employed fathers spend increased hours in early child care (Bianchi & Milkie, 2010). Finally, burgeoning research on child development highlights the first year of life as a particularly sensitive period in infant development (Gunnar & Fisher, 2006). The first year of life is a time when work conditions may play a critical role in shaping employed parents’ ability to be engaged and sensitive caretakers, ultimately influencing children’s development. The current study aims to address these gaps by examining how the conditions of both mothers’ and fathers’ working-class jobs, across the first year of parenthood, are directly, and indirectly via parenting, related to their children’s behavior problems and adaptive social skills in the first grade.

Work Experiences and Child Development

Although a great deal of research has examined the relationship between maternal employment, primarily hours and schedules, and children’s development (Brooks-Gunn, Han & Waldfogel, 2010), less is known about how employees’ experiences at work, such as autonomy and self-direction, are related to child development. In the current study we focus in on workers’ autonomy on the job, defined by Kohn and Schooler (1982), as “the use of initiative, thought, and independent judgement at work” as it shapes aspects of family life. Kohn posits that work experiences are echoed in parenting styles, such that parents working under conditions of high supervision and limited autonomy are likely to use more controlling and punitive parenting styles in the home, which in turn, predict more negative child outcomes. In an extension of Kohn’s work, Rogers, Parcel and Menaghan (1991) theorized that workers who experience a sense of control at work are more likely to promote supportive and less controlling home environments for children. By extension, more overreactive and controlling parenting is linked to more behavior problems (e.g., aggression, hyperactivity) in children (Meteyer & Perry-Jenkins, 2009).

During the 1990s, a number of empirical studies documented that more self-directed and complex work was more important than work status alone in relation to the quality of parenting (Kohn & Schooler, 1973; Parcel & Menaghan, 1994; Presser, 2003). Parcel and Menaghan (1994) found that higher occupational complexity in mothers’ jobs predicted more positive home environments, which provided greater cognitive stimulation, emotional support and safety for children. Other studies have shown that mothers and fathers with more complex and autonomous jobs use more positive, warm and responsive parenting skills with their 5–7 year-old children (Greenberger, O’Neil and Nagel (1994), use less restrictive parenting (Mason, Cauce, Gonzales, Hiraga, & Grove, 1994), and use less harsh parenting by fathers (Whitbeck et al., 1997). Luster, Rhoades and Hass (1989) found that mothers who valued self-direction scored higher on maternal involvement and warmth than those who valued conformity.

A number of studies have linked autonomous work to child outcomes. Roeters, van der Lippe and Kluwer (2010) found that parents who worked fewer hours and had more engaging jobs spent more time with their children, that, in turn, predicted higher quality parent-child relations. Wheeler, Updegraff and Crouter (2015) found that, among Mexican-origin families, fathers’ occupational self direction was linked to less parent-adolescent conflict and, in turn, better adolescent adjustment. Taken together, research suggests that more autonomy at work translates into more proactive and supportive parenting behaviors at home. Or, in contrast, parents who worked in controlling environments, with little autonomy, would be more likely to bring those behaviors home in the form of being more controlling and rigid. In the current study, we test two hypotheses emerging from a work socialization perspective. First, parents’ who experience more autonomy at work will engage in more positive parenting that will predict fewer behavior problems and greater social skills in children. In addition, parents with little job autonomy will engage in more negative (e.g., overreactive, controlling) parenting styles that will, in turn predict more behavior problems and poorer adaptive skills in children.

An important caveat to much of the research linking parental work conditions to child development is the practice of assigning characteristics of occupations, as defined by the Dictionary of Occupational Titles (National Academy of Sciences, 2003), to workers’ experiences of their jobs (Parcel & Menaghan, 1994; Raver, 2003; Spector, Jex, & Chen, 1995). Although this is a creative approach to linking work characteristics to workers’ jobs in the absence of subjective reports, it assumes that all individuals working in a particular occupation experience the same levels of complexity and autonomy at work. Such an approach would code the majority of working-class jobs as low in autonomy, an assumption that research has shown to be unfounded. In short, jobs within particular social class strata are not experienced in the same way by all workers (Perry-Jenkins, Goldberg, Smith & Logan, 2011). From a policy perspective, it is important to understand what conditions of low prestige jobs can be experienced as autonomous, as this knowledge could improve the development of effective interventions focused on enhancing job conditions for low-income, working parents.

Work, Parenting and Timing: The Transition to Parenthood as a Sensitive Period

More recent research has begun to focus on the issue of timing, specifically examining the impact of parental employment during the first years of a child’s life. The theoretical argument for how and why early parental employment, most often conceptualized as maternal employment, might affect child development stems from an extensive literature in developmental psychology that points to the quality of the early parent-child relationship in shaping children’s socio-emotional and cognitive development (Bowlby, 1958; Macoby, 1992; Brooks-Gunn, Han, & Waldfogel, 2010). Parents provide a secure base for early infant development as well as providing support and scaffolding to enrich social relationships and cognitive development. Given the critical role of parents in early development, concerns that early maternal employment may impinge upon mothers’ time and ability to develop secure and nurturing relationships with their children spurred the latest focus on maternal employment in the first years of a child’s life.

A number of studies provide support for the early life experience hypothesis. Costigan, Cox and Cauce (2003) found that mothers’ reports of more negative interpersonal relationships at work predicted decreases in positive parenting of their infant from 9 to 12-months of age. Looking at much longer time frames, Yetis-Bayraktar, Budig, and Tomaskovic-Devey (2013) found that mothers’ concurrent occupational complexity, defined as autonomy and supervisory responsibilities, was positively related to their 6–13 year old children’s academic achievement. In addition, mothers’ occupational complexity in the first 3 years of the child’s life was related to children’s later positive academic outcomes in school, with enhanced advantages for boys’ math abilities. Fathers work conditions were unrelated to parenting. Brooks-Gunn, Han and Waldfogel (2010) found that early maternal employment in the first three months of a child’s life predicted more externalizing behaviors at 4.5 years; however, maternal sensitivity and a positive home environment buffered this negative relationship. A considerable research base provides evidence that early parental employment may be critical in setting the stage for later developmental outcomes in children, and that process may be mediated through parenting styles.

