Mexican-Origin Parents’ Work Conditions and Adolescents’ Adjustment

Abstract

Mexican-origin parents’ work experiences are a distal extra-familial context for adolescents’ adjustment. This two-wave multi-informant study examined the prospective mechanisms linking parents’ work conditions (i.e., self-direction, work pressure, workplace discrimination) to adolescents’ adjustment (i.e., educational expectations, depressive symptoms, risky behavior) across the transition to high school drawing on work socialization and spillover models. We examined the indirect effects of parental work conditions on adolescent adjustment through parents’ psychological functioning (i.e., depressive symptoms, role overload) and aspects of the parent-adolescent relationship (i.e., parental solicitation, parent-adolescent conflict), as well as moderation by adolescent gender. Participants were 246 predominantly immigrant, Mexican-origin, two-parent families who participated in home interviews when adolescents were approximately 13 and 15 years of age. Results supported the positive impact of fathers’ occupational self-direction on all three aspects of adolescents’ adjustment through decreased father-adolescent conflict, after controlling for family socioeconomic status and earner status, and underemployment. Parental work pressure and discrimination were indirectly linked to adolescents’ adjustment, with different mechanisms emerging for mothers and fathers. Adolescents’ gender moderated the associations between fathers’ self-direction and girls’ depressive symptoms, and fathers’ experiences of discrimination and boys’ risk behavior. Results suggest that Mexican-origin mothers’ and fathers’ perceptions of work conditions have important implications for multiple domains of adolescents’ adjustment across the transition to high school.

Parents’ work serves as an important extra-familial context that distally shapes adolescents’ adjustment (Bronfenbrenner & Crouter, 1982). Positive work conditions have been found to foster parents’ well-being (Perry-Jenkins & Wadsworth, 2013), parent-child relationships (Greenberger, O’Neil, & Nagel, 1994), and children’s psychosocial adjustment (Perry-Jenkins & Gillman, 2000). Conversely, parents’ experiences of work conditions that may be deemed stressful in nature are linked to parents’ feelings of overload and distress, poor parent-child relationships, and child adjustment problems (see Crouter & Bumpus, 2001, for a review). Much of the evidence on work-family linkages has been found with European American families; we know little about the effects of work conditions on Mexican origin families (Perry-Jenkins & Wadsworth, 2013), a large ethnic subgroup (U.S. Census Bureau, 2014).

Advancing our understanding of work-family linkages for Mexican-origin families is critical for two reasons. First, ecological perspectives (Bronfenbrenner & Crouter, 1982; García Coll et al., 1996) highlight the larger sociocultural context in which development unfolds, including the distal and proximal contexts that touch on youth’s lives. In particular, Mexican-origin adolescents are at high risk for mental health problems (Centers for Disease Control and Prevention [CDC], 2012) and school dropout (U.S. Census Bureau, 2014), and are overrepresented among delinquent youth (Piquero, 2008). Second, these disparities have been attributed to Mexican-origin families’ disproportionate exposure to contextual risk factors, such as poverty, discrimination, and stressful work conditions (Gonzales, Germán, & Fabret, 2012; Yoshikawa, 2011). Thus, findings from other ethnic groups may not generalize to youth from Mexican sociocultural backgrounds; they must be studied directly.

This study was informed by work socialization (e.g., Kohn & Schooler, 1982) and work-family spillover models (e.g., Repetti, 1987) that theorize about the ways that parents’ work, psychological functioning, and family processes jointly affect youth adjustment. Integrating these perspectives, this study examined indirect links between parents’ work conditions (i.e., self-direction, work pressure and discrimination) and changes in adolescents’ adjustment (i.e., educational expectations, depressive symptoms, risk behavior) through the intervening processes of parents’ psychological functioning (i.e., depressive symptoms, role overload) and parent-adolescent relationships (i.e., parental solicitation, parent-adolescent conflict). We focused on the transition from junior high to high school as a critical turning point associated with declines in achievement (Dotterer, McHale, & Crouter, 2009), and increases in externalizing (Boyer, 2006) and internalizing (Newman, Newman, Griffen, O’Connor, & Spas, 2007) behaviors. Drawing on research on the salience of gender dynamics in Mexican-origin families (e.g., Raffaelli & Ontai, 2004) and gender intensification perspectives (e.g., Maccoby, 1998) that suggest the importance of same-gender parent-youth dyads for development, we examined youth gender as a moderator of the proposed associations.

Parents’ Work Conditions and Adolescents’ Adjustment

This study conceptualized multiple indicators of parents’ work conditions as critical features of the extra-familial context that shape youth adjustment (Bronfenbrenner & Crouter, 1982). Prior studies with primarily European American samples have shown that parents’ work has both positive and negative implications for adolescents’ adjustment (Perry-Jenkins & Wadsworth, 2013). The degree to which the same associations hold for Mexican-origin parents is unknown. Differences in adolescents’ adjustment risks and parents’ work experiences in Mexican-origin families as compared to those included in previous studies reinforce the importance of testing the associations within this group. For Mexican-origin parents, self-direction (Yoshikawa, 2011), work pressure (Updegraff, Crouter, Umaña-Taylor, & Cansler, 2007), and workplace discrimination (Yoshikawa, 2011) may be particularly salient.

Scholars have defined self-direction as the extent that the workplace offers autonomy, complex tasks, and minimal supervision (Kohn & Schooler, 1982). Workplace self-direction often characterizes professional more so than unskilled positions, but research with ethnic minority parents has found variability in levels of self-direction across low-wage jobs (Yoshikawa, 2011). Across varying work contexts, parents’ occupational self-direction has been positively associated with adolescents’ expectations for educational success (primarily African American families; Gardner Neblett & Schnabel Cortina, 2006) and mental health (fathers and pre-adolescent daughters, but not mothers, in dual-earner working-class families; Perry-Jenkins & Gillman, 2000), and protective against later problem behavior (for fathers of young children in nationally representative sample; Parcel & Menaghan, 1993).

