Introduction

Historical shifts in immigration and increased diversification of the US population have many real-world implications including the redrawing of societal color lines and a resurgence of interest in the developmental impact of skin color more generally speaking (Lee and Bean 2007). In light of xenophobia, anti-immigrant sentiments, bullying, and “veiled racism” in the modern-day US rhetoric (Shattell and Villalba 2008; Southern Poverty Law Center 2017), issues of race and phenotype are highly salient. A collection of race-related processes exist (e.g., unfair treatment, stereotypes, phenotypical perceptions), and such experiences are intricately linked. For example, the underlying core of racism and discrimination stem from biases related to “phenotypicality” or the degree to which individuals resemble prototypical members of their ethnic or racial group based on race-related physical characteristics such as skin color (Maddox 2004). However, it is surprisingly unclear how these factors work together to shape youth development, especially in light of those from Latinx backgrounds who represent one of the fastest growing immigrant groups in the US and for whom much of the current social derogation, disparagement, and increase in school bullying have been targeted (Ee and Gándara 2020; Huang and Cornell 2019). The overarching goal of the present study is to address these understudied associations among this marginalized group, comprised of individuals who are also understudied in the psychological literature in and of themselves, and to determine how different forms of discrimination (e.g., foreigner objectification, peer and adult discrimination) might be related to skin color satisfaction. In turn, satisfaction with skin color is examined as one possible mechanism that links such differential, unfair treatment to other important psychological and academic outcomes in adolescents’ lives.

Differential Treatment and Adolescent Adjustment

A vast theoretical and empirical literature focusing on the negative consequences of discrimination exists (see meta-analyses by Pascoe and Smart Richman 2009 and Schmitt et al. 2014). Conceptual models, such as those stemming from critical race theory (e.g., Delgado and Stefancic 2017) and developmental science (e.g., García Coll et al. 1996), have established clear ways in which social stratification and discrimination can hinder the progress and well-being of minoritized youth. Empirical support for such perspectives consistently point to detrimental associations between discrimination and diverse outcomes including low self-esteem and well-being, high depressive symptoms, and poor peer relationships, academic outcomes, and social and physical health (Pascoe and Smart Richman 2009; Schmitt et al. 2014; Williams et al. 2003). Yet, despite this ever-growing body of work, the specific mechanisms that link experiences with discrimination and outcomes could still use further clarification. Moreover, it is unclear whether different forms of negative interactions have similar mediators and outcomes, or if there might be nuances depending on the nature of the discrimination experience.

Much of the existing literature on perceived discrimination has centered on situations in which peers or adults are the perpetrator. For example, some of the field’s early, foundational work has pointed to important differences in trajectories and correlates of peer- versus adult/teacher-based discrimination (e.g., Greene et al. 2006; Wong et al. 2003), and more recent research continues to establish the utility of unpacking these distinct experiences (e.g., Niwa et al. 2014). Above and beyond such traditionally measured peer/adult discrimination experiences (e.g., being treated unfairly, not respected), Latinx adolescents must also contend with stereotypes and bias that are specifically tied to their assumed immigration status (Hwang and Goto 2008; Rosenbloom and Way 2004).

Foreigner objectification refers to the experience of being negatively treated due to assumptions about one’s “foreigner” status (Armenta et al. 2013), and has been increasingly studied as a salient context with which Latinx youth must cope (Chavez 2013; Kiang et al. 2019). A range of daily experiences can be captured by foreigner objectification, including more subtle microaggressions that presume foreigner status with ostensibly harmless assumptions and remarks (e.g., you speak English well; where are you from) (Juang et al. 2016; Sue et al. 2007). Among mostly adult samples, but also among adolescent and young adults, including those from Latinx backgrounds, foreigner objectification has been linked with a host of negative consequences, similarly to more general reports of interpersonal discrimination (Huynh et al. 2011; Kiang et al. 2019; Tran and Lee 2014). Indeed, among a sample of Latinx adolescents, Cavanaugh et al. (2018) found that foreigner objectification and peer discrimination were independently associated with both greater internalizing and externalizing symptoms. In terms of their overall position in society, existing work therefore suggests that Latinx youth must cope with subordination both in terms of presumed inferiority as well as with respect to cultural foreignness (Zou and Cheryan 2017).

The current study considers how each of these insidious experiences (i.e., foreigner objectification, peer discrimination, adult discrimination) is linked with adjustment among Latinx American youth. Diverse psychological outcomes (i.e., self-esteem, depressive symptoms), and a meaningful indicator of academic adjustment (i.e., value or importance attributed to academic success) were examined to further add to a comprehensive picture. In addition, given that race-related discrimination and stereotyping are often incited by easily observable visual cues (e.g., skin color, physical features) (Keith et al. 2017; Maddox 2004), satisfaction with skin color was implicated as one possible linking mechanism between perceptions of differential treatment and outcomes.