Social Context, Parental Employment and Child Development

In the current research, the question is whether the social ecological niche of being in a low-income, dual-earner family uniquely shapes the ways in which work-family processes play out. Lucas-Thompson, Goldberg and Prause (2010) conducted a meta-analysis of 69 studies to synthesize the research linking maternal employment during infancy and early childhood to children’s developmental outcomes. They found that, with few exceptions, “early employment was not significantly associated with later achievement or internalizing/externalizing behaviors” (p. 915) in children. Follow-up analyses revealed, however, that social class moderated these findings such that early employment had modest but beneficial effects on children’s outcomes in low-income families, and slightly more negative effects in high SES families.

When looking within samples of low-income families, a number of studies have found positive associations with maternal employment early in children’s lives and positive child outcomes (Coley & Lombardi, 2013; Dunifon, Kalil, & Danziger, 2003; Secret & Peck-Heath, 2004). It has been hypothesized that the benefits of maternal employment in low SES contexts, where jobs bring economic stability and financial resources, may outweigh the potentially negative effects of decreased maternal time with children. For example, Coley and Lombardi (2013) found, in a low-income, working sample of African-American and Hispanic families, that children whose mothers were employed early had enhanced socioemotional functioning in comparison to their peers with non-employed mothers. Utilizing the same data, Lombardi and Coley (2013) found that mothers with high work quality (i.e., income and insurance) and job stability had children with higher reading and math skills and fewer emotional and behavioral problems than mothers with low quality jobs. These authors’ conceptualization of work quality, focusing on wages and insurance, clearly moves research beyond the past emphasis on maternal work status or work hours; however, questions remain as to how workers’ experiences of low-income jobs, such as perceptions of autonomy, are related to parent quality and child well-being.

What about Fathers’ Employment and Child Development?

We know surprisingly little about how fathers’ employment, both hours and social conditions of work, is related to the development of their children. Gerson (2010) argues that the current culture of fatherhood calls for men to not only be workers but also involved fathers and equal partners with their partners. Two studies from the Family Life Project found that greater paternal work stress was related to poorer quality parenting with infants. In addition, when fathers worked more hours on a nonstandard shift they were less likely to be sensitive and engaged parents (Goodman, Crouter, Lanza, Cox, 2008; Goodman et al. 2011). In contrast, Costigan, Cox and Cauce (2003) found that fathers’ job autonomy and co-worker relations were unrelated to parenting quality with infants. More research is needed to better understand the long term implications of fathers’ work, especially during the transition to fatherhood when childcare and involvement patterns are just being established, on children’s later development.

The current study addresses some of the aforementioned gaps in the literature by focusing on one aspect of parents’ work during the first year of parenthood, job autonomy, as it predicts children’s behavior problems and social skills in the first grade. These issues are investigated in a unique sample of working-class parents who were followed longitudinally from their third trimester of pregnancy through the child’s first year of school. Given that concerns have been raised about early parental employment (NICHD-SECC, 1997, 2001) the current study examines how both mothers’ and fathers’ work conditions in the first year of the child’s life predict social outcomes in children in the first grade. The use of teacher reports of child outcomes is a strength of the study design since much research relies solely on parent reports of child outcomes.

Research Questions and Hypotheses

Our primary research questions examine whether working-class, employed parents’ experiences of job autonomy during the first year of their child’s life predict children’s adaptive skills and behavior problems in the first grade. We control for parents’ work hours in the models because we are interested in whether, above and beyond parents’ work hours, job autonomy has a unique effect on children’s development. We also examine the role of parenting quality as a mediator between work conditions and child outcomes. We test the following hypotheses:

Hypothesis 1:

Net of mothers and fathers work hours, greater job autonomy reported by parents across the first year of parenthood will predict more adaptive skills in children in the first grade. This effect will be mediated by higher quality parenting.

Hypothesis 2:

Net of mothers’ and fathers’ work hours, greater job autonomy reported by parents across the first year of parenthood will predict fewer behavior problems in children in the first grade. This effect will be mediated by higher quality parenting.

Method

Participants

Data for this project came from a longitudinal study of 153 working-class couples interviewed on five occasions across the transition to parenthood between 1999 and 2002 and again on a sixth occasion in 2006–07 when the target child was in the first grade. Data were obtained during in-depth, two- to three-hour interviews where each parent completed survey questionnaires and answered open-ended questions about their experiences of new parenthood. Heterosexual couples in their third trimester of pregnancy were recruited from prenatal classes at four hospitals and five clinics in Western New England. Based on data from the hospitals where we recruited parents from, approximately 73% of first-time expectant parents, who were either married or partnered, attended prenatal classes at the hospital. The clinics did not have similar data on participation rates, but reported that first-time parents were more likely to attend classes. Thus, we captured a wide swath of first-time, coupled parents with this procedure. Criteria for eligibility included the following: (a) both members of the couple were employed full-time (32+ hours per week) prior to the baby’s birth, (b) both members of the couple planned to return to full-time work within six months of the baby’s birth, (c) both members of the couple were “working-class” (defined by educational level and work characteristics defined below), (d) both members of the couple were expecting their first child, and (e) the couple was either married or cohabiting (for at least one year) at the time of inclusion in the study. We focused on first-time parents because we were interested in how new parents addressed the challenges of work and parenting for the first time, without the skills parents develop with multiple children.

