Racial Discrimination and Conduct Problems Among Black American Youth: The Moderating Effect of Ethnic Racial Socialization

Abstract

Purpose:

Experiencing racial discrimination in everyday life can have lasting detrimental effects on Black American youth. Conduct problems are one of the consequences of discriminatory experiences. The present study examined whether the effects of discrimination on conduct problems vary by the levels of ethnic-racial socialization youth experience.

Methods:

A sample of 472 Black American youth provided five waves of survey data from ages 11-15 (June 2013 – December 2017). We tested the moderating effect of different types of ethnic-racial socialization (preparation for bias, cultural socialization, promotion of mistrust) on the association between discrimination and conduct problems using logistic regression analyses.

Results:

We found that cultural socialization significantly moderates the effects of racial discrimination on conduct problems. Promotion of mistrust was directly associated with increased risk for conduct problems.

Discussion:

Findings suggest that high levels of cultural socialization are beneficial to mitigate the effect of discrimination on conduct problems. Promotion of mistrust needs to be avoided but future research is warranted to investigate if promotion of mistrust can be helpful to youth in certain contexts.

Conduct problems are characterized by aggressive or delinquent behaviors during childhood and/or adolescence. For the Black American youth who are the focus of the present study, conduct problems are particularly concerning due to disproportionately negative consequences they experience from these behaviors. Consequences range from harsher disciplinary measures in school [1] to greater risk of wrongful conviction and longer and harsher sentences in the juvenile justice system compared to White youth [2,3].

Racial discrimination, unfair treatment based on one’s race [4], is a robust risk factor for conduct problems among Black American youth [5-8]. Institutional discrimination is caused by formal or informal organizational practices which treat people differently based on race. Examples include harsher school discipline for Black American youth, and less opportunities for Black American youth to take college preparation or honors class compared to their White counterparts. Interpersonal racial discrimination–also called “everyday racism,” which is the focus of the present study, often happens ‘between people’ and occurs frequently in everyday life. Examples of interpersonal discrimination include being closely monitored by a security guard at a store, treated disrespectfully by service people, or getting complimented for doing ordinary things (e.g., speaking well, being clean). Black American adolescents experience approximately, 5.21 individual or vicarious discriminatory events per day on average [9]. These experiences can have cumulative effects on Black American youths’ mental and physical health [10], potentially greater than the influence of major, more episodic experiences [11]. The psychological consequences of interpersonal racial discrimination can be severe–inducing symptoms of anxiety and trauma such as nightmares, hypervigilance, and mood disturbances [12,13]. For youth, exposure to racial discrimination can increase risk for conduct problems (e.g., stealing, physical aggression, truancy) as discrimination induces feelings of frustration, anger, and hopelessness which undermine youth self-regulation and increase risk taking behavior [8,14].

The pervasive and destructive influence of racial discrimination on youth development underscores the critical importance of changing American society in ways that provide Black American youth with the opportunity to develop and thrive. Given the slow pace of progress, however, investigations are needed that identify malleable psychosocial processes that protect Black American youth from the effects of racial discrimination on conduct problems. This information forms the foundation for developing programs and policies that support Black American youths’ development and the prevention of conduct problems in a society that systemically devalues them.

Accumulating evidence suggests that family ethnic-racial socialization (ERS) can affect the impact of interpersonal racial discrimination on youth conduct problems [5,15]. ERS refers to parenting practices that influence racial identity development of youth and their response to negative experiences that they may encounter due to their race or ethnicity [15]. ERS is particularly important during early adolescence which is when youth’s experience of discrimination increases [16], parents increase ERS messages [17], and teen’s ethnic identities are particularly malleable [18]. According to the foundational work of Hughes [19], ERS includes (a) cultural socialization, (b) preparation for bias, and (c) promotion of mistrust. Cultural socialization refers to parental practices which teach their child about the value of their race, culture, history, or traditions [15,19]. Examples include celebrating cultural holidays or taking children to places or events that reflect racial heritage. Cultural socialization predicts positive youth adjustment and is linked to adaptive coping styles and a positive racial identity during early and mid-adolescence [20,21]. Although direct, compensatory effects on externalizing problems have been observed [5], no studies document the moderating influence of cultural socialization on the link between racial discrimination and conduct problems. Based on priori research, youth bolstered by a sense of racial self-esteem induced by cultural socialization will be less likely to internalize implicit and explicit messages communicated during experiences of interpersonal racial discrimination [22]. We thus hypothesized that the effects of discrimination on conduct problems will be weakened, when cultural socialization is high.

