Racial/ethnic discrimination, a practice of oppression, exclusion, and unequal treatment of stigmatized groups, has gained increased attention as a salient source of stress (Carter, Lau, Johnson, & Kirkinis, 2017; Goosby, Cheadle, & Mitchell, 2018; Williams & Mohammed, 2009). A survey conducted by the Pew Research Center (Neal, 2017) found that 58 percent of individuals in the United States reported believing that racial discrimination “is a big problem,” an increase of 10 percent from 1995.
While 67 percent of white/European Americans reported never experiencing racial/ethnic discrimination, a survey conducted in 2019 found that 76 percent of black/African Americans, 75 percent of Asians, and 58 percent of Hispanics indicated that they had been discriminated against as a result of their race or ethnicity (Horowitz, Brown, & Cox, 2019). This same survey also noted 65 percent of Americans uniformly acknowledged that it has become more common for people to express racist or racially insensitive views, suggesting that even for those who may not classify this as “discrimination” per se, racially charged rhetoric remains prevalent in American society. Notably, these numbers reflect the racial landscape of the United States prior to the killing of George Floyd on May 29, 2020. In the weeks after that time, McCaskill (2020) summarized a variety of public opinion polls which pointed to “a seismic quake” in the public’s recognition of the disparate treatment of people in racial and ethnic minority groups in the United States. While the long-term implications of these shifts in terms of people’s experiences of discrimination—and willingness to disclose these experiences—remains to be seen, what is clear is that racial/ethnic discrimination remains an important part of public discourse in the United States and across the world, as similar racial awakenings have also occurred internationally.
Perceived racial/ethnic discrimination has implications for health, with the extant literature providing evidence for an association between subjective appraisal of a discriminatory event and health outcomes. Specifically, perceived racial/ethnic discrimination is correlated with mental health outcomes such as depression, anxiety, and substance use (Clark, 2014; Paradies, 2006; Pascoe & Smart Richman, 2009; Priest & Williams, 2017; Unger, Soto, & Baezconde-Garbanati, 2016; Williams, Lawrence, & Davis, 2019) as well as physical health outcomes such as hypertension, cancer, obesity, high blood pressure, higher mortality rates, and overall self-reported health (Pascoe & Smart Richman, 2009; Williams et al., 2019). Further, perceived discrimination among racial/ethnic minority populations has been associated with underutilization of health services. In fact, researchers noted that underutilization of health services persisted even after controlling for sociodemographic characteristics, health care access, and physical and mental health (Burgess, Ding, Hargreaves, van Ryn, & Phelan, 2008).
A recent meta-analytic review of studies examining the association between discrimination and health services utilization found that one’s satisfaction with health services and perceived quality of care was inversely associated with the experience of racial discrimination. Further, those with higher self-reported racial discrimination reported fewer positive relationships with health care providers. Notably, although reduced treatment adherence and treatment seeking were associated with higher self-reported discrimination, the authors noted these associations may have been due to small sample sizes and publication bias (Ben, Cormack, Harris, & Paradies, 2017).
Trust in health care professionals and the health care system are often cited as potentially influencing service use amongst racial/ethnic minority populations (Hausmann, Kwoh, Hannon, & Ibrahim, 2013; Musa, Schulz, Harris, Silverman, & Thomas, 2009). The health care system reflects a microcosm of the larger society, with its own patterns of discrimination and prejudice; thus, racial/ethnic minority populations often fear discrimination by health care professionals, limiting their trust in the system and their likelihood of using mental health services (Dovidio et al., 2008; Hausmann et al., 2013; Klonoff, 2009; Musa et al., 2009). Even when mental health treatment is sought, perceived discrimination impacts treatment outcomes with racial/ethnic minority populations being more likely to prematurely discontinue treatment than their white counterparts, particularly in the United States (de Haan, Boon, de Jong, & Vermeiren, 2018; Mays, Jones, Delany-Brumsey, Coles, & Cochran, 2017). To cope with experiences of discrimination, individuals may turn to unhealthy coping behaviors such as substance use (Gee, Spencer, Chen, Yip, & Takeuchi, 2007; Hunte & Barry, 2012; Metzger et al., 2018). Further, the association between experiences of racial discrimination and unhealthy coping behaviors may be impacted by race/ethnicity. For example, Borrell et al. (2007) found that for black participants, but not white participants, racial discrimination was associated with marijuana, tobacco, and alcohol use.