Over the first year of the study 12 families dropped out, an attrition rate of approximately 8%. These families did not differ on key demographic variables such as work hours, income, marital status, or age from those families who remained in the study. Of the 12 families who did not complete the first year of data collection, four were either separated or divorced by the end of the first year, and eight families withdrew. By the sixth interview, all participants in the original study (N = 153) were invited to participate in a follow-up study, and 79% (N = 121) consented. Of the 32 families dropped, 20 had separated or divorced, 5 had moved out of the area and were not interested, and 7 were uninterested. These families did not differ on key demographic variables including work hours, income or age. One additional family was dropped because both partners were unemployed. The final sample size for the current study was thus 120 families.

Average age for men at the time of the pregnancy was 29.1 (mdn 28.7), and women was 27.8 (mdn 27.5). Nearly 80% (79.7%) of the couples were married, and the average length of relationship was 3.3 years. Cohabiting couples had been together for and average of 2.1 years. The majority of the sample was White (92.2% of women, 90.6% of men) with only 1.5 % African American women and men, 1% Latino women, 3% Latino men, and 5% Asian women and men. For 16% of women and 22.5% of men, the highest degree held was a high school diploma or GED; a majority of the sample (44.8% of women and 63.4% of men) had some type of additional schooling or vocational training after high school (e.g., beautician’s school, refrigeration mechanic training). Only 31% of women and 14.1% of men held a one- or two-year Associate’s Degree. None of the parents had a college degree.

In the current study, working class was defined in part based on educational level, participants could have an associate’s degree or lower, the majority in our sample were high school educated. In addition, individuals had to be working in jobs where they were considered “order takes” versus “order givers”, thus, if you supervised others in any capacity you were excluded from the study. Our sample mirrors current statistics on families considered “working-class” in the U.S. Data from the 2012 American Community Survey indicates that jobs within the range of $22K – 31K (the range of working mothers and fathers in our sample) include nursing aides, secretaries, truck drivers, cooks, janitors, customer service, retail sales, laborers and cashiers. In our sample the most common jobs for mothers were nursing aides, food service workers and beauticians and for fathers were truck drivers, laborers and maintenance workers. Median salaries were $30,237 ($14.53 an hour) and $25,320 ($12.17 an hour) for men and women respectively, and the median family income was $52,607. These estimates represent gross income; after taxes take home pay is significantly less. In addition, a third of the fathers and ten percent of mothers had a second job that contributed to the higher individual incomes and longer hours reported by some parents (second job average hours for fathers was 8.6 and mothers 3.2) All partners were employed full-time at the outset of the study; men worked an average of 47 hours per week and women averaged 40 hours. It is important to note that the majority of parents remained in the same job across all times of measurement (85% of fathers; 80 % of mothers). A majority of children in the sample were girls (n = 69, 57.5%).

Procedure

The material and procedures used in this study were approved by the University of Massachusetts Amherst Review Board; the title of the project was “The Sociocultural Context of New Parenthood.” The PI and graduate students were given 5 minutes at the beginning of prenatal classes to describe the study to expectant parents. All parents filled out a short demographic form with basic information on age, relationship status, income, type of job, work hours, and intent to return to work after the baby’s birth. Interested and eligible families were contacted and scheduled for an interview; all families received a total of $150 for their participation in the first five interviews that occurred in the first year, and another $100 for the sixth interview that took place when the child was in the first grade. In comparing our sample to the broader population of first-time parents in prenatal classes, the present sample, as expected given the selection criteria, was less educated, had lower family income and worked more hours than the full sample. Our sample was also older, on average, than national norms for first time parents, due to the fact that couples had to be married or long-time cohabiters to participate.

Fathers and mothers participated in a series of five interviews across the first year of parenthood: (a) during the couples’ third trimester of pregnancy (hereafter, Time 1); (b) one-month after the baby’s birth (Time 2); (c) one month after mothers returned to full-time employment (15 weeks postpartum, on average (Time 3); (d) a mail survey conducted when the baby was 6 months (Time 4); (e) a face-to-face interview when the baby was one year old (Time 5). The timing of the third interview was based upon mothers’ return to work; thus, there is variability in timing of this assessment occasion. Once mothers were back at work for approximately one month, the Time 3 interview was completed. The majority of mothers returned to full-time work by 6 months post-partum. If mothers returned to only part-time work or did not plan to return by 6 months, they still completed a Time 3 interview. Data on work hours and job autonomy were assessed during this first year and used in our statistical modeling.

A final, Time 6 interview was conducted when the target child was entering the first grade. Face-to-face interviews were conducted with both mothers and fathers where they completed the Behavioral Assessment System for Children – Parent Rating Scale (BASC-PRS), assessing a range of their child’s behavior (e.g., social skills, behavior problems), as well as self-report surveys on their discipline style as a parent and their level of involvement with their six-year old. Upon receiving consent from the children’s parents, we contacted the first-grade teachers of the children via email or mail and got their consent to participate. Given that parents had consented and given us teachers’ contact information, only one teacher did not participate. Teachers completed the Behavioral Assessment System for Children (BASC-TRS).

Measures

Work hours.

During each in-home interview, both partners provided detailed information on work schedule and hours. Since all mothers were employed full-time at Time 1, and unemployed (due to baby’s birth) at Time 2, work hours at Times 3, 4 and 5 were used as indicators of the latent variable representing average work hours in our structural equation models. For fathers, work hours at each of the 5 time points in the first year of parenthood were used as indicators of the latent variable representing average work hours.

Job Autonomy.

Job autonomy was assessed at Times 1, 3, and 5 using a scale developed by Greenberger, O’Neil, and Nagel (1994). Respondents indicated how well each item described the way they spent their day at work. Eighteen items from the challenge/autonomy subscale described one’s job as challenging, varied and self-directed. The questionnaire uses a 5-point Likert scale ranging from strongly disagree (1) to strongly agree (5). Sample items include: “I’m frequently expected to solve challenging problems at work” and “I have a great deal of influence in the decision-making process at work.” Authors found ratings of autonomy were unrelated to level of education, income, or job prestige, but were positively correlated with job satisfaction. Across the three time points, Cronbach’s α reliability for autonomy ranged from .85 to .87 for men and from .83 to .89 for women. See Appendix A for all measures.