The Role of Skin Color

The social significance of skin color has its roots in colorism, which refers broadly to a system of privileges that traditionally holds phenotypically “white” skin (and more Eurocentric features) as being the most favored (Burke 2008; Glenn 2009; Hall 2005). This system of benefits and advantages allocated to those with lighter skin over darker skin can be found across diverse ethnic/racial groups and societies, including Latinx (Bonilla-Silva 2009), and has multifaceted implications for individuals and families, especially those from minoritized backgrounds (Burton et al. 2010). Such ramifications run the gamut of being external and structurally-driven (e.g., racial profiling; Glover 2009), as well as internally imposed, such as with respect to satisfaction with the color of one’s skin.

The present study focused on adolescents’ skin color satisfaction as one salient dimension of phenotype. Despite such a seemingly rudimentary assessment, asking youth to directly indicate how happy they are with their skin tone is high in face validity and also reflects an important marker that is largely visible and permanent (Thompson and Keith 2001). It can also be seen as an indicator of internalized racism, which could have particularly meaningful implications for well-being and adjustment (Hunter 2007). Moreover, existing research on African Americans suggests that, as opposed to more objective perceptions of actual skin color, skin color satisfaction is the key driving force that predicts associations with subsequent adjustment (Maxwell et al. 2015). Based on the long, world-wide historical legacy of colorism that affects not only African Americans but many societies and cultural groups including those from Latinx backgrounds (Dixon and Telles 2017), greater skin color satisfaction likely promotes positive development and hinders negative outcomes among Latinx youth. Above and beyond such possible main effects, it is also possible that skin color satisfaction signals ways in which individuals might cope with oppression (e.g., representing a form of internalized colonialism; Quiros and Dawson 2013) and serve to explain or link such discrimination experiences with subsequent outcomes.

To be sure, lower levels of skin color satisfaction could be associated with negative race-related interactions. Drawing on social identity perspectives, individuals who encounter group threat could choose to dis-identify from that group in an attempt to protect their sense of self (Tajfel and Turner 1979; Yip 2016). Given that one way in which to dis-identify with one’s group is to perceive oneself as a less prototypical member of the group (Arndt et al. 2002; Elsbach and Bhattacharya 2001), encounters with discrimination could heighten the salience of one’s perceived prototypicality with the group, which includes their phenotypicality (Ma et al. 2018), and lead to less satisfaction with such phenotypic characteristics. That is, to the extent that targets of discriminatory experiences attribute their negative encounters to their physical characteristics and visual markers that link them to their group, self-evaluations of one’s physical appearance could decline. Notably, as the current study aims to address, it is also possible that such processes related to skin color satisfaction differ based on the specific type of differential treatment that is experienced (e.g., objectification, peer-based and adult-based discrimination).

Considering Gender and Perceived Skin Color

Two variables that are potentially relevant to the associations between discrimination, skin color satisfaction, and outcomes are gender and self-perceived skin color. Both of these variables were included as controls in the current study. Gender is important to consider given prior work suggesting that skin tone biases reflect highly gendered processes (Kiang and Takeuchi 2009; Hunter 2005). Social standards of beauty often subject women to a singular focus on appearance (Harter 1999), and phenotypic implications do appear to be more profound for women than for men (Codina and Montalvo 1994; Thompson and Keith 2001).

Gendered differences in psychological and academic adjustment, as well as in discriminatory experiences have been also fairly consistently found in the literature (Davis et al. 2016; Kiang et al. 2012; Nolen-Hoeksema et al. 1999; Perreira et al. 2013). For example, women from immigrant families tend to report higher levels of depression than do men (Dion and Dion 2001). Also, among immigrant families including Latinx Americans, sons are often granted more freedom than daughters, which could result in how much they are exposed to American culture and, in turn, this has implications for the possibility of experiencing discrimination (Nesteruk and Gramescu 2012).

To isolate the effect of affective evaluations related to skin color satisfaction, another covariate that was considered in the current study is individuals’ perceptions or evaluations of their actual skin color. Although such differentiations have not been examined among Latinx Americans, some research with African Americans supports links between skin color satisfaction and adjustment above and beyond objective skin color perceptions (Maxwell et al. 2015). To build on this scant, emerging literature, the focus of the current study is on how happy or satisfied adolescents feel about their phenotype, while also controlling for their perceptions of their actual skin color.

The Current Study

Adolescents, particularly those from minoritized backgrounds, face a confluence of race-related experiences, such as perceived discrimination, and appearance-laden self-evaluative concerns. The primary aim of the current study is to investigate how these developmentally-relevant process work together to shape diverse outcomes among youth. Within this understudied research area, the specific focus is on early adolescents from Latinx backgrounds in emerging immigrant communities, who also represent an understudied population in terms of age as well as ethnic and geographic background (Perreira et al. 2013; Stein et al. 2016).