Preparation for bias refers to helping youth to understand that racial bias exists and discussing strategies to deal with future encounters. For example, parents may assist youth in recognizing racial discrimination and support the identification of strategies to deal with it [23]. Findings regarding the effects of preparation for bias on youth adjustment are mixed [15]. Some studies found preparation for bias to decrease the effects of discrimination on conduct problems among fifth-grade youth [5,24] but other studies found no association [25]. Other research with adolescents age 12-18 finds that the effects of preparation for bias vary based on communication and trust in the mother-adolescent relationship [26] suggesting the need to consider other parentchild relationship dimensions in investigations of ERS. We hypothesized that when considered independently of other parenting practices, preparation for bias will attenuate the negative effects of discrimination on conduct problems among Black American youth when discrimination is elevated.

Promotion of mistrust involves encouraging youth to be wary and cautious about people from different racial/ethnic groups. This includes advising youth to keep distance from kids of other races or telling children about negative qualities people of other races have [27,28]. Studies of promotion of mistrust are scarce. Extant findings link promotion of mistrust to increased maladaptive behaviors–such as depressive symptoms, substance use, low academic achievement, and aggressive behavior from adolescents to young adults [15,25,29]. Nevertheless, parents tend to increase promotion of mistrust when they or their children experience elevated discrimination [15]. Whether promotion of mistrust might have beneficial effects for children in the context of elevated racial discrimination remains to be tested. Based on the existing literature, youth may find these messages distressing, increasing the salience of discriminatory events in their lives. On the other hand, in the context of high levels of discrimination, promotion of mistrust may encourage youth to attribute unfair treatment to wrongness of people from other races rather than to themselves and internalize racism. Therefore, we made no directional hypotheses regarding the moderating influence of promotion of mistrust on the link between discrimination and conduct problems.

Methods

Participants and Procedure

Black American youth and their caregivers were recruited from eight rural Georgia counties representative of areas of persistent rural poverty in the Southeast US [30]. Based on public school lists of fifth-grade students, we randomly selected families to recruit. Families with an 11- or 12-year-old youth who identified themselves as African American or Black American were eligible to participate. In total, 825 families were screened for eligibility, 625 were eligible, and 472 enrolled in the study (a 76% recruitment rate).

After consent and assent, a baseline assessment (Time [T] 1: June 2013 – August 2014) was conducted followed by four additional assessments: T2 (May 2014 – May 2015); T3 (July 2015 – July 2016); T4 (May 2016 – July 2017) and T5, 42 months post-baseline (February 2017 – December 2017). For each assessment, Black American researchers made home visits to collect data using audio computer-assisted self-interviews on laptop computers. Retention at T5 was 86.9%. Participation incentives were $100 for caregivers and $40 for youth at each assessment. All study protocols were approved by the University Institutional Review Board.

At baseline, youth’s mean age was 11.61 (SD = .62) and 49.2% of them were boys. Primary caregivers’ mean age was 37.2 years, and the majority of primary caregivers were the youths’ biological mothers (88.6%) followed by grandparents (5.7%) and fathers (2.8%). Regarding caregiver education, 18.2% had less than a high school education, 26.5% graduated from high school or passed the General Educational Development Test, 44.5% had some college education, trade school diploma or certificate or associate degree, and 10.8% had a bachelor’s degree or higher; 64.4% had incomes below the federal poverty threshold.

Measures

Conduct problems.

At T1, T2 and T5, youth completed a 15-item scale from the National Youth Survey [31]. Participants indicated how often in the past 6 months they engaged delinquent behaviors such as “taken a weapon to school,” “hit, or threatened to hit an adult like a teacher or parent,” “shoplifted something from a store,” or “purposefully damaged or destroyed property that did not belong to you.” Cronbach alphas were .72 (T1), .96 (T2), and .98 (T5). To measure presence of conduct problems, we recoded the items so that any exhibition of conduct problem was coded as 1 (yes) and no conduct problem was coded as 0 (no).

ERS.