Work pressure is common in job environments characterized by deadlines, demands, and a fast pace. Few studies have specifically examined work pressure among Mexican-origin samples, though research has found that work stress is higher among Latin American as compared to European American professionals (Rodriguez-Calcagno & Brewer, 2005). Research among European American dual-earner working- and middle-class families found parents’ work pressure linked to higher levels of adolescents’ depressive symptoms (Crouter, Bumpus, Maguire, & McHale, 1999) and problem behaviors (Galambos, Sears, Almeida & Kolaric, 1995).

As work-family scholarship extends to diverse populations, workplace discrimination is a particularly critical aspect of the work experience to consider (Hughes & Dodge, 1997). Mexican-origin workers in the U.S. report high rates of workplace discrimination (Roberts, Swanson, & Murphy, 2004). Few studies have specifically focused on the implications of Mexican-origin parents’ perceived workplace discrimination for adolescents’ adjustment. In our own research, we found fathers’ workplace discrimination to be linked with adolescents’ depressive symptoms (Crouter, Davis, Updegraff, Delgado, & Fortner, 2006). More generally, discrimination is an important correlate of Mexican-origin adolescents’ adjustment, including depressive symptoms, problem behavior, and educational expectations (Gonzales et al., 2012).

Intervening Processes of Work-Adolescent Adjustment Links

Work socialization and spillover perspectives provide hypotheses about how work conditions relate to family dynamics and individual adjustment. The work socialization literature argues that opportunities for autonomy and complexity in the workplace shape social competencies, behaviors, and outlooks that workers generalize to life off the job (Kohn & Schooler, 1982). In particular, parents’ experiences of workplace self-direction may shape their beliefs and values that influence parenting behaviors, and ultimately affect children’s behavior. A theoretically distinct, but related line of research, an affect spillover perspective (e.g., Repetti, 1987), suggests that one’s mood at work influences mood and behaviors in the family domain. Parents’ adverse work experiences may elicit stress reactions, depressive symptoms and feelings of overload, in turn taxing parents’ resources and coping mechanisms and distracting parents from the needs of their children (Crouter & Bumpus, 2001). Conversely, work characterized by positive conditions and experiences may lead to less distress overall, enhancing mood, contributing positively to parenting, and, in turn, to better adolescent adjustment.

Prior research guided by work socialization and spillover perspectives supports the role of parents’ psychological functioning and parent-youth relationships as important intervening factors that link parents’ work to youth adjustment. Parents’ work pressure has been indirectly linked to adolescents’ poor adjustment (depressive symptoms, low general self-worth; problem behaviors) through low levels of parents’ psychological functioning (depressive symptoms, role overload) and parent-adolescent relationship quality (conflict) among working- and middle-class European American (Crouter et al., 1999) and Canadian families (Galambos et al., 1995). Informed by these two traditions, we extended prior research by simultaneously examining multiple domains of parents’ psychological functioning and parent-adolescent relationships to gain an understanding of the indirect links between positive and negative parental work conditions to adolescent adjustment among Mexican-origin families.

Parents’ psychological functioning

There is some evidence suggesting that parents’ psychological functioning as linked to parent-youth relationships is one explanatory mechanism of the indirect relation between parents’ work and youth adjustment. With the current sample, we found cross-sectional evidence of parents’ work pressure linked to parent-adolescent conflict through parents’ depressive symptoms and role overload (i.e., the feeling there is too much to do and not enough time to do it; Wheeler, Updegraff, & Crouter, 2011). These results are consistent with research using European American (Crouter et al., 1999) and national (Vandewater & Lansford, 2005) samples. Research also has found positive effects of self-direction on parent-youth relationships through parents’ psychological functioning. Specifically, research has found cross-sectional links between parents’ self-direction and parent-youth relationship quality via low levels of parents’ depressive symptoms for mothers of adolescents (Wheeler et al., 2011) and middle-class (Greenberger et al., 1994) families. Thus, we examined parents’ depressive symptoms and role overload simultaneously.

Parent-adolescent relationships

The current study focused on aspects of parent-adolescent relationships that may be particularly salient to working parents in Mexican-origin families. Given Mexican-origin parents’ tendency toward a protective and controlling parenting style (e.g., Halgunseth, Ispa, & Rudy, 2006), they may rely more on active methods of acquiring knowledge, such as parental solicitation (i.e., a component of parental monitoring tapping parents’ direct efforts to learn about their children’s experiences; Stattin & Kerr, 2000), than on passive methods, such as youth self-disclosure. Moreover, parental solicitation as an active parenting approach may be amenable to the effects of parents’ work. For example, parents with demanding jobs may have less knowledge of children’s activities, possibly because work stress may lead parents to withdraw from family interactions (Crouter & Bumpus, 2001). Furthermore, conflict with parents is a key process in adolescence that may be especially salient for family members’ well-being in Latino families. There is evidence that conflictual family relations are particularly detrimental to Mexican-origin youth’s adjustment, possibly because their cultural context places a strong emphasis on cohesive and harmonious family relationships (Cauce & Domenech-Rodríguez, 2002). Further, findings show that conflict is linked to multiple dimensions of Mexican-origin adolescents’ adjustment (e.g., problem behaviors, depressive symptoms, academic problems; Pasch et al., 2006), and has more negative effects on adolescent adjustment in Latino as compared to non-Latino families (Mechanic & Hansell, 1989). This suggests conflict may be a particularly important predictor of adjustment in Mexican-origin families. Based on this research, we targeted parents’ solicitation and parent-youth conflict.

To our knowledge, no prior studies have examined parental solicitation or parent-youth conflict as possible links between Mexican-origin parents’ work and adolescent adjustment. There is empirical support for an indirect link from negative work experiences through parent-youth conflict to adolescent problem behaviors in a Canadian sample (Galambos et al., 1995) and to adolescents’ negative attitudes toward education in a Finnish sample (Sallinen, Kinnunen, & Rönkä, 2004). Research has linked lower parental knowledge to higher levels of risky behavior (e.g., Crouter, Bumpus, Davis, & McHale, 2005) and less positive educational outcomes (Crouter, MacDermid, McHale, & Perry-Jenkins, 1990) among dual-earner families.

