A Prospective Examination of Neighborhood Social and Cultural Cohesion and Parenting Processes on Ethnic-Racial Identity among U.S. Mexican Adolescents

Abstract

Neighborhood social processes may have important implications for parenting processes and ethnic-racial identity (ERI) processes and content in adolescence. Past research suggests that adolescents whose parents engaged in more cultural socialization, an important aspect of parental racial socialization, had higher levels of ERI processes and content. Parenting, however, is also situated within neighborhood contexts and can be influenced by resources available in neighborhoods. For example, having neighbors who share mutual values, trust one another, and appreciate/celebrate one’s heritage culture may be a resource that promotes parents’ efforts to engage in cultural socialization. We prospectively examined (from ${\overline{x}}_{\mathrm{age}}=10.9-15.8\text{ years}$) a model in which U.S. Mexican parents’ perceptions of neighborhood social and cultural cohesion supported parents’ engagement in higher levels of cultural socialization and, in turn, promoted adolescents’ ethnic-racial identity processes and content. We tested a longitudinal mediation model with a sample of 749 U.S. Mexican adolescents (30% Mexico-born; 48.9% female) and their parents. Mother-adolescent models suggest mothers’ perception of neighborhood social and cultural cohesion in late childhood promoted middle adolescents’ ERI affirmation via intermediate increases in maternal cultural socialization. Similar patterns were observed for ERI resolution, but only for adolescents whose mothers were born in the U.S. We did not find evidence for mediation in the father-adolescent models. Findings are discussed in the context of the promoting nature of socially and culturally supportive neighborhood environments for U.S. Mexican families and adolescents.

Adolescence is a formative period in which youth are developing their self-concepts (Erikson, 1968). Ethnic-racial identity (ERI) refers to the meaning and significance that individuals place on their ethnic-racial backgrounds, in addition to the process by which adolescents develop meaning and significance over time (Sellers et al., 1998; Umaña-Taylor et al., 2014). Culturally and contextually informed sociological (Portes & Rumbaut, 2001) and developmental (García Coll et al., 1996) theories highlight that familial and extrafamilial contexts shape youth development, including ERI, but there is little research that examines their combined effects (c.f., Saleem et al., 2016; Thornton et al., 1990), especially among Latino youth (c.f., White, Knight, et al., 2018). Prior research has established that parents’ cultural socialization, practices that parents use to teach their children about their ethnic-racial backgrounds, influences adolescent development generally (Umaña-Taylor & Hill, 2020) and ERI development specifically (Hughes et al., 2006; Umaña-Taylor et al., 2014). However, according to the most recent decade review of research, limited extant work actually examines how extrafamilial contexts, especially neighborhood social processes, shape parents’ engagement in cultural socialization (see Umaña-Taylor & Hill, 2020, for a review).

In the current study, we examined if perceived social and cultural cohesion in U.S. Mexican families’ neighborhoods supported mothers and fathers engagement in higher levels of cultural socialization, and in turn, promoted adolescents’ ERI (see Figure 1 for conceptual model). Our conceptual model is situated within neighborhood and cultural-developmental theories. According to mainstream theories of neighborhood effects, one of the ways that neighborhood contexts influence development is through parenting processes (Leventhal et al., 2015; Noah 2015). Building from these broader neighborhood theories, the integrative model of minority youth development specifically proposes that neighborhoods can be promoting for family processes (including socialization) and, in turn, youth development (including ERI) if the neighborhood is supportive, consonant, and displays congenial and comprehensible norms (García Coll et al., 1996). For example, having neighbors who share mutual values, trust one another (Sampson et al., 1997), and appreciate/celebrate one’s heritage or culture (Nair et al., 2013) may be a resource that promotes parents’ efforts to engage in cultural socialization. Furthermore, we focus on parents’ perceptions of neighborhood social and cultural cohesion, to incorporate their phenomenology, which may be particularly important in research with minoritized families (Spencer, 2007). The current study focused on U.S. Mexican Latino families, the largest Latino subgroup in the United States (Ennis et al., 2011), in which cultural socialization is an important and normative parenting process (Hughes et al., 2006). Additionally, we focused on late childhood to middle adolescence, as cultural socialization and ERI are salient during this developmental period because social and cognitive maturation allows adolescents to think more abstractly about ethnicity and race (Umaña-Taylor et al., 2014).

Figure 1.

Conceptual model describing the association among perception of neighborhood social and cultural cohesion, mothers’ and fathers’ cultural socialization, and adolescent ERI.

ERI in Adolescence

ERI is comprised of two components: processes and content. Processes include exploring one’s ethnic-racial background (i.e., exploration) and gaining clarity about what one’s ethnic-racial background means to the individual (i.e., resolution). Exploration includes attending events that celebrate one’s heritage culture, talking to others about one’s ethnic-racial background, or reading books about one’s ethnic-racial group (Syed et al., 2013; Umaña-Taylor & Fine, 2004). Resolution is the degree to which individuals have a clear understanding of what their ethnic-racial background means to them (Umaña-Taylor et al., 2013). ERI content involves attitudes and beliefs individuals have toward their ethnic-racial group. An adolescent can have a high degree of ethnic-racial affirmation, which is the degree to which individuals have pride and positive feelings towards their ethnic-racial background (Umaña-Taylor et al., 2014). Because ERI is a salient aspect of development for adolescents of color, including Latino adolescents (Umaña-Taylor et al. 2013), it is important to consider how familial and extrafamilial contexts influence ERI processes and content, especially during adolescence.

Culturally-informed developmental theories emphasize the family context as a main socializing agent that shapes youth’s development, including ERI (García Coll et al., 1996; Spencer, 2007). Families play a critical role in providing tools and skills to help youth be successful in their culturally bounded contexts (Fuller & García Coll, 2010; White, Nair, Bradley, 2018). For ethnic-racial minoritized parents, teaching their children about ethnicity, race, and culture is an important part of ethnic-racial socialization (Hughes et al., 2006). Cultural socialization, an important component of ethnic-racial socialization, involves practices in which parents teach their children about ethnic-racial heritage through promoting cultural traditions and values, talking to children about prominent historical events and figures, celebrating cultural holidays, and speaking the culture’s native language (Hughes et al., 2006). In the current study, we focused on cultural socialization in early adolescence, as it is particularly important for Latino families to expose their early adolescents to values, traditions, and behaviors of their culture of origin because these are distinctly different from values, traditions, and behaviors of the dominant or mainstream U.S. culture (Umaña-Taylor & Yazedjian, 2006). Parental and familial engagement in cultural socialization during adolescence has been positively associated with youths’ ERI. For instance, Latino family engagement in more cultural socialization in adolescence has been linked to higher levels of exploration, in middle and late adolescence, (Umaña-Taylor & Fine, 2004) and resolution and affirmation, in late adolescence (Umaña-Taylor & Guimond, 2012). Moreover, in their longitudinal study, Umaña-Taylor & Guimond (2012) found that cultural socialization in middle adolescence was associated with higher levels of ERI exploration and resolution during later adolescence, and these associations were stronger among Latina girls than Latino boys.