Parenting Involvement.

Involvement in everyday parenting activities was assessed with the 10-item involvement subscale of Alabama Parenting Questionnaire (Shelton, Frick, & Wootton, 1996) when children were in the first grade. Parents rated how often interactions typically occur, such as having a friendly talk with their child, playing games, or doing other fun things with their child. According to the authors, the involvement subscale is an indicator of how engaged, flexible and accessible parents are with their children, as opposed to being directive. Participants reported their involvement using a five-point scale ranging from 1 = Never to 5 = Always. Cronbach’s alpha was .79 for mothers and .74 for fathers. Sample items include: “You have a friendly talk with your child,” “You help your child with his or her homework.”

Parent Overreactivity.

Overreactivity was measured using the overreactivity subscale of The Parenting Scale (Arnold, O’Leary, Wolff, & Acker, 1993) when children were in the first grade. The 10-item scale had two anchors: an effective parenting response to a child behavior and an ineffective one. An effective anchor of one item is I am no more picky than usual, an ineffective anchor of this item is When I’m upset or under stress, I am picky and on my child’s back. Parents used a scale of 1 to 7 to indicate which anchor was more like them. Higher values represented more overreactivity. Cronbach’s alpha was .80 for mothers and .76 for fathers.

Child behavioral outcomes.

Children’s adaptive skills and behavior symptoms were measured using the Behavioral Assessment System for Children – Teacher Rating Scale (BASC-TRS; Reynolds & Kamphaus, 1992) and Parent Rating Scales (BASC-PRS; Reynolds & Kamphaus, 1992). The BASC–TRS is a 131-item and the BASC-PRS is a 138-item comprehensive rating scale that assesses a broad range of psychopathology in children ages 2 ½ years and older (Reynolds & Kamphaus, 1992). We used the BASC-TRS and -PRS for the age group 6 – 11 years. Teachers and parents are asked to rate on a 4-point scale (never, sometimes, often, almost always) the degree to which each item describes the child. The BASC–TRS and – PRS demonstrate good reliability and validity with school-aged children (Reynolds & Kamphaus, 1992). T scores (based on general, not gender-specific, norms) for Adaptive Skills composites and the Behavioral Symptoms Index were used. Adaptive Skills represents a composite variable averaging across adaptability, leadership, social skills, and study skills (TRS only). The Behavioral Symptoms Index is a combination of central scales including hyperactivity, aggression, anxiety, depression, attention problems, and atypicality, which reflect the overall level of problem behavior. Norms for the BASC are based on a sample of more than 13,000 nationally representative cases. The ranges for adaptive skills were 35 – 73 (TRS) and 25 – 74 (PRS) and behavioral symptoms ranged from 37 – 62 (TRS) and 30 – 74 (PRS).

Control Variables.

As Yetis-Bayraktar et al. (2013) argue, “mothers’ education is the strongest alternative explanation to any observed association between occupational complexity (e.g., autonomy) or hours worked and child outcomes” (p. 44). Thus, we controlled for educational level. Family income and child gender have also been linked to child outcomes and were controlled for in our models. (Yetis-Bayraktar, Budig, & Tomaskovic-Devey, 2013).

Analytic Strategy

A series of models were estimated to test the influence of parental work quality in the child’s first year of life on children’s behavior problems and adaptive skills in the first grade. We used structural equation modeling (SEM) as our analytic strategy for two reasons: 1) to capitalize on the information available in the multiple measures of each construct; enabling us to model measurement error and estimate true score relationships among the latent constructs, and 2) to obtain model fit statistics to test that our overall theoretical model was supported by the data. We assessed the early influence of work hours and work quality measured during infancy on child behavioral problems and adaptive skills 6 years later. For each predictor/outcome combination, we fit two models. Model 1 contained only parent work hours as predictors. Model 2 added parent work quality variables. Model comparison tests provided evidence for whether the more complex model (accounting for work quality) improved the fit compared to the model with work hours alone. We used one-tailed tests of significance, in line with our directional hypotheses. Finally, we tested whether aspects of overreactivity and involvement mediated the link between work variables and child outcomes. Model fit was evaluated by four indicators: Model χ2 relative to model df ; Root-Mean-Square Error of Approximation (RMSEA; values of .08 and lower represent acceptable model fit), Bentler’s Comparative Fit Index (CFI; values higher than .90 indicate acceptable model fit), and Standardized Root Mean Square (SRMR; values lower than .08 indicate adequate model fit) (Kline, 2004).

Results

As shown in Table 1, all parents were employed at Time 1, the prenatal visit, since that was a criterion for inclusion in the study. Although all parents planned to return to work full-time by 6-months postpartum, mothers’ average work hours dropped by 6–8 hours in the postpartum year. This decline in work hours was due to approximately 20% of the mothers who dropped to part-time hours for at least some of the year. On average, parents reported relatively high levels of involvement (4.04 out of 5) with their six-year olds, and low levels of overreactivity (2.86 out of 5).

Table 1

Descriptive Statistics for Predictors and Outcomes

Table 2 provides both intra- and inter-correlations among predictor and outcome variables. Correlations between parents’ job autonomy, parent involvement and overreactivity were weaker for mothers than fathers. Although in the direction hypothesized, significant correlations emerged in only two of 8 cases for mothers and half the cases for fathers. Mothers’ and fathers’ reports of work conditions were unrelated, while reports of parenting styles were modest (r = .25, p < .01 for overreactivity and r = .23, p < .01 for involvement). Parents reports of behavior problems correlated at r = .44, p < .01; and adaptive skills at r = .50, p < .01.