Adolescents’ Gender as a Moderator of Work-Adolescent Adjustment Links

In families of Mexican origin, one source of variation in how larger social and economic forces may influence family dynamics and adolescents’ adjustment is gender. Gender is as an organizing feature in Mexican culture that may have implications for the potentially different roles of mothers and fathers and experiences of daughters and sons (Cauce & Domenech-Rodríguez, 2002). Little is known about the role of gender in the links between parents’ work experiences, family processes, and adolescents’ adjustment among Mexican-origin families, however. One possibility is that in the context of more traditional attitudes toward gender, parents play a more central role in the lives of same-gender offspring (Raffaelli & Ontai, 2004). Likewise, gender intensification perspectives (e.g., Maccoby, 1998) suggest that youth are more involved and receptive to socialization efforts by their parents of the same gender. Thus, the effects of parents’ work spillover may be more pronounced for same-gender dyads. A second possibility stems from literature indicating that traditional Mexican-origin families may place a greater emphasis on daughters’ family responsibilities and involvement as compared to sons’ (Cauce & Domenich-Rodríguez, 2002); thus, the hypothesized linkages may be more apparent for daughters than for sons, and, in particular, for mother-daughter dyads.

Current Study

Consistent with work socialization (Kohn & Schooler, 1982) and spillover (Repetti, 1987) models, we expected that more positive experiences at work, as indexed by more self-direction and less work pressure and discrimination, would be linked to more positive parent psychological functioning (lower levels of parental depressive symptoms and role-overload), that in turn, would be linked to more positive parent-youth relationships (more solicitation by parents and less conflicts with them), which would be in turn linked to better adolescent adjustment (higher academic expectations and lower depressive symptoms and risky behavior). Furthermore, as guided by cultural (Cauce & Domenech-Rodríguez, 2002) and gender intensification perspectives (e.g., Maccoby, 1998), we explored gender as a moderator, hypothesizing more pronounced spillover for same-gender dyads, and, in particular, for mother-daughter dyads. Covariates included prior levels of parent-adolescent relationship quality, adolescents’ adjustment and gender, and family socioeconomic status (e.g., Gonzales et al., 2012), as prior research has linked them to differences in adolescents’ adjustment. To hold constant salient conditions of Mexican-origin parents’ work (Updegraff et al., 2007; Yoshikawa, 2011), we also controlled for underemployment and earner status (i.e., single-earner versus dual-earner family).

Method

Participants

Data came from a longitudinal study of 246 Mexican-origin families (Updegraff, McHale, Whiteman, Thayer, & Delgado, 2005) who were recruited through five school districts and five parochial schools that served ethnically and linguistically diverse communities in a southwestern metropolitan area. Eligible families included those with: (a) mothers of Mexican origin, (b) a 7^th grader and an older sibling living with their biological mother and biological or long-term adoptive (i.e., a minimum of 10 years) father, and (c) fathers working at least 20 hours/week. Most fathers (93%) were also of Mexican origin. There were 421 eligible families (23% of initial rosters; 32% of those contacted and screened for eligibility); 67% agreed to participate, 23% refused, and 10% were unreachable.

The current study included mothers, fathers, and young adolescents in families that represented a range of socioeconomic levels from poverty (18.3%) to upper class (median family income = $40,000) at Time 1 (T1). Most fathers (98%) and mothers (66%) were employed, with 65% being in dual-earner families. The majority of parents worked the day shift (65% fathers, 72% mothers), and worked an average of 41 hours per week (fathers’ M = 46.88, SD = 11.52; mothers’ M = 35.96, SD = 11.98). Mothers’ occupations ranged in prestige from dishwasher to teacher, with the modal occupation of housekeeper; for fathers, occupations ranged from car detailer to psychiatrist, with the modal occupations of maintenance and construction workers. Parents were primarily born outside the U.S. (70%); this subset of parents had lived in the U.S. an average of 12.38 (SD = 8.86) and 15.18 (SD = 8.78) years for mothers and fathers, respectively. Mothers and fathers reported an average of 10 years of education (mothers’ M = 10.34, SD = 3.74; fathers’ M = 9.88, SD = 4.37) and were on average 40 years of age (mothers’ M = 39.00, SD = 4.63; fathers’ M = 41.70, SD = 5.78). Most parents completed interviews in Spanish (almost 70%), whereas most youth completed interviews in English (84%). Adolescents on average were 12.77 (SD = .58) years of age, 51% female, and 62% U.S.-born.

The second set of interviews (Time 2; T2) were conducted two years after T1 when adolescents averaged 15.10 years (SD = .46) and were in the 9^th grade. The retention rate was 91% of T1 families (n = 222). Attrition analyses between non-participating (n = 24) and participating families revealed a difference in fathers’ employment status, χ²(1) = 5.62, p < .05, with a greater percentage of fathers working in participating (98.6%) as compared to non-participating families (1.4%). This variable was included in analyses to improve estimation under conditions of missing data. There were no other differences in demographic or study variables.

Procedure

Trained bilingual interviewers conducted structured in-home computer-assisted interviews with parents and adolescents in their preferred language (English or Spanish) at T1, and structured phone interviews with adolescents at T2. Families received a $100 honorarium at T1, and adolescents received $40 at T2. The Institutional Review Board approved all procedures.

Measures

Separate individuals forward translated to Spanish and back translated to English all measures for the local Mexican dialect. The research team resolved discrepancies. For all measures, higher scores indicate more of the construct named (e.g., higher work pressure).

Demographic/control variables (T1; parent report)

Parents answered demographic questions, including annual family income and educational attainment (0 = no schooling to 21 = advanced graduate degrees). Family socioeconomic status (SES) was created from the standardized average of both mothers’ and fathers’ educational attainment and family income (transformed to correct for positive skew; α = .76). Family earner status was indexed by creating two groups: both parents working for pay (dual-earner) or one parent working for pay (single-earner). Underemployment was measured using an average of seven items created for this study to capture parents’ perceptions of their jobs tapping their full earning and skill potential (e.g., “Given my skills, education, and experience, I should be in a better job than my current job”). Parents rated items on a 1 = strongly disagree to 5 = strongly agree scale (parents’ αs = .92).