Prior work has focused on both familial and parental (usually maternal) cultural socialization (Priest et al., 2014). For instance, respondents may report on familial cultural socialization. In this case, fathers’ cultural socialization efforts may be captured as part of adolescents’ responses to items like, My family teaches me about our family’s ethnic/cultural background (Umaña-Taylor et al., 2009, p. 51). Though a focus on familial cultural socialization is important, it is impossible to distinguish contributions among family members. Alternatively, respondents may also report on a specific parents’ cultural socialization behaviors, responding to items like, How often does your mother/father tell you about important and famous Mexican or Mexican American people in history (Knight et al., 2011). When parent-specific assessments are available, it is important to consider mothers’ and fathers’ contributions separately. For example, in many U.S. Mexican families, mothers are the carriers of culture and the main agents in culturally socializing children (Umaña-Taylor & Updegraff, 2013). Fathers, on the other hand, are typically the primary economic provider for the household and thus, may spend less time at home with the children (Baca Zinn & Wells, 2000; Delgado et al., 2011). However, recent empirical work finds that fathers’ cultural socialization may also have implications for youth’s ERI. In a longitudinal study examining U.S. Mexican youth in early adolescence, researchers found that mothers’ and fathers’ cultural socialization predicted youths’ ethnic pride (Hernández et al., 2014). In another longitudinal study, Knight and colleagues (2017) found that maternal cultural socialization in late childhood (but not early adolescence) and paternal cultural socialization in early adolescence (but not late childhood) were associated with youths’ ERI exploration in middle adolescence. The effects of mothers’ and fathers’ cultural socialization practices on ERI, therefore, may depend on developmental timing and the component of ERI (White et al., 2018). Thus, it is important to examine the unique effects of mothers’ and fathers’ cultural socialization practices on diverse components of ERI during adolescence.

Ethnic-Racial Socialization and Identity in Neighborhood Contexts

Though prior work with African American and U.S. Mexican families has situated parental ethnic-racial socialization within neighborhood social contexts (Thornton et al., 1990; Saleem et al., 2016; White et al., 2018), the current study tests a specific hypothesis whereby parenting processes mediate the association between neighborhood contexts and youth development (Leventhal et al., 2015; Noah, 2015). Substantial work has examined parenting and family processes that mediate the effects of neighborhood characteristics on adolescents’ behavioral and emotional health (e.g., Deng et al., 2006; Mrug & Windle, 2008; Taylor, 2000; White et al., 2015). This work consistently finds evidence of parenting and family processes mediating the association between neighborhood contexts and adolescent development with some limitations. First, the studies focused on negative aspects of neighborhood environments (i.e., risky, dangerous, disadvantaged) alongside costs to family processes and adolescent development. Second, there was a focus on family processes derived primarily from work with European American families (e.g., parental warmth, cohesion). Lastly, prior work has largely been conducted with cross-sectional data (Deng et al., 2006; Mrug & Windle, 2008; Stevenson et al., 2005; Taylor, 2000; Thornton et al., 1990) rather than longitudinal data. Our use of a longitudinal design is capable of testing true mediation and permits temporal sequencing (Leventhal et al., 2015).

In addition, there is limited research that has examined how neighborhood environments support parents’ engagement in ethnic-racial socialization practices specifically, and, in turn, adolescent ERI. Indeed, in the most recent review of ethnic-racial socialization research, Umaña-Taylor & Hill (2020) concluded that examination of neighborhood social processes would provide a more nuanced understanding of the extrafamilial contexts that shape parents’ cultural socialization. In a cross-sectional study, Stevenson et al., (2005) found that the sociocultural context of the neighborhood – in particular living in a predominantly Black neighborhood – was associated with higher ethnic-racial socialization of Black youth, but that study did not assess neighborhood social processes. In one longitudinal study with Black and African American families, Saleem et al., (2016) found that perceived neighborhood social cohesion, an important social process, interacted with parents’ experiences of racial discrimination to predict cultural socialization of boys. Our study moves beyond this limited extant work by examining the degree to which parents’ perception of neighborhood social and cultural cohesion, an important neighborhood social process, predicts parents cultural socialization and, in turn, adolescents’ ERI among U.S. Mexican families. We argue that having neighbors who share mutual values, trust one another (Sampson et al., 1997), and support or participate in Mexican culture (Nair et al., 2013) may be a resource that supports parents’ engagement in cultural socialization. In addition, these same neighborhood resources may also have implications for ERI, especially for U.S. Mexican adolescents (White, Zeiders, & Safa, 2018; Pasco & White, 2019). Parents’ perceptions of neighborhood social and cultural cohesion may promote adolescents’ ERI via parents’ cultural socialization. It is also possible that higher levels of neighborhood social and cultural cohesion will influence adolescents’ ERI directly, because nurturing and culturally cohesive contexts support adolescents to develop a healthy self-concept (Feinauer & Whiting, 2012).

The Current Study

Integrating mainstream neighborhood (Leventhal et al., 2015; Noah, 2015) and cultural-developmental (García Coll et al., 1996) theories, we prospectively examined whether U.S. Mexican mothers’ and (separately) fathers’ perceptions of neighborhood social and cultural cohesion in late childhood supported mothers’ and fathers’ engagement in early adolescent cultural socialization, and in turn, their middle adolescents’ ERI. The study addresses prior research that relied on cross-sectional data with Black and African American families (e.g., Stevenson et al. 2005; Thornton et al., 1990) by testing a prospective mediational model across three waves of data from late childhood to middle adolescence among U.S. Mexican families. Developmentally, as youth are transitioning out of childhood and into adolescence, mothers and fathers may be paying particular attention to neighborhood social and cultural norms and resources, leaning on extra-familial local resources to support their efforts at youth socialization (Yoshikawa, 2011). Additionally, middle adolescence is a particularly active time in ERI development (Umaña-Taylor et al., 2014). Overall, the current study addressed the following research question: Can neighborhoods that are high in social and cultural cohesion promote adolescent ERI, including exploration, resolution, and affirmation, by supporting mothers’ and fathers’ efforts to engage in cultural socialization? We hypothesized that mothers and fathers who perceived more neighborhood social and cultural cohesion in late childhood would engage in more cultural socialization practices in early adolescence which, in turn, would increase their children’s exploration, resolution and affirmation in middle adolescence.