In terms of main effects, the current study drew on prior theoretical and empirical work (e.g., Armenta et al. 2013; Pascoe and Smart Richman 2009; Schmitt et al. 2014) and expected that multiple forms of social stratification in the form of negative, differential treatment (e.g., foreigner objectification, peer and adult discrimination) would be associated with poorer outcomes (e.g., low self-esteem, high depressive symptoms, low academic adjustment), as well as less satisfaction with skin color. Based on existing work on the legacy and implications of colorism (e.g., Bonilla-Silva 2009; Burton et al. 2010), greater skin color satisfaction was expected to promote benefits for psychological and academic adjustment and hinder negative outcomes. Skin color satisfaction was also examined as a mediating link between perceptions of unfair treatment and outcomes. Consistent with prior work suggesting that race-related discrimination and stereotyping are often incited by easily observable visual cues (e.g., skin color, physical features) (Keith et al. 2017; Maddox 2004), the current study expected that such experiences might be associated with individuals feeling unhappy with their phenotypic characteristics which, in turn, would be associated with more negative outcomes. Moreover, an additional study aim was to explore whether similar direct and mediating effects would be found across all types of treatment, considering different forms of discrimination (e.g., adult, peer, foreigner-based objectification). Also, gender and self-assessments of actual skin color were included as covariates in all analyses.

Methods

Participants

Data were collected from 175 adolescents from Latinx backgrounds (51.4% female; Mage = 12.86 years, SD = 0.68, range = 10.33–15.23 years; 86.9% US-born). For those not born in the US, the average age of immigration was 4.25 years old. The majority of adolescents were from families with Mexican heritage (89%). Other countries of origin included Columbia, Dominican Republic, Ecuador, El Salvador, Guatemala, Honduras, Nicaragua, and Puerto Rico. Mean family annual income for the sample as a whole was $23,020 (SD = $12,390) with an average of 4.72 (SD = 1.10) people living off that yearly income.

Procedures

Project staff visited two public middle schools in a semi-rural area of the Southeastern US to discuss the study and distribute study information to school personnel. The study was initiated in 2013 within an area that is considered an emerging immigrant community having recently experienced new and fast growth in its immigrant population. Per IRB guidelines, and in collaboration with the schools, flyers and letters about the study were mailed home to students. Using school call lists of 7th and 8th graders who were identified as Latinx by the school, project staff recruited interested and eligible families based on the following criteria: (a) both biological parents were also from Latinx backgrounds, (b) the mother was the resident caregiver of the participating child, and (c) youth ranged between 11 and 14 years of age. An additional phase of recruitment included door-to-door home visits to target families who could not be reached via phone. This process was repeated twice yearly, with the recruitment of rising 7th graders and new 8th graders in the schools. Eligibility criteria (e.g., age, ethnicity) was predominantly based on recent calls in the literature to better understand the developmental contexts of Latinx youth living in emerging immigrant communities, and particularly those in early to middle adolescence for whom race-related interactions and identity development are salient (Stein et al. 2016). The focus on mothers was based on the broader study’s goals to target the primary caregiver who, in these communities, tend to be the maternal caregivers (e.g., Kiang et al. 2019).

Across both recruitment strategies, the current study attempted to reach a total of 597 families via phone or door-to-door. Of these, 16 families had moved (3%) and 217 were not located (e.g., disconnected numbers, families not home; 36%). Of the families who were successfully contacted (n = 364), 47 were not eligible (13%), 125 declined (34%), 16 consented but did not complete interviews (4%), and 176 families consented and completed interviews (48%). Upon study enrollment, research assistants (including at least one Spanish speaking RA) visited families’ homes to administer consent forms and questionnaires. Data from mothers were collected as part of the larger study, but the current paper focuses only on the data collected from adolescents. All assessments were available in Spanish and English and administered according to participants’ preference (all but two youth chose to complete the materials in English). Youth were administered questionnaires presented via a laptop computer that was provided by the researchers. A research assistant was in the room to answer questions, but youth completed questions independently using this computer-assisted interview format, which lasted 1.5–2 h. Following survey completion, adolescent were given a $10 gift card.

Measures

Skin color

Developed by Bond and Cash (1992), one item was used to measure adolescents’ satisfaction with their skin color. Adolescents were asked, “How satisfied are you with the shade (lightness or darkness) of your own skin color?” Responses are scored a nine-point scale ranging from 1 = Extremely Dissatisfied to 9 = Extremely Satisfied. To isolate the specific effect of satisfaction, we included an additional item by Bond and Cash (1992) to control for subjective reports of actual skin color. In comparison to “most Black people,” adolescents were asked to rate their own skin color on a nine-point scale ranging from 1 = Extremely Light to 9 = Extremely Dark. This self-perceived skin color item was modestly correlated with skin color satisfaction (r = 0.16, p < 0.05; M = 4.20, SD = 1.78), and included in our models as a covariate.