Parents’ work conditions (T1; parent report)

Self-direction of working mothers’ and fathers’ jobs was assessed with Lennon’s (1994) 20-item Work Dimensions Scale designed to measure specific tasks or behaviors on a 4-point scale from 1 = not at all to 4 = very much (α = .87, mothers and fathers). Items (e.g., “You decide on your own how to go about doing the work”) captured the three components of occupational self-direction: lack of closeness of supervision, low routinization, and high complexity. The items were averaged to create the scale score. Working parents rated their work pressure with nine items (Moos, 1986) on a 4-point scale from very true to very untrue (mothers’ α = .81; fathers’ α = .72). The scale items were averaged to assess the degree that parents’ work environments are dominated by work and time demands (e.g., “There is constant pressure to keep working”). A measure of workplace discrimination assessed the extent that working parents experienced discrimination and bias in the workplace. Using a combination of Hughes and Dodge’s (1997) measures of Institutional Discrimination and Interpersonal Prejudice in the Workplace, a 12-item scale was created by averaging items (α = .89 for mothers and .88 for fathers; e.g., “Mexicans/Mexican Americans get the least desirable assignments”) on a scale of 1 = strongly disagree to 4 = strongly agree.

Parents’ psychological functioning (T1; parent report)

The 20-item Center for Epidemiological Studies Depression Scale (CES-D; Radloff, 1977) was used to measure parents’ depressive symptoms on a 4-point scale (0 = rarely or none of the time to 3 = most of the time; mothers’ α = .88; fathers’ α = .86). Scores were created by summing items (e.g., “I had crying spells”). Parents’ role overload was assessed with an adapted version of the 13-item Role Overload Scale (House & Rizzo, 1972; Reilly, 1982) on a 4-point scale (1 = very untrue to 4 = very true). Items were averaged to measure parents’ sense that there is too much to do and not enough time to do it (e.g., “There are too many demands on my time”; parents’ α = .92).

Parent-adolescent relationships (T1, T2; adolescent report)

We assessed parental solicitation with Stattin and Kerr’s (2000) measure of parents’ efforts to acquire knowledge about their children’s daily activities. Adolescents reported the frequency (1 = almost never to 5 = almost always) of their mothers’ and fathers’ (both parents’ α = .78 T1, .77 T2) efforts to find out about their daily activities (6 items; e.g., “Does your mom/dad ask you about your mood or feelings?”). With an adapted version of measures by Smetana (1995) and Harris (1992), we assessed conflict in the parent-adolescent relationship by averaging adolescents’ reports of the frequency (1 = not at all to 6 = several times a day) of conflict between parents and youth on 12 topics (e.g., “How often in the past year have you had disagreements with your mom/dad about how late you stay up/out?”; mothers’ αs = .81 T1, .85 T2, fathers’ α = .86 at both time points).

Adolescents’ adjustment (T1, T2; adolescent report)

Adolescents reported on their educational expectations by responding to: “How far do you really think you’ll go in school?” Responses were made on a continuous scale representing the total number of years of education (e.g., 12 = high school diploma, 16 = bachelor’s degree). Adolescents rated the frequency of their depressive symptoms (20-items; CES-D; Radloff, 1977) on a 4-point scale (0 = rarely or none of the time to 3 = most of the time; αs = .84 for T1 and T2). We created the scale by summing items (e.g., “I had crying spells”). Adolescents rated the frequency they engaged in 23 risk behaviors (Eccles & Barber, 1990) during the past year (e.g., “skip a day of school”) on a 4-point scale (1= never to 4 = more than 10 times). Items were averaged (T1 α =.92; T2 α = .89).

Results

We tested hypothesized pathways using structural equation modeling (SEM) in Mplus 7.11 (Muthén & Muthén, 1998–2013). First, confirmatory factor analysis was used to evaluate the fit of the measurement models for all constructs (available from the first author). Second, we estimated separate SEMs for each parent to examine the influence of parental work on adolescents’ adjustment. Correlations are presented in Table 1.

Table 1.

Bivariate Correlations Among Latent Factors (Paternal Model above the Diagonal; Maternal Model below the Diagonal; N = 246 Families)