Method

Participants

Data are from three waves (5^th, 7^th, 10^th grades) of a study investigating culture, context, and U.S. Mexican youth and their parents (Roosa et al., 2008). These waves corresponded with late childhood, early, and middle adolescence, respectively. The three waves of data, collected between Fall 2004 to Spring 2011, derive from a larger study of U.S. Mexican families of 5^th graders in a large metropolitan area of the southwest. The metropolitan area at the time was both an established area for large number of immigrants and an area of rapid growth in immigrant settlement, especially for U.S. Mexicans (Singer, 2004). Over the course of data collection, anti-immigrant elected officials, vigilante groups, and legislation received regular and repeated national attention (e.g., Sherriff Joe, the Minutemen, and SB1070, respectively; see Romero, 2011). Additionally, the state has been comprised primarily of non-Latino Whites (59%) and Latinos (30%). The remaining 11% consists almost evenly of non-Latino Blacks, non-Latino Native Americans and Alaskan Natives, and other races; the majority (91%) of Latinos were of Mexican-origin (Brown & Hugo Lopez, 2013).

U.S. Mexican families (N = 749) were recruited from 5^th grade classes as part of a study titled Culture, Context, and Mexican American Mental Health, which was approved by the Arizona State University Social and Behavioral IRB (protocol # 0905004020). Complete study procedures are detailed elsewhere (Roosa et al., 2008). Stratified random sampling was used to identify communities served by 47 public, religious, and charter schools throughout the metropolitan area. All study materials were available in English and Spanish. Recruitment materials were sent home with all fifth graders in the sampled schools and interested families were screened for eligibility. Of the eligible families, 73% of them participated. Informed consent and assent were obtained from the parents and youth, respectively. Participating family members completed computer assisted interviews and were paid $45 for participating at 5^th grade, $50 at 7^th grade, and $55 at 10^th grade. Of the original sample of families (N = 749) that participated in 5^th grade, which was shown to be a good representation of the demographic make-up of U.S. Mexican families in the metropolitan area (Roosa et al., 2008), 710 (95%) participated in 7^th grade, and 638 (85%) participated in 10^th grade.

Families (579 were two-parent and 170 were single-mother households) were screened according to the following criteria: 1) they had a target fifth grader attending one of the sampled schools; 2) the participating mother was the biological mother, lived with the youth, and identified as Mexican or Mexican American; 3) the youth’s biological father was Mexican-origin; 4) the youth was not learning disabled; and 5) no step-father or mother’s boyfriend was living with the youth unless he was the biological father. Father participation from two-parent households was optional; 467 (83%) of the 579 eligible (biological and living in the same household as the target youth) fathers participated at 5^th grade. The full sample of mothers and their youth (749 = 579 two-parent and 170 single-parent) represents one of the largest and most diverse samples of U.S. Mexican families (Roosa et al., 2008). The opportunity to use the subsample of fathers in two-parent families (n = 579) to study fathers’ cultural socialization is valuable, as fathers are underrepresented in research on cultural socialization (Priest et al., 2014).

In the sample of mother-youth dyads (and the subsample of father-youth dyads), 48.9% (48.4%) of the youth were female, the mean age at 5^th grade was 10.9 (10.8) years and the SD was 0.46 (0.47). In terms of youth nativity, generational status, and language, 70.2% (66.6%) of youth were born in the U.S. (ranging from 2^nd, to 3^rd generation and beyond) with the remaining of the youth being 1^st generation immigrants who were born in Mexico; and 82.4% (81.6%) of youth in the study were interviewed in English (the remaining in Spanish). In terms of parent nativity, generational status, and language, 74.4% (79.9%) of mothers (fathers) were first generation immigrants born in Mexico, with the rest being 2^nd and 3^rd generation immigrants born in the U.S.; and most 69.9% (76.8%) chose to interview in Spanish. Mean age was 35.9 (SD = 5.81) for mothers and 38.1 (SD = 6.26) for fathers. First generation immigrant mothers were on average 22.57 years (SD = 8.74) and fathers were 21.42 years (SD = 7.81) old when they came to the U.S. The average family income on a scale of 1 ($0,000 -$5,000) to 20 ($95,001+) in the sample of mother-youth dyads was 6 ($25,001-$30,000/year) and subsample of father-youth dyads was 7 ($30,001 - 35,000/year).

Measures

We assessed a range of relevant demographic variables. Parents reported on adolescents’ country of birth (0 = Mexico, 1 = United States) and sex (0=female, 1 = male). Mothers and fathers reported on annual family income (1 = $0,000 - $5,000 to 20 = 95,001+) and their own countries of birth (0 = Mexico, 1 = United States). Additionally, see Supplemental Materials Table S1 for a list of all items used to assess each construct.

Parents’ perception of neighborhood social and cultural cohesion (5^th Grade).

Mothers and fathers reported on their own perceptions of neighborhood social and cultural cohesion using a 6-item scale. Perceptions of neighborhood social and cultural cohesion assessed the degree to which neighbors shared mutual values and goals and trusted one another (e.g., “People around here are willing to help their neighbors”; Sampson et al., 1997) and support Mexican families, culture, and traditions (e.g., “People in this neighborhood appreciate Mexican culture and people”; Nair et al., 2013) using a 5-point scale, 1 (not at all true) to 5 (very true) and each parent was assigned a mean score. Factor structure, reliability, and evidence of construct validity are established elsewhere (Nair et al., 2013). Cronbach’s alpha was .84 for mothers and .82 for fathers. In the subsample of two-parent families, the correlation between mothers’ and fathers’ report of neighborhood social and cultural cohesion was small (r = .24, p < .01).

Parents’ cultural socialization (5^th grade, 7^th grade).

Mothers and fathers reported on their own cultural socialization practices by responding to the 10-item Ethnic Socialization Scale (Knight et al., 1993). The factor structure, reliability, and evidence of construct validity of the ethnic socialization scale are established elsewhere (Knight et al., 2011). Parents separately reported the frequency of their behaviors (e.g., “How often do you take your child to Mexican celebrations like Quinceñeras, Mexican weddings, or baptisms?”) using a 5-point scale, 1 (almost never or never) to 5 (a lot of the time), and each parent was assigned a mean score. Cronbach’s alpha was .74 in 5^th grade and .76 in 7^th grade for mothers and .75 in 5^th grade and .77 in 7^th grade for fathers.

Ethnic-racial identity (7^th grade, 10^th grade).