Table 2

Intra- and Interparental Correlations Among Study Variables

Measurement Models

Each latent variable had multiple indicators. Mothers’ and fathers’ work hours and job autonomy measured at multiple times across their child’s first year were used as indicators to create latent variables of work hours and work quality. Latent variables for child outcomes were measured by reports from the child’s mother, father, and teacher. Latent variables of parenting were indicated by each parent’s self-reported parenting overreactivity and involvement. These latent parenting variables were measured by a single indicator, so the loadings were fixed to one and the error variances were fixed to zero. We used the two-step approach suggested by Kline (2004) where we first fit a measurement model to the data; all factor loadings and error variances except one were significantly different from zero, indicating a sound measurement model. The exception was the factor loading for the teacher-reported variable of child adaptive skills. It was not significantly different from zero, suggesting a weak or zero correlation with the latent factor, and so was excluded from the analysis. For several indicators, the error variances were fixed to zero when the initial fit showed them to be negative. These were: mothers’ job autonomy (first measure, Time 1) in the both the child adaptive skills and child behavioral problems mediational models, and fathers’ job autonomy (first measure, Time 1) in the child behavioral problems mediation model only. All other error variances were estimated by the model.

Parents’ Job Autonomy and Children’s Behavior Problems and Adaptive Skills

Results from main effects models showed that parental work hours predicted fewer behavior problems and more social skills (see M1 estimates in Table 3). Model comparison tests indicated that maternal job autonomy provided additional explanatory power over and above the effect of work hours (see M2 estimates in Table 3). There was a significant negative association between maternal autonomy and child behavioral problems, indicating that, controlling for work hours, greater autonomy in the child’s first year of life was associated with fewer behavioral problems in the first grade. In addition, controlling for work hours, mothers’ job autonomy in the child’s infancy was associated with higher adaptive skills five years later. These two models provided a good fit to the data as assessed by several fit indices (reported in notes below each figure). Unstandardized and standardized coefficient estimates, associated standard errors, and p values are presented in Table 3. Model fit indices and comparison tests are presented below each set of findings. The addition of parental work quality variables significantly improved model fit (“M2” columns in Table 3) as compared to models that only included parental work hours (“M1” columns in Table 3). Fathers’ job autonomy was not a direct predictor of either child outcome.

Table 3

Parental Job Autonomy Predicting Children’s Behavior Problems and Adaptive Skillsa

Parenting as a Mediator between Job Autonomy and Child Outcomes

In our final models, we tested whether parenting mediated the associations found between work quality and child outcomes. Following Hayes (2013), we used evidence of a significant indirect effect to bolster our claim of mediation. In summary, for work autonomy variables, we found significant mediational effects for both fathers’ and mothers’ overreactivity. Parent involvement was also a significant mediator of children’s adaptive skills, but not behavioral problems. In contrast, the association of work hours and child outcomes was not mediated by parenting for fathers or mothers. In sum, parenting variables mediated the effect of autonomy on child outcomes but did not mediate the effect of work hours on child outcomes.

The covariance matrix of indicator variables and statistics for the direct and indirect effect parameters for each of the mediational models for work hours are available in Supplemental Materials Tables B, C, and D. Standardized coefficients for the mediational models without and with covariates are available in Supplemental Materials Table E. Model comparison statistics indicated that the more complex, mediational process models provided a significantly better fit to the data than a model with only direct effects from work variables to child outcomes. Figures 1 and ​2 depict the mediational process by which work quality predicts parenting, which in turn is associated with better child outcomes. This was a partial mediational finding, with parenting accounting for 23% of the total effect of mothers’ work quality on children’s behavior problems, and 67% of the total effect of mothers’ work quality on children’s adaptive skills (percentages calculated by dividing the standardized indirect effect by the standardized total effect). Interestingly, while in prior models we found no direct effect of fathers’ work quality (i.e., autonomy) on child outcomes, once parenting was included as a mediator, fathers’ work quality variables were significantly associated with child outcomes, both indirectly and directly. In these models, the direct paths in the model without the mediator (M1) were non-significant. Recent scholarship argues that mediational processes can occur even in the absence of a direct relationship between the predictor and outcome (Hayes, 2009; MacKinnon et al., 2000; Shrout & Bolger, 2002). In such cases it is possible that there is a suppression effect, because of opposite signs of the two paths (MacKinnon et al., 2000). This is the case for the effect of fathers’ autonomy on adaptive skills. Specifically, the path from autonomy to overreactivity is negative; the path from overreactivity to adaptive skills is negative. Therefore the indirect effect, which is the product of these two coefficients, is positive. The direct effect from autonomy to adaptive skills is negative. Therefore the direct effect and the indirect effect are of opposite signs, a classic instance of suppression.

Parenting as a Mediator of the Effect of Job Autonomy on Children’s Behavior Problems.

Note. Absolute model fit indices indicated a good model fit: χ2/df = 1.52; RMSEA = .06; CFI = .92; SRMR = .08. Standardized coefficients are presented. Solid lines indicate significant effects, dashed lines indicate non-significant paths. * p < .05,** p < .01, *** p < .001.

Parenting as a Mediator of the Effect of Job Autonomy on Children’s Adaptive Skills

Absolute model fit indices indicated adequate model fit: χ2/df = 1.77; RMSEA = .06; CFI = .89; SRMR = .08. Standardized coefficients are presented. Solid lines indicate significant effects, dashed lines indicate non-significant paths. * p < .05,** p < .01, *** p < .001.