Variables	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19
1. T1 P. SD	-	−.08	−.09	−.23*	.02	.01	−.22*	.19*	.03	−.02	.04	−.22*	.16*	−.17*	−.13†	.67*	−.32*	.07	−.17*
2. T1 P. WP	−.17†	-	.30*	.22*	.39*	.07	.06	.01	.01	−.04	.02	.19*	−.03	.04	.08	−.25*	.20*	−.17*	−.02
3. T1 P. WD	−.24*	.31*	-	.26*	.20*	−.08	.07	.03	−.02	−.07	.03	.18*	−.03	−.03	−.02	−.16†	.22*	−.01	−.03
4. T1 P. DS	−.14†	.50*	.44*	-	.30*	−.09	.10	−.14†	.11	.13†	−.22*	.24*	−.13†	.19*	.25*	−.31*	.31*	−.06	.19*
5. T1 P. RO	.04	.52*	.30*	.53*	-	.18*	.10	−.13†	.08	−.08	−.06	.09	−.12†	.00	−.04	−.09	.20*	−.07	.01
6. T2 solicit	.07	.08	−.10	−.10	−.07	-	−.01	.18*	−.18*	−.27*	.50*	.05	.09	−.02	−.12	−.05	−.22*	−.17*	−.07
7. T2 conflict	−.05	.05	.07	.28*	.07	−.22*	-	−.19*	.34*	.34*	.05	.48*	−.10	.31*	.30*	−.20*	.01	−.01	.08
8. T2 A. EE	.08	.00	.02	−.25*	−.02	.12	−.16*	-	−.08	−.28*	.24*	−.14*	.26*	−.05	−.20*	.25*	−.25*	.09	−.31*
9. T2 A. DS	.01	.23*	.15†	.28*	.21*	−.15†	.35*	−.09	-	.28*	−.09	.16*	−.07	.40*	.22*	−.04	−.03	−.00	−.29*
10. T2 A. RB	.15*	−.11	.02	.22*	.04	−.23*	.53*	−.30*	.28*	-	−.11	.26*	−.12†	.22*	.63*	.09	.03	.09	.19*
11. T1 solicit	.02	−.13	.00	−.12	.02	.40*	−.13†	.15*	−.07	−.11	-	.10	.12†	−.15*	−.14†	.12	−.12†	.07	−.07
12. T1 conflict	−.08	.03	.04	.27*	−.02	.06	.49*	−.09	.23*	.35*	−.13	-	−.15*	.49*	.40*	−.29*	.12	−.09	.17*
13. T1 A. EE	.16*	.06	.02	−.12†	.04	.08	−.07	.26*	−.06	−.12†	.26*	−.27*	-	−.23*	−.28*	.27	−.21*	−.01	−.17*
14. T1 A. DS	−.03	.07	.17*	.33*	.23*	−.06	.25*	−.07	.41*	.21*	−.13†	.51*	−.23*	-	.59*	−.28*	.07	−.03	−.08
15. T1 A. RB	.01	.01	.04	.29*	.13†	−.17*	.35*	−.22*	.22*	.62*	−.19*	.49*	−.28*	.59*	-	−.12	.09	.09	.15*
16. T1 SES	.34*	−.27*	−.05	−.26*	−.03	.04	−.09	.20*	−.02	.12	.16*	−.22*	.26*	−.27*	−.08	-	−.48*	.38*	−.13†
17. T1 P. UE	−.48*	.24*	.44*	.33*	.21*	−.05	.14†	−.10	.05	.12	.04	.21*	−.08	.12	.06	−.20*	-	.03	.13*
18. T1 ES	−.18*	.06	.48*	.09	.20*	−.09	.01	.09	−.01	.09	.00	−.06	−.01	−.03	.09	.33*	.31*	-	.01
19. A. gender	−.11	−.15†	−.09	.07	−.18*	−.18*	.14*	−.31*	−.30*	.19*	−.23*	.11	−.17*	−.08	.15*	−.08	−.02	.01	-

Note. T1 = time 1 (7^th grade). T2 = time 2 (9^th grade). P = parent report. A = adolescent report. SD = self-direction. WP = work pressure. WD = workplace discrimination. DS = depressive symptoms. RO = role overload. EE = educational expectations. RB = risk behavior. UE = underemployment. WH = work hours. ES = earner status. Adolescent gender coded as: 0 = girls, 1 = boys. Earner status coded 0 = single-earner, 1 = dual-earner.

^†p < .10.

^*p < .05.

Links between Parents’ Work Conditions and Adolescents’ Adjustment

We examined indicators of parents’ work conditions (i.e., T1 self-direction, work pressure, workplace discrimination) as predictors, parents’ psychological functioning (i.e., T1 depressive symptoms, role overload) and the quality of parent-adolescent relationships (i.e., T2 parental solicitation, parent-adolescent conflict) as intervening variables, and adolescents’ adjustment (i.e., T2 educational expectations, depressive symptoms, risk behavior) as outcomes. The control variables (all T1) were adolescent gender (0 = girls, 1 = boys), family SES and earner status (0 = single-earner, 1 = dual-earner), parental underemployment, parent-adolescent relationships, and adolescent adjustment. We allowed exogenous, endogenous, and within-time variables to correlate. Model fit was assessed using fit indices as recommended for longitudinal studies (Little, 2013), including the chi-square statistic, root mean square error of approximation (RMSEA ≤ .08), the CFI (≥ .90), and the Tucker-Lewis Index (TFI ≥ .90). We used full information maximum likelihood and auxiliary variables (i.e., parents’ work status) to maintain power and improve estimation under conditions of missing data (Enders, 2010).

To test for indirect effects (referred to as b₁b₂b₃ or b₁b₂), we used the product of coefficients method using bias-corrected bootstrapping with 1000 resamples to calculate the confidence intervals (CI; Taylor, MacKinnon, & Tein, 2008). With this method, it is not necessary to have a significant total or direct effect of X on Y to have a significant indirect effect (Taylor et al.). To test for moderation by adolescent gender, we estimated multiple group models, first allowing estimation of the path coefficients to vary freely across girls and boys, and then constraining paths to be equal for girls and boys when a structural path coefficient of interest was significant for one group and not the other group. We conducted model comparisons using the χ² difference test (i.e., Δχ², p < .05 indicating moderation).

We present results first for the direct associations, followed by indirect associations, and lastly moderated associations. The father and mother models demonstrated adequate fit to the data and accounted for significant variance in intervening variables and outcomes (Figures 1 and 2, respectively). For fathers (Figure 1), all three of the hypothesized work conditions were associated with fathers’ psychological functioning. Self-direction was associated with lower levels of fathers’ depressive symptoms, whereas discrimination was associated with higher levels of fathers’ depressive symptoms. Work pressure was associated with higher levels of role overload. Additionally, fathers’ self-direction was associated with decreased father-adolescent conflict two years later. Turning to the results for psychological functioning, only fathers’ role overload was significant and related to increased solicitation and decreased education expectations two years later. The results for father-adolescent relationships revealed that solicitation linked with adolescents’ increased educational expectations, and decreased depressive symptoms and risk behavior. Father-adolescent conflict linked with adolescents’ decreased educational expectations, and increased depressive symptoms and risk behavior.

Figure 1.

Structural Equation Model Linking Fathers’ Work Experiences to Adolescents’ Adjustment (N = 246 Families).

Note. ¹For figure simplification, only significant (p < .05) standardized path estimates (standard errors) presented; the measurement model, the exogenous (X1 – X3) and endogenous (Y1 – Y3) covariance, within-time covariance, and control variable paths were not included.

²Model fit: χ2(1023) = 1562.21, p < .001. RMSEA = .05(.04, .05); CFI = .92; TLI = .91.