In consultation with the original developer of the Ethnic Identity Scale (personal communication, Umaña-Taylor, 2006), we first administered this scale during the early adolescent wave, as it was not developmentally appropriate for 5^th graders. Using a 4-point scale, 1 (does not describe me at all) to 4 (describes me very well), adolescents responded to seven ERI exploration items (e.g., “You have attended events that have helped you learn more about your background”) and four ERI resolution items (e.g., “You understand how you feel about your background”) from the Ethnic Identity Scale (Umaña-Taylor et al., 2004). The factor structure, reliability, and evidence of construct validity of ERI exploration and resolution are detailed elsewhere (White et al., 2011). Because prior work found that early adolescents had difficulty with the affirmation subscale of the Ethnic Identity Scale (White et al., 2011), adolescents responded to four ERI affirmation items (e.g., “You have a lot of pride in your Mexican roots”) using the Mexican American Ethnic Pride scale that assessed adolescents’ sense of affirmation and positive attitudes toward their ethnic-racial group (Umaña-Taylor et al., 2012). The scale was assessed on a 5-point scale, 1 (not true at all) to 5 (very true). A total score was computed by calculating the mean across all items of the subscale. The factor structure, reliability, and evidence of construct validity of ERI affirmation are detailed elsewhere (Umaña-Taylor et al., 2012). For ERI exploration, Cronbach’s alpha was .73 in 7^th grade and .81 in 10^th grade. For ERI resolution, Cronbach’s alpha was .74 in 7^th grade and .86 in 10^th grade. For ERI affirmation, Cronbach’s alpha was .70 for 7^th grade and .78 for 10^th grade.

Data Analytic Plan

Preliminary Analysis

Prior to conducting primary analyses, we conducted attrition analyses to examine whether families who participated in 7^th and 10^th grade interviews differed on 5^th grade child demographic (i.e., age, nativity, gender, household structure), mother demographic (i.e., marital status, age, annual family income, nativity, education), and father demographic variables (i.e., age, education) from those who did not participate. Next, we examined descriptive statistics (i.e., frequencies, means, standard deviations, correlations, skewness, kurtosis) of observed study variables using SPSS 24 (IBM Corp, 2016). Furthermore, because some families were geographically clustered within the same neighborhoods, which we defined as census tracts, we estimated intraclass correlation coefficients (ICC) to examine the proportion of neighborhood-level shared variance in all main study variables. We used maximum likelihood estimation with robust standard errors (MLR; Enders, 2013) to account for non-normality of the data.

As an additional preliminary step, we fit a multi-group structural equation model to examine if the hypothesized model fit equally well across four social position factors that can theoretically set parents and adolescents of color on different developmental pathways (García Coll et al., 1996), including adolescent gender, adolescent nativity, parent nativity, and, in the mother-adolescent sample only (because all fathers were from two-parent households), household structure (one-parent vs. two-parent households). In these preliminary analyses, we examined each social position variable as separate grouping variables in the multi-group framework. First, we estimated a model constraining all mediational paths to be equal across the grouping variable. Then, we estimated a model where all paths were free. We compared the relative fit of the nested models using Satorra-Bentler scaled χ² difference tests (Satorra, 2000). A non-significant χ² difference test indicated that adding model constraints did not lead to misfit. A significant χ² difference test indicated that one or more model constraints contributed to misfit. In the latter case, we freed each path one by one to examine which path constraint or constraints contributed to misfit to determine where differences among the grouping variable in the model occurred (Mackinnon, 2008). Any observed differences were carried into the primary analytic models.

Primary Analysis

After examining preliminary analyses, we examined longitudinal mediation models using path analysis in structural equation modeling framework using Mplus 8 (Muthén & Muthén, 2017). Based on attrition analyses, variables were included as auxiliary variables in the analytical models to reduce bias attributed to missingness (Enders, 2010). We examined whether perceived neighborhood social and cultural cohesion in 5^th grade supported mothers’ and fathers’ engagement in higher levels of cultural socialization in 7^th grade and, this, in turn, promoted adolescents’ ERI exploration, resolution, and affirmation in 10^th grade. We controlled for prior levels of mothers’ and fathers’ cultural socialization and adolescents’ ERI exploration, resolution, affirmation. Missing data were handled using a Full Information Maximum Likelihood estimator (Acock, 2005), which minimizes bias in parameter estimates but retains original sample size (Enders, 2013). All exogenous variables were allowed to covary in these models. Model fit was assessed using chi-square tests (Hu & Bentler, 1999); root mean square error of approximation (values less that .06 indicate acceptable fit; RMSEA; Hu & Bentler, 1999); standardized root mean square residual (values less than .08 indicate acceptable fit; SRMR; Hu & Bentler, 1999); and comparative fit index (values close to 1 indicate acceptable fit; CFI; Hu & Bentler, 1999). We tested mediation using the test of joint significance (Leth-Steetson & Gallitto, 2016). Prior research suggests that the test of joint significance holds more power than bias-corrected bootstrapping and produces reasonable Type I errors rates (Leth-Steetson & Gallitto, 2016). Because the mother sample and the father subsample derive from different underlying populations (e.g., mothers are from a population of U.S. Mexican families with both two-parent and single-parent households; fathers derive from a population of U.S. Mexican families with only two-parent households), we tested the hypothesized model separately.

Results

Preliminary Analyses

Attrition analyses examined whether families who participated in interviews in 7^th and 10^th grades differed on 5^th-grade child demographic (i.e., age, nativity, gender, household structure), mother demographic (i.e., marital status, age, annual family income, nativity, education), father demographic (i.e., age, education), and study variables from those that did not participate. The majority of demographic comparisons were non-significant, however families who participated in 10^th grade (n = 640) reported higher 5^th grade annual family income [t(730) = −2.986, p = .003] and children were less likely to be born in Mexico [x²(1) = 4.681, p = .031] compared to those who did not participate in 10^th grade (n = 109). On study variables, fathers who participated in 10th grade (n = 395) reported higher baseline perceptions of neighborhood social and cultural cohesion than fathers who did not participate [t(464) = 2.092, p = .042]. Because all other variables related to attrition were incorporated into the model testing framework and, therefore, accounted for in the FIML estimates, 5^th grade annual family income was included as an auxiliary variable to further address missingness.