Foreigner objectification

The Foreigner Objectification Scale (Armenta et al. 2013) asks participants to indicate whether they experienced specific events in the past year. Originally validated with ethnically diverse college students, including Latinx, sample items include, “Asked by strangers, ‘where are you from?’ because of your ethnicity/race” and, “Had someone speak to you in an unnecessarily slow or rude way.” The current study used the four-item version of the scale, as recommended by Armenta et al. (2013) and Pituc et al. (2009). Items are scored on a 1 = Never to 4 = Five or More Times scale (α = 0.85).

Peer/adult discrimination

Peer and adult discrimination was each assessed using a seven-item adapted version of Way’s (1997) school-based discrimination measure. For peer discrimination, on a scale from 1 = Never to 5 = All The Time, youth rate how often they have experienced different types of mistreatment due to their ethnicity/race (e.g., being treated with less respect, being insulted or called names) from “other students at your school.” Adult discrimination was assessed using the same items and the same scale, but asking adolescents to respond with respect to their experiences with “adults at your school.” Items were averaged with higher scores reflecting higher perceptions of discrimination (peer: α = 0.87; adult: α = 0.69).

Self-esteem

The ten-item Rosenberg (1979) self-esteem scale was used to assess self-esteem. Items are rated on a five-point scale ranging from 1 = Strongly Disagree to 5 = Strongly Agree, with higher values indicating higher self-esteem. A sample item reads, “I feel that I have a number of good qualities” (α = 0.83).

Depressive symptoms

Symptoms of depression were assessed using the Moods and Feelings Questionnaire (Angold et al. 1995). There are 33 items scored on a scale from 0 = Not True to 2 = Mostly True. Youth indicate how often certain statements applied to them in the last 2 weeks (e.g., felt miserable or unhappy, cried a lot). Higher scores reflect higher levels of depressive symptoms and were summed (α = 0.94).

Academic importance

Adolescents’ perceptions regarding the value or importance of doing well and succeeding in school was assessed through a six-item scale. Based on research from Eccles (1983) and used successfully in prior work (e.g., Fuligni et al. 2005), adolescents reported how important the following things are to them: that you do well in school, that you get good grades, that you get an “A” on almost every test, that you go to college after high school, that you be one of the best students in your class, that you go to the best college after high school. Responses are scored on a scale ranging from 1 = Not Important to 5 = Very Important with higher values reflecting higher importance (α = 0.81).

Results

Preliminary Analyses and Descriptives

Means, SDs, and ranges for all primary study variables and bivariate correlations can be seen in Table 1. Overall, participants reported experiencing relatively low levels of foreigner objectification (M = 1.57, SD = 0.61), peer discrimination (M = 1.67, SD = 0.64), adult discrimination (M = 1.27, SD = 0.41), and depressive symptoms (M = 11.04, SD = 11.04). Average levels of satisfaction with skin color (M = 6.92, SD = 1.95) and perceived academic importance (M = 4.45, SD = 0.604) were notably above the midpoint of the scales. In terms of basic associations, foreigner objectification was negatively associated with skin color satisfaction (r = −0.283, p < 0.001) and depressive symptoms (r = 0.333, p < 0.001). Links with academic importance approached significance (r = −0.141, p = 0.067), but links with self-esteem were not statistically significant (r = −0.118, p = 0.131). Skin color satisfaction was positively associated with self-esteem (r = 0.214, p < 0.01) and academic importance (r = 0.346, p < 0.001), and, only marginally, related to depressive symptoms (r = −0.137, p = 0.08).

Table 1 Means (SDs) and bivariate correlations among primary study variables
Full size table

Analysis Plan

Prior to conducting primary analyses, each variable was standardized by z-scoring to ease interpretation. The role of skin color satisfaction as a possible mediator in the relations between discriminatory experiences and outcomes was examined using Model 4 with 5000 bootstrapped samples in the PROCESS macro (Hayes 2018). Due to standard recommendations and the current study’s relatively small sample size, each outcome (i.e., self-esteem, depressive symptoms, academic importance) was tested in a separate model (e.g., Baron and Kenny 1986; MacKinnon 2000). Each discriminatory experience was also included in a separate model to avoid any clouding of the interpretation of the results given that each additional variable added to a multivariate model could absorb variance and reduce the power to detect a unique effect.