Although educational levels in our sample were constrained due to selection criteria, we still had mothers with less than high school, high school degree, and associate degrees. We conducted the mediational models controlling for parental education, parental income, and child gender. Other variables that have been related to child outcomes include marital status, number of children and age of children, all constructs controlled for in the current study through sample selection (e.g., first baby, all coupled). Of note, child gender was significantly associated with each child outcome, with boys having higher reported behavioral problems (β = .5, p = .001) and lower reported adaptive skills (β = −.3, p = .001) than girls. Family income and education levels did not show significant associations, likely due to the homogenous nature of the sample via inclusion criteria. Mediational findings remained significant even with the addition of these covariates to the model. While findings held when including covariates in the model, fit statistics fell below acceptable levels. For the behavioral problems child outcome, these results were χ2/df = 2.02; RMSEA = .07; CFI = .81; SRMR = .09, with only the RMSEA falling within index values for a well-fitting model. For the adaptive skills child outcome, these results were χ2/df = 1.96; RMSEA = .08; CFI = .83; SRMR = .09, with only the χ2/df falling within index values for a well-fitting model. The poorer fit found in the model including covariates is likely due to the increased complexity of the models with the addition of these covariates with a relatively small sample size. Given that our primary findings of interest hold with or without covariates in the model, and the better fitting model is that without the covariates, we thus report findings from the model without covariates in Figures 1 & 2.

Discussion

This study is among the first to examine both mothers’ and fathers’ work conditions in working-class jobs during the first year of their child’s life in relation to both parenting quality and children’s behavior problems and children’s adaptive skills in the first grade. Although other studies have documented the effects of early maternal employment, specifically hours and schedules, on children’s social and cognitive outcomes (Coley & Lombardi, 2013; Dunifon, Kalil, & Danziger, 2003; Secret & Peck-Heath, 2004), few have documented the effects of fathers’ work hours or work conditions in the first year of the child’s life on child outcomes. More importantly, it was hypothesized that above and beyond more objective conditions of employment, such as hours, the quality of work experiences, such as autonomy, would have positive effects on children’s social development, as well as effects mediated through parenting.

Findings support a work socialization model proposed by Kohn (1977) and Parcel and Menaghan (1994). Results revealed that mothers’ experiences of autonomy and self-direction predicted fewer behavior problems and more adaptive skills in their children. Although job autonomy was not directly related to child outcomes for fathers, fathers’ work hours positively predicted greater adaptive skills for children. It is likely that stable, full-time work is protective in working-class families in that stable jobs provide stability and key resources for children.

Looking past direct relationships between work and child outcomes, and based on work socialization theory, it was hypothesized that greater autonomy at work would predict more sensitive and involved parenting that would, in turn be related to greater social skills and fewer behavior problems in children. Results showed that both mothers’ and fathers’ reports of greater job autonomy across the child’s first year of life predicted less overreactive parenting and more involved parenting that, in turn, predicted greater adaptive skills in children. The mediation model linking job autonomy and children’s behavior problems only held up for overreactive parenting as a mediator, not parenting involvement. Consistent with the work of Kohn (1977), Parcel and Menaghan (1994), and Yetis-Bayraktar, Budig, and Tomaskovic-Devey (2013), our data suggest that values and skills from work, focused on self-direction and autonomy, are transmitted to children through less overreactive parenting styles and more engaged parent involvement. It is of note that these results emerged in a sample of working-class parents who reported significant variability regarding experiences of autonomy on the job, a finding that would have been lost had we simply attached job characteristics using the DOT. These results highlight the importance of assessing workers’ experiences on the job as opposed to linking indicators of job quality from the Dictionary of Occupational Titles (DOT) to job categories.

We turned to our qualitative data to gain insight into what a low status job with high levels of autonomy looks like. The participant in our study who had the highest score on job autonomy was “Linda” who worked in a candle-packing factory as an order packer. As Linda described her work, we learned that when she first started packing orders for customers she would slip in some new candle scents with each order with a note to customers about how they might enjoy this new product. Slowly customers began to specifically request her services. Her boss recognized her creativity and “autonomy” and had her train new workers on how to connect with customers. She felt valued and respected for her work. Others parents described autonomy as having their opinions valued and having a voice in decision-making Had we relied on DOT descriptors of Linda’s factory job it have been coded as low in autonomy. These results suggest that future research would benefit from examining the range of job experiences categorized as low SES, including gathering workers’ reports of the quality of their work conditions.

Findings that both mothers’ and fathers’ job conditions predicted children’s adaptive skills and behavior problems highlight the importance of considering how both mothers’ and fathers’ work experiences shape child development. Fathers have been virtually ignored in the literature on employment and child development, unless of course they are unemployed. The persistent focus on mothers’ employment, and the exclusion of fathers’, highlights the gendered ideology that continues to shape our understanding of early child development and the role of mothers. Findings from this study indicate that both mothers’ and fathers’ work and parenting are related to children’s social development and our educational, prevention and intervention strategies to support working families should include all members in the family system.

Perhaps most importantly, these findings suggest that parental work conditions can enhance children’s development. When jobs provide parents the ability to solve problems, have a voice in the system, in essence, feel valued at work, those values are translated into how workers approach their parenting. In short, they use less controlling and overreactive parental styles and remain engaged and open to their child’s lead. The question of why the direct effects of fathers’ work hours in the first year (but not mothers’) and mothers’ autonomy in the first year (but not fathers’) predicted children’s behavior problems and adaptive skills is perplexing. It may be that the stability provided by fathers’ employment is a protective variable for the family relieving financial strain and stress. In contrast, mothers work experiences in that first year may directly impact the child in others ways, besides through her parenting, such as her mental health, availability or self-esteem, all processes that deserve attention in future research. The direct and positive effect of fathers’ work hours on children’s positive outcomes disappears in the behavior problems model when fathers parenting is added to the model, suggesting that fathers work is entirely mediated through parenting in these models. In terms of the adaptive skills model, however, fathers’ work becomes directly and negatively linked to adaptive skills when parenting is added to the model. It may be the case that when fathers work long hours, in the absence of positive parenting, children do not fare as well in terms of their social and adaptive skills.