Correlations, means, standard deviations, Cronbach’s alpha, skewness, and kurtosis are presented in Table 1. In the full sample of mother-adolescent dyads, mothers’ 5^th grade perceptions of neighborhood social and cultural cohesion were positively associated with their cultural socialization in the 5^th and 7^th grades. Maternal cultural socialization in the 7^th grade was positively associated with ERI exploration, resolution, and affirmation in the 7^th and 10^th grades. Adolescent ERI exploration, resolution, and affirmation in the 10^th grade were positively associated with one another. In the father-adolescent dyad sub-sample, fathers’ 5^th grade perceptions of neighborhood social and cultural cohesion was positively associated with their cultural socialization in the 5^th and 7^th grades. Paternal cultural socialization in the 7^th grade was positively associated with adolescent ERI exploration in the 7^th and 10^th grades, but not with ERI resolution or ERI affirmation. Adolescent ERI exploration, resolution, and affirmation in the 10^th grade were positively associated with one another. We examined ICCs for main study variables. The highest ICC was .059 (.004 - .059), which indicates that about 6% of the variance for fathers’ 5^th grade perceptions of neighborhood social and cultural cohesion was shared among individuals in the same geographic neighborhood (i.e., census tract). To address potential non-independence of observations nested within census tract neighborhoods, therefore, we used the CLUSTER IS command and TYPE = COMPLEX, which adjusted the standard errors across individuals living in the same census tract (Muthén & Muthén, 2017).

Table 1.

Summary of Descriptives and Correlations for Study Variables

Variables	1.	2.	3.	4.	5.	6.	7.	8.	9.	M	SD	α	S	K
1. Parents’ perceptions of neighborhood social and cultural cohesion (5th)	-	.21**	.18**	−.02	−.02	.004	−.03	.02	.09	2.95	.75	.82	.26	−.23
2. Cultural Socialization (5th)	.23**	-	.68**	.063	.10	.03	.073	.07	.09	3.00	.54	.75	−.41	−.14
3. Cultural Socialization (7th)	.22**	.66**	-	.13**	.17**	.10	.097	.09	.10	3.13	.51	.77	−.63	.11
4. ERI Exploration (7th)	.04	.18**	.17**	-	.48 **	.53**	.36**	.38**	.28**	3.75	.72	.73	−.33	−.47
5. ERI Exploration (10th)	−.07	.17**	.13**	.42**	-	.29**	.51**	.27**	.37**	3.77	.76	.81	−.40	−.08
6. ERI Resolution (7th)	.05	.14**	.13**	.54**	.28**	-	.38**	.52**	.34**	4.30	.73	.74	−1.13	.91
7. ERI Resolution (10th)	.03	.17**	.12**	.30**	.53**	.33**	-	.35**	.52**	4.31	.67	.86	−.75	−.21
8. ERI Affirmation (7th)	.03	.15**	.13**	.41**	.23**	.53**	.28**	-	.47**	4.54	.53	.70	−1.46	2.33
9. ERI Affirmation (10th)	.04	.18**	.17**	.21**	.39**	.28**	.53**	.39**	-	4.61	.53	.78	−1.78	3.52
M	2.79	3.01	3.189	3.73	3.74	4.31	4.27	4.51	4.60
SD	.83	.52	.50	.73	.77	.73	.70	.57	.52
α	.84	.74	.76	.73	.81	.74	.86	.70	.78
Skewness	.25	−.44	−.61	−.31	−.29	−1.17	−.73	−1.55	−1.61
Kurtosis	−.29	−.21	.10	−.48	−.33	1.15	−.21	2.65	2.68

Notes. Descriptive and correlation analyses were conducted in SPSS using listwise deletion. ERI = Ethnic-Racial Identity; S = Skewness; K = Kurtosis. Mothers’ descriptives and correlations are below the diagonal (N = 749) and fathers’ descriptives and correlations are above the diagonal (n = 579).

^**p < .01

The results of the preliminary multi-group structural equation models examining whether the hypothesized model fit the data equally well across key social position variables are presented in Supplemental Materials Table S2. For the mother-adolescent dyad sample, nested model comparisons indicated that the hypothesized model fit equally well across adolescent gender, household structure, and adolescent nativity, but not across maternal nativity. There was a difference in the association between maternal cultural socialization and ERI resolution across adolescents of U.S. born vs. Mexico-born mothers. Primary analyses for mothers, therefore, allowed this association to vary across these groups. For the father-adolescent dyad sample, nested model comparisons indicated that the hypothesized model fit equally well across adolescent gender, adolescent nativity, and father nativity.

Effects of Mothers’ Perception of Neighborhood Social and Cultural Cohesion on Adolescents’ ERI via Mothers’ Cultural Socialization

Based on the preliminary analyses with the mother and adolescent data, we tested a multi-group longitudinal structural equation model in Mplus 8 (Muthén & Muthén , 2017) and used maternal nativity as a grouping variable, constraining all mediational paths to be equal across maternal nativity, except the path from maternal cultural socialization to adolescent ERI resolution, which was free to vary across groups. Results, including model fit statistics, are presented in Figure 2. Model fit was considered acceptable. Mothers’ 5^th grade perceptions of neighborhood social and cultural cohesion predicted higher 7^th grade cultural socialization (B = .043 (.022), p < .05), controlling for prior levels of cultural socialization. Mothers’ 7^th grade cultural socialization positively predicted adolescents’ 10^th grade ERI exploration (B = .122 (.057), p < .05), and affirmation (B = .108 (.036), p < .05) controlling for adolescents’ 7^th grade ERI exploration and affirmation, respectively. For Mexico-born mothers, 7^th grade cultural socialization did not predict adolescents’ 10^th grade ERI resolution [B = −.014 (.072), ns]. For U.S.-born mothers, 7^th grade cultural socialization predicted increases in adolescents’ 10^th grade ERI resolution (B = .286 (.079), p < .001). Finally, mothers’ 5^th grade perceptions of neighborhood social and cultural cohesion had a negative direct effect on adolescents’ 10^th grade ERI exploration (B = −.098 (.041), p < .05), but did not have direct effects on resolution or affirmation.

Figure 2.

Test of hypothesized model for the association between 5^th grade mothers’ perception of neighborhood social and cultural cohesion and 10^th grade ethnic-racial identity via 7^th grade mothers’ cultural socialization (N=749). ERI = Ethnic-Racial Identity. Solid lines represent significant paths (p< .05) and dotted lines represent non-significant paths. Covariates are in grey. The value before the slash (/) is the estimate for U.S.-born mothers and the value after the slash (/) is the estimate for Mexico-born mothers. Unstandardized coefficients reported. Model fit: x²(34) = 78.288, p<.01; CFI = .969; TLI = .934; RMSEA= .059; SRMR = 0.075. *p<.05, **p<.01, ***p<.001

Using the test of joint significance (MacKinnon et al., 2002; Leth-Steensen & Gallitto, 2016), mothers’ 5^th grade perceptions of neighborhood social and cultural cohesion had a positive mediated effect on adolescents’ 10^th grade ERI exploration and affirmation through mothers’ 7^th grade cultural socialization, as the a and b paths in the path analyses were significantly different from zero. For 10^th grade ERI exploration only, the positive mediated effect of mothers’ perception of neighborhood social and cultural cohesion, however, was contrasted by a negative direct effect (B = −.098 (.041), p < .05). Finally, for 10^th grade ERI resolution, according to the same test of joint significance, mediation was present for youth with U.S.-born mothers, but not Mexico-born mothers.