Tests of Mediation

Foreigner-based objectification

Results testing the mediational pathway with self-esteem as the outcome indicated that experiencing foreigner-based objectification was negatively related to skin color satisfaction (β = −0.269, p < 0.001; R2 = 0.108). As shown in Table 2, experiencing foreigner-based objectification was unrelated to self-esteem (β = −0.104, p = 0.138; R2 = 0.050). Skin color satisfaction was positively related to self-esteem (β = 0.192, p < 0.01) and, when taken into account, the relation between foreigner-based objectification and self-esteem was further weakened (β = −0.052, p = 0.466; R2 = 0.092). Moreover, a significant indirect effect of foreigner-based objectification on self-esteem through skin color satisfaction was found, (β = −0.052, SE = 0.024; z = −2.081, p = 0.037, 95% CI = [−0.104, −0.015]; see Fig. 1). Self-perceived skin color was (β = −0.152, p = 0.029), and gender was not (β = 0.229, p = 0.102) a significant covariate in the full model.

Table 2 Skin color satisfaction as a mediator of foreigner objectification and outcomes
Full size table
Fig. 1
figure 1

Significant indirect effect of foreigner based objectification on self-esteem through skin color satisfaction, β = −0.052, [−0.104, −0.015]

Full size image

In testing the mediational pathway with depression as the outcome, results indicated that experiencing foreigner-based objectification was negatively related to skin color satisfaction (β = −0.262, p < 0.001; R2 = 0.104). Experiencing foreigner-based objectification was positively related to depressive symptoms (β = 0.328, p < 0.001; R2 = 0.178). Skin color satisfaction was unrelated to depressive symptoms, (β = −0.031 p = 0.691). When skin color satisfaction was entered into the model, the link between experiences with foreigner-based objectification and depressive symptoms was statistically unaffected (β = 0.320, p < 0.001; R2 = 0.178), failing to provide evidence of a significant indirect effect of skin color satisfaction (β = 0.008, SE = 0.021; z = 0.380, p = 0.704, 95% CI = [−0.027, 0.045]). Self-perceived skin color was not (β = 0.114, p = 0.132), and gender was (β = −0.483, p = 0.002) a significant covariate in the full mediation model.

Similar to other models, results for academic importance indicated that experiencing foreigner-based objectification was negatively related to skin color satisfaction (β = −0.262, p < 0.001; R2 = 0.104). Experiencing foreigner-based objectification was negatively related to academic importance (β = −0.164, p = 0.035; R2 = 0.031). Skin color satisfaction was positively related to academic importance, (β = 0.332 p < 0.001) and, when entered into the model, rendered the link between foreigner-based objectification and academic importance non-significant (β = −0.077, p = 0.311; R2 = 0.135), providing evidence of a significant indirect effect (β = −0.087, SE = 0.034; z = −2.645, p = 0.008, 95% CI = [−0.161, −0.029]; see Fig. 2). Neither self-perceived skin color (β = −0.106, p = 0.158) nor gender (β = −0.050, p = 0.741) were significant covariates in the mediation model.

Fig. 2
figure 2

Significant indirect effect of foreigner based objectification on academic importance through skin color satisfaction, β = −0.087, [−0.161, −0.029]

Full size image

Peer discrimination

Results testing the mediational pathway with self-esteem as the outcome indicated that experiencing peer discrimination was marginally, negatively related to skin color satisfaction (β = −0.133, p = 0.099; R2 = 0.055). As shown in Table 3, experiencing discrimination from peers was significantly negatively related to self-esteem (β = −0.357, p < 0.001; R2 = 0.185), and skin color satisfaction was positively related to self-esteem (β = 0.174, p < 0.01). When skin color satisfaction was entered into the model, the link between peer discrimination and self-esteem was statistically unaffected (β = −0.334, p < 0.001; R2 = 0.220), failing to provide evidence of an indirect effect (β = −0.023, SE = 0.018; z = −1.334, p = 0.182, 95% CI = [−0.066, 0.006]). Neither self-perceived skin color (β = −0.113, p = 0.093) nor gender was (β = 0.227, p = 0.089) a significant covariate in the full model.

Table 3 Skin color satisfaction as a mediator of peer-based discrimination and outcomes
Full size table

Results testing depressive symptoms as the outcome indicated that experiencing discrimination from peers was marginally and negatively related to skin color satisfaction (β = −0.133, p = 0.099; R2 = 0.055). Peer discrimination was positively related to depressive symptoms (β = 0.514, p < 0.001; R2 = 0.317). Skin color satisfaction was unrelated to depression (β = −0.051, p = 0.459) and did not statistically affect the link between peer discrimination and depression (β = 0.507, p < 0.001; R2 = 0.320), failing to provide evidence of a significant indirect effect of experiencing discrimination from peers on depressive symptoms through skin color satisfaction (β = 0.007, SE = 0.011; z = 0.594, p = 0.553, 95% CI = [−0.015, 0.032]). Self-perceived skin color was not (β = 0.059, p = 0.397), and gender was (β = −0.479, p < 0.001), a significant covariate in the full mediation model.