A limitation of this study is the inability to draw causal conclusions whereby work conditions “affect” parenting and child outcomes. Selection effects of workers into certain types of jobs is also a potential confound. We did test our models controlling for potential third-level variables that may explain our results, namely education level and income, and the findings still held up. The idea that workers who report more autonomy at work are reflecting their own personality characteristics that allowed them to secure more autonomous occupations is a possibility. Another shortcoming is the gap between the assessment of parents’ work conditions in the first year of life and child outcomes in the first grade. Ideally, we would have had work and parenting data measured at yearly time points to better examine pathways linking work and parenting to child outcomes. Parents’ reports of both parenting and child outcomes is a limitation of the data due to shared variance in reports; future work looking at observed behavioral outcomes or child clinical assessments would be a great strength. Finally, the lack of racial and ethnic diversity is a limitation; a second study with a more diverse sample is currently underway.

The implications of these findings for policymakers point to the worksite as an important place for intervention. The notion that lower status jobs can be experienced as autonomous and self-directive and that these experiences positively impact workers and their children leads to the question of what aspects of lower SES jobs lead to autonomy. The one example we provided above about the woman who worked at a candle packing factory and experienced high levels of autonomy points to the critical role of the supervisor. A supervisor who recognizes creative thinking and new ways to do things and also rewards workers for ingenuity creates autonomy for workers. Future qualitative work and focus groups at work sites with employees who can share the ways in which their jobs are routine and mundane or meaningful and self-directed could provide insight into possible ways that workplaces can support worker autonomy. In addition, these results point to a key area ripe for intervention in working-class occupations, namely providing workers with self-direction and autonomy on the job, whereby workers feel valued and engaged---values that are transferred to their relationships with their children.

Supplementary Material

Supplemental Material

Acknowledgments

This research was supported by grants from the National Institute of Mental Health (R29-MH56777, R01-MH56777)).

Footnotes

Author Note: Prior dissemination of these ideas and data were presented at a 2016 symposium at the National Council on Family Relations.