Although a number of adverse health outcomes have been associated with exposure to racial/ethnic discrimination, some questions still remain. Of importance, there is no clear consensus on how best to measure racial/ethnic discrimination. This is in part because racial/ethnic discrimination is a unique and multifaceted stressor that can be measured at different biopsychosocial levels (Goosby et al., 2018; Todorova, Falcón, Lincoln, & Price, 2010). However, type and appraisal of racial/ethnic discrimination may differentially influence health outcomes. With this in mind, adequate operationalization and measurement are needed to enhance the accuracy and quality of research examining prevalence and associated outcomes of racial/ethnic discrimination.
A fundamental issue regarding measurement of racial/ethnic discrimination in the mental health field is lack of definitional clarity and inadequate construct representation in previous research. For instance, Boutwell et al. (2017) used data from the National Representative Longitudinal Study of Adolescent to Adult Health (Add Health), a prominent study receiving funding from over twenty-three federal agencies and yielding over 7,500 publications to date. Boutwell et al. reported that 25.2 percent of all participants reported discriminatory experiences due primarily to “unique and perhaps situationally specific factors other than race, gender, sexual orientation, and age” (2017, p. 5). Breaking down racism and discrimination by racial group, these authors found that 23.5 percent of whites, 31.9 percent of blacks, 27.2 percent of Hispanics, 11.6 percent of American Indians, 18.7 percent of Asians, and 27.0 percent of mixed-race participants reported having experienced discrimination. Discrimination was measured by a single item: “In your day-to-day life, how often do you feel you have been treated with less respect or courtesy than other people?” with response options ranging from “never” to “often.”
In response, Lee, Perez, Boykin, and Mendoza-Denton (2019) published a study questioning the validity of these findings on three premises:
Using data from the Pew Research Center’s 2016 Racial Attitudes in America Survey, these researchers found much higher rates, with 50 to 75 percent of black, Hispanic, and Asian respondents reporting discriminatory treatment. Further, they manipulated the framing of the question from one capturing respect (“In your day-to-day life, how often do you feel you have been treated with less respect or courtesy than other people?”); one capturing any type of discrimination (“In your day-to-day life, how often do you feel you have been treated with less respect or courtesy than other people because of your race, ethnicity, gender, disabilities, sexual orientation, or age?”); and one capturing race-based discrimination (“In your day-to-day life, how often do you feel you have been treated with less respect or courtesy than other people because of your race or ethnicity?”). The framing of the question influenced participant responses, with non-white participants reporting higher rates of race-based discrimination (d = 1.11), marginally higher rates of any type of discrimination (d = .44), and equivocal rates of less respect (d = .02). Thus, questions framed to specifically reflect race-based discrimination resulted in higher prevalence of discrimination among non-white participants than those simply suggesting experiences of disrespect.
An additional potential problem, which has thus far been unaddressed by previous research, is the impact of the race/ethnicity of the assessor and the match, or lack thereof, with the race/ethnicity of the participant. This information is rarely collected in substance use research. However, research on race/ethnicity matching in substance use treatment has indicated that treatment has resulted in better outcomes when therapists share the same race/ethnicity as their clients (Chapman & Schoenwald, 2011; Flicker, Waldron, Turner, Brody, & Hops, 2008). It is possible a similar dynamic may result in higher or more accurate response rates for discrimination when the assessor and research participant share the same characteristics. This also implies that racially/ethnically discordant participant-assessor dyads may create a tendency for participants to be less disclosive about experiences of discrimination. Coupled with psychometric inadequacy on the part of the assessment tools, this likely leads to significant underreporting of experiences of discrimination.
The confluence of these factors makes accurate examination of the impact of discrimination on treatment outcomes challenging to assess. This phenomenon is well-illustrated in the data example presented next, which was identified while attempting to conduct a larger study examining longitudinal associations among racial/ethnic minority emerging adults receiving substance use treatment.