Effects of Fathers’ Perception of Neighborhood Social and Cultural Cohesion on Adolescents’ ERI via Fathers’ Cultural Socialization

Based on preliminary analyses with the father and adolescent data, the hypothesized model fit equally well across key social position variables (i.e., adolescent gender, adolescent nativity, and father nativity), so a multi-group framework was not used. Results, including model fit statistics, are presented in Figure 3. Model fit was considered acceptable. Fathers’ 5^th grade perceptions of neighborhood social and cultural cohesion did not predict 7^th grade cultural socialization, controlling for previous levels of cultural socialization (B = .036 (.024), ns). In addition, fathers’ 5^th grade perceptions of neighborhood social and cultural cohesion was not associated with adolescents’10^th grade ERI exploration (B = −.028 (.050), ns), resolution (B = −.045 (.040), ns), or affirmation (B = .048 (.032), ns). Fathers’ 7^th grade cultural socialization did not predict adolescents’ 10^th grade ERI resolution (B = .112 (.060), ns) or affirmation (B = .069 (.046), ns). However, fathers’ 7^th grade cultural socialization predicted higher levels of adolescents’ 10^th grade ERI exploration (B = .180 (.071), p < .05). The pattern of results precluded further examination of any mediated effect.

Figure 3.

Test of hypothesized model for the association between 5^th grade fathers’ perception of neighborhood social and cultural cohesion and 10^th grade ethnic-racial identity via 7^th grade fathers’ cultural socialization (n=579). ERI = Ethnic-Racial Identity. Solid lines represent significant paths (p< .05) and dotted lines represent non-significant paths. Covariates are in grey. Unstandardized coefficients are reported. Model fit: x²(12) = 47.272, p<.01; CFI = .966; TLI = .897; RMSEA= .071; SRMR = 0.063. *p<.05, **p<.01, ***p<.001

Discussion

Addressing recommendations made in the most recent review of ethnic-racial socialization research (Umaña-Taylor & Hill, 2020), we examined whether U.S. Mexican parents’ perceptions of neighborhood social and cultural cohesion during their youth’s late childhood promoted their engagement in early-adolescent cultural socialization practices, and, in turn, youths’ ERI in middle adolescence. Consistent with study hypotheses, our findings showed that mothers’ perceptions of neighborhood social and cultural cohesion in late childhood promoted middle adolescents’ ERI affirmation via intermediate increases in maternal cultural socialization in early adolescence. Similar benefits were observed for ERI resolution, but only for adolescents whose mothers were born in the U.S. Findings for ERI exploration were equivocal. Contrary to study hypotheses, however, fathers’ perceptions of neighborhood social and cultural cohesion in late childhood did not promote middle adolescents’ ERI through paternal cultural socialization in early adolescence, regardless of adolescent gender, adolescent nativity, or father nativity. Addressing limitations in prior research that relied on cross-sectional methods (e.g., Deng et al., 2006; Mrug & Windle, 2008; Stevenson et al., 2005; Taylor, 2000, Thornton et al., 1990), we examined a prospective mediation model from late childhood to middle adolescence. Contributing to neighborhood and family research, the current study underscores how socially and culturally supportive neighborhood environments and mothers’ cultural socialization both support adolescents’ ERI affirmation and for some, ERI resolution.

The Role of Mothers’ Perceptions of Neighborhood Social and Cultural Cohesion on Adolescents’ ERI via Mothers’ Cultural Socialization

Consistent with the study hypothesis, mothers who perceived higher levels of neighborhood social and cultural cohesion in late childhood engaged in more cultural socialization practices in early adolescence which, in turn predicted increases in middle adolescents’ ERI affirmation, and, for those with U.S. born mothers, their resolution. Overall, our findings are consistent with theoretical models of neighborhood and family processes on youth development, which implicate parenting processes as mechanisms via which neighborhood environments influence youth development (García Coll et al., 1996; Leventhal et al., 2015; Noah et al., 2015). Parenting processes, specifically mothers’ cultural socialization, may be an indirect pathway by which neighborhoods influence adolescent outcomes over time, including ERI. This work advanced beyond several prior studies that examined parenting processes mediating the association between neighborhood contexts and adolescent development because we (a) focused on neighborhood social and cultural cohesion (instead of neighborhood deficits; Deng et al., 2006; Mrug & Windle, 2009), (b) examined parental cultural socialization (instead of parental warmth or family cohesion; White et al., 2015), and (c) employed a longitudinal design capable of testing true mediation (instead of a cross-sectional one; Stevenson, 2005; Taylor, 2000; Thornton et al., 1990). Lastly, we addressed a major gap identified in a recent review of ethnic-racial socialization by examining neighborhood social and cultural cohesion, a key neighborhood social process, as a predictor of parents’ cultural socialization (Umaña-Taylor & Hill, 2020).

We found that when U.S. Mexican mothers perceived higher levels of social and cultural cohesion in their neighborhoods, they engaged in more cultural socialization, an important aspect of ethnic-racial socialization that involves practices in which parents teach their children about their ethnic-racial heritage (Hughes et al., 2006). This finding is consistent with prior theoretical and empirical work that suggests that neighborhood sociocultural resources support normative parenting processes among immigrant populations (e.g., Portes & Rumbaut, 2001; Yoshikawa, 2011). Mothers, especially those from immigrant backgrounds, may rely on informal networks to help socialize their children (Yoshikawa, 2011). When mothers perceive their neighbors to have shared goals, mutual trust and support Mexican culture (Nair et al., 2013), they engaged in higher levels of cultural socialization. Thus, these neighborhood resources may be key to supporting U.S. Mexican mothers’ gendered roles to culturally socialize their adolescents. The current study extends prior research (e.g., Deng et al., 2006; Mrug & Windle, 2008; Taylor, 2000; Saleem et al., 2016; Thornton et al., 1990), such that our study included both social and cultural cohesion and how these neighborhood processes support parents’ cultural socialization practices. Prior work suggests that neighborhood social processes, like shared culture and values, have implications for parenting processes (see Leventhal et al., 2015 for review). Furthermore, cultural developmental theory specifically posits that neighborhoods can be promoting for parenting processes when the neighborhood is supportive, consonant, and provides adequate resources (García Coll et al., 1996). Drawing from both neighborhood (Noah, 2015) and cultural developmental (García Coll et al., 1996) perspectives, the current study shows that mainstream (social cohesion) and culturally-specific (cultural cohesion) neighborhood resources positively predict U.S. Mexican mothers’ cultural socialization, an aspect of socialization that is understudied as it relates neighborhood processes (Umaña-Taylor & Hill, 2020). Future research can consider examining how neighborhood social and cultural cohesion may support other parenting processes and domains.