References

• Arnold DS, O’Leary SG, Wolff LS, & Acker MM (1993). The Parenting Scale: A measure of dysfunctional parenting in discipline situations. Psychological Assessment, 5(2), 137–144. doi: 10.1037/1040-3590.5.2.137 [CrossRef] [Google Scholar]
• Baron RM, & Kenny DA (1986). The moderator–mediator variable distinction in social psychological research: Conceptual, strategic, and statistical considerations. Journal of Personality and Social Psychology, 51(6), 1173–1182. doi: 10.1037/0022-3514.51.6.1173 [PubMed] [CrossRef] [Google Scholar]
• Bianchi SM, & Milkie MA (2010). Work and family research in the first decade of the 21st century. Journal of Marriage and Family, 72(3), 705–725. doi: 10.1111/j.1741-3737.2010.00726.x [CrossRef] [Google Scholar]
• Bowlby J (1958). The nature of the child’s tie to his mother. The International Journal of Psychoanalysis, 39, 350–373 [PubMed] [Google Scholar]
• Brooks-Gunn J, Han W, & Waldfogel J (2010). First-year maternal employment and child development in the first 7 years: Monographs of The Society for Research in Child Development, 75(2). [PMC free article] [PubMed] [Google Scholar]
• Bronfenbrenner U & Crouter AC (1982). Work and family through time and space In Kamerman SB & Hayes CD (Eds.), Families that work: Children in a changing world (pp. 39–83). Washington, D.C.: National Academy Press. [Google Scholar]
• Chase-Lansdale PL, Moffitt RA, Lohman BJ, Cherlin AJ, Coley RL, Pittman LD, & Votruba-Drzal E (2003). Mothers’ transitions from welfare to work and the well-being of preschoolers and adolescents. Science, 299(5612), 1548–1552. doi: 10.1126/science.1076921 [PubMed] [CrossRef] [Google Scholar]
• Coley RL, & Lombardi CM (2013). Does maternal employment following childbirth support or inhibit low‐income children’s long‐term development? Child Development, 4(1), 178–197. doi: 10.1111/j.1467-8624.2012.01840.x [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Costigan CL, Cox MJ, & Cauce AM (2003). Work-parenting linkages among dual-earner couples at the transition to parenthood. Journal of Family Psychology, 17(3), 397–408. doi: 10.1037/0893-3200.17.3.397 [PubMed] [CrossRef] [Google Scholar]
• Dunifon R, Kalil A, & Danziger SK (2003). Maternal work behavior under welfare reform: How does the transition from welfare to work affect child development? Children and Youth Services Review, 25(1–2), 55–82. doi: 10.1016/S0190-7409(02)00266-9 [CrossRef] [Google Scholar]
• Gassman‐Pines A (2011). Associations of low‐income working mothers’ daily interactions with supervisors and mother‐child interactions. Journal of Marriage and Family, 73(1), 67–76. doi: 10.1111/j.1741-3737.2010.00789.x [CrossRef] [Google Scholar]
• Gerson K (2010). The unfinished revolution: How a new generation is reshaping family, work and gender in America. New York, NY: Oxford University Press. [Google Scholar]
• Goodman WB, Crouter AC, Lanza ST, & Cox MJ (2008). Paternal work characteristics and father-infant interactions in low-income, rural families. Journal of Marriage and Family, 70(3), 640–653. doi: 10.1111/j.1741-3737.2008.00511.x [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Goodman WB, Crouter AC, Lanza ST, Cox MJ, & Vernon‐Feagans L (2011). Paternal work stress and latent profiles of father–infant parenting quality. Journal of Marriage and Family, 73(3), 588–604. doi: 10.1111/j.1741-3737.2011.00826.x [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Greenberger E, O’Neil R, & Nagel SK (1994). Linking workplace and homeplace: Relations between the nature of adults’ work and their parenting behaviors. Developmental Psychology, 30(6), 990–1002. doi: 10.1037/0012-1649.30.6.990 [CrossRef] [Google Scholar]
• Hayes A (2013). Introduction to mediation, moderation, and conditional process analysis. Guilford Press: NY. [Google Scholar]
• Huston AC, & Aronson SR (2005). Mothers’Time With Infant and Time in Employment as Predictors of Mother-Child Relationships and Children’s Early Development. Child Development, 76(2), 467–482. doi: 10.1111/j.1467-8624.2005.00857.x [PubMed] [CrossRef] [Google Scholar]
• Kline Rex B. , (2004). Beyond significance testing: Reforming data analysis methods in behavioral research. , (pp. 61–91). Washington, DC, US: American Psychological Association, xii, 325 pp. 10.1037/10693-003 [CrossRef] [Google Scholar]
• Kohn ML, (1977). Class and conformity: A study in values with a reassessment. Chicago, IL: University of Chicago Press. [Google Scholar]
• Kohn ML, & Schooler C (1973). Occupational experience and psychological functioning: An assessment of reciprocal effects. American Sociological Review, 38(1), 97–118. doi: 10.2307/2094334 [CrossRef] [Google Scholar]
• Klute MM, Crouter AC, Sayer AG, & McHale SM (2002). Occupational self-direction, values, and egalitarian relationships: A study of dual-earner couples. Journal of Marriage and the Family, 64, 139–151. [Google Scholar]
• Laughlin L (2011). Maternity leave and employment patterns of first-time mothers: 1961–2008 Current Population Report, 70–128. U.S. Census Bureau, Washington, DC. [Google Scholar]
• Lombardi CM, & Coley RL (2013). Low-income mothers’ employment experiences: Prospective links with young children’s development. Family Relations: An Interdisciplinary Journal of Applied Family Studies, 62(3), 514–528. doi: 10.1111/fare.12018 [CrossRef] [Google Scholar]
• Lucas-Thompson RG, Goldberg WA, & Prause J (2010). Maternal work early in the lives of children and its distal associations with achievement and behavior problems: A meta-analysis. Psychological Bulletin, 136(6), 915–942. doi: 10.1037/a0020875 [PubMed] [CrossRef] [Google Scholar]
• Luster T, Rhoades K & Hass B (1989). The relation between parental values and parenting behavior: A test of Kohn’s hypothesis. Journal of Marriage and the Family, 51, 139–47. [Google Scholar]
• MacKinnon DP, Krull JL, & Lockwood CM (2000). Equivalence of the mediation, confounding and suppression effect. Prevention Sciences, 1(4), 173–181. [PMC free article] [PubMed] [Google Scholar]
• McGill BS (2014). Navigating new norms of involved fatherhood: Employment, fathering attitudes, and father involvement. Journal of Family Issues, 35(8), 1089–1106. doi: 10.1177/0192513X14522247 [CrossRef] [Google Scholar]
• Meteyer KB, & Perry-Jenkins (2009). Dyadic parenting and children’s externalizing symptoms. Family Relations, 58 (3), 289. [PMC free article] [PubMed] [Google Scholar]
• Parcel TL, & Menaghan EG (1994). Parents’ jobs and children’s lives. New York: Aldine de Gruyter. [Google Scholar]
• Perry-Jenkins M, Goldberg A, Smith JZ, & Logan J (2011) Working-class jobs and new parents’ mental health. Journal of Marriage and Family, 73(5), 1117–1132. [Google Scholar]
• Perry-Jenkins M, Repetti RL, & Crouter AC (2000). Work and family in the 1990s. Journal of Marriage and the Family, 62(4), 981–998. doi: 10.1111/j.1741-3737.2000.00981.x [CrossRef] [Google Scholar]
• Presser HB (2003). Working in a 24/7 economy: Challenges for American families. New York, NY, US: Russell Sage Foundation. [Google Scholar]
• Raver CC (2003). Does Work Pay Psychologically as Well as Economically? The Role of Employment in Predicting Depressive Symptoms and Parenting Among Low-Income Families. Child Development, 74(6), 1720–1736. doi: 10.1046/j.1467-8624.2003.00634.x [PubMed] [CrossRef] [Google Scholar]
• Reynolds CR, & Kamphaus RW (1992). Behavior Assessment System for Children: Manual. Circle Pines, MN: American Guidance. [Google Scholar]
• Roeters A, Van der Lippe T, & Kluwer ES (2010). Work characteristics and parent‐child relationship quality: The mediating role of temporal involvement. Journal of Marriage and Family, 72(5), 1317–1328. doi: 10.1111/j.1741-3737.2010.00767.x [CrossRef] [Google Scholar]
• Secret M, & Peck-Heath C (2004). Maternal labor force participation and child well-being in public assistance families. Journal of Family Issues, 25(4), 520–541. doi: 10.1177/0192513X03257761 [CrossRef] [Google Scholar]
• Shelton KK, Frick PJ, & Wootton J (1996). Assessment of parenting practices in families of elementary school-age children. Journal of Clinical Child Psychology, 25(3), 317–329. doi: 10.1207/s15374424jccp2503_8 [CrossRef] [Google Scholar]
• Shrout PE, & Bolger N (2002). Mediation in experimental and nonexperimental studies: New procedures and recommendations. Psychological Methods, 7(4), 422–445. [PubMed] [Google Scholar]
• Spector PE, Jex SM, & Chen PY (1995). Relations of incumbent affect-related personality traits with incumbent and objective measures of characteristics of jobs.” Journal of Organizational Behavior,16. 59–65. [Google Scholar]
• Wheeler LA, Updegraff KA, & Crouter A (2011). Work and Mexican American parent–adolescent relationships: The mediating role of parent well-being. Journal of Family Psychology, 25(1), 107–116. doi: 10.1037/a0022440 [PMC free article] [PubMed] [CrossRef] [Google Scholar]
• Yetis-Bayraktar A, Budig MJ, & Tomaskovic-Devey D (2013). From the shop floor to the kitchen floor: Maternal occupational complexity and children’s reading and math skills. Work and Occupations, 40(1), 37–64. doi: 10.1177/0730888412465879 [CrossRef] [Google Scholar]