At T1-T3, youth reported on their ERS with a slightly adapted version of the 15-item Racial Socialization Scale [19,32]. Adaptions involved modifying items to clarify their meaning based on focus group feedback. The scale asks youth to indicate how often during the past month their caregivers engaged in ERS behaviors on a three-point scale ranging from 1 (never) to 3 (three to five times). Cultural socialization measured how often caregivers teach about culture, history, and heritage of their race (e.g., “taken you to places or events that reflect racial heritage?”) with five items (α = .76 [T1], .78 [T2], .86 [T3). The six-item preparation for bias subscale measured how often primary caregivers teach their child about prejudice and discrimination against their own race (e.g., “talked with you about the possibility that some people might treat you badly or unfairly because of your race?”) (α = .72 [T1], .82 [T2], .88 [T3]). Promotion of mistrust assessed how often caregivers encourage weariness and distrust of people of other race (e.g., “told you not to trust kids from other racial or ethnic groups?”) using four items (α = .74 [T1], .72 [T2], .86 [T3]). Items were summed, and higher scores indicate more frequent ERS. To index cumulative youth exposure to ethnic racial socialization from T1 to T3 (age 11–13), we averaged the summed scores for each subscale across waves. Alpha for the cumulative scores exceeded .80 for each subscale.

Perceived racial discrimination.

At T3, a 13-item measure of racial discrimination was included in the study protocols [33]. Youth indicated how often during the past six months they experienced racial discrimination on a scale from 1 (never) to 4 (frequently). Example items included, “How often did others respond to you as if they were afraid because of your race?” and “How often have you been called a name or harassed because of your race?” Chronbach’s alpha was .95.

Covariates.

A socioeconomic (SES) risk index was constructed using four binary variables at T3. Based on caregiver report, we assessed whether the family is in poverty per federal guideline; caregiver was unemployed; family received Temporary Assistance for Needy Families; and caregiver had less than high school education. We counted the number of risk factors a family had and made an index (ranged from 0 to 4). Higher scores reflected higher SES risk.

Caregivers’ use of monitoring, consistent discipline, and nurturant parenting was measured using the 18-item Involved-Vigilant Parenting scale [32]. Youth reported the frequency of their caregiver’s parenting behaviors on a scale from 1 (never) to 5 (always). Example items included “How often does your caregiver discipline you by reasoning, explaining, or talking to you?” and “How often do you talk to your caregiver about things that bother you?” After reverse coding negatively weighted question items, scores were summed (higher scores indicated higher level of involved vigilant parenting). The scores from T1 to T3 were summed to index exposure to involved vigilant parenting from age 11-13 (α = .90 for the cumulative index).

The current dataset was originally collected for randomized prevention trial [34]. In this trial, youths were assigned to a control group or either one or two family-centered prevention programs. To address potential variability introduced by random assignment, experimental condition was controlled in all tests of study hypotheses.

Statistical analysis

Logistic regression analyses were conducted for each interaction term controlling for SES risk, involved vigilant parenting, baseline conduct problems, and ERS subscales. Models with significant interaction effects were examined further using Johnson-Neyman plots [35]. Little’s Missing Completely at Random test indicated that our data were missing completely at random (χ²[31] = 15.78, p = .99). We thus used full-information maximum likelihood estimation to deal with missing data. The logistic regression models and plotting were conducted using the Mplus version 8 (Muthen & Muthen, 1998–2017).

Results

Per Table 1, 31.6% of the youth reported engagement in conduct problems in the past 6 months at T5. Compared to youth who did not have conduct problems, youth who reported conduct problems at T5 were more likely to be from families with greater SES risk at T3 (χ²4] = 10.82, p = .03) and low levels of involved vigilant parenting from T1 to T3 (F[1] = 3.86, p = .05), and reported elevated promotion of mistrust from their caregiver T1 to T3 (F[1] = 8.03, p < .01).

Table 1.