Our findings also indicate that mothers’ cultural socialization levels in their youths’ early adolescence predicted higher levels of youths’ ERI exploration and affirmation, and for youth with U.S. born mothers, higher ERI resolution in middle adolescence. This finding is consistent with literature that shows that family-wide cultural socialization during adolescence supports middle and late adolescents to explore their ERI (Umaña-Taylor & Fine, 2004) and predicts higher levels of resolution and affirmation among Latino youth in late adolescence (Umaña-Taylor & Guimond, 2012). An important difference, however, is that the current findings document the influence of maternal cultural socialization specifically, not cultural socialization in the broader family context (Umaña-Taylor & Fine, 2004; Umaña-Taylor & Guimond, 2012). Our findings suggest that U.S. Mexican mothers, who are the carriers of culture (Umaña-Taylor & Updegraff, 2013), can help promote both ERI exploration and affirmation (and for some youth resolution) by providing their children with resources to learn about one’s ethnic-racial background, talking about historical events or prominent figures, celebrating cultural events (e.g., quinceñeras), and speaking in Spanish (Umaña-Taylor et al., 2014). These findings highlight the key role that mothers have in U.S. Mexican families to pass heritage culture onto their children.

The positive association between mothers’ cultural socialization in early adolescence and ERI resolution in middle adolescence was for youth with U.S.-born mothers only. This finding contrasts prior work that found that youths’ report of familial cultural socialization in late adolescence predicted higher levels of ERI two years later for youth with Mexico-born mothers, but not youth with U.S.-born mothers (Umaña-Taylor et al., 2013). That work, however, differs from the current study in important ways. Umaña-Taylor and colleagues (2013) utilized adolescents’ report of familial cultural socialization and related familial cultural socialization to an ERI composite of exploration and resolution. Consequently, the difference in our current findings from that of Umaña-Taylor and colleagues’ (2013) findings may be due to either parent-specific versus family-wide nature of cultural socialization practices, or our separate examination of exploration and resolution. Notably, subsequent theoretical and conceptual advancements strongly supports approaching each ERI component separately (see Umaña-Taylor et al., 2014).

In terms of the current study’s findings, it is important to consider that U.S.-born mothers may be able to couple cultural socialization messages with specific references to gaining clarity relative to what it means to be Mexican in the U.S. because these mothers also experienced their adolescent ERI development in the U.S. These U.S. born mothers and their adolescents, therefore, have the shared developmental experience of gaining clarity about what it means to be Mexican while growing up in a minoritized U.S. context. In contrast, Mexico-born mothers, who, on average in the current sample, experienced their adolescence in Mexico, may not be coupling their cultural socialization messages with specific references to gaining clarity about being Mexican in a minoritized U.S. context. These mothers do not share, with their children, the experience of gaining adolescent clarity about what it means to be Mexican in the U.S. This non-shared developmental experience may erect a cultural gap, of sorts, between the Mexico-born mothers’ cultural socialization efforts and the needs their adolescents have relative to developing ERI resolution in an American context. The same gap may not exist for mother-adolescent pairs who both grew up in the U.S. More research is needed to understand U.S.-born mothers’ cultural socialization practices, specifically, and the role mothers’ nativity plays in such processes, generally.

Findings for ERI exploration in the mother model were equivocal. After accounting for the positive indirect effect, there was a residual negative direct association between mothers’ perceptions of neighborhood social and cultural cohesion and adolescent ERI exploration. This finding was unexpected and merits further investigation. First, mothers’ perceptions may not match adolescents’ perceptions of neighborhood social and cultural cohesion. Thus adolescents may view their neighborhoods differently, and, possibly, view the shared values and Mexican customs as inhibiting to their opportunities to explore what it means to be both Mexican and American, which can both be parts of their ethnic self-concepts (Flores-Gonzalez et al., 2014, Knight et al., 2018; Safa et al., 2018). Second, and related to the first, the meaning of the 5^th grade neighborhood context for adolescents’ 10^th grade ERI exploration may be reduced or altered by developmental changes that take place between late childhood and middle adolescence. By middle adolescence, youth are gaining more autonomy and are likely exploring their neighborhood more independently (Leventhal et al., 2015). Adolescents are making their own meaning of their neighborhood and choosing what spaces (even within broader neighborhood contexts) are relevant for exploring their ethnic-racial background. Third, it could be that socially and culturally cohesive neighborhoods promoted ERI exploration earlier in adolescence and, consequently, youth in such neighborhoods were doing less exploration by 10^th grade. Finally, given the inter-neighborhood mobility of this sample (White et al., 2014), this finding could reflect, to some degree, confounds associated with mobility between 5^th and 10^th grade. Additional research documenting adolescents’ perceptions of neighborhood social and cultural cohesion and ERI is needed to explicate this unexpected finding.

The Role of Fathers’ Perceptions of Neighborhood Social and Cultural Cohesion on Adolescents’ ERI via Fathers’ Cultural Socialization

Contrary to study hypotheses, fathers’ perceptions of neighborhood social and cultural cohesion in their youths’ late childhood did not promote middle adolescents’ ERI through paternal cultural socialization in early adolescence, regardless of adolescent gender, adolescent nativity, or father nativity. This is because fathers’ perceptions of neighborhood social and cultural cohesion in youths’ late childhood did not predict paternal cultural socialization in early adolescence. One explanation for these non-significant findings is that fathers may be responding, parentally, to a different set of neighborhood signals. Prior research, for example, has found consistent significant associations between fathers’ perceptions of neighborhood danger and their parenting processes, but has documented inconsistent associations between mothers’ perceptions of neighborhood danger and their parenting processes (White et al., 2009; White et al., 2015). The father’s gendered role in U.S. Mexican family systems may be to attend and respond, parentally, to signals of neighborhood danger because of their sense of obligation, as the main authority figure, to protect the family (Umaña-Taylor & Updegraff, 2013). Neighborhood resources, on the other hand, do not pose a threat to the family. Therefore, neighborhood resources, including neighborhood social and cultural cohesion, may not be the neighborhood signals to which fathers’ parenting processes are responsive. Research that simultaneously examines neighborhood danger and social and cultural cohesion as predictors of parents’ cultural socialization could examine potential gender differences in parents’ responses to different features of the neighborhood environment.