Similar to other models, experiencing discrimination from peers was marginally, negatively related to skin color satisfaction (β = −0.133, p = 0.099; R2 = 0.055). Experiencing discrimination from peers was negatively related to academic importance (β = −0.158, p = 0.049; R2 = 0.026) and skin color satisfaction was positively related to academic importance (β = 0.351 p < 0.001). When skin color satisfaction was entered into the model, the association between experiences with discrimination from peers and academic importance became non-significant (β = −0.111, p = 0.132; R2 = 0.148). However, no evidence of a significant indirect effect of experiencing discrimination on academic importance through skin color satisfaction was found, (β = −0.047, SE = 0.033; z = −2.534, p = 0.125, 95% CI = [−0.123, 0.010]). Neither self-perceived skin color (β = −0.071, p = 0.346) nor gender (β = −0.073, p = 0.627) was a significant covariate in the full mediation model.

Adult discrimination

Results testing the mediational pathway with self-esteem as the outcome indicated that experiencing discrimination from adults was unrelated to skin color satisfaction (β = −0.047, p = 0.557; R2 = 0.038). Discrimination from adults was negatively related to self-esteem (β = −0.192, p < 0.01; R2 = 0.183) (see Table 4) and skin color satisfaction was positively related to self-esteem (β = 0.194, p = 0.007). Taking skin color satisfaction into account, the association between adult discrimination and self-esteem was statistically unaffected (β = −0.184, p < 0.01; R2 = 0.126), failing to provide evidence of an indirect effect (β = −0.009, SE = 0.017; z = −0.543, p = 0.587, 95% CI = [−0.043, 0.030]). Self-perceived skin color was not (β = −0.124, p = 0.085) and gender was (β = 0.292, p = 0.042) a significant covariate in the full mediation model.

Table 4 Skin color satisfaction as a mediator of adult-based discrimination and outcomes
Full size table

Results testing depression as the outcome indicated that experiencing discrimination from adults was marginally and negatively related to skin color satisfaction (β = −0.042, p = 0.597; R2 = 0.038). Experiencing discrimination from adults was positively related to depressive symptoms (β = 0.251, p = 0.001; R2 = 0.129). Skin color satisfaction was unrelated to depressive symptoms (β = −0.114, p = 0.459) and did not affect the link between adult discrimination and depressive symptoms (β = 0.246, p = 0.002; R2 = 0.141), failing to provide evidence of a significant indirect effect (β = 0.005, SE = 0.012; z = 0.422, p = 0.673, 95% CI = [−0.024, 0.028]). Self-perceived skin color was not (β = 0.084, p = 0.285), and gender was (β = −0.506, p = 0.001), a significant covariate in the full mediation model.

Similar to other models, experiencing adult discrimination was unrelated to skin color satisfaction (β = −0.042, p = 0.597; R2 = 0.038) when testing academic importance as the outcome. Experiencing discrimination from adults was negatively related to academic importance (β = −0.157, p = 0.048; R2 = 0.026) and skin color satisfaction was positively related to academic importance (β = 0.354, p < 0.001). When skin color satisfaction was entered into the model, the relation between discrimination from adults and academic importance became non-significant (β = −0.142, p = 0.056; R2 = 0.152). However, no evidence of a significant indirect effect was found, (β = −0.015, SE = 0.031; z = −0.516, p = 0.606, 95% CI = [−0.076, 0.052]). Neither self-perceived skin color (β = −0.077, p = 0.308) nor gender (β = −0.045, p = 0.762) was a significant covariate in the full mediation model.

Sensitivity Analyses/Alternative Models

As one way to address to robustness of the two indirect effects that were found, a separate set of mediational analyses tested for the possibility that foreigner-based objectification served as a mediator between skin color satisfaction and outcomes (e.g., self-esteem, academic importance), rather than the other way around. No significant indirect effects of skin color satisfaction on self-esteem (β = 0.014, SE = 0.023; z = 0.690, p = 0.490, 95% CI = [−0.029, 0.065] nor academic importance (β = 0.021, SE = 0.024; z = 0.940, p = 0.347, 95% CI = [−0.019, 0.083] through foreigner-based objectification was found.

An additional series of alternative models whereby gender was included not as a covariate but as a moderator of associations between discrimination experiences, skin color satisfaction, and outcomes was also conducted. These analyses tested the possibility of moderated mediational models. No evidence for such moderation by gender was found.