Sample Characteristics

Characteristics	Overall		No conduct problem (n = 252)		Conduct problem (n = 149)		χ2/F2	p value
	N	%	n	%	N	%	χ2
Conduct problem (T5)
No	252	53.39
Yes	149	31.57
Gender
Girls	240	50.85	123	48.81	87	58.39	3.45	.063
Boys	232	49.15	129	51.19	62	41.61
SES risk factors (T3)
0	75	15.89	46	18.70	23	15.54	10.82	.029
1	159	33.69	99	40.24	46	31.08
2	108	22.88	58	23.58	40	27.03
3	65	13.77	35	14.23	24	16.22
4	29	6.14	8	3.25	15	10.14
	M	SD	M	SD	M	SD	F
Involved vigilant parenting (T1-T3)	67.67	8.59	68.54	8.51	66.81	8.52	3.86	.050
Racial socialization (T1-T3)
Preparation for bias	8.75	2.18	8.70	2.17	8.93	2.15	1.13	.288
Cultural socialization	8.84	2.19	8.71	2.31	8.97	1.95	1.42	.235
Promotion of mistrust	5.05	1.26	4.90	1.17	5.27	1.36	8.03	.005
Perceived racial discrimination (T3)	20.13	9.19	19.65	9.49	20.92	8.71	1.77	.184

Note. N = frequency, % = percentage, M = mean, SD = standard deviation.

Table 2 presents study results. Models 1-3 present interaction effects on conduct problems for preparation for bias, cultural socialization, and promotion of mistrust, respectively. Model two presents a significant interaction effect indicating that the association between perceived racial discrimination and conduct problems varied by the extent of cultural socialization (B = −.01, standard error = .01, p = .02). The Johnson-Neyman plot in Figure 1 presents the adjusted effect of discrimination on conduct problems (y-axis) along with 95% confidence intervals across different values of cultural socialization (x-axis). The shaded area on the left side of the plot indicates that for youth who received little cultural socialization, increases in discriminatory experiences are associated with elevated risk for conduct problems. The shaded area on the right side suggests that for youth who received high levels of cultural socialization from their caregivers, elevated discrimination is related to lower risk for conduct problems. Model three revealed that increases in promotion of mistrust were directly associated with heightened risk for conduct problems.

Table 2.

Logistic regression models (n = 472)

	B	SE	p value
Model 1
Gender	−.65	.24	.01
SES risk	.10	.10	.30
Involved vigilant parenting	−.01	.01	.65
Cultural socialization	.04	.07	.60
Promotion of mistrust	.18	.11	.11
Preparation for bias	−.07	.07	.34
Perceived racial discrimination	.01	.02	.59
Interaction: Preparation for bias x Perceived racial discrimination	−.01	.00	.27
Model 2
Gender	−.67	.24	.01
SES risk	.10	.10	.34
Involved vigilant parenting	−.01	.01	.66
Preparation for bias	−.06	.07	.39
Promotion of mistrust	.20	.11	.09
Cultural socialization	.02	.07	.82
Perceived racial discrimination	.01	.02	.46
Interaction: Cultural socialization x Perceived racial discrimination	−.01	.01	.02
Model 3
Gender	−.60	.24	.01
SES risk	.11	.10	.28
Involved vigilant parenting	−.01	.01	.62
Preparation for bias	−.09	.07	.24
Cultural socialization	.04	.07	.57
Promotion of mistrust	.24	.12	.04
Perceived racial discrimination	.01	.02	.45
Interaction: Promotion of mistrust x Perceived racial discrimination	−.01	.01	.06

Note. Logistic regression models assessing the association between youth’ perceived racial discrimination and conduct problems moderated by preparation for bias (model 1), cultural socialization (model 2), and promotion of mistrust (model 3). B = unstandardized beta; Gender (0 = girls, 1 = boys); T1 and T2 conduct problems and experimental conditions were controlled in all three models.

Figure 1.

Johnson-Neyman plot. The shaded areas indicate regions of significance. The x-axis indicates the score of subscales of ERS centered to the mean and y-axis represents the adjusted effect of discrimination on conduct problem. Solid line represents the effects of discrimination on youth conduct problem corresponding to the values of cultural socialization. Dashed lines show 95% confidence interval.