³Significant control variable paths (p < .05): Adolescents’ gender → T2 educational expectations, β = −.26 (.28). Adolescents’ gender → adolescents’ T2 depressive symptoms, β = −.31 (.06). Adolescents’ T1 depressive symptoms → adolescents’ T2 depressive symptoms, β = .29 (.10). Adolescents’ T1 risk behavior → T2 risk behavior, β = .55 (.08). Father-adolescent T1 conflict → father-adolescent T2 conflict, β = .47 (.07). Fathers’ T1 solicitation → fathers’ T2 solicitation, β = .51 (.08).

⁴Significant indirect effects (bold arrows): work pressure (WP) → role overload (RO) → expectations (EE), b₁b₂ = −.22, 95% CI [−.558, −.003]. WP → RO → solicitation (SO) → EE, b₁b₂b₃ = .05, 95% CI [.005, .165], depressive symptoms (DS), b₁b₂b₃ = −.02, 95% CI [−.044, −.004], and risk behavior (RB), b₁b₂b₃ = −.01, 95% CI [−.031, −.002]. Self-direction → conflict → EE, b₁b₂ = .09, 95% CI [.006, .236], DS, b₁b₂ = −.03, 95% CI [−.061, −.004], and RB, b₁b₂ = −.02, 95% CI [−.044, −.003].

Figure 2.

Structural Equation Model Linking Mothers’ Work Experiences to Adolescents’ Adjustment (N = 246 Families).

Note. ¹For figure simplification, only significant (p < .05) standardized path estimates (standard errors) presented; the measurement model, the exogenous (X1 – X3) and endogenous (Y1 – Y3) covariance, within-time covariance, and control variable paths were not included.

²Model fit: χ2(1024) = 1519.15, p < .001. RMSEA = .04(.04, .05); CFI = .92; TLI = .91.

³Significant control variable paths (p < .05): Adolescents’ gender → T2 educational expectations, β = −.27 (.29). Adolescents’ gender → adolescents’ T2 depressive symptoms, β = −.32 (.06). Mothers’ T1 underemployment → T2 risk behavior, β = .21 (.02). Adolescents’ T1 depressive symptoms → adolescents’ T2 depressive symptoms, β = .33 (.33). Adolescents’ T1 risk behavior → T2 risk behavior, β = .47 (.07). Mother-adolescent T1 conflict → mother-adolescent T2 conflict, β = .48 (.10). Mothers’ T1 solicitation → mothers’ T2 solicitation, β = .30 (.07).

⁴Significant indirect effects (bold arrows): workplace discrimination (WD) → depressive symptoms (DS) → expectations (EE), b₁b₂ = −.24, 95% CI [−.665, −.023]. WD → solicitation (SO) → DS, b₁b₂ = .03, 95% CI [.001, .078].

For mothers, all three of the hypothesized work conditions were associated with mothers’ psychological functioning (Figure 2). Self-direction and work pressure were both associated with higher levels of role overload. Work pressure and discrimination were associated with higher levels of depressive symptoms. Additionally, mothers’ workplace discrimination was associated with decreased maternal solicitation two years later. Turning to the results for psychological functioning, mothers’ depressive symptoms were associated with increased mother-adolescent conflict and decreased education expectations two years later. The results for mother-adolescent relationships revealed that solicitation was associated with adolescents’ decreased depressive symptoms. Conflict linked to adolescents’ increased depressive symptoms and risk behavior.

Indirect effects

Figures 1 and 2 contain the results for the indirect effects for fathers and mothers, respectively. There were different patterns for fathers and mothers. For fathers, there was evidence of role overload, solicitation, and father-adolescent conflict as intervening variables, with variation in indirect pathways for each work condition. Work pressure was indirectly associated with increased adolescent adjustment in all three domains two years later through high levels of role overload and, in turn, increased paternal solicitation. Work pressure also indirectly linked to adolescents’ decreased educational expectations through higher levels of paternal role overload. Self-direction was indirectly associated with increased adolescent adjustment in all three domains through decreased father-adolescent conflict. Turning to mothers, there was evidence of maternal depressive symptoms and solicitation as intervening variables. Discrimination indirectly linked to decreased education expectations through higher levels of maternal depressive symptoms. Workplace discrimination indirectly linked to adolescents’ increased depressive symptoms through decreased maternal solicitation.

Adolescent gender moderation

To test adolescent gender as a moderator, we first estimated the hypothesized models unconstrained across gender and then compared them to models invariant by gender. The overall chi-square difference test was significant for fathers, Δχ² (37) = 72.87, p < .001, and mothers, Δχ² (37) = 73.98, p < .001, suggesting variation for boys versus girls. Follow-up analyses indicated that for the father model adolescent gender moderated two indirect pathways. The indirect pathway linking fathers’ self-direction to adolescents’ increased depressive symptoms through higher levels of paternal role overload was significant for girls, b₁b₂ = .03, 95% CI[.002, .084], but not boys, b₁b₂ = .00, 95% CI[−.016, .028]; Δχ² (3) = 9.87, p < .05. Fathers’ workplace discrimination was linked to decreased risk behavior through higher levels of paternal role overload and, in turn, increased paternal solicitation for boys, b₁b₂b₃ = −.02, 95% CI[−.070, −.001], but not girls, b₁b₂b₃ = .00, 95% CI[−.005, .008]; Δχ² (6) = 20.47, p < .01. For mothers, youth gender did not moderate any paths of interest.