The Global Appraisal of Individual Needs (GAIN) dataset originated in 1993 as a collaborative effort between clinicians, researchers, and policymakers to create a comprehensive and standardized biopsychosocial assessment tool; it has been a data source for over 560 publications (Chestnut Health Systems, 2021). The complete dataset comprises records from over thirty thousand adolescents and adults and includes data from several types of assessments (e.g., initial screenings, routine monitoring assessments, etc.). The current example included data from the GAIN Initial (GAIN-I), which is the baseline measurement conducted to identify diagnoses, inform placement recommendations, and assist in treatment planning decisions. Validated in both adolescents and adults, the GAIN-I’s main scales have good internal consistency (alpha over 0.90 on main scales, 0.70 on subscales) and test-retest reliability (rho over 0.70 on number of days and problem counts, kappa over 0.60 on categorical measures). Substance use scales are highly correlated with measures of timeline follow-back measures, urine tests, collateral reports, treatment records, and blind psychiatric diagnoses (Dennis et al., 2002).
Data for the originally planned study include self-reports of substance use, perceived discrimination, developmental life events (e.g., establishing committed relationship, birth of a child, obtaining full-time employment), perceived social support, socioeconomic status, and treatment retention from 2,780 racial/ethnic minority emerging adults (i.e., ages eighteen to twenty-five). Participants were 21 percent female and 47 percent self-identified as Hispanic, 26 percent African American, 22 percent mixed race/ethnicity, and 5 percent identified as “other” race/ethnicity.
Descriptive analysis of the data yielded only 208 participants endorsing experiences of discrimination (7.5 percent total: 4.0 percent African American, 6.3 percent Hispanic, 9.7 percent mixed race/ethnicity, 4.6 percent “other”), which is substantially lower than rates of discrimination measured elsewhere (e.g., 25 percent by Boutwell et al., 2017; 50 to 75 percent by Lee et al., 2019). This caused us to more closely examine the methodological context in which the data were obtained. Specifically, the GAIN battery includes the following question regarding perceived discrimination: “During the last twelve months, have you been under stress because of discrimination in [community, work, school, or transportation]?” With respect to Lee et al.’s (2019) framing questions, it is similar to the broadest item aimed at capturing any experiences of discrimination. Despite this, the GAIN item establishes a specific time frame for these experiences, whereas both Lee et al. (2019) and Boutwell et al. (2017) asked participants to reflect on their “day-to-day life.” The GAIN question further may limit participants’ frame of reference by asking them to narrow their responses to the categories of “community, work, school, or transportation.”
Unfortunately, comparable data from white/European American participants was not available as part of our larger study, thus we were not able to assess rates of endorsement of the discrimination item among white/European American emerging adults. Nevertheless, the comparison of rates of endorsement across studies suggests that the way the question is framed may be influencing the observed lower prevalence rates. Additionally, information related to the race/ethnicity of the assessor was not available, making it impossible to conduct analyses to determine the impact of race/ethnicity matching on the rate of endorsement of discrimination experiences.
While these theoretical explanations help us make sense of why the disparity in rates of reported discrimination might be occurring, they do not address the implications of such low rates of endorsement. Inadequate linguistic representation not only leads to underreporting, it also compromises external validity (i.e., the portion of the population for which the results may generalize). Substantively, the lack of definitional clarity and accuracy in discrimination-related data means that researchers and treatment providers are unable to form a clear understanding of how experiences of discrimination are impacting mental health, service utilization, and treatment outcomes. The implications of this are that assessment and intervention programs are likely missing a crucial piece of the puzzle when thinking about how to make services accessible and maximize effectiveness with racial/ethnic minority populations.
Given the widespread negative physical and mental health outcomes resulting from the discrimination summarized in this article, and the international shifting tides supporting public recognition of discrimination, the time for the field of substance use assessment and treatment to bolster our assessment of discrimination is now. In setting forth recommendations for improving measurements of discrimination in substance use research, Lee et al. (2019) provides an excellent starting point in their call for more precision in framing discrimination-related questions. As Lee et al. (2019) elucidates, questions about discrimination must be specific regarding the status-based characteristic being examined. Beyond this, discrimination is a multifaceted construct and as such is unable to be defined in one to two questions.