Discussion

Using a longitudinal sample of Black American youth from rural Georgia communities, we examined moderating effects of different types of ERS on the association between discrimination and risk for conduct problems. Consistent with past research investigating the influence of cultural socialization on related outcomes [15,36], we found that higher level of cultural socialization decreases the effect of discrimination on conduct problems. Interestingly, our results showed that, in the region of significant interaction, there was a negative association between discrimination and conduct problems. This suggests that when both discrimination and cultural socialization are high, some youths are galvanized in a manner that supports optimal functioning. Such resilient responding is consistent with research on post-traumatic growth [37] where subgroups of individuals who experience stress or trauma are able, with the presence of protective processes, to evince positive adaptation. Cultural socialization is distinct from preparation for bias and promotion of mistrust in that its focus is on the self (i.e., one’s own racial identity) rather than others (i.e., how people of other races might think and behave). Indeed, prior research found that cultural socialization promotes adaptive development by fostering self-esteem [38].

Contrary to extant research, we did not find a direct effect of racial discrimination on conduct problems. This may be consequence of power, measurement issues, or the longitudinal study design where we controlled for the baseline conduct problems. Other research suggests that the effect of discrimination on emotional distress is attenuated in low SES samples [39]. Additional research is warranted with diverse samples of Black youth.

Regarding promotion of mistrust, previous studies found it to be associated with elevated risk of maladaptive youth behaviors [40]. Consistent with previous studies [40], we found a direct effect of promotion of mistrust on youth conduct problems. Promotion of mistrust increased the risk for conduct problems when the levels of cultural socialization and preparation for bias were held constant and the effects of other covariates were controlled (Table 2). The moderating effect of promotion of mistrust did not meet conventional levels for statistical significance (p = .06). However, given that sample size may have limited power, we probed the moderating effect. Youth who received high levels of mistrust messages evince reduced risk for conduct problems when discrimination is high. We speculate that the parenting behaviors identified by Hughes et al. [17] – parents increasing promotion of mistrust messages when their children experience frequent racial discrimination–might have beneficial effects in extreme circumstances. Additional research, however, is warranted to confirm this speculation.

We did not find significant moderating effects for preparation for bias, which is inconsistent with previous studies [5,24]. This discrepancy could be due to power differences. Previous studies used larger samples (367 girls and 306 boys) than ours (240 girls and 232 boys) which could have affected p values. Also, the operationalization of conduct problems could be a source of the discrepancy. Previous studies also included a broader set of conduct problems which included behaviors such as alcohol or marijuana use.

In practice, our results along with past research emphasize the potential protective power of cultural socialization [15]. Caregivers should proactively promote racial pride for their child. In fact, a lack of cultural socialization may be harmful–in our study, it increased the negative effect of discrimination on conduct problems. Caregivers should exercise caution in promoting mistrust of other groups in conversations with their youth as our results, as well as other studies, showed that it is associated with increases in youth maladaptive behaviors [40]. It may be the case, however, that in extreme circumstances promotion of mistrust could be helpful in mitigating the effects of racial discrimination on youth maladaptive behaviors.

Our findings have some limitations which shed light on future direction. We relied on youth’s self-report to measure parents’ ERS which is subject to recall and social desirability biases. Future studies will benefit from including caregivers’ report on ERS as well. Although prior studies suggest that subpopulations of Black Americans (African vs. Caribbean Americans, immigrants vs., native-born Americans) can show different characteristics in ERS [27], we did not have data on this variability nor sufficient power to conduct subgroup comparisons. Future studies might benefit from investigating the differences in subpopulations. Other ERS strategies such as cultural pluralism or egalitarianism [41] may affect youths’ vulnerability to conduct problems. A post-hoc power analysis using Monte Carlo simulation [42] revealed that power to detect interaction effects (α = .05) was .26 for preparation for bias, 60 for promotion of mistrust and .71 for cultural socialization (.71). Additional studies are needed with larger samples. Lastly, the present study investigated interpersonal racism. Given that youth’s experience of institutional or structural racism can affect their risk of conduct problems, future studies would need to account for different forms of racism. Despite these limitations, the present study adds to the literature by providing empirical evidence that could inform Black American families of the impact of ERS on the link between discrimination and youth conduct problems. Future studies should continue to find protective mechanisms and processes (e.g., racial identity, emotion regulation) which could attenuate negative effects of racial discrimination on conduct problems.

IMPLICATIONS AND CONTRIBUTIONS.

The present study found that high levels of cultural socialization decrease the impact of racial discrimination on conduct problems whereas, low cultural socialization increases the influence of discrimination on conduct problems. Promotion of mistrust independently increases the risk for conduct problems.