Discussion

Our prospective test of the mechanisms linking Mexican-origin parents’ work conditions to adolescents’ adjustment across the transition from junior high to high school focused on an underrepresented group who face disproportionate risk for both stressful work conditions (Yoshikawa, 2011) and adolescent adjustment problems (e.g., Gonzales et al., 2012). This study advanced the current literature in two important ways. First, we examined the unique contributions of three work conditions (self-direction, work pressure, discrimination) simultaneously, rather than in isolation from one another. By examining multiple work conditions simultaneously, we were able to identify specific pathways of risk or protection for each aspect of work, while controlling for the influence of the other dimensions of work. Such an approach provides a more comprehensive understanding of the challenges facing this group. Second, we considered both mothers’ and fathers’ work experiences in two-parent families, as reviews have emphasized the importance of fathers in the lives of children, especially Latino children (e.g., Lamb & Tamis-Lemonda, 2004), and such research is particularly relevant in a study of Mexican-origin families, for whom two-parent families are prevalent (U.S. Census Bureau, 2014). Finally, we considered three key aspects of youth adjustment (i.e., depressive symptoms, risky behaviors, and educational expectations), across a critical educational transition period from junior high to high school when risks of adjustment difficulties are pronounced (Boyer, 2006; Dotterer et al., 2009; Newman et al., 2007).

Mexican-Origin Maternal and Paternal Work Experiences and Adolescents’ Adjustment

The results of this study underscore the significance of mothers’ and fathers’ work as important distal developmental contexts for Mexican-origin adolescents’ adjustment over the transition to high school. For mothers, after controlling for other work conditions, perceptions of workplace discrimination had the strongest effects on adolescents’ adjustment. Consistent with our hypothesis that negative work experiences would link to poor parental psychological functioning, parent-youth relationships, and adolescent adjustment, results supported a negative spillover mechanism linking maternal workplace discrimination to adolescents’ adjustment indirectly through parenting behavior (solicitation) or mothers’ depressive symptoms, and not a three-stage process. Particularly, mothers’ perceptions of workplace discrimination were associated with adolescents’ increased depressive symptoms via less solicitation and to adolescents’ decreased educational expectations via mothers’ increased depressive symptoms. Notably, the associations between mothers’ perceptions of workplace discrimination with solicitation and depressive symptoms were powerful enough to emerge even after accounting for the other dimensions of work. To our knowledge, this is the first study to suggest that Mexican-origin mothers’ perceptions of workplace discrimination may precipitate negative family relationship dynamics and youth adjustment. The findings, however, are consistent with literature linking adults’ experiences of general discrimination to depressive symptoms and parenting behaviors (Brody et al., 2008), and with views that ethnic minority youths’ adjustment is affected by incidents of discrimination even when they do not directly experience such events (Simons et al., 2002). These results suggest that as Mexican-origin adolescents are transitioning to high school, working mothers’ experiences of workplace discrimination may place youth at higher risk for increased depressive symptoms and decreased expectations for their future education during this critical time, when risks of adjustment difficulties are already pronounced, and may indicate one potential point of intervention for these families.

For fathers, experiences of work pressure and self-direction had the strongest effects on adolescents’ adjustment. The pattern of findings supports two distinct mechanisms relating fathers’ work experiences to the family domain. For work pressure, we found evidence for a spillover process, but it only partially supported our hypotheses. Fathers’ experiences of work pressure indirectly related to adolescents’ increased educational expectations and decreased depressive symptoms and risky behaviors via the links with high levels of role overload and adolescents’ perceptions of increased solicitation. In contrast to findings with European Americans that suggest that workplace stress may leave fathers fatigued and therefore less engaged in parenting (e.g., Crouter et al., 1999; Repetti, 1987), here greater work pressure and paternal role overload was associated with fathers’ increased solicitation of adolescents’ daily activities. Fathers of Mexican origin have been characterized as taking on the social roles of provider and authority figure for the family (Cauce, & Domenech-Rodríguez, 2002). Thus, the positive link between fathers’ work pressure and role overload and their efforts to learn about their children’s daily lives may reflect busy, overloaded fathers’ determination to make sure their children are on the right track in an attempt to fulfil their traditional social roles. Alternatively, it could reflect a compensatory mechanism, with fathers increasing their solicitation efforts to buffer their experiences of pressure and overload in their other roles. As the current findings are novel, it will be important in future research to replicate these results. In addition, the roles of cultural values and orientations (e.g., familism, machismo, orientation toward Mexican culture) will be important to consider in an effort to clarify further the linkage between work pressure, role overload, and increased solicitation among Mexican-origin fathers.

In partial support of our hypothesis that positive work experiences would relate to positive parental psychological functioning, parent-youth relationships, and adolescents’ adjustment, the findings for fathers’ occupational self-direction were consistent with a work socialization mechanism, such that fathers’ self-directed work experiences were linked to adolescents’ adjustment through less conflictual father-adolescent relationships. This is consistent with prior research with European American families finding positive features of work, including complexity (one dimension of self-direction), to be related to positive parent-child relationships (Greenberger et al., 1994). Our results suggest fathers’ experiences with autonomy and complexity may shape how they relate to their children and, thus, indirectly improve adolescents’ adjustment. Fathers’ self-directed work (Kohn & Schooler, 1982) may act as a form of social capital by shaping home environments that stimulate cognitive functioning, increase educational resources, and provide a model for future pursuits that translate into adjustment that is more positive. This social capital may be particularly important at a time when youth are transitioning to high school, forming concrete ideas about their future educational plans and making decisions about involvement in risky behaviors.

Taken together, the pattern of findings suggests that, in these two-parent families, the role of mothers’ work in adolescents’ adjustment operated primarily through negative spillover, whereas fathers’ work operated through positive mechanisms, although the separate analyses conducted for mothers and fathers did not allow for direct comparisons between the two. This pattern is consistent with evidence that Latina women have higher rates of negative work-family spillover as compared to Latino men, a pattern attributed to gender role and acculturation differences in work experiences (Roehling, Jarvis, & Swope, 2005). Our results highlight that mothers’ negative work-family spillover was a result of a contextual interpersonal factor (i.e., workplace discrimination) and not a result of intrapersonal work factors (i.e., work pressure, self-direction). This pattern may be attributed, in part, to gender differences in reactions to stressful experiences, such as discrimination, with women experiencing more psychological distress than men (Matud, 2004). Thus, mothers’ reactions to workplace discrimination may contribute to higher levels of depressive symptoms that compromise parenting practices and adolescents’ adjustment. For Mexican-origin mothers, who are likely to assume primary responsibility for the care of children and needs of the family (Roopnarine, Krishnakumar, & Xu, 2009), spillover from experiences of workplace discrimination may be particularly detrimental. Conversely for fathers, their sense of obligation to family as providers in this cultural context (Cauce & Domenich-Rodríguez, 2002) may dictate the kinds of experiences that spill over to the family context, possibly shielding the family from or counteracting negative work experiences by being involved with their children. Future research is needed that examines processes that may explain the differential effects of mothers’ and fathers’ interpersonal versus intrapersonal work conditions on adolescents’ adjustment.