Though perceptions of neighborhood social and cultural cohesion in their youths’ late childhood did not predict fathers’ cultural socialization in early adolescence, their cultural socialization did positively predict youths’ ERI exploration in middle adolescence. This pattern was also observed in the mother-adolescent models. Extending upon prior work on cultural socialization that focused mostly on mothers (see Priest et al., 2014 for review) or all family members (Umaña-Taylor & Guimond, 2012), the current finding underscores the importance of fathers in U.S. Mexican adolescents’ cultural development. For example, Zeiders and colleagues (2016) found that U.S. Mexican fathers’ familism values in early adolescence were associated with increases in youths’ familism values in middle adolescence and, in turn, set the tone for family members’ daily activities, which may be consistent with their authority figure roles (Umaña-Taylor & Updegraff, 2012). Thus, the gendered role of U.S. Mexican fathers in cultural socialization may be to encourage their children to spend time and engage in activities that help them to learn about their ethnic-racial background and get involved in their heritage culture, thus promoting ERI exploration. Whether fathers influence on ERI exploration has later implications for ERI resolution remains to be seen.

Fathers’ cultural socialization in early adolescence, however, did not predict youths’ ERI resolution or affirmation in middle adolescence. These patterns were unique from those patterns found in the mother-adolescent models. These father-specific findings are unique from prior work that focused on cultural socialization during middle adolescence in the broader family context and found that it promoted ERI resolution and affirmation in late adolescence (Umaña-Taylor & Guimond, 2012). Thus, our work raises important questions about fathers’ gendered roles with regard to cultural socialization in the broader family system and encourages future work to tease apart caregivers’ unique gender roles relative to cultural socialization of adolescents (White, Nair, Bradley, 2018). In one prior examination that considered maternal and paternal cultural socialization separately, the effects of mothers’ and fathers’ cultural socialization practices on ERI depended on developmental timing (Knight et al., 2017), a dimension not teased apart in the current study. Prior work combined with the current work suggests that cultural socialization from mothers and fathers may play unique roles at unique times for unique components of adolescent ERI. These patterns raise important questions for developmental scientists regarding how timing, parent gender roles, and specific developmental competencies play out in the associations between cultural socialization and ERI. Importantly, it is not possible to make a direct statistical comparison between the father findings and the mother findings in the current study because we estimated them in separate models and the samples derived from distinct underlying populations. That is, the father subsample is constrained to two-parent family contexts and the mother sample includes both two-parent and single-mother families. Thus, our discussion of the patterns of differences between the mother-model findings and the father-model findings must be interpreted with caution. Still, the mother-model findings generalized across household structure, thus the observed difference in the pattern of findings for mothers and fathers is not likely due to differences in household structure.

Summary, Limitations, and Future Directions

Prior theoretical models of neighborhood effects emphasized the importance of including family processes as mechanisms explaining associations between neighborhood contexts and youth development (Leventhal et al., 2015; Noah, 2015). Drawing from neighborhood (Noah, 2015) cultural developmental theoretical models (García Coll et al., 1996), we found that mothers who perceived more neighborhood social and cultural cohesion in their youths’ late childhood increased their cultural socialization practices in early adolescence, which increased youths’ ERI affirmation and, for youth with U.S. born mothers, their ERI resolution in middle adolescence. These findings did not extend to father-adolescent models. The current study addresses limitations of prior work regarding how neighborhood and family processes combined to influence adolescent development. First, we employed a longitudinal design to examine a true test of mediation. We focused on a culturally-informed positive neighborhood social process (i.e., social and cultural cohesion). The majority of prior work focuses on the negative aspects of neighborhood environments (Deng et al., 2006; Mrug & Windle, 2009; Taylor, 2000). We examined parent-specific cultural socialization and found that for mothers, it is a key mechanism through which neighborhoods indirectly influence adolescent outcomes, especially ERI affirmation and, for youth with U.S. born mothers, ERI resolution. Finally, our focus on parent-specific cultural socialization vs. family-wide cultural socialization raised important questions for future research.

The contributions must be considered alongside limitations. Although we focused on perceptions of neighborhood social and cultural cohesion, consistent with the importance of phenomenology (Spencer, 2007), future research should also consider various census and other neighborhood structural characteristics to incorporate objective aspects of the neighborhood context. The current study had a large sample of U.S. Mexican families that was a good representation of key demographic characteristics of the U.S. Mexican population in the metropolitan area (Roosa et al., 2008), including diversity on households structure. By design, however, only fathers from two-parent households were invited to participate in the study, which means that biological fathers of youth living in single-mother households were excluded. For these reasons, we were not able to test both mothers and fathers in the same model (because not all households had fathers). Future research may want to study two-parent families specifically and examine maternal and paternal processes in the same model. Although we were able to control for earlier assessments of parents’ cultural socialization in our models given our longitudinal design, which minimizes the impact of common method variance, we could not completely eliminate the influence of common method variance. Additionally, though we drew from multiple reporters across multiple waves, we did not capture adolescent reports of neighborhood social and cultural cohesion, and their reports may be especially important for considering any direct associations between neighborhood contexts and ERI. Future research should capture adolescents’ perceptions of neighborhood social and cultural cohesion, especially in middle adolescence, a time where youth gain more autonomy and are exploring and experiencing their neighborhoods independently.

In regards to generalizability, it is unclear how our findings would generalize across other regions of the U.S. The area of the southwestern U.S. in which this study was conducted was both an already established immigrant settlement area (making it comparable to other established areas), but also experienced rapid growth in immigrant settlement in the last 20 years (making it comparable to emerging immigrant destination; Singer, 2004). The state’s openly hostile political climate at the time of data collection (Romero, 2011) offers some parallels to today’s openly hostile anti-immigrant climate (Roche et al., 2018; Roche et al., 2020), thus supportive, consonant and congenial neighborhoods may continue to be important sources of support to U.S. Mexican families, especially mothers’ cultural socialization and, in turn, and adolescent ERI. Future work should explore such hypotheses.

Our findings show that U.S. Mexican mothers are accessing and relying on neighborhood resources (i.e., neighbors who appreciate heritage culture and have shared values) to support their cultural socialization efforts and this, in turn, supports adolescents’ ERI affirmation, and for some, resolution. This focus on the promoting aspects of neighborhood environments could inform future policies in terms of how to make cultural resources more accessible to Latino residents and identifying ways to promote social and cultural cohesion in the neighborhood. Government officials may reevaluate policies that divide neighborhoods, or disrupt their social and cultural cohesion processes. Such policies, including zoning and public disinvestment, could undermine parents’ ability to engage in cultural socialization and support ERI development in youth. Thus, efforts should be garnered at supporting existing cultural and social cohesion in U.S. Mexican families’ neighborhood, as these processes support maternal cultural socialization, and, in turn, youth ERI.