Discussion

Race-related discrimination and colorism have long pervaded the American landscape, and continued xenophobia and a divisive US sociopolitical climate endure to shape the importance that individuals attribute to skin color and cultural background (Lee and Bean 2007; Shattell and Villalba 2008). For example, reports suggest that anti-immigrant sentiments and hate groups are on the rise, and hate crimes against Latinx Americans have tripled since 2012 (Southern Poverty Law Center 2017). The current study examined the impact of such lived experiences among Latinx American youth, for whom issues of racial interactions, self-concept, and adjustment are highly salient. More specifically, associations between different discriminatory experiences (e.g., foreigner objectification, peer/adult discrimination), skin color satisfaction, and diverse outcomes (e.g., self-esteem, depressive symptoms, academic importance) were examined, along with additional tests of whether skin color satisfaction might explain or mediate some of the expected links between discrimination and outcomes, thereby providing useful insight into possible points of intervention.

As hypothesized and consistent with prior work (e.g., Huynh et al. 2011; Pascoe and Smart Richman 2009; Schmitt et al. 2014), foreigner objectification, peer discrimination, and adult discrimination experiences were generally associated with poorer reported outcomes. Foreigner objectification was directly linked with higher depressive symptoms and lower importance attributed to academic success. Negative associations between foreigner objectification and self-esteem were also evident, but did not reach statistical significance. Both peer and adult discrimination were each associated with lower self-esteem and greater depressive symptoms, and negative links with academic importance were not statistically significant but in expected directions. These results corroborate a steadily growing research base documenting the many and varied harmful consequences associated with perceived unfair treatment and discrimination (Williams et al. 2003).

In terms of main effects between discrimination experiences and satisfaction with skin color, it is notable that only the paths from foreigner objectification were statistically significant. Assumptions about foreigner status are typically based on phenotypic characteristics and easily observable markers (Armenta et al. 2013). Although more traditional definitions of discrimination, such as those also measured in the current study, are also based on such physicality, it appears that being directly targeted as an outsider could have particularly strong implications for one’s appearance satisfaction. Given that the nature of the objectification items themselves are largely based on stereotypes and external qualities, perhaps individuals more easily attribute such experiences to what they look like, and internalize negative thoughts about their appearance when they do have such experiences. Similarly, the specific items in the foreigner objectification measure reflect incidents that tend to be perpetrated by strangers. There are particularly notable implications of these results with respect to the specific dangers faced by Latinx communities (e.g., xenophobia, assumed documentation status) (Stacey et al. 2011; Yakushko 2009) that are largely based on easily observable markers such as skin color. Despite the lack of significance in associations between peer and adult discrimination and skin color satisfaction, it is notable that effects were in expected directions with greater discrimination being linked with lower satisfaction.

Also in line with prior work (Burton et al. 2010; Maxwell et al. 2015), skin color itself was found to be an important factor to consider in adolescents’ lives due to its direct links with outcomes. Although the significance levels of these direct paths varied depending on the specific model, the more adolescents reported feeling satisfied with the color of their skin, the more they also reported higher self-esteem, lower depressive symptoms, and higher importance of academic success. It is important to note that these effects were found after controlling for gender and also for self-perceived skin color. Hence, the consequences of skin color demonstrated in the current study appear to be due to subjective and affective evaluations related to satisfaction, rather than the actual color of one’s skin. Reflective of a global, pandemic phenomenon (Dixon and Telles 2017; Quiros and Dawson 2013), issues related to racial markers and physical appearance therefore continue to be salient aspects of Latinx adolescents’ lives. In light of these findings, it appears particularly worthwhile to better understand the processes by which Latinx youth might internalize aspects of colorism and societal values regarding skin color, especially given the current sociopolitical rhetoric and the public’s increased engagement in profiling based on race/ethnicity and other cultural markers (Zarate 2018).

In addition to the direct effects of skin color that were found, two significant indirect effects were also evident, which point to additional ways in which skin color plays a key role in youth adjustment. The negative effects of foreigner objectification to self-esteem and academic importance can be explained, in part, by mediation through skin color satisfaction. That is, foreigner objectification is associated with adolescents’ lower satisfaction with the color of their skin which, in turn, is linked to lower self-evaluations and lower reported importance of academic success. Global self-evaluations, such as overall self-esteem, have long been linked with evaluations that are more domain-specific, such as physical appearance (Harter 1999). Our results extend such work by implicating skin color satisfaction as a possible linking mechanism between negative social or contextual experiences that can hinder self-esteem. Moreover, perhaps adolescents are so preoccupied with their appearance and esteem that they are distracted from finding importance or meaning from school. Collectively, these results are noteworthy because they point to a specific target for possible intervention or prevention efforts. That is, perhaps adolescents’ health and well-being could be protected if their appearance satisfaction could be preserved in the face of race-related stereotypes and bias. Given that some of the negative effects of such treatment seems to filter through skin color satisfaction, greater attention towards promoting phenotypic satisfaction could be worthwhile.