The Role of Adolescents’ Gender

Adolescent gender moderated the work-family models tested here in two important ways. First, consistent with our hypothesis of more pronounced work-family spillover for same-gender dyads, we found that for sons, but not for daughters, fathers’ perceptions of discrimination related to decreased risky behavior through role overload and increased solicitation. Consistent with the pattern for fathers’ work pressure, these findings may also point to overloaded fathers’ determination to make sure their children are on the right track or may reflect fathers’ efforts to compensate for negative work experiences. Research by Brody and colleagues (2006) with African American adolescents suggests that the link between general perceived discrimination and adjustment is stronger for boys’ conduct problems than for girls’, possibly because boys more often express anger or frustration through behavior (Hetherington, 1989). Thus, fathers experiencing or perceiving high levels of discrimination may be particularly vigilant with their sons as they are making the transition to high school, a time of increased risk for problem behavior. Second, in contrast with our hypothesis, we found for daughters, but not sons, fathers’ perceptions of self-direction related to a greater sense of role overload, and, in turn, to increased depressive symptoms. This was the only instance of fathers’ work experiences relating to adolescent adjustment through fathers’ own psychological functioning. This is consistent with research suggesting that daughters may be more sensitive to family environment factors, and in particular, daughters may be more sensitive to fathers’ emotional states (Perry-Jenkins & Gillman, 2000). Girls are also more likely to internalize negative feelings arising from relational and life stress (Nolen-Hoeksema, 2004). Thus, daughters may be more apt to notice fathers’ feelings of role overload, which they then may internalize.

In contrast with our expectation of more pronounced spillover for same-gender dyads, particularly for mother-daughter dyads, we found no evidence of adolescents’ gender moderating the links between mothers’ work experiences and adolescents’ adjustment. This highlights the importance of mothers’ work for both sons and daughters. In contrast to the gender-intensification perspective that highlights the role of same-gender parents as key socialization agents, these results are consistent with literature on traditional role expectations for mothers in Mexican-origin families to be the primary caregivers, even when working (Cauce & Domenech-Rodriguez, 2002). Our own research has indicated that mothers in the current sample spend more time with both sons and daughters than fathers (Updegraff, Delgado, & Wheeler, 2009). Because of these parental role expectations for mothers, mothers’ negative work experiences may be salient for both daughters and sons. Overall, our findings on the moderating role of adolescents’ gender on work-family linkages add a more nuanced understanding of when the same-gender parent may be more or less salient for adolescents’ development. As prior research on work spillover to child adjustment has rarely attended to gender (Crouter & Bumpus, 2001) or focused on ethnic minority families, there is a need for replication of these findings. Future research with Mexican-origin families is needed to examine the role of cultural values and beliefs related to parenting practices with daughters versus sons to understand further the role of gender on work-family linkages.

Limitations and Future Directions

This study is not without limitations and points to avenues for future research. First, our findings document how mothers’ and fathers’ work conditions for primarily dual-earner two-parent Mexican-origin families are linked to adolescents’ adjustment. Though a large percentage of Mexican-origin family households in the U.S. include two parents (65%; U.S. Census Bureau, 2014) that are working (43%; U.S. Census Bureau, 2014), it is important to extend this work to examine the influence of parents’ work on youth adjustment in different family structures (e.g., single-parent, remarried) and work situations (e.g., single-earner). For instance, in single-parent families, parents and youth could be more vulnerable to the lone parent’s negative work experiences (Perry-Jenkins & Gillman, 2000). Second, the sample was drawn from one geographic region, characterized by a strong and established presence of Latinos (of predominantly Mexican heritage), and thus, a broader array of jobs may have been available. This may have implications for the strength of our observed effects. It will be important to test the role of parents’ work on adolescents’ development using a larger, nationally representative sample that includes geographic regions that vary in jobs available to parents of Mexican origin.

Conclusions

Our findings point to the importance of understanding the ecology of Mexican-origin adolescents’ educational expectations, depressive symptoms, and risky behaviors particularly because these youth are at increased risk for school dropout and externalizing behaviors (Gonzales et al., 2012). This investigation extended our knowledge of Mexican-origin mothers’ and fathers’ workplaces as distal ecological contexts that make a difference in terms of adolescents’ adjustment across the transition from junior high to high school. This study has significant implications for practice and policies related to family health and well-being over this transitional period. The findings suggest that fathers’ involvement in their jobs and their relationships with their adolescents are potential protective factors and mothers’ negative work experiences and psychological distress are possible points of risk for Mexican-origin family dynamics, and in turn, adolescents’ adjustment. Mental health practitioners, interventionists, and researchers working with communities of Mexican-origin will benefit from considering the contextual and social precursors that lead to family and adolescent adjustment as highlighted by this study. The current study suggests that programs promoting positive links between work and family should target parents’ distress and parenting practices and use the opportunity to alert parents to the ways that work may impinge on their well-being, family dynamics, and adolescents’ adjustment. Furthermore, the White House Summit on Working Families (White House Press Secretary, 2014) called for family-friendly workplace policies such as promotion of workplace flexibility and empowerment of workers, and access to paid leave and equal pay. Extrapolating from our findings, it would be important for effective family-friendly workplace policies to target reducing workplace discrimination and maximizing workers’ opportunities to exert autonomy and experience job complexity to have positive effects that not only improve workers’ functioning at work but family dynamics and youth adjustment. In sum, these findings have implications for work policy and targeted programs that promote positive development for this large and rapidly growing group of working, two-parent families.