In order to adequately measure discrimination, researchers must begin from an adequate operationalization of discrimination and design questions that map onto each of the multiple facets. One example of how to concisely and reliably measure discrimination is the Everyday Discrimination Scale-Short Version (EDS), a five-item survey assessing potential experiences of discrimination in daily life (Sternthal, Slopen, & Williams, 2011) used in the Chicago Community Adult Health Study (CCAHS). However, the EDS uses the same “day-to-day life” verbiage, which may be confusing to participants. Therefore, it is important to ask individuals about specific contexts in which discrimination may occur (e.g., housing, employment, educational, or social settings) while also keeping in mind the specific context a study will examine in order to avoid overburdening participants with similar questions regarding various settings. Additionally, the severity of discrimination varies substantially from microaggressions to hate crime victimization. Therefore, the severity of discrimination events is another important facet to include in measurement. Temporal information, including information related to the frequency of discrimination events, is also important in developing a complete understanding of the experience of discrimination. Parsing apart these variables will allow researchers to develop a more robust understanding of how discrimination is experienced as well as a more nuanced understanding of the consequences of discrimination.
Finally, it is unclear whether ethnicity matching between interviewers and interviewees would impact the likelihood of individuals reporting discrimination. However, given the improved outcomes associated with ethnicity matching in substance use treatment, it is important to begin collecting information about the ethnicity of interviewers as well as participants in order to determine whether matching race or ethnicity might improve the likelihood of individuals reporting previous discriminatory experiences.
The field of substance use largely employs assessment tools that inadequately measure the extent to which discrimination occurs among ethnic minority populations. This is disappointing given the renewed public recognition of the pervasiveness of discrimination and its possibility for limiting access to and adapting substance use prevention and treatment programs to suit the needs of racial/ethnic minority populations. Unfortunately, discrimination and its sequelae are not a uniquely American problem and the events surrounding George Floyd’s death have led to calls for deeper examination of the negative impacts of discrimination and systemic racism on a global scale (Weine et al., 2020). Thus, while the current commentary focuses on substance use and discrimination in the United States, the same framework and recommendations for improved measurement of discrimination are equally applicable internationally.
Elise M. Yenne, PhD, is a forensic psychology postdoctoral fellow at Patton State Hospital in Highland, California. Dr. Yenne completed her MA and PhD in clinical psychology with a forensic emphasis at Sam Houston State University. Her research interests include bias in forensic and clinical assessment, while her clinical interests include forensic assessment and trauma-informed care, particularly with highly stigmatized groups.
Temilola K. Salami, PhD, is an assistant professor in the Department of Psychology and Philosophy and the director of The Health and Resilience Initiative for Vulnerable and Excluded Groups (weTHRIVE) lab. Dr. Salami received her BA in psychology from McGill University in Montreal, Canada and completed her MS and PhD in clinical psychology from the University of Georgia. She has received broad clinical training with her main clinical and research interests focusing on the psychological sequela of trauma and discrimination.
Tessa A. Long, MA, is a rising sixth-year doctoral candidate in the clinical psychology program at Sam Houston State University. She began her predoctoral internship in July 2021 at the University of Kansas Medical Center in the underserved populations track. Long’s research and clinical interests broadly encompass culturally responsive personality assessment and intervention practices.
Lauren J. Ryan, MA, is a doctoral candidate in the clinical psychology program at Sam Houston State University. Her research interests encompass issues related to multiculturalism/diversity, substance abuse, and juvenile competency. Ryan’s clinical interests involve providing therapy and assessment services to underserved groups, specifically individuals from ethnic minority backgrounds, juveniles, and justice-involved adults.
Amanda C. Venta, PhD, is an associate professor of psychology at the University of Houston. Her research focuses on parent-adolescent attachment with a subfocus on Latinx immigrant families. Dr. Venta’s research has been funded by the National Institute of Mental Health and the National Institute of Minority Health and Health Disparities.
Craig Henderson, PhD, is professor of psychology at Sam Houston State University and a licensed psychologist in the State of Texas. Dr. Henderson’s research interests concern the health behaviors of adolescents and emerging adults. He is chair of the SHSU clinical psychology program diversity committee and is committed to improving the diversity of the field of psychology.