Yet, it is important to note that skin color did not always play a mediating role. For example, in the regression models, neither direct nor mediating effects of skin color satisfaction were found with respect to depression. Although these null effects need replication, it could be that the strong direct effect of foreigner objectification, as well as the effect of gender, already explain so much variance in depressive symptoms that there is little left to predict from other variables. It could also be that skin color satisfaction is more relevant to outcomes that reflect internalized self-evaluations and academic outcomes that are tied to race-based stereotypes and expectations, rather than on more affective and mood-based outcomes. Again, more research is needed to better differentiate such possibilities.

It was important to take gender into account given prior work suggesting that implications related to skin color and phenotypic bias might be particularly strong for women as compared to men (Codina and Montalvo 1994; Thompson and Keith 2001). However, sensitivity analyses suggested that the direct and indirect effects of discrimination and skin color satisfaction on youth adjustment were not moderated by gender. That said, consistent with some prior work among older samples (e.g., Dion and Dion 2001; Nolen-Hoeksema et al. 1999), main effects of gender were found, most notably with respect to girls reporting higher levels of depressive symptoms than did boys. Any intervention efforts to improve youth adjustment and mitigate potential risk factors should therefore consider whether treatment or messaging might be most effective if specifically geared toward individuals’ gender. Addressing how gender might interact with discrimination, stereotyping, and broader processes of identity development would be consistent with recent calls to be more intentional in surfacing issues of intersectionality (Rogers et al. 2020; Syed and McLean 2016).

Despite the current study’s contributions, it is not without limitations. For example, the data were cross-sectional in nature. Although sensitivity analyses provided some confidence that the proposed models were arguably robust, it should be noted that causal links cannot be fully determined by the current work. It would be worthwhile for future research to incorporate longitudinal or even experimental (e.g., manipulating exposure to possible discrimination or bias) methods to better detangle issues of directionality. The present study’s measure of skin color satisfaction was high in face validity and was also consistent with prior work demonstrating the importance of subjective satisfaction with skin color rather than actual skin color itself (Maxwell et al. 2015), but this single-item measure is another notable limitation. For example, inquiring about not only skin color but also racial/ethnic features could extend the field’s understanding of phenotypic bias even further (Kiang and Takeuchi 2009).

While this study is unique in its focus on early adolescents from Latinx American backgrounds in an emerging immigrant community, who are understudied in terms of ethnic and geographic background in addition to age, the sample is limited in generalizability and more work with more diverse groups that span a larger age range could allow for greater insight on how discrimination, skin color, and outcomes come together for other ethnicities, for Latinx youth in more traditional areas of settlement, as well as with respect to possible developmental trends. For example, perhaps skin color itself is particularly relevant to early adolescents who are in the throes of identity development and for whom social acceptance and physical appearance are of the utmost priority (Harter 1999). It is also possible that the ramifications of phenotypes are widespread and equally intense and influential across the entire lifespan. Alternatively, it could be that the effects found here do not apply to older adolescents or, more likely, the effects could be even stronger as youth experience more opportunities for race-related interactions and possible bias.

In future work, it would also be important to examine multiple dimensions of influence, such as those stemming from the family. Parents play a critical role in helping children navigate discrimination experiences and build resilience in the face of stereotypes and bias (Hughes et al. 2006). Socialization messages that emphasize not only preparation for bias but also cultural pride and self-acceptance could help youth withstand appearance-related pressures and promote their feeling comfortable in their own skin (Stein et al. 2018). In doing so, it could also be important to determine whether negative messaging with respect to colorism might be transmitted within families. In light of recent work showing that race-related bias can have different implications for parents versus children (Kiang et al. 2019), it would be useful to examine phenotypic bias and discrimination at the family level. For example, research could investigate whether parents and older family members are affected in similar or different ways, and whether or how adolescents might absorb or model their family members’ strategies in coping with such experiences.

Conclusion

By examining the interplay between differential treatment (e.g., discrimination, stereotyping), skin color satisfaction, and adjustment, the results of the present study highlight how the long legacy of colorism continues to affect adolescents from Latinx backgrounds and contemporary US society. Among this ever-growing population, the negative experiences of discrimination remain highly consequential to adolescent development, with skin color satisfaction identified as one mechanism to explain its links with youth outcomes. As such, targets for intervention and for the proactive promotion of youth development in the face of group threat could rest in not only ameliorating the sources of threat itself, but also in boosting adolescents’ awareness of colorism and maintaining positive self-image with respect to phenotype and physical appearance. Among early adolescents, for whom fitting in and conforming to societal expectations of attractiveness are developmentally-salient, it appears particularly urgent to continue gaining insight into these insidious processes and to more systematically work towards boosting youth adjustment in the face of race